Wikipedia:Village pump (technical)/Archive 108

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Page note[edit]

Note: This archive has a talk page at Wikipedia talk:Village pump (technical)/Archive 108


New feature: guided tours[edit]

I've been writing a lot about this lately, so I am going to be brief, but I just wanted to give folks a heads up about the first, preliminary release of a new feature from the editor engagement experiments team: interactive guided tours. You can read more about this in the following places...

We're only using this feature in a couple experimental places right now, but it has a lot of potential to do good, if used carefully. Please let me know if you have any questions, especially about creating tours. Me and the developers on the team would be happy to walk you through the process. Steven Walling (WMF) • talk 18:21, 1 February 2013 (UTC)

Your test tour goes off the edge of the screen when it points to the search box (Monobook, Firefox 18, 1920x1080). MER-C 06:31, 4 February 2013 (UTC)
Thank you for the report. I've filed it as bugzilla:44635. Superm401 - Talk 09:30, 4 February 2013 (UTC)

How to use this html code on wikipedia?[edit]

I want to use this html code on wikipedia, however it doesn't seem to recognise even the most simple html. How to make wikipedia to interpretate it as html and not text ..?

<div style="position: relative; left: 0; top: 0;">
  <img src="a.jpg" style="position: relative; top: 0; left: 0;"/>
  <img src="b.jpg" style="position: absolute; top: 30; left: 70;"/>

The goal is a track map. Electron9 (talk) 02:38, 5 February 2013 (UTC)

<img> tags are not allowed. Use wikimarkup instead, and wrap the image markup in spans if necessary to apply positioning. Anomie 02:46, 5 February 2013 (UTC)
If you want to overlay a marker on a map (as indicated on your test page), there is probably already a template to do the job. For example:
  • 1: {{Location map|USA|width=250|lat=33.755|long=-84.39|label=Atlanta}}
  • 2: {{Location map|USA Georgia|width=150|lat=33.755|long=-84.39|label=Atlanta}}
If you need to display uploaded images directly, type the file title in double square brackets:
  • 3: [[File:Map of USA GA.svg|thumb|75|Map with Georgia highlighted]]
1 2 3
Atlanta is located in USA
Atlanta (USA)
Atlanta is located in Georgia (U.S. state)
Atlanta (Georgia (U.S. state))
Map with Georgia highlighted
See Help:Wiki markup for a general introduction to wikitext (most of which is not like HTML), or Help:HTML in wikitext for information about the permitted markup that is similar to HTML (including style="..." attributes which you can use to apply complex inline formatting to individual elements).
Richardguk (talk) 04:18, 5 February 2013 (UTC)
I want several markers like a track on the same image. Even lines if possible. Electron9 (talk) 12:23, 5 February 2013 (UTC)
Lines I don't know about; but multiple markers is fairly simple, like this. --Redrose64 (talk) 12:51, 5 February 2013 (UTC)
You can overlay any two images using {{superimpose}}, or up to 50 images using {{superimpose2}}. – PartTimeGnome (talk | contribs) 22:10, 5 February 2013 (UTC)
{{Attached KML}} and an appropriate KML file can be used to draw a path. For an example, look at M-553 (Michigan highway). That article has a KML file which plots the route of the highway on the WikiMiniAtlas that pops up from the globe in the upper right corner. That route can also be displayed on either Google or Bing Maps as well. Imzadi 1979  00:57, 7 February 2013 (UTC)

Detecting sysop bit; selective page content.[edit]

Over at WT:CSD, some editors are looking at a way of streamlining the SPEEDY process for attack pages. We obviously want to allow an editor to immediately hide that kind of content from the general public, but we also want admins to be able to evaluate the CSD tag efficiently and either delete the page or remove the tag. The way to do that would be to have the {{db-attack}} template hide page content for viewers unless they have the sysop bit. So my question is whether there is any way for content to be dependent on the viewer. Possibly a parser function so you could have code like {{#ifeq: {{UserRight}} | Sysop | Page content | --blank-- }}. One of the contributors seems to remember having seen an experiment on conditional content before, but doesn't remember where. Anybody have any leads on such a thing? VanIsaacWS Vexcontribs 04:20, 5 February 2013 (UTC)

You can't modify static page content based on viewer settings (other than language, but that is a whole different ball of wax). This would be broken by cache, as the last person to cause the article content to re-render would set it for everyone who saw the cached version afterwards. This could be done with CSS and/or JS, but I think that is frowned upon on the English Wikipedia. There has been admin-only css in the past, but I think that has all been removed for various reasons. Probably your best bet is a collapsed section or simply commenting it out or blanking it (the history and edit pages are readily available, and what you suggest can be bypassed by non-admins this way anyway). --Splarka (rant) 08:08, 5 February 2013 (UTC)
This was something I suggested, at template talk:db-meta. I posted a link to the page at WT:CSD, but wasn't aware that discussion was taking place there as well as that page isn't on my watchlist. I'll try to gather it all together rathether than have different forums taking place. Incidentally, it does seem technically possible to hide text from nonm-admins, as discussed recently on this VP, using <span class="sysop-show">An optimist on the run! 10:19, 5 February 2013 (UTC)
That CSS right there is exactly the sort of thing I was thinking of. Thanks. VanIsaacWS Vexcontribs 00:04, 6 February 2013 (UTC)
For the record, this comes from MediaWiki:Group-sysop.css. Helder 13:19, 6 February 2013 (UTC)

Are Commons files transcluded into File: namespace?[edit]

{{Rename media}} is used to request renames of image files, but it's only designed for renaming images actually stored in the File: namespace. If someone attempts to request a rename of a file stored on Wikimedia Commons, the template complies and puts the page in Category:Wikipedia files requiring renaming. I would like to add some intelligence to this template, which checks if the file is (transcluded?) from Commons, and if so, writes an appropriate error message. Is that possible? Is there a magic word that can be checked, or something like that? See this diff for an example. Thanks, Wbm1058 (talk) 15:27, 5 February 2013 (UTC)

There is not a way to fix the template to detect whether the image is originally on Commons or on Wikipedia, I believe. The "transclusion" of the image is not transclusion as you might speak of in a template but is more closely a mirror of the image and its description page. However, all images which are from Commons do have a standard boilerplate box which appears which identifies the image as being from Commons. Did you miss that? --Izno (talk) 15:52, 5 February 2013 (UTC)
No, I didn't miss the message, but I can see how others might. I found this because a less technically sophisticated editor made an error, which I found, and I also determined that what they wanted to do had not been done. So I am working through the process of figuring out how to help such editors help themselves to accomplish their desired changes, because often editors just revert their attempts, say something like, "no you can't do it this way," but then the change never happens. This is interesting. Someone tagged this page for rename back on 19 September 2011, and that request was reverted 21 September 2011. Then over a year later, another editor tries to rename it, but makes an obvious error. This is where I come in. Every now & then, I check for transclusions of {{subst:requested move}}, a template that must be substituted. I put myself in the novice editor's shoes, and walk through the process of following the error messages generated by "my" template (I wrote the edit checks that write these errors), and the trail leads to advice to use {{Rename media}}. So, curious to see what would happen, I tried this edit, and, a bit to my surprise, rather than tell me I have to do that at commons, a file mover complied with my request! So, now, I'm wondering if I can fix this by just "uploading" the file that is stored at commons under a different name? Wbm1058 (talk) 16:47, 5 February 2013 (UTC)
Somehow I don't think re-uploading a commons file is a good idea. Why did the "file mover" software let this move happen? Couldn't the "file mover" software see that the file was on Commons? Now, if someone who has been entrusted with the file mover bit can't get this right, how does the foundation expect any average Joe editor to understand this complex process? Help! please. Can someone who understands WMF files make this right?? Wbm1058 (talk) 17:14, 5 February 2013 (UTC)
Now wait a minute, there is a history here, which seem to actually be physically located on English Wikipedia, not commons, right? An there is a "file" here too, it's just a single line transcluding {{DYKfile}}. And this "file" is on English Wikipedia, and it needs to be moved too? So moving a Commons file on English Wikipedia is really a two-step process, and I have just successfully requested one of two steps, moving everything except the image file itself. Now I need to go over to Commons, and make my first-ever edit there, to request the second step, moving the image itself. Is that right? Wbm1058 (talk) 17:35, 5 February 2013 (UTC)
I just discovered Wikipedia:File namespace noticeboard. I'm still playing detective on this issue, but let's continue the conversation over there. Perhaps I'll find some wp:file movers who respond there. Wbm1058 (talk) 21:51, 6 February 2013 (UTC)
  • What you really want is a bot that can move such tags to the appropriate Commons: file description page, I think. Don't know how hard that would be to set up. – Philosopher Let us reason together. 03:34, 7 February 2013 (UTC)

Help with {{BillboardURL}}[edit]

I'm pulling what's left of my hair out. The two references in User talk:Kww/BillboardURL should be identical, but obviously they are not. I cannot figure out why getting the text string from {{BillboardURL}} breaks the reference. I can't use plain <ref></ref> markup inside of {{singlechart}}, so I have to get the tag ref version to work.—Kww(talk) 06:04, 6 February 2013 (UTC)

Not sure what you're aiming for or how all the templates are meant to interact. But if you edit User talk:Kww/BillboardURL to replace
[{{BillboardURL|artist=Gotye|chartnum=305}} Gotye Album & Song Chart History]
[{{trim|{{BillboardURL|artist=Gotye|chartnum=305}}}} Gotye Album & Song Chart History]
then the external link will work. But I can't say I understand what is intended and why. — Richardguk (talk) 07:43, 6 February 2013 (UTC)
Template:BillboardURL had a space between the </includeonly> and <noinclude> tags. Peter James (talk) 10:32, 6 February 2013 (UTC)
That did it. Someone will have to explain to me precisely what that extra space did some day. Thanks.—Kww(talk) 14:42, 6 February 2013 (UTC)
It wasn't a space (which in this case would have been harmless) but a line break, see here. It's sometimes forgotten that when a template is transcluded into another page, the whole template is transcluded - the <noinclude>...</noinclude> sections are not stripped until later. Therefore, if any spaces or line breaks occur between the "real" template code and the <noinclude>, those spaces or line breaks end up in the final wikicode:
 Gotye Album & Song Chart History]}}
External links cannot contain line breaks, but a superfluous space here would have been harmless. --Redrose64 (talk) 17:44, 6 February 2013 (UTC)

WikiProject cleanup listing missing some categories[edit]

I am using tools like and I've noticed that it seems to be missing some categories:

Any idea how to include those categories in the toolserver report? --Piotr Konieczny aka Prokonsul Piotrus| reply here 17:00, 6 February 2013 (UTC)

Category:Articles that need to be wikified, which was populated by {{wikify}}, was deprecated and replaced with Category:Articles with too few wikilinks, which is populated by {{underlinked}}, among other more specific templates. I'd contact the author and ask them to update their tool. Where is the tool described and documented? Wbm1058 (talk) 18:43, 6 February 2013 (UTC)
The report was created by Smallman12q and Svick according to the footer, may be you could ask them for comment. Keith D (talk) 18:53, 6 February 2013 (UTC)
Thanks. See User:Svick/WikiProject cleanup listing and User talk:Svick/WikiProject cleanup listing#New Category:Articles with too few wikilinks needs added to reports. Wbm1058 (talk) 19:35, 6 February 2013 (UTC)

Hover over refs while editing[edit]

So I like the Reference Tooltip gadget that is now a default feature. However, it would be pretty helpful if it was expanded to be active on the editing screen as well, not just while viewing the article; how feasible an update would this be? (and I have no clue where or who to ask) EVula // talk // // 17:11, 6 February 2013 (UTC)

Try Wikipedia:Tools/Navigation popups. One of the better-kept "secrets" on Wikipedia. I really like this tool. Hovering over refs and links works when in "show preview" editing mode. – Wbm1058 (talk) 17:38, 6 February 2013 (UTC) {{Navigation popups topicon}}
While that is certainly a fantastic feature (and I agree with the secrets comment), I'd still like for the default feature to be improved, if only because it would also work for logged-out editors as well. (I also think the default tool looks better, but that's just my personal taste) EVula // talk // // 18:19, 6 February 2013 (UTC)
the cite-tooltip *is* available in edit mode, with some restrictions:
when pressing "preview" you can see the references tooltips, but only if the edit actually includes the "references" section. this means it would usually be available when editing the whole article, but not when editing a section. i believe this limitation can't be overcome without, practically, rewriting the whoel tool, so i doubt User:Yair rand (the author of cite tooltip) will implement it, but you can talk to him and see what *he* says about it...
without pressing "preview", cite tooltip is, of course, not available, but i can't imagine what the tool *would* do outside of "preview".
peace - קיפודנחש (aka kipod) (talk) 19:51, 6 February 2013 (UTC)

Talk:Bali Safari and Marine Park[edit]

For some reason Talk:Bali Safari and Marine Park is showing up at the toolserver list as "Unknown" class despite having been classified several months ago. This has been happening on and off for a while now, no matter what we do to the templates on the talk page. It doesn't seem to be affecting any other pages. — Crisco 1492 (talk) 06:14, 7 February 2013 (UTC)

The Toolserver's copy of the English-language Wikipedia database is corrupted. Most of the errors surround edits from early December 2012. The Toolserver admins are aware of this, but have yet to give a date to fix it. = TB (talk) 11:30, 7 February 2013 (UTC)
  • Would clearing the classification (blanking the page) and then reclassifying help? — Crisco 1492 (talk) 12:07, 7 February 2013 (UTC)
According to the toolserver list, which shows the class in the "Quality" column, Talk:Bali Safari and Marine Park isn't unknown-class, it's Stub-class and unknown-importance. The latter is because although the {{WikiProject Zoo}} here has |importance=low, the template is not set up to recognise the |importance= parameter. An edit like this would fix that; seven categories would also need to be created. --Redrose64 (talk) 14:08, 7 February 2013 (UTC)
(edit conflict) Scratch that. I now see that I was misled because I was looking at the assessment changes over the last five months - which have all been to the {{WikiProject Zoo}} banner, so Toolserver has not been updating the tables for any other project. Since 12 September 2012‎, the line
{{WikiProject Indonesia|class=stub|importance=low}}
has not been altered at all, thus Toolserver does not know that its data should be updated. If you alter the |importance=low in {{WikiProject Indonesia}} to something else, and save it, and some time later alter it back, Toolserver should pick up the change. --Redrose64 (talk) 14:23, 7 February 2013 (UTC)
  • Doubt the Zoo project wants to add the parameter. Would removing "importance=low" from that template affect the Indonesia one? If so, why? — Crisco 1492 (talk) 14:17, 7 February 2013 (UTC)

DEFAULTSORT problem ?[edit]

Found A Bell Is a Cup in Category:All articles with dead external links sorted at A but I have checked the article itself and it has DEFAULTSORT:Bell is a Cup coded so I would expect to find it appearing at B. It appears to be correctly sorted in other categories in which is appears (for example Category:1988 albums so the problem appears to be with the category rather than the article itself...

Any ideas?, Thanks GrahamHardy (talk) 13:35, 7 February 2013 (UTC)

It is also sorted at A in Category:Articles with dead external links from January 2012, so maybe(!) these hidden container categories do not regard defaultsort. Jared Preston (talk) 14:25, 7 February 2013 (UTC)
Is this the same problem described at Help talk:Category#DEFAULTSORT inconsistency (?) --Redrose64 (talk) 14:27, 7 February 2013 (UTC)
Yes it is ! the category is within a reference and the DEFAULTSORT comes after Reflist, Thanks GrahamHardy (talk) 17:35, 7 February 2013 (UTC)

Loss of session data[edit]

It seems I'm getting a "Sorry, we could not save your edit due to a loss of session data" error every day now. Why am I getting so many of these? Ten Pound Hammer(What did I screw up now?) 14:34, 7 February 2013 (UTC)

I have been getting this a lot for at least 2–3 months now. Rarely does it happen when making an edit to an article, but quite often when creating one. Jared Preston (talk) 14:38, 7 February 2013 (UTC)

Cite toolbar not working: admin needed[edit]


After this revision the cite toolbar hasn't been working because the user forgot to add a comma after the '. Could an admin please fix this? Please change

'cite-work-tooltip' : 'Name of journal, magazine, newspaper, periodical or website'


'cite-work-tooltip' : 'Name of journal, magazine, newspaper, periodical or website',

The Anonymouse (talk | contribs) 18:02, 7 February 2013 (UTC)

Done. HTH, - Jarry1250 [Deliberation needed] 18:05, 7 February 2013 (UTC)
Thanks! The Anonymouse (talk | contribs) 18:09, 7 February 2013 (UTC)
Facepalm3.svg mea maxima culpa --— Gadget850 (Ed) talk 18:24, 7 February 2013 (UTC)

Editors affected by this may need to purge their browser's cache (e.g., shift-reload on an edit page) before things will work again. Jason Quinn (talk) 22:39, 7 February 2013 (UTC)

Updates to "your edit was saved" feature[edit]

Hey all, this is just a note that we've deployed a group of bug fixes to the "Your edit was saved." notification feature. They are...

  1. Displays on new tab after submit
  2. Expand onclick region for close
  3. Increase duration from two to three seconds
  4. Covers Monobook tabs

If anyone uncovers anything new, please speak up. Steven Walling (WMF) • talk 21:24, 7 February 2013 (UTC)

It still hides the Monobook tabs, and the text is still too small, as is the little "close" cross. It does stay up for three seconds though, AFAICT. --Redrose64 (talk) 22:50, 7 February 2013 (UTC)
If you have extra tabs like for TW, VisualEditor etc. that definitely means some would be covered. In any case, the vertical offset is better than it was at least. The text and icon size issue is in part a fault of the Monobook CSS text sizing, IIRC, but the expansion of the close region should mean that the icon size shouldn't matter as much. Thanks for testing it out. Steven Walling (WMF) • talk 23:01, 7 February 2013 (UTC)
As an admin, I get a few extra tabs, yes. Immediately before I went for the "[edit]" link on this section, those tabs that were on this page were: project page; talk; edit this page; new section; history; delete; move; change protection; unwatch. The upper left corner of the "your edit was saved" box, when present, just obscures the right-hand end of the "history" tab. There is also one custom tab (enabled at Preferences → Gadgets → Add a "Purge" option to the top of the page, which purges the page's cache when followed. - which shows as a "*"); it occurs to me that since that tab can work out where it can be placed so that it doesn't obscure anything else (see MediaWiki:Gadget-purgetab.js), a popup message should be able to do that too. --Redrose64 (talk) 23:24, 7 February 2013 (UTC)

Grabbing the lede text?[edit]

I want to do some data mining on (en) wikipedia articles, with Python. I'm going to be using NLTK to infer a few basic facts from the lede of each article of interest. I'm using mwclient for the basic access, but all that gets me is a blob of text which I then have to parse to find the lede. I've come up with the heuristic that "the intro is everything before the first section break, and the lede is everything in the intro after the last template". In python:

host = ''
path = '/wiki/'
site = mwclient.Site(host, path = '/w/')
p = site.Pages['My Page Name']
text = p.edit(readonly=True)

# The intro is everything up to the first line that starts with "="                                                                     
intro_text = text.split("\n=", 1)[0]

# Eliminate all the templates that might be at the beginning                                                                            
lede = intro_text.rsplit('}}', 1)[-1]

Does this seem like a plausible way of doing things? -- RoySmith (talk) 02:29, 8 February 2013 (UTC)

Maybe strip the templates before the headers. Otherwise a header template laid out like this:

Would catch you out. Andy Dingley (talk) 02:12, 8 February 2013 (UTC)

Also, you can't ignore everything before the last closing template brackets in the intro, because many templates are used inline; for example: {{Lang}}, {{IPA}}, {{Convert}}, {{Citation needed}}. You would therefore lose foreign phrases, pronunciations and distances, as well as cleanup tags.
Could you parse the article in HTML form instead of the raw wikitext? That way, you could extract everything before the first <h#> heading that is not in a <div> (this could be refined to also exclude other block or floating elements). You would then be left with the paragraphs of text in the <p> elements of the intro.
If you still want to rely only on the raw wikitext, an improvement would be to strip only those templates which (after stripping whitespace) both begin and end at a linebreak. But putting block-level templates on separate lines is only a common practice, not a firm rule, as wikitext permits infoboxes in the same paragraphs as ordinary text. You would also have to take account of nested templates and consider how to treat tables and magic words.
Richardguk (talk) 04:19, 8 February 2013 (UTC)
MZMcBride has written which is pretty much exactly what you're looking for. Legoktm (talk) 04:28, 8 February 2013 (UTC)

diff not working right in this case[edit]

This diff has a problem. After the part that says Line 29 it looks like a paragraph was deleted on the left, and a new paragraph added on the right. But it's the same paragraph, only a single line of whitespace was actually changed. This of course is confusing to the person viewing the diff. Silas Ropac (talk) 02:35, 8 February 2013 (UTC)

My point is the diff algorithm is not working right, not that this particular file history has something wrong with it. Silas Ropac (talk) 17:11, 8 February 2013 (UTC)
The diff is correct, even though misleading. There is no universally optimal way of computing diffs, the algorithm needs to rely on some heuristics as to which minimal set of changes to choose, and it is bound to get it wrong (from the human point of view) sometimes. You can try enabling the wikEdDiff gadget in your Preferences, which will give you access to an alternative diff algorithm.—Emil J. 17:53, 8 February 2013 (UTC)
Thanks wikEdDiff does it better in this case. I'm used to the side-by-side presentation more, but as a backup for the very few times the normal one "fails" it's great to have a 2nd option. Silas Ropac (talk) 18:28, 8 February 2013 (UTC)

Editor toolbar[edit]

Was there a change in the editor toolbar? There was a drop-down citation template with four options which I found very useful, which now seems to be absent? JMK (talk) 13:40, 8 February 2013 (UTC)

See about 4 threads above...? (✉→BWilkins←✎) 13:44, 8 February 2013 (UTC)
Thanks! JMK (talk) 13:49, 8 February 2013 (UTC)

"An error occurred while changing your watchlist settings"[edit]

I tried adding User talk:Dhastingsabaarsotech to my watchlist. I got a popup saying that an error occurred. I supposed I should just wait and try again later, but this seems like a bug. I can add and remove other pages from my watchlist just fine. Any ideas? • Jesse V.(talk) 18:18, 8 February 2013 (UTC)

I just tried again and it went through this time. Odd. • Jesse V.(talk) 18:20, 8 February 2013 (UTC)

A strange malfunction[edit]

I nominated an article for deletion using NewPageFeed. My contribs show this, but incorrectly list the article as being $1, not the correct one, Reuben Sarin. The relevant wording is here. dci | TALK 20:40, 8 February 2013 (UTC)

  • In case anyone is interested, the following wording is that which has confused me. (pagetriage-del-talk-page-notify-summary: Parse error at position 44 in input: Notifying author of deletion nomination for $1) Thank you, dci | TALK 20:41, 8 February 2013 (UTC)
It's MediaWiki:Pagetriage-del-talk-page-notify-summary without the argument. {{int:Pagetriage-del-talk-page-notify-summary}} gives: Notifying author of deletion nomination for $1. {{int:Pagetriage-del-talk-page-notify-summary|Reuben Sarin}} gives: Notifying author of deletion nomination for Reuben Sarin. PrimeHunter (talk) 21:55, 8 February 2013 (UTC)
This is bug 44459. – PartTimeGnome (talk | contribs) 23:19, 8 February 2013 (UTC)


How do I make a map with openstreetmap? — Preceding unsigned comment added by (talk) 22:04, 3 February 2013 (UTC)

Symbol move vote.svg Have you tried the Computing section of Wikipedia's Reference Desk? They specialise in answering knowledge questions there; the page you are on is only for discussion of technical matters that relate to Wikipedia. I hope this helps. – PartTimeGnome (talk | contribs) 21:25, 4 February 2013 (UTC)
OK, how do I make a .svg map with openstreetmap that is compatible with Wikipedia / Wikimedia? Graeme Bartlett (talk) 11:29, 9 February 2013 (UTC)
Would be good to have a set of step by step instructions created so that people can be pointed to it when they want to create maps. Keith D (talk) 23:14, 9 February 2013 (UTC)

a heads-up -- may go dark[edit]

There is a red notice on today -- they too need donors.

WebCite will stop accepting new submissions end of 2013, unless we reach our fundraising goals to modernize and expand this service.
Please support our crowdfunding campaign. If you are interested in keeping this service alive, please give generously - or at least share our campaign on Facebook. Funders supporting us with $250 or more will be acknowledged by name on our redesigned website.

Archive sites like, that will archive web pages to protect us from those sites going dark, are an important resources. But, occasionally, when I have gone to the trouble of adding an archive link, I have wondered -- wouldn't it be a drag if the archive site itself went permanently dark?

Ideally,, (the wayback machine) and any other mirroring/archiving sites out there were either making sure they mirrored all the pages mirrored by their opposite numbers. Failing that, it would be wonderful if the other sites stepped in and started making copies of the pages archived by -- in case it went dark.

They seem to be indicating that even if they stopped accepting new links they would still serve copies of pages already in their archive. But the notice is a very broad hint that the whole site might go dark.

Is there any value in checking to see how many pages archived at we are using are also archived at

FWIW I tried to use external links special pages to get an idea of how many links we have to It only reported 808 links, while the Citizendium, a wiki site about one 10000ths the size of the wikipedia has 3200 links. I think I may have made 800 links to just by myself.

WMF projects may make so many links to that it should consider stepping in and adopting if the alternative were to be it going dark?

Cheers! Geo Swan (talk) 18:20, 5 February 2013 (UTC)

  • Your counting method is off, Ive seen 50k links to and Im still counting. Werieth (talk) 18:45, 5 February 2013 (UTC)
    • There are 527,788 links to, Im getting the count for now. Werieth (talk) 19:27, 5 February 2013 (UTC)
    • There are 182,368 links to Werieth (talk) 19:40, 5 February 2013 (UTC)
  • WebCite is really useful for news sites, as they often can become unavailable online and their robots.txt files prevent from crawling the page. Chris857 (talk) 18:49, 5 February 2013 (UTC)
  • Comment I have used WebCite numerous times for archiving Wikipedia citations and many citation links only exists there, since the original has died. I also think the WMF should consider a cooperation with WebCite and perhaps support them with part of their funds? If WebCite dies, this would definitely also hurt Wikipedia. Our content is only as good as the sources supporting it. -- Toshio Yamaguchi 18:54, 5 February 2013 (UTC)
Can someone give a clear indication that WebCite will be genuinely nonprofit? I'm looking at [1] and seeing "An increasing number of journal editors and publishers are currently joining the growing WebCite® Consortium. We are currently in the process of incorporating WebCite® as a distinct entity." This worries me because we don't want to be funding it if it turns out simply to be an Internet company. However, if it is nonprofit, then we should definitely find some way to help it continue to exist as a separate entity, because we wouldn't want to take all of its inevitable copyright squabbles with publishers and lump them together with our own Fair Use issues, infringing uploads, etc. Wnt (talk) 20:10, 5 February 2013 (UTC)
The last paragraph under Copyright and Long-Term Preservation Issues at says "All WebCite code is Open Source, and all documentation is licensed under Creative Commons licenses. Secondly, through collaborations with libraries and consortia interested in preservation of digital material, who may act as a curator, custodian or trustee for the WebCite project. These long-term preservation partners may have agreed to hold backups of the service and to legally assume the domain name, all intellectual property such as trademarks, and the service itself, should for any reason the original WebCite service go out of business. " -- Toshio Yamaguchi 20:22, 5 February 2013 (UTC)
The information on Gunther Eysenbachs userpage (the initiator of the project, see this) says "WebCite is not a company, but a non-profit consortium of publishers (mostly open access publishers). WebCite in turn is member of an international community of archiving services, of which the Internet Archive, the Library of Congress and other institutions are a part of." -- Toshio Yamaguchi 20:29, 5 February 2013 (UTC)
  • The higher citation counts linked above sound more realistic; good search-fu! Note that the number of WebCites is even higher than that, because not all URLs are inserted in refs yet. I'm a prolific WebCiter using the bookmarklet, though my insertion rate into refs is low (frequently deferring insertion until a URL dies). ;)
    It's odd that they don't actually link to their "Facebook campaign". FYI donation info is at . --Lexein (talk) 21:46, 5 February 2013 (UTC)

I made a proposal on Meta regarding WebCite, which can be found here under Open proposals. -- Toshio Yamaguchi 14:45, 9 February 2013 (UTC)

Thank you for making a formal proposal. I added myself to the People interested over there.
Below is a user box template for the like minded people to be used at their User page. --DancingPhilosopher my talk 17:04, 9 February 2013 (UTC)
WebCite This user believes in archiving sources to prevent link rot.

Bug 15434: Disabled special pages are refreshed[edit]

Hi bugzilla:15434 (Periodical run of currently disabled special pages) has had some activity lately, resulting in a once-off refresh of the special pages. e.g. Special:Wantedpages. It looks like periodical refreshes may be coming, so I expect that there are some on-wiki prep work to be done to ensure these tools are able to be useful once again.

For example, I noticed that wanted pages is less useful than it could be because it is full of items from /todo pages. e.g. Special:WhatLinksHere/The Sibley Guide to Bird Life and Behavior is high on the list because it is listed at Template:Birds tasks. The result is the ranking of wantedpages is caused by the number of pages claimed by a WikiProject, rather than the number of incoming links from content-space. We have two options here; either remove these 'wanted page listings' appearing on thousands of talk pages, or ask for a linked-from-namespace parameter on special:wantedpages.

Another example is "Wikipedia:WikiProject Blah/Assessment‏‎" appears many times because of the logic deep within Template:WPBannerMeta that automatically looks for an /Assessment subpage. John Vandenberg (chat) 01:13, 6 February 2013 (UTC)

bugzilla:12019 --MZMcBride (talk) 09:37, 6 February 2013 (UTC)
I have manually generated this report at Wikipedia:Most_Wanted for the last few years. Eliminating common red links arising from other namespaces, in particular templates is a non-trivial task. The most useful single number for 'wantedness' of a red link seems to be the number of times the text of the red link appears as a wikilink in the raw source of all pages in namespace 0, or any page in namespace 10 that is transcluded into namespace 0 at least once. This is rather painful to calculate. A nearly-as-good approximation is for any red link that is not present in namespace 10, the number distinct namespace 0 articles that link to it. This is far easier to calculate and nobody's complained yet ;) - TB (talk) 10:06, 6 February 2013 (UTC)
I personally see no reason why we should ignore the inclusion of red links on talk pages via WikiProjects... I would personally be happy for a filter just to show redlinked namespace 0 articles; I don't particularly care which namespace the redlinks are in.... --Izno (talk) 15:09, 6 February 2013 (UTC)
The first 20,000 entries on a naive 'most linked to red links' arise from navigation templates like Template:Towns in West Gippsland and Template:Music of Asia. Such templates are typically included on many pages, making any red links they contain appear on each. The 'wantedness' of a title that has been linked manually from a handful of different articles is generally greater than one linked from a single but widely used template.
I am interested in your idea of incorporating red links arising in the Wikipedia_talk pages - can you point me towards some examples where interesting new titles are widely linked there please? - TB (talk) 09:42, 9 February 2013 (UTC)

I have found the best response to such templates is to delink and remove the red link entries myself, per the reasons listed at WP:NAV (an essay). However, general "wantedness" is more difficult to measure, though I understand where you come to that conclusion... I of course also understand that there are many editors who would disagree with the conclusions at WP:NAV. YMMV. Somebody does want those pages though, else they would not be listed in the templates.

A number of the pages that I viewed were linked from Talk pages, via the WikiProject banners. Many such banners have /to do lists (or similar), which usually include redlinks that someone put there. I couldn't point you to them now. --Izno (talk) 01:31, 10 February 2013 (UTC)

Is the toolserver broken?[edit]

I just went there to try and find out how to register an account so I can try building something, but the front page isn't loading. — Hex (❝?!❞) 13:07, 7 February 2013 (UTC)

I can't say anything specific, but it isn't loading for me either, I just see an empty white page. shows INFO Replication disrupted for some services and INFO MNT-1286 Was re-setuped for two of the database clusters. -- Toshio Yamaguchi 13:13, 7 February 2013 (UTC)
Replication lag is high... again! Jared Preston (talk) 14:27, 7 February 2013 (UTC)
14 hours and growing... at a rate of one minute per minute. Face-wink.svg ‑Scottywong| confess _ 17:44, 7 February 2013 (UTC)
as to the original message: it is my understanding that it's probably preferable to create an account on "wikimedia labs" server ([2]). supposedly it has similar capabilities to the TS, and, being a project of wikimedia foundation rather than dewiki, it's likely to be better resourced. it is my understanding that part of TS woes stem from the fact that the foundation, which funded significant portion of the TS operational costs, stopped doing so this year, with the rationale that this kind of stuff belongs on wikimedia labs, and dewiki is not sure it wants to pick the slack. projects should be migrating from TS to wikimedia labs, so building a new service/tool on TS seems like the wrong thing to do, IMO. peace - קיפודנחש (aka kipod) (talk) 18:45, 7 February 2013 (UTC)
Thanks, but... blurf. I've just spent a while reading around on Labs from your link; for what I want to do (a simplistic tool relating to presenting user contributions) it seems like using an atomic bomb to kill a fly. I guess I'll go to their IRC channel at some point and see if someone can explain it to me in terms that make sense to someone who only occasionally writes tiny Perl programs and bits of SQL. — Hex (❝?!❞) 19:30, 7 February 2013 (UTC)
I've accounts on both systems. The toolserver is mature and convenient to use, but lacks capacity and has regular problems with database replication. Labs has lots of capacity, but is complex to get into and lacks precedent for almost all actions, limiting progress to the speed of discussion. - TB (talk) 09:47, 9 February 2013 (UTC)
Yeah. I think I'm going to take my chances with the toolserver, frankly, the Labs environment appears to be a massive impediment to the casual hacker. — Hex (❝?!❞) 20:33, 9 February 2013 (UTC)
See DaB's email from this morning: [3]. Legoktm (talk) 19:44, 7 February 2013 (UTC)

Rendered text not matching wikitext (water is never pure...)[edit]

Wikimedia Foundation received a report of a problem at Musicology.

Note that the first sentence starts:

Musicology (water is never pure in a chemical sense

Yet in edit mode:

Musicology (from Greek ??????? (mousike-), meaning "music", and -????? (-logia), meaning "study of-") is the scholarly study of music.

I think the problem is broader, as Laparoscopy will attest:

Laparoscopy (water is never pure in a chemical sense.

--SPhilbrick(Talk) 12:13, 9 February 2013 (UTC)

Should be fixed after this revert at {{Ety}}. -- John of Reading (talk) 12:19, 9 February 2013 (UTC)
(edit conflict) I made a null edit to the page, so it should render correctly now. -- Toshio Yamaguchi 12:21, 9 February 2013 (UTC)
For future reference: at Musicology I clicked "Edit" at the top, then scrolled down to "View templates on this page" at the bottom, and that showed me which template had recently been edited. -- John of Reading (talk) 12:24, 9 February 2013 (UTC)
Just out of interest, how do you see from that which one of them had recently been edited? I see nothing indicating recent edits to one of them. -- Toshio Yamaguchi 12:29, 9 February 2013 (UTC)
They are displayed with the most recently-modified template at the top. I see lots of detail about the most recent edit to each template - do you? If not, you might need to copy the first two lines from User:John of Reading/vector.js. -- John of Reading (talk) 12:35, 9 February 2013 (UTC)
Even after copying those lines, all I see is
View templates on this page
Template:About (view source) (protected)
Template:Ambox (view source) (protected)
Template:Ambox/category (view source) (protected)
Template:Ambox/core (view source) (protected)
Template:Basepage subpage (view source) (protected)
Template:Category handler (view source) (protected)
Template:Category handler/blacklist (view source) (protected)
Template:Category handler/numbered (view source) (protected)
Template:Citation/core (view source) (protected)
Template:Citation/identifier (view source) (protected)
Template:Citation/make link (view source) (protected)
Template:Cite book (view source) (protected)
Template:Column-count (view source) (protected)
Template:Commons (view source) (protected)
Template:Commons category (view source) (protected)
Template:Dead link (view source) (protected)
Template:Div col (view source) (protected)
Template:Div col end (view source) (protected)
Template:Ety (edit)
Template:Etymology (edit)
Template:Fix (view source) (protected)
Template:Fix/category (view source) (protected)
Template:Hatnote (view source) (protected)
Template:Hide in print (view source) (protected)
Template:ISO 639 name gre (edit)
Template:Ibid (view source) (protected)
Template:If pagename (view source) (protected)
Template:Iso2language (edit)
Template:Longlabel (edit)
Template:Main (view source) (protected)
Template:Music topics (edit)
Template:Namespace detect (view source) (protected)
Template:Navbar (view source) (protected)
Template:Navbox (view source) (protected)
Template:Only in print (view source) (protected)
Template:Other uses (view source) (protected)
Template:Please check ISBN (edit) (semi-protected)
Template:Portal (view source) (protected)
Template:Portal/Images/Music (edit) (semi-protected)
Template:Portal/core (view source) (protected)
Template:Reflist (view source) (protected)
Template:Rellink (view source) (protected)
Template:Side box (view source) (protected)
Template:Sister (view source) (protected)
Template:Small (view source) (protected)
Template:Template other (view source) (protected)
Template:Transclude (view source) (protected)
Template:World topic (edit)
View hidden categories on this page
-- Toshio Yamaguchi 12:53, 9 February 2013 (UTC)
That's curious. My list begins
2013-02-09 12:18 (hist) Template:Ety (edit) John of Reading: Reverted 2 edits by (talk): Rv irrelevant text, broke many articles. (TW)
2013-02-07 18:29 (hist) Template:Longlabel (edit) CsDix: line-height tweak
Are there any scripting gurus about? -- John of Reading (talk) 13:01, 9 February 2013 (UTC)
Yes, John, what you are seeing is due to my script. Toshio, are you using the Vector skin? Anomie 13:14, 9 February 2013 (UTC)
Yes, I am using vector. -- Toshio Yamaguchi 13:19, 9 February 2013 (UTC)
I found that I only needed to add one line to Special:MyPage/common.js. This works in Monobook. --Redrose64 (talk) 14:14, 9 February 2013 (UTC)
I tried but it has no effect. The only thing I can see is whether the page is protected or not. -- Toshio Yamaguchi 15:47, 9 February 2013 (UTC)

'Related changes' in the left toolbox is also useful. --— Gadget850 (Ed) talk 14:22, 9 February 2013 (UTC)

In VisualEditor red links show blue[edit]

As I was correcting a typo in Goldberg Variations discography with VisualEditor I noticed that the red links appeared blue while you're in the editor. Signed: Basemetal (write to me here) 17:40, 9 February 2013 (UTC)

Please discuss at Wikipedia:VisualEditor/Feedback. --— Gadget850 (Ed) talk 19:47, 9 February 2013 (UTC)

Using {pagename} to link to[edit]

I'm trying to make a clickable link in a template, to send people to the page for the relevant article. However, the pagename uses spaces instead of underscores, and when the information passes to the website, the grok site replaces them with plus signs and the link doesn't work. Is there a way to grab a page name using underscores, or any other way to make the template function as intended? --NickPenguin(contribs) 04:18, 10 February 2013 (UTC)

The link in page histories is made with MediaWiki:Histlegend which says{{FULLPAGENAMEE}}. PrimeHunter (talk) 04:54, 10 February 2013 (UTC)
I made the change to the template.[4] PrimeHunter (talk) 05:01, 10 February 2013 (UTC)
Excellent, thank you so much. --NickPenguin(contribs) 05:16, 10 February 2013 (UTC)

Insidious problem with pdf rendering[edit]

A reader wrote in to Wikimedia Foundation to report a problem with pdf exports. I know there are some general problems, but this one is troubling, because it appears to work, yet will produce results which are off by a factor of ten.

I'll copy the person's own words:

I would like to report a problem concerning exporting articles from Wikipedia to pdf. The problem lies with the following code:

1.7 - 3.1 MeV/c2 (and all variations of it)

Which should translate to 1.7 - 3.1 ?eV/c2 but when translated to pdf it reads

1.7 - 31 ?eV/c2 thus killing the "." mark. (See attached image from the pdf)

This, as said, is also true for all values in the function. I.e.

2.01±0.14 MeV/c2 should read 2.01 plus or minus 0.14 MeV/c2 but is shown as:

The generated pdf can be found in the following address.[..]

Since confusion can arise from this situation I think that a fix of the problem should be found as soon as possible.

I didn't find the example above in the link provided, but this line: "the masses of elementary particles become visible at energies above 1.4 TeV" in Standard Model

renders in the pdf (see page 6) as:

"the masses of elementary particles become visible at energies above 14 TeV"

Problem in bold.

I agree this needs to be addressed. — Preceding unsigned comment added by Sphilbrick (talkcontribs) 12:43, 6 February 2013 (UTC)

Template:Val/delimitnum contains: {{formatnum:{{#expr:trunc(abs({{{1|0}}}))}}.}} - formatnum containing a dot at the end produces a dot in the wiki page but not in pdf. Could the dot be moved outside of the formatnum? Peter James (talk) 13:33, 6 February 2013 (UTC)
I've made this change to the template. Peter James (talk) 00:13, 11 February 2013 (UTC)


It says the Wikipedia:RefToolbar/2.0 page that the new Toolbar should be default for all English users(and I used to too) but for some reason I seem to have lost the "cite" option on my toolbar.

I don't seem to be using the "RefToolbar 2.0a" either because I have no {{}} icon featured. Could anyone tell me how I can get it back?! The cite tool is the only thing on there I actually need!!!

Many thanks! ----Brigade Piron (talk) 10:04, 9 February 2013 (UTC)

See the section #Cite toolbar not working: admin needed, above, and see if bypassing your browser cache fixes the problem or you. -- John of Reading (talk) 10:11, 9 February 2013 (UTC)
Try logging out and logging in. --— Gadget850 (Ed) talk 10:54, 9 February 2013 (UTC)
Thank you both! --Brigade Piron (talk) 11:06, 9 February 2013 (UTC)
Which one worked? --— Gadget850 (Ed) talk 15:32, 9 February 2013 (UTC)
I tried a combination and between them, they seemed to work. Just for note, how can I stop the above happening in future? The same has occurred today and I've been wiki-ing for 9 months+ and nothing like this happened before... ---Brigade Piron (talk) 09:38, 10 February 2013 (UTC)
As you will have read, someone made a minor error. Your browser kept that error-version in its cache. You're the only one in charge of your browser, and nobody can predict the errors of others. So - first rule of thumb anywhere in the world when you have problems with a website: clear your cache (✉→BWilkins←✎) 10:37, 10 February 2013 (UTC)

Rename a page[edit]

Good morning. I don't speak English very well. The problem is to rename Jean Pierre Sauvage to Jean-Pierre Sauvage. Thank you! --Cjp24 (talk) 02:42, 11 February 2013 (UTC)

I requested the move. Thanks! Vacation9 02:52, 11 February 2013 (UTC)

Template:efn and <ref> tag error when no <ref> tags are used?[edit]

I don't understand the error generated shown here. Biosthmors (talk) 19:22, 10 February 2013 (UTC)

{{Efn}} transcludes the ref tags onto the article - if you're going to use it, you need to have a {{notelist}} at the bottom where you want the note to show. See the template's documentation at Template:Efn for more. – Philosopher Let us reason together. 19:31, 10 February 2013 (UTC)
Maybe that went over my head. {{notelist}} and {{efn}} are being used exclusively at User:West.andrew.g/Popular pages header, so I don't understand why the edit I made generated an error about missing ref tags. Biosthmors (talk) 19:36, 10 February 2013 (UTC)
(edit conflict)Read that in too big of a hurry. Striking some. The problem is the URLs that use the = key. From the template documentation: "Remember that "=" cannot appear in an unnamed template parameter and "|" cannot appear in any template parameter. For "|", {{!}} can be used. For "=", {{=}} can be used or the parameter can be named 1= explicitly:
{{efn|1=Converting at a rate of Kr 20 = £1.}}" – Philosopher Let us reason together. 19:40, 10 February 2013 (UTC)
{{efn}} includes a <ref> tag. --— Gadget850 (Ed) talk 10:02, 11 February 2013 (UTC)

Constant server error when editing one particular article[edit]

I have an interesting problem with the article List of South Korean idol groups. Whenever I try to edit the page or a section and I push submit, I get the server error message of Wikimedia, the one we usually see when the servers are down. It doesn't happen with any other article, so I started to doubt this is a general problem with the website. The interesting thing is that my edits are saved when I go back to the page and check the history tab. I still get the server error screen right after saving. A friend suggested that i ask here as he thinks maybe a particular template is causing the problem, or perhaps the page is too big in size? 200,286 bytes at the moment, with almost 500 {{cite web}} templates. Any help or suggestion would be great小龙 (Timish) # xiǎolóng de xìnxiāng 10:25, 11 February 2013 (UTC)

Seems to load for me, but this is one massive article :/ ·Add§hore· Talk To Me! 10:40, 11 February 2013 (UTC)
Loading is not a problem for me, the problem occurs when saving. Try a null edit or a small corection somewhere, to see what I mean. 小龙 (Timish) # xiǎolóng de xìnxiāng 10:52, 11 February 2013 (UTC)
I've occasionally seen this happen elsewhere on very large pages: edits do go through (as evidenced by the history), but it takes a long time and often you see a server error page. I just made a small edit and got the same result.
The best solution is to split up the page (perhaps by decade); it's just too big. Ucucha (talk) 12:35, 11 February 2013 (UTC)

Thanks! 小龙 (Timish) # xiǎolóng de xìnxiāng 12:57, 11 February 2013 (UTC)

Bug in Recent Changes filtering by namespace[edit]

Here is my usual recent changes feed for the portal namespace: [5]. At the top, currently, is this entry...

...which is nothing to do with the portal namespace. Also that first link, to Special:Log/eparticle, doesn't seem correct. -- John of Reading (talk) 11:04, 11 February 2013 (UTC)

For me it says "added article Computer graphics" with a piped link to Portal:Computer graphics. The "Articles" column for Natbrock at Education Program:University of Hull/Interaction, Experience and Engagement (Spring 2013) says Portal:Computer graphics with the namespace displayed. I guess the students are only supposed to add articles and mw:Extension:Education Program isn't programmed to handle a Portal page properly. PrimeHunter (talk) 13:34, 11 February 2013 (UTC)
You're quite right - it is only the wording that is incorrect. Never mind. -- John of Reading (talk) 14:08, 11 February 2013 (UTC)

Centralising social network-sharing templates?[edit]

Would it be possible for any template which uses social networking sites' tools for interacting with external websites (e.g. "Share this on Twitter!"-type links/buttons) to automatically be put in some Category; and for there to be a piece of user-JS/-CSS code that automatically hides all members of said Category? Current examples of such templates on en-wiki are Template:Share and Wikipedia:Signpost/Template:Signpost-article-start. I've recently added some CSS code to my page that hides {{Share}}, but what I'd really like is code that hides invitations to share WP content on social networks ipso facto, due to the potential for equivalent templates to appear under other names (as the Signpost example demonstrates). Thanks in advance for your help, It Is Me Here t / c 13:41, 11 February 2013 (UTC)

Please nominate NAS4Free for deletion[edit]

I cannot find instructions at WP:AFD for starting a second deletion ("2nd nomination"). This article was deleted in 2012, recreated recently and I think it still fails GNG. Oh, and please add clear instructions to AFD what to do in such a case. Thanks! --Piotr Konieczny aka Prokonsul Piotrus| reply here 16:51, 11 February 2013 (UTC)

Is WP:AFD#How to nominate a single page for deletion unclear? What can I do to make it better? Prodego talk 16:54, 11 February 2013 (UTC)
Use {{subst:afdx|2nd}} instead of {{subst:afd1}} at the top of the article. --Malcolmxl5 (talk) 20:25, 11 February 2013 (UTC)
Alternatively, enable the Twinkle gadget which takes care of all this automagically.--ukexpat (talk) 20:39, 11 February 2013 (UTC)
I'll do it. Go post your rationale at Wikipedia:Articles for deletion/NAS4Free (2nd nomination). UltraExactZZ Said ~ Did 20:53, 11 February 2013 (UTC)

Password requirements discussion regarding MediaWiki[edit]

I've started an RFC for the MediaWiki software. The goal is to make it more practical to change the password requirements, and to allow configuring different requirements for different groups (e.g. admins, stewards, etc.)

No immediate change to Wikipedia is proposed, but this would make such changes possible. The goal is to have stronger requirements for accounts with extra rights, without inconveniencing regular users (though it is still advisable for everyone to choose a strong password). Superm401 - Talk 06:05, 12 February 2013 (UTC)

Statistics not working?[edit]

Resolved: site was updated today --Nemo 09:46, 11 February 2013 (UTC) may be delaying yesterday's and today's statistic reports. --George Ho (talk) 04:15, 9 February 2013 (UTC)

I was thinking it might have been a glitch yesterday, but nothing is up today either. Ryan Vesey 17:10, 9 February 2013 (UTC)
Does anyone know who runs the page views tool? Ryan Vesey 01:37, 10 February 2013 (UTC)
Yes, Henrik runs it, and the OP recently posted a message about this problem to his talk page. Graham87 08:03, 10 February 2013 (UTC)
I'm still having problems with accessing * I can't connect from my site but can connect from others. Any reason why this would be? --Lydgate (talk) 10:42, 12 February 2013 (UTC)

Wikidata deployment phase 1[edit]

I was asked to also post a note on this village pump in case you have not seen Wikipedia:Village_pump_(miscellaneous)#Wikidata_deployment_phase_1. The deployment is scheduled for later today. I hope everything goes smoothly. I am happy to answer questions you might have. --Lydia Pintscher (WMDE) (talk) 14:04, 11 February 2013 (UTC)

We unfortunately ran into issues. We'll have to reschedule the deployment. Currently it looks like we'll do this on Wednesday. Sorry folks. --Lydia Pintscher (WMDE) (talk) 22:03, 11 February 2013 (UTC)
We'll do another attempt later today (probably around 17:00 UTC). --Lydia Pintscher (WMDE) (talk) 11:56, 12 February 2013 (UTC)
There were unfortunately too many other issues unrelated to Wikidata so we also had to call off this one. Sorry. --Lydia Pintscher (WMDE) (talk) 19:34, 12 February 2013 (UTC)

Multiple article stats[edit]

Hello, I'm familiar with how to get article stats at, however, I can only use that tool to view one article's stats at a time. Is there any way to feed multiple article names, say 25-30 article titles, to view them all at once? (That way I can know which is getting the most hits at a glance, rather than going thru them one by one.) Thank you for answer! Ihardlythinkso (talk) 15:29, 11 February 2013 (UTC)

As well, can anyone suggest a method of getting hits on a larger timescale than day-by-day? Before the system was revamped a while ago (a couple years?), would give stats by month in the same way that gives stats by day. I'm guessing that Ihardlythinkso would be open to another stat counter, and I definitely would be. Nyttend (talk) 15:47, 11 February 2013 (UTC)
Yes, affirmed. (And from your followup I understand there's no way currently to do what I wanted. With all the technical programmers at the Pedia, and the importance of stats ... I don't get it!) Ok, thank u. Ihardlythinkso (talk) 19:39, 12 February 2013 (UTC)

Prevent Signatures in Mainspace?[edit]

Feature Request. Make it so that rather than converting three, four or five tildes to the appropriate signature/time values, reject the edit if it takes place on a mainspace article (there are probably other spaces where this would be equally useful like template).Naraht (talk) 22:04, 11 February 2013 (UTC)

This can't be done here; you would need to file a feature request at bugzilla. --Redrose64 (talk) 22:17, 11 February 2013 (UTC)
This could be done through an edit filter. Some time ago I asked about having a filter that would block this kind of thing, as well as many other features (e.g. '''Bold Text''' or <ref>Insert footnote text here</ref>) but I was reminded that it would be a bad idea because there are apparently many more times than I imagined that such things are needed in articles. Nyttend (talk) 22:30, 11 February 2013 (UTC)
What if someone wrote a large section or made a lot of modifications to an article but wrote "~~~~" at the end? The user would be discouraged by the edit filter warning, and might not try to add the text (s)he wrote again. πr2 (tc) 19:24, 12 February 2013 (UTC)

Articles with a "/" character in their name[edit]

As we know, the "/" character tends to signify a change of directory or folder to computers. In its simplest form an article entitles A/B creates B as a sub page of A. Whether it is stored thus is immaterial. BUt this gives rise to some anomalies.

I noticed these with articles about the incidents known as "9/11 (things)".

Look, for example at 9/11 conspiracy theories, not for article content, but for article naming. It is a free standing article and does not present itself as if 11 conspiracy theories is a subpage of 9. Now navigate to the talk page with the talk tab. we arrive happily at Talk:9/11 conspiracy theories.

On line three, right at the top, just under From Wikipedia, the free encyclopedia is a link: < Talk:9. That takes us to Talk:9

Now I have no idea if Mediawiki has an error or a restriction, and Bugzilla is a place I am not headed towards. But I felt it to be interesting.

That least me to a question: "Do we consider this to be something that needs to be solved? If so, shoudl it be solved at Mediawiki level or by moving the articles to '/' free names?"

Over to the experts, who, as we know, are those who are no longer pert :) Fiddle Faddle (talk) 23:19, 11 February 2013 (UTC)

Subpages are disabled in main namespace, that is why in the case of 9/11 conspiracy theories, 11 conspiracy theories is not a subpage of 9. The example of Talk:9/11 conspiracy theories shows that subpages are enabled in Talk namespace. I agree that subpages maybe should be disabled in Talk namespace as well. Does anybody know whether there are any valid uses of subpages in talk namespace that disabling this would break? -- Toshio Yamaguchi 23:35, 11 February 2013 (UTC)
GA noms, various drafts and comments pages off the top of my head. --Izno (talk) 23:37, 11 February 2013 (UTC)
Ah, yes, correct. And all archives of article talk pages are also subpages of the respective article talk page. -- Toshio Yamaguchi 23:45, 11 February 2013 (UTC)
See WP:Subpages also. --Izno (talk) 23:35, 11 February 2013 (UTC)
Talk pages have many subpages: talk archives, todo lists, centralized discussion lists, drafts, reference work pages and more. --— Gadget850 (Ed) talk 23:47, 11 February 2013 (UTC)
One example of why the subpage feature is useful for this: The move form for admins on articles with talk subpages has a "Move subpages of talk page" check box. PrimeHunter (talk) 23:56, 11 February 2013 (UTC)
In the early days of Wikipedia, we used subpages for subtopics, and many of these titles are still around as redirects for historical purposes. See Albania/History, for example. The problem with subpages in mainspace is the question of hierarchies: should History of Albania be located at Albania/History (i.e. Albanian history is a sub-topic of Albania in general) or History/Albania (i.e. Albanian history is a subfield of history in general)? This is largely why we developed the category system, since History of Albania can be put both in a general Albania category and in a general history category. Nyttend (talk) 03:23, 12 February 2013 (UTC)
Would this be technically possible: disable subpages of a talk page Talk:Foo/Bar if Foo/Bar exists? πr2 (tc) 19:21, 12 February 2013 (UTC)
Clarification:by "disable subpages" I meant "disable breadcrumb links at the top of the page and treat as if it didn't have "/" in its title. πr2 (tc) 22:11, 12 February 2013 (UTC)

"My" Watchlist[edit]

This is just a query. Why was "My" got rid of on Special:Watchlist? Simply south...... catching SNOWballs for just 6 years 12:45, 12 February 2013 (UTC)

Seconded. Ks0stm (TCGE) 12:48, 12 February 2013 (UTC)
👍 Thirded. ‹See TfD› — PinkAmpers&(Je vous invite à me parler) 12:51, 12 February 2013 (UTC)
I'm sure there was a discussion - can't imagine how obvious the change would have been anyway. The word "my" was a redundant waste of space (✉→BWilkins←✎) 12:56, 12 February 2013 (UTC)
It was several months ago, for the reasoning that it was inconsistent with the rest of the Internet to add the personal words of ownership. Which, it's hard to disagree with that as a statement of fact. --Izno (talk) 13:13, 12 February 2013 (UTC)
I absolutely don't care if it as a "my" or not. But... the fact is that page is *my* watchlist, not some global WP watchlist, not your watchlist, not John Doe's watchlist. It is my watchlist, the one I built and chose to have. - Nabla (talk) 13:31, 12 February 2013 (UTC)
The change to the "personal tools" links in the upper right corner was in November 2012 (see Wikipedia:Village pump (technical)/Archive 105#Small change to the top right menu); but the H1 heading at Special:Watchlist (which comes from the MediaWiki software, but can be customised for this site at MediaWiki:Watchlist) changed very recently, no more than two or three days ago. --Redrose64 (talk) 15:53, 12 February 2013 (UTC)
Yep. Liked that change. Don't like this one. Anyone know of a CSS trick I can use to switch it back? — PinkAmpers&(Je vous invite à me parler) 16:01, 12 February 2013 (UTC)
Yeah...lemme see if i can remember where I left it... Writ Keeper 16:05, 12 February 2013 (UTC)
Here we go:
#pt-watchlist a:before {content: "My "}
I think that should work. (The site CSS is going seriously wonky right now (wasn't me), so I can't really test it to make sure.) Writ Keeper 16:09, 12 February 2013 (UTC)
Thanks WK! Though I'm curious if anyone would jump on a bandwagon to edit MediaWiki:Watchlist back to the old version. (And of course it was you. It's always you. Unless you feel like challenging ANI 2.0 himself.) — PinkAmpers&(Je vous invite à me parler) 16:17, 12 February 2013 (UTC)
Bugzilla:44415 changed the default English-language MediaWiki:Watchlist page title from "My watchlist" to "Watchlist", apparently to be consistent with the re-titled user links at the top of the page. — Richardguk (talk) 16:43, 12 February 2013 (UTC)
Oh, were you talking about the title of the Special page? My CSS thingy changes the link at the top of the page. The title in the top of the content window can be changed with:
body[class~=page-Special_Watchlist] #firstHeading span:before{content:"My ";}
If you also want to change the html page title (the text that shows up on tab labels and the top bar of the browser), I don't think that can be done through CSS. It would require Javascript. Should be something to the effect of:
if(wgPageName == "Special:Watchlist"){$("head>title").html("My " + $("head>title").html());}

Writ Keeper 17:46, 12 February 2013 (UTC)

Page loading issues[edit]

Maybe it's something on my end, but often as not, various Wikimedia pages aren't completely loading. I'm seeing this both here and at Commons. Sometimes the "skin" doesn't load, sometimes the Twinkle script doesn't load, sometimes the clock doesn't load etc. Doing a purge (server or local) sometimes clears up the issue. -- Gyrofrog (talk) 16:27, 12 February 2013 (UTC)

Also happening to me but only with my watchlist ·Add§hore· Talk To Me! 16:29, 12 February 2013 (UTC)
Seconded, with every page. Purge is useless.—cyberpower OnlineBe my Valentine 16:30, 12 February 2013 (UTC)
See above "Wikidata deployment 1"-section. --Saddhiyama (talk) 16:31, 12 February 2013 (UTC)
Gaaah. Can't these things be test deployed first before deploying it and destroying the interface of Wikipedia?—cyberpower OnlineBe my Valentine 16:41, 12 February 2013 (UTC)

This is unrelated to the Wikidata deployment. That has not happened yet. --Lydia Pintscher (WMDE) (talk) 16:47, 12 February 2013 (UTC)

Yes, the issues are WMF-wide including wikis that don't have wikidata yet. The servers that serve the JS and CSS are down and ops are looking into it right now. --Aude (talk) 17:14, 12 February 2013 (UTC)

Displaying Issues[edit]

Today, many pages seem to be using a different skin. Every time I tried to go to my Watchlist in the past couple of hours, it displayed it in a weird, bland skin (with just a bunch of links on the side, rather an spaced around the page like usual. (I always use the Vector skin.) This just fixed itself, but I'm still have the same issue with Wikidata:Wikidata:Project chat. Also, Special:NewPagesFeed is working on-and-off, and even when it's on I can't access the page curation tools. What is going on? (I'm using Firefox on a Dell laptop running Windows 7.) -- Ypnypn (talk) 16:29, 12 February 2013 (UTC)

I just pressed save page, and it went to that weird skin, but then I clicked edit (to add this comment) and I'm back on Vector. Help! -- Ypnypn (talk) 16:30, 12 February 2013 (UTC)
See above "Wikidata deployment 1"-section. --Saddhiyama (talk) 16:31, 12 February 2013 (UTC)
I see :) ·Add§hore· Talk To Me! 16:32, 12 February 2013 (UTC)
So did they abort the attempt, or what? -- Ypnypn (talk) 16:43, 12 February 2013 (UTC)

This is unrelated to the Wikidata deployment. That has not started yet. --Lydia Pintscher (WMDE) (talk) 16:47, 12 February 2013 (UTC)

I can confirm there is some sort of problem. My vector.css is not taking effect as of today (it was working fine yesterday, and has been working fine for years). —SeekingAnswers (reply) 17:02, 12 February 2013 (UTC)

So what is the issue? And is it being worked on? Thanks. --StarcheerspeaksnewslostwarsTalk to me 17:03, 12 February 2013 (UTC)
I don't know what the issue is but people on IRC are on it, yes. --Lydia Pintscher (WMDE) (talk) 17:06, 12 February 2013 (UTC)
It's not a "different skin", but your normal skin where some or all of the CSS and JS files have either failed to be sent by the server, or have failed to be loaded by your browser. A hard refresh of the page may fix it; however, when I tried this, I found that a different combination of CSS/JS were failing. --Redrose64 (talk) 17:04, 12 February 2013 (UTC)
It's being caused by the server returning HTTP 503: Service Unavailable errors in response to the HTTP GET requests. Reaper Eternal (talk) 17:06, 12 February 2013 (UTC)

Just seconding the notion that my stuff has gone all wonky loading taking *forever*, not loading completely (all pages missing the tab with the RPP/Welcome/Warning/Tag/etc), the rollbacker/'restore this page' feature is gone, hovering the cursor over editors' names doesn't bring up any of their information, and so on. Shearonink (talk) 17:10, 12 February 2013 (UTC)

It's not fixed yet; the GET requests that RE mentioned are failing at random, making each bit of Wikipedia (skin, user JS, site JS, whatever) work or fail independently from the others. What works and what doesn't changes from pageload to pageload, as different requests succeed or fail. Writ Keeper 17:24, 12 February 2013 (UTC)
Ops are still working on it. Hopefully not too much longer before it's fixed. :( --Aude (talk) 17:40, 12 February 2013 (UTC)
Looks like JS and CSS are working again, although perhaps some people still see problems? --Aude (talk) 17:53, 12 February 2013 (UTC)
Looks good so far. Writ Keeper 17:55, 12 February 2013 (UTC)
It seems fixed to me. -- Ypnypn (talk) 18:26, 12 February 2013 (UTC)
Stylesheets are loading again now, but Pop-ups are working intermittently. Rivertorch (talk) 19:28, 12 February 2013 (UTC)
────────────────────────────────────────────────────────────────────────────────────────────────────I'm not noticing any issues now (I wasn't online when the problem was reported). Per the standard drill, please bypass your cache before reporting that you're still seeing a problem. You might be seeing it because of something cached by your browser that has since been fixed on the Wikimedia servers. – PartTimeGnome (talk | contribs) 22:05, 12 February 2013 (UTC)

Is the default skin borked?[edit]

About every third page load today, Wikipedia refuses to load the default skin, and instead loads as raw HTML. It's been doing this for several hours now; not every time, but often enough to be annoying. Is this sort of intermittent problem happening for anyone else? --Jayron32 20:06, 12 February 2013 (UTC)

This is covered at #Page loading issues and #Displaying Issues above. --Redrose64 (talk) 20:29, 12 February 2013 (UTC)

removing userpages from Category:Wikipedia[edit]

Unfortuntely, I cannot figure out how to do it; I guess they are transcended in some templates. Category:Wikipedia reports 4 user pages: User talk:Хорошинда, User:Tohey, User:Diremarc/Sandbox/Archive and User:Dairiki. Hopefully somebody can figure out where they are and remove those categories. --Piotr Konieczny aka Prokonsul Piotrus| reply here 18:23, 12 February 2013 (UTC)

Here we go...
So they all transclude Wikipedia:Userboxes policy pages, which are in Category:Wikipedia. DMacks (talk) 18:39, 12 February 2013 (UTC)
I came to the same conclusion, and since none of these users have edited in years, I invoked WP:NOBAN, see User:Dairiki, User:Diremarc/Sandbox/Archive, User:Tohey and User talk:Хорошинда. --Redrose64 (talk) 20:16, 12 February 2013 (UTC)
Also, since Category:Wikipedia is for encyclopedia articles, project pages like Wikipedia:Userboxes/Using existing shouldn't be in it anyway. I've removed the category from the page. (It was only added 10 days ago.) – PartTimeGnome (talk | contribs) 22:29, 12 February 2013 (UTC)

"Email this user" edit notice[edit]

A discussion has started at Wikipedia talk:Emailing users about user customized edit notices when using the "Email this user" functionality. Please feel free to discuss and provide feedback. Hasteur (talk) 20:19, 12 February 2013 (UTC)

many flag icons in a big table[edit]

I have a less than entirely favorable view of the quality of the article Daisaku Ikeda, but quality aside, it's a slug to load. It contains a long table, which itself contains many flags, and which first grows, slowly, and then is collapsed in front of me while I wait for the page to load. I could experiment to see where the problem lies, but guessed that people here would immediately know from experience. (Of course, my particular choices of browser may make particularly heavy weather of the page.) Comments?

(Not strictly a technical matter, but I'm also not convinced that there's a rationale for the flag for each institution. [Adding them was not my idea.] Most of these institutions aren't national, in the narrower sense. Though OR tells me that even in the moderately democratic nations of Asia, municipal, privately financed and other institutions are thought of as representing their nations, or anyway as doing so outside those nations.) -- Hoary (talk) 02:15, 13 February 2013 (UTC)

The use of flags in this situation fails WP:MOSFLAG. They should be removed. --MASEM (t) 02:20, 13 February 2013 (UTC)

Ratings at bottom of pages[edit]

Have there been any technical problems reported regarding the ratings at the bottom of pages? --Bob K31416 (talk) 23:22, 8 February 2013 (UTC)

Do you have something specific in mind? After your post there is a report below at #"Rate this page" broken? PrimeHunter (talk) 17:23, 10 February 2013 (UTC)
I am having problems as well. There is html code in { brackets (like a template) under the ratings. ―Rosscoolguy 22:24, 10 February 2013 (UTC)
On Calculus of variations the ratings changed abruptly from 3's and 4's to all 5's with no, or no significant, change in the the number of ratings. On Functional derivative the ratings changed from 3's and 4's to no ratings, and then to all 1's. --Bob K31416 (talk) 11:58, 11 February 2013 (UTC)
It's happening to me as well. It says something like this: 31 ratings. Does anyone know why? Supercuty27 (talk) 06:48, 13 February 2013 (UTC)

"Rate this page" broken?[edit]

I'm seeing something rather odd in the "Rate this page" box at the bottom of pages - instead of the numbers of ratings being displayed, something along the lines of "{{PLURAL:60|one rating|60 ratings}}" is being displayed. Prioryman (talk) 17:03, 10 February 2013 (UTC)

It's a code regression. Legoktm (talk) 17:13, 10 February 2013 (UTC)
Here is the displayed text with nowiki tags so others can see what you mean (it has now been fixed above): {{PLURAL:60|one rating|60 ratings}}. It renders as "60 ratings" here in wikitext but apparently not in the ratings box. The magic word plural is documented at Help:Magic words#Formatting. PrimeHunter (talk) 17:19, 10 February 2013 (UTC)
Yes, I am seeing the same problem. That number shows how many people have voted, but it is appearing as what PrimeHunter said in the previous comment. Supercuty27 (talk) 06:43, 13 February 2013 (UTC)

I'd like to Wikilink the date in my signature. How?[edit]

I noticed an editor in Talk:Confirmation#Capitalization has a 2006 and a red-linked 1 Sep. I'm pro-Wikilink and would enjoy the same (but with the full month name). It appears that editor has retired, so I can't ask him how. Sorry if this is painfully obvious or an old question, but what should I do? InedibleHulk (talk) 19:05, 12 February 2013 (UTC)

To be clear, I want this to happen automatically when I type four tildes. I don't want to manually Wikilink the dates after I post something. InedibleHulk (talk) 19:08, 12 February 2013 (UTC)

Well, I think what you'd have to do would be to bypass the addition of the automatic timestamp. You could do this by simply typing three tildes instead of four. You could then amend your signature to end with something like {{subst:CURRENTTIME}}, [[{{subst:CURRENTMONTH}}|{{subst:CURRENTDAY}} {{subst:CURRENTMONTH}}]] [[{{subst:CURRENTYEAR}}]] (UTC). Of course, I have no idea if the incredibly temperamental signature thingy will allow you to actually do that. There might also be a simpler way to do this... the first person I can think of who might know is X! (talk · contribs), who keeps that cool .beat time in his signature. — PinkAmpers&(Je vous invite à me parler) 19:28, 12 February 2013 (UTC)
Thanks, I'll give it a whirl. InedibleHulk (talk) 19:42, 12 February 2013 (UTC)

Took a small bit of tweaking, but I like it. Thanks for the help. InedibleHulk (talk) 20:08, February 12, 2013 (UTC)

Please note that date linking has been discouraged for some four years now (since December 2008, IIRC); also, dates in signatures normally follow the day-month-year format, so your month-day-year format may confuse certain scripts and bots (that's if they're not confused already by the linking). --Redrose64 (talk) 20:27, 12 February 2013 (UTC)
Duly noted. I've fixed a redundant pipelink, hopefully that makes the bots' work easier. If any major problem arises, I'll gladly change it back. InedibleHulk (talk) 22:01, February 12, 2013 (UTC)
I've pinged Misza13, the owner of User:MiszaBot I and its ilk, which can be quite fussy about the format of timestamps. MiszaBot II never archived this message, for instance. As implied at User:MiszaBot/Archive FAQ, this problem could be fixed by adding the timestamp in a normal format in an HTML comment (<!-- -->); for this you'd have to create a template like {{!}} to generate a less than sign, or something. The whole idea sounds like far more trouble than it's worth. Graham87 04:37, 13 February 2013 (UTC)
Concerning “far more trouble than it’s worth”, I’d also like to remind about WP:SIG#Length: “Keep signatures short, both in display and in markup” (emphasis mine).—Emil J. 13:04, 13 February 2013 (UTC)

Scribunto extension roll out planned for February 18[edit]

Wikimedia Foundation has been working on improving performance and adding features to template pages within MediaWiki. As a result of this WMF developed a new Scribunto extension for MediaWiki which enables Lua scripting language for templates. The developers are confident about the extension's maturity and it was decided that Scribunto is going to be deployed to this wiki in the first batch. The deployment will begin on February 18 and will simply add a feature. Please be so kind and spread the word about the deployment on your wiki. If you are interested in converting current templates to Lua, please see more information and submit your feedback to Lua page on Meta. Regards, Kozuch (talk) 18:04, 13 February 2013 (UTC)

Login/create account and watchlist now live on the mobile web[edit]

Screenshot of the mobile watchlist view

Exciting update from the Wikimedia mobile web team: we've just released the ability to log in or create an account, watchlist/unwatchlist pages, and see a mobile-friendly view of your watchlist on the mobile web ( – and the ".m" version of every other mobile Wikimedia project). See screenshot of my watchlist to the right :) The watchlist view comes with a full list of articles you've starred, as well as a filterable "modified" view; you can see the diff by tapping on any change. Check it out on your Internet-enabled mobile device of choice, and feel free to leave feedback for us on Meta, or visit us on our IRC channel, #wikimedia-mobileconnect

Next up, we're working on contributory features for mobile, including photo uploads and editing, so stay tuned for the mobile revolution, already in progress! :) Maryana (WMF) (talk) 20:50, 13 February 2013 (UTC)

Template question[edit]

For Audie Murphy an IP tried to insert a legitimate website address into the infobox. It didn't work for the IP address. I looked at the website coding for Template:Infobox military person, and it looks identical to my lay eyes as the website coding on Template:Infobox person. As a test, I switched out what the IP had added, and instead inserted the website URL template from Bill Gates, and the Bill Gates website did not show on Audie's infobox. I've changed Audie Murphy's to the URL template. It still doesn't work. Can anyone look at the Audie Murphy infobox and figure out why the website url does not work there? — Maile (talk) 20:59, 13 February 2013 (UTC)

I think the problem is in {{Infobox military person}}, where it uses header20 for the "Website" label and data20 for the url itself. According to the documentation at {{Infobox}}, the use of header20 means that data20 is ignored. Compare this with {{Infobox person}}, which uses header70 and data71 for its "Website" field - consecutive numbers. -- John of Reading (talk) 22:19, 13 February 2013 (UTC)
I discovered the same and fixed it.[6] It had been broken since 30 December. PrimeHunter (talk) 22:24, 13 February 2013 (UTC)
Thank you both. — Maile (talk) 22:34, 13 February 2013 (UTC)
I was sure that I'd coded the addition of the website properly (see Template talk:Infobox military person/Archive 2#URL parameter) - it seems that somebody else has since screwed with my code. I deliberately chose rows 70/71 because I was aware that some people would want a merge into {{infobox person}} at some point; that used rows 70/71 (and still does, rows 64-69 being unused).
Despite what the {{infobox}} documentation says about leaving gaps, there are people who insist on having all the rows in unbroken sequence from 1 - which makes it a pain to insert additional rows later on, you first have to renumber those that are already there. Anybody here remember programming in BASIC with no RENUMBER command? We did everything by tens, because we knew that at some point, statements would have to be inserted.. --Redrose64 (talk) 23:57, 13 February 2013 (UTC)
Tens, you were being optimistic :-) I used to use 20 or even 50 and still hit problems on occasions! NtheP (talk) 14:57, 14 February 2013 (UTC)

Extra breaks in template[edit]

Hello :) I'm doing a new template with collapsible rows. Something went wrong and I've got

<p><br /></p>

as many times as rows in the template (I noticed them in the page's code). Here's the template in my noteboook. Where's an error? Help, please! (talk) 23:53, 13 February 2013 (UTC)

I believe I fixed that problem for you; see this edit. You didn't have the newlines enclosed in the parser functions, so all of the newlines in the code were showing up no matter what. However, I reverted myself because that broke the example in the documentation, probably because MediaWiki is stripping those newlines while parsing, which would break the template. You need to find a way to include the newlines in the parser functions, but not have them begin or end a parameter in the parser function, which would cause them to be stripped. jcgoble3 (talk) 00:48, 14 February 2013 (UTC)
I still haven't found the way to include the newlines in the parser functions.
Feel free to edit the template if you got some idea. Thank you very much for taking time and help! (talk) 01:13, 14 February 2013 (UTC)
Maybe the if parametr for building the "tr" or line should be different? At the moment it's the same parametr ("8") as in the line itself:
|}}{{#if:{{{8|}}}|{{!}} style="{{{styleleft}}}" {{!}} {{{7|}}} {{!!}} style="{{{styleright}}}" {{!}} {{{8|}}}

So, maybe this would help:

|}}{{#if:{{{newline8|}}}|{{!}} style="{{{styleleft}}}" {{!}} {{{7|}}} {{!!}} style="{{{styleright}}}" {{!}} {{{8|}}}

Hence the documentaion:


So, it will built logically. Right? (talk) 01:36, 14 February 2013 (UTC)

I've made several changes, but empty rows still showing as
<p>  </p>
Don't know what to do... (talk) 02:41, 14 February 2013 (UTC)
Started working with the following:
{{!}}colspan=2 style="{{{stylesubtitle}}}" {{!}} {{{2}}}
| }} {{#if:{{{4|}}}|<br>
{{!}} style="{{{styleleft}}}" {{!}} {{{3}}} {{!!}} style="{{{styleright}}}" {{!}} {{{4}}}
| }}
I've added <br> after parser's slash. Also, added space characters. Have a look at the result and also check out the page's code. Will it work correctly in all browsers? And is it the right coding after all? All I know — it works with my Chrome browser. (talk) 05:39, 14 February 2013 (UTC)
Parser functions are independent of the browser as they are processed on the WMF servers, and once they're parsed you're left with a standard wikitable, which the servers convert to basic HTML table code before feeding it to your browser, so it will work in all browsers. I removed some unnecessary stuff, which killed the last case of <p><br /></p>. The template should work now. jcgoble3 (talk) 06:44, 14 February 2013 (UTC)
Yes, it works! By the way, my question was about that <br> stuff, will it expand cells in some browsers? Thanks a lot for taking time and help me! -- (talk) 16:15, 14 February 2013 (UTC)
I tested it in four browsers (Safari for Windows 5.1.7, Chrome 24.0.1312.57 m, Firefox 18.0.2, and IE 10), with no problems in any of them. jcgoble3 (talk) 19:48, 14 February 2013 (UTC)
Thank you! Now I'm getting expand the limit of rows. Already publish it in Ru.Wikipedia by the name Template:Collapse. Cheers! (talk) 05:56, 15 February 2013 (UTC)
Well done at getting this working! For future reference, here's a trick I like to use to stop trailing newlines and spaces being stripped from the end of a parser function: Place <nowiki/> after the line break(s), before the closing }}. For example:
{{#if:{{{show1|}}}|This is the first para.

<nowiki/>}}{{#if:{{{show2|}}}|This is the second para.

PartTimeGnome (talk | contribs) 00:01, 15 February 2013 (UTC)
Thanks mate! Nice trick of yours! That things I always enjoy :) I think I'll use it somehow, too. (talk) 05:56, 15 February 2013 (UTC)

Constant error messages[edit]

Will Wikipedia pay for my therapy when I suffer a major nervous breakdown as a result of having to post every edit several times against error messages and edit conflicts (even when postings are minutes apart)? The random loss of character spaces and appearance of unwanted linespaces after edits is also driving me crazy. Operations that use to be performed in a few seconds now take anything up to ten minutes. Please expedite ridding the system of its current gremlins, as they must be a serious deterrent to many people willing to contribute time and effort to improving Wikipedia. Kim Traynor 00:34, 14 February 2013 (UTC)

I've been having issues with editing large pages for the past few days (and probably longer). For example, New York City. I takes 30-40 seconds simply to load the page. When I edit, it takes 40+ seconds then I get a timeout error and it fails to reload the page w/ my change. My edit *does* get saved though. I'll try poking the tech people about this though not sure of a simple, easy fix for this. Maybe introducing the Lua thing for templates will help or something. --Aude (talk) 04:22, 14 February 2013 (UTC)
I think it is the newer operating systems. Why I edit with my old 286 and go online with a 9600 baud modem then things move much faster.--Canoe1967 (talk) 04:26, 14 February 2013 (UTC)
Hmm.... well, I chatted with tech staff and they suggest 1) when we get Lua for templates (next week, planned), that will help a lot. 2) editors can try to remove some of the templates. There might be certain ones that are especially "heavy" (with lots of parser functions, recursively / nested additional templates) Or it might be a sign that some of the article content needs to be split into subarticles, to reduce the number of cite templates. 3) for template editors, some of the templates can probably be improved. They can't change the timeout period. For edit conflicts, there might be some additional issues going on and I am not sure the solution. --Aude (talk) 11:40, 14 February 2013 (UTC)
One work-around is, where practical, to use section edits rather than editing the whole article. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:49, 14 February 2013 (UTC)
Yes, well even this edit threw the error in question, as did this one. The problem on that page is the massive RDT up the right-hand side, generated by {{Cross Country Route}} in the lead section. Even if you edit only one section, the MediaWiki software must still rebuild the whole page. --Redrose64 (talk) 21:04, 14 February 2013 (UTC)
  • I just ran some numbers New York City is 1491874 bytes (excluding images), at 1.4MB + files that is an extremely large page. With 389 references and a template depth of 32 I am surprised the page even renders. A good size for article size is about 1/3 that size. Werieth (talk) 21:06, 14 February 2013 (UTC)
Probably running up on Wikipedia:Template limits; see that page to check the NewPP limit report. --— Gadget850 (Ed) talk 00:09, 15 February 2013 (UTC)

Wingnut idea #2[edit]

Is it possible for a bot to cross check articles in dead people categories with infoboxes that don't have images? This way we could create a category to search for articles that we are allowed to add fair use images to.--Canoe1967 (talk) 02:43, 14 February 2013 (UTC)

This is technically possible. You only need to find a bot expert willing to program the bot. Ruslik_Zero 12:12, 14 February 2013 (UTC)

Proposal about search[edit]

Hello all: sorry if this is impossible, but does anyone else think that it would be a good idea to implement a search from the end of an article's name (e.g. I could search for all articles that ended in "Persei")? For what I'm working on with astronomy articles, it would be really nice to have such a function. StringTheory11 (t • c) 05:30, 14 February 2013 (UTC)

There's a Toolserver tool for this called "grep" (named after the Unix utility program). You can use the regular expression Persei$ (the metacharacter $ represents the end of the page title). The MySQL documentation describes the supported regular expression syntax. PleaseStand (talk) 08:13, 14 February 2013 (UTC)
A few of us have working implementations of this on the Toolserver, the problem being that queries take an age to run (suffix searches being non-indexed in the database). - Jarry1250 [Deliberation needed] 10:41, 14 February 2013 (UTC)
OK, thanks! StringTheory11 (t • c) 22:23, 14 February 2013 (UTC)

Merging Collapsibles[edit]

Hello :) Is it possible to merge 2 tables like it's one? Without row between and with working [hide] [show] buttons?

1 2 3
1*1 2*1 3*1
1*2 2*2 3*2
1 2 3
1*1 2*1 3*1
1*2 2*2 3*2
{| class="wikitable collapsible" style="margin-bottom:0;"
!colspan="3" style="background:#ccc"|Table1
{|class="wikitable collapsible" style="margin-top:0;"
!colspan="3" style="background:#ccc"|Table2

-- (talk) 16:22, 14 February 2013 (UTC)

Do you mean something like above? Ruslik_Zero 18:49, 14 February 2013 (UTC)
Yes, something like that. I wanted my tables here look like one, but there are still bold line between them. I think this comes from the table global styles (wikitable, tiles). Is it possible to make the borders between them more thin, like it's one table? Thanks a lot for taking time and help! -- (talk) 03:06, 15 February 2013 (UTC)

custom code to prevent "Archived from the original" string from displaying?[edit]

I'd like to add some custom code to my common.css / common.js so that if a reference has an "active" archiveurl parameter (i.e. the deadurl parameter, if it exists, is set to "yes"), the "Archived from the original on [date]" string appended to the reference is not displayed -- I feel it adds unnecessary clutter to the reference. For example, I already have a bit of custom CSS that doesn't display the accessdate parameter: span.reference-accessdate { display: none; } Does anyone know how to do this? The best thing I can think of is some custom JS that looks for the text "the original" or "Archived from" in an element and just removes it. Thanks! Wingman4l7 (talk) 23:14, 14 February 2013 (UTC)

You can hide accessdate in that manner because {{citation/core}} wraps it in <span class="reference-accessdate">. You can make a request at {{citation/core}} to do something similar for archiveurl. --— Gadget850 (Ed) talk 23:39, 14 February 2013 (UTC)

Special:BookSources update needed[edit]

The entry for in the Book swapping websites section of Special:BookSources needs to be deleted. The WP article is deleted and the website is gone. -- Alan Liefting (talk - contribs) 01:53, 15 February 2013 (UTC)

And while we are at it can we have a hatnote on the page to Help:ISBN, a reader help page that I have presently thrown together? Cheers. -- Alan Liefting (talk - contribs) 02:04, 15 February 2013 (UTC)

Do you mean at Wikipedia:Book sources#Book swapping websites? Suggestions can be posted to Wikipedia talk:Book sources, but the page is not protected. The top of Special:BookSources is from MediaWiki:Booksources-summary. PrimeHunter (talk) 02:55, 15 February 2013 (UTC)
Ah. That's right. Forgot about that page. Will try and sort it out. -- Alan Liefting (talk - contribs) 03:16, 15 February 2013 (UTC)

A script on this page may be busy, or it may have stopped responding.[edit]

Sorry if this is a FAQ but I seem to have been getting this message recently when I look at diffs.

A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete.

Admittedly my browser is old (Mozilla/5.0 (X11; U; Linux i686; en-US; rv: Gecko/2008070206 Firefox/3.0.1). Any ideas? --Northernhenge (talk) 19:59, 12 February 2013 (UTC)

This is covered at #Page loading issues and #Displaying Issues above. --Redrose64 (talk) 20:29, 12 February 2013 (UTC)
Thanks Redrose. I guess you've concluded I'm seeing a different symptom of the same problem. --Northernhenge (talk) 20:44, 12 February 2013 (UTC)
I've merged the sections mentioned by Redrose64 (and also #Is the default skin borked?). I've not merged this section, however, since I'm not convinced this is the same issue. The "A script on this page may be busy" warning can only appear for scripts that have been loaded, and that section is about scripts (and CSS) not loading.
I'm not ruling out that the problems could be related, though. It is possible that a successfully loaded script was malfunctioning because of another script that failed to load. – PartTimeGnome (talk | contribs) 22:05, 12 February 2013 (UTC)
That was my thought - along the lines of: page loads script a.js which depends on script b.js but server fails to return b.js; and browser doesn't want to wait forever, so throws error instead stating that a.js has hung. --Redrose64 (talk) 23:00, 12 February 2013 (UTC)

I've now tried a few other computers and can't reproduce this error elsewhere, so given the lack of people saying "me too" I suggest we blame my computer and move on. Apologies. --Northernhenge (talk) 13:31, 15 February 2013 (UTC)

Searching for files in all the wrong places[edit]

When processing OTRS permission requests people often provide the name of the image, without a link so I have to search for it.

For example, someone just emailed regarding File:Member for Fremantle Melissa Parke MP Profile Pic.jpg

Their email simply included Member for Fremantle Melissa Parke MP Profile Pic.jpg. If I enter that into the search box, I get the response:

There were no results matching the query.

The default search is articles only, but I have added a check box next to File. I thought that should search for a file with that name. I thought this did work, but I may be dreaming.

While it may seem simple to add "file:" to the search, I do this dozens of times a day. I think searching for a file named Member for Fremantle Melissa Parke MP Profile Pic.jpg should find that file, especially when I check the File box. Is there something wrong, or something I am doing wrong?--SPhilbrick(Talk) 01:10, 14 February 2013 (UTC)

presuming i understand the question, you probably want to click Preferences => Search and tell it to search in the File: namespace by default. קיפודנחש (aka kipod) (talk) 01:43, 14 February 2013 (UTC)
I had already done that (That's what I meant when I said I added a check box next to "Files" I didn't mean every time, I did it via preferences.)--SPhilbrick(Talk) 14:04, 14 February 2013 (UTC)
File:Member for Fremantle Melissa Parke MP Profile Pic.jpg was uploaded yesterday and hasn't been indexed by the search function yet. See Help:Searching#Delay in updating the search index. If you enter an exact page name including namespace in the search box then you go to the page even if search hasn't indexed it yet. PrimeHunter (talk) 01:56, 14 February 2013 (UTC)
I think that's it. Which is very disappointing, because editors often file a permission statemnt the same day they upload, which mean it is quite common to have to struggle with the search.--SPhilbrick(Talk) 14:04, 14 February 2013 (UTC)
Sadly, that means it will be easier to process if we ignore it for a day. Too bad.--SPhilbrick(Talk) 14:05, 14 February 2013 (UTC)
i find the most convenient place to look for stuff i did myself recently is Contributions. for stuff i did/uploaded longer time ago, regular search should do the trick. peace. קיפודנחש (aka kipod) (talk) 20:28, 15 February 2013 (UTC)

How to track the use of Australian War Memorial photos on Wikipedia[edit]

The Australian War Memorial is in the process of increasing the size of the images it makes available on its collection database. Many of these images are in the public domain and are used in Wikipedia articles, with the great majority being hosted on Wikimedia Commons and tagged with {{AWM-image}} there (which in turn places them into ). The AWM is currently in the process of increasing the size of the images on its database by about a third, and when the Imperial War Museum made a similar change to its image database last year it (anecdotally, but clearly in my experience) lead to a large expansion in the use of IWM images on En-Wiki. Is there any way to measure the extent to which this change influences the use of Australian War Memorial images? (eg, can we take a snapshot of the usage of images with the {{AWM-image}} tag now and compare it to the figure in, say, six months time?). Please let me know if this is a question best asked at Commons. Thanks, Nick-D (talk) 09:54, 14 February 2013 (UTC)

Commons is probably the best place to ask.--ukexpat (talk) 20:11, 14 February 2013 (UTC)
Actually just make a request at someone with toolserver access can make quick work of this. Werieth (talk) 21:13, 14 February 2013 (UTC)
Once it's in the tracking category on Commons, it's relatively trivial to get comprehensive usage data - well, as comprehensive as we offer. The best tool for this is glamourous. Feel free to drop me an email if you've any problems with it.
However, changes over time are pretty hard to do. Your best bet is to take a copy of the report now and save it for future reference; you can't easily go back and say "show me as of last January". Andrew Gray (talk) 21:57, 14 February 2013 (UTC)
That's an excellent tool - thank you for pointing me to it. Regards, Nick-D (talk) 10:49, 15 February 2013 (UTC)

What is going on with the block log?[edit]

I specified a number of days (1001) when I blocked this IP, but the block log shows an amazingly precise 2 years, 270 days, 12 hours, 21 minutes and 36 seconds. I didn't specify an exact number of seconds, so I'm not sure where those 36 seconds have come from (1001 days is 3,603,600 seconds according to my calculator) nor why the log can't just display the duration I specified. I assume this is a software change; can somebody tell me where and/or why it was changed? HJ Mitchell | Penny for your thoughts? 21:13, 14 February 2013 (UTC)

The mean length of a year in the Gregorian calendar when all leap year rules are considered is 365.2425 days. Your numbers, including the 36 seconds, match exactly if the 2 years are each 365.2425 days. PrimeHunter (talk) 22:24, 14 February 2013 (UTC)
  • Which leads to further confusion. In many cases, 2-year blocks are now represented as 1 year, 364 days, and a number of minutes.--Jasper Deng (talk) 02:02, 15 February 2013 (UTC)
So the answer would be that, rather than programming the software to know what day it is and to calculate what day it will be in 1001 days or whether or not there are any leap years involved, it just sees everything as a portion of a year and bases it on strict mathematics? Is that about right? Makes sense in a purely mathematical sort of way even if it is a bit weird. Then again I've never seen a block for 1001 days before either, so Harry is a b it weird too. Beeblebrox (talk) 03:18, 15 February 2013 (UTC)
Presumably caused by Gerrit:41209 (MW1.21wmf7) changing $durationIntervals in Language.php which affects the output from translateBlockExpiry() in LogPage::actionText(). User:Liangent's change refined year length from 365.25 to 364.2425 days, to reflect the long-term Gregorian calendar. Does the quirk only affect the display of existing blocks (a transitional problem), or are new blocks also being displayed inconsistently? — Richardguk (talk) 09:03, 15 February 2013 (UTC)

Bypassing the no page creations for anons?[edit]

As is likely well known, since the Seigenthaler incident unregistered editors have been prohibited from creating new articles on Wikipedia. However, I came across an apparent method for bypassing that policy the other day, where Aristasia had been converted from a redirect - which it had been since it was moved and changed to a bio on the movement's main proponent in 2007 - into a full-blown (and highly promotional) article by an IP editor [7]. It has since been reverted back to a redirect (and protected as such), but I was wondering if this possibly needed to be looked at, as it essentially provides a way for unregistered editors to "end-around" the Seigenthaler rule. - The Bushranger One ping only 06:13, 15 February 2013 (UTC)

Technically, it's not bypassing, it's a loophole in our rules. A redirect is a page, and they aren't 'creating' a page, just adding to the redirect (and removing it in the process). Afaik, there's no technical way to bar new users from changing the redirect to an article (as of now), so it may be good to just have all redirects semi'ed until it's possible to implement a change like that? Or just leave it as is, and revert if necessary? gwickwiretalkedits 06:21, 15 February 2013 (UTC)
  • Special:AbuseFilter/342 monitored for redirects becoming articles, but it had far too many false positives. Something like 99% false positives, so it was turned off 2 years ago. (talk) 09:34, 15 February 2013 (UTC)
Adding, when redirects become articles they are automatically included in Special:NewPagesFeed, so NPPers doing Page Curation will see them. See also Wikipedia talk:Page Curation/Archive 3#ON second thought, do we actually want to patrol redirects?. Cheers. (talk) 10:30, 15 February 2013 (UTC)
There's a 1% joke in there somewhere... Face-wink.svg Semi-ing all redirects would be impractical (and prevent legit IP editors from fixing them as needed). I guess just keeping a sharp eye out is the answer, thanks. - The Bushranger One ping only 13:56, 15 February 2013 (UTC)

Self-generated edit conflict[edit]

The software appears to have changed in an unexpected way for me: if you examine my two most recent contributions before this request (accessible in my sig), you will see I was on a talk page. I entered a new section headline, wrote the entry, and got an edit conflict message with the usual two-part screen 1) actual text on wikipedia, 2) my attempted text.

My problem is that I was the only editor on that talk page at that time; so how is the software thinking there were two separate processes? To my thinking, that is new behavior on the part of the software. __Ancheta Wis   (talk | contribs) 06:20, 15 February 2013 (UTC)

Your mouse button apparently 'bounced'. Whenever I acidentally double-click the save button, I always get an edit conflict with myself. Edokter (talk) — 10:55, 15 February 2013 (UTC)

Wikidata office hour on IRC[edit]

I'll be available on IRC to answer any question you might have around Wikidata in an office hour tomorrow, Feb 16 20:00 UTC. Further details are here. I'd be delighted to see many of you there. --Lydia Pintscher (WMDE) (talk) 21:30, 15 February 2013 (UTC)

setting of Huggle's diff colour/monobook.css[edit]

Hello. I have following problem: I haven't found out yet how I can change the color of Huggles diffview between versions in the English Wikipedia. It's actually dependent on the monobook.css and on Wikipedia appearance I thought. I'm using the same monobook.css here as in the German Wiki, but for some reason it doesn't have any effect here on Huggle. The monobook.css changes the look of Wikipedia here aswell, however, it doesn't have any effect on Huggles diffview whereas in the German Wiki it has. It's actually this default view which I want to have changed and make it look this way. I would be very grateful if someone could help me. Thanks, --Avoided (talk) 21:39, 15 February 2013 (UTC)

Undeployment of MoodBar/Feedback Dashboard[edit]

Hi guys!

This is just a heads up to let you know that we have undeployed Moodbar and the Feedback Dashboard from the English Wikipedia. As an experiment, Moodbar was a fair success but we have come to the conclusion that it will require a fair chunk of development work (on the Feedback Dashboard side) to make it fully usable as a mechanism for new user engagement.

Additionally, we have new tools that we are planning to roll out and test that will operate in the same space and we want to avoid conflicts.

We still have faith in the Feedback Dashboard, and expect to fully productize it as part of the upcoming Flow initiative.--Jorm (WMF) (talk) 23:09, 5 February 2013 (UTC)

What about the moodbar dashboard on nederlands Wikipedia? It's still there, active. And I don't see anyone giving them notice about this Undeployment or about the reasoning for that. I guess the appropriate forum would be nl:Wikipedia:De kroeg, and they do understand english. --Atlasowa (talk) 08:32, 7 February 2013 (UTC)
Atlasowa, thanks for posting a notice at wpnl. @Jorm: if it turns out we at wpnl would like to keep using Moodbar in its current state until the new tools become available, would that be possible? MrBlueSky (talk) 22:27, 9 February 2013 (UTC)
Absolutely. We are only undeploying it from the English Wikipedia at this time. It is conflicting with another set of work that we hope will bear greater fruit.--Jorm (WMF) (talk) 19:38, 11 February 2013 (UTC)
Very nice. Thanks! MrBlueSky (talk) 22:06, 12 February 2013 (UTC)
Except... apparently moodbar stopped working 3 days ago on nlwp telling everyone that they are violating 'the rules' when they try to post feedback or try answering feedback (bugzilla:45068). What magic was done during that time? Sumurai8 (talk) 11:54, 16 February 2013 (UTC)

non-infoboxes, but *boxes[edit]

I understand the concept of the Infobox, and happy to see them, but i feel the need to complain about the apparently random naming, for i have found: Campaignbox, chembox, drugbox, Geobox, hockeybox, infobox, Kent_College_Infobox, Mtnbox, navbox, Pfam_box, Planetbox, Pwcompanybox, Quote_box, Reactionbox, starbox, superherobox, Superteambox, taxobox, WPMIXInfobox, and wrugbybox. Note i mention the "infobox" as distinct from "Infobox". I feel that making up new names for an Infobox is sabotage against sharing the data in the boxes, because in the search for the "Infobox" on a topic, "superherobox" is going to be missed.

I'd also like to rail against mentioning in the <comment> tag that a specific {{Infobox template was used, as a text search for the "{{Infobox" header will trip up on that mention.

Also, there is random use of the "|" divider, sometimes many on one line, some at the end of a line, some at the beginning of a line, and this makes parsing somewhat difficult if you are trying to not ignore the "\n" used inappropriately too.

I am also not fond of the nesting of {{ }} in each other, perhaps {2, {3, or {{ and {{{ can be used? I can parse the {{ blah Template:Foo Template:Blarrg bar bloof }} (i did not write Template:Blarrg in that nest, wikipedia added it there) out of the page ok, but the nesting gives me the crawly skin that something is going to go wrong with an unclosed }} somewhere, because after all, there's already unclosed "{{Infobox" within <comment></comment>.

I'm not into chasing statements like "this is the wrong place blah blah blah", so if anyone else thinks this is important enough to repost in the right place, feel free. (talk) 06:22, 14 February 2013 (UTC) aikat (talk) 19:45, 14 February 2013 (UTC) aikat

  • Many things to ponder: I can understand the frustration, but many people have promoted the "*box" names, such as "taxobox" or "navbox", as meaning specific types, or subtypes, of infoboxes. Typically, a wp:navbox is a horizontal box near the bottom of an article. As for the placement of the pipe/bar ("|"), I recommend to use bar-space "| " at the start of lines in an infobox, then indent by two spaces for nested infoboxes inside of other infoboxes. Much confusion has been caused by several Kafkaesque days of warped parser changes, over the years, when newlines were added to template parameters unless followed by an end-bar "x=45|"; however, I think people have realized to preserve end-spaces or newlines by passing a null-nowiki tag "<nowiki/>" around a parameter value (such as "x=<nowiki/> space-enclosed text <nowiki/>"). Also, URLs can be passed inside an wp:empty-name parameter (such as "{{{|}}}") to allow an equals-sign in a web address. As for the overuse of multiple braces, a fix could be to extend the parser to treat bracketed "#[x]" as an optional syntax to get parameter {{{x}}}, while "#]" would be a literal right square-bracket symbol inside any "#[name]" parameter. Also, the parser could be extended to allow "#then" and "#else" to be "|" and "#endif}}" would be a token to ensure the matching end of a #ifeq (or such), whereas simple "}}" would be the typical pot-luck end of a parser function. Perhaps we need more essays about these issues. -Wikid77 (talk) 18:07, 15 February 2013 (UTC)
But why "navbox", when "Infobox nav" works better? Isn't "Kent_College_Infobox" absolutely going overboard on new names, when "Infobox Kent_College" might fit in better? Same for all the rest, they are infoboxes, their name on the page is Infobox, using a template for a subtype of thing is "Infobox subtypename", not making up a new *boxname! (talk) 21:44, 15 February 2013 (UTC) aikatt
Navboxes are not infoboxes. Infoboxes (information boxes) contain quick information and stats about a subject, whereas navboxes (navigation boxes) contain links to related subjects. The {{navbox}} template does not use the {{infobox}} template. However, your point is a good one for things like {{taxobox}} and {{chembox}}, which do function as infoboxes. On a technical point, they don't use the {{infobox}} base template either. I suspect there's some history to this. – PartTimeGnome (talk | contribs) 22:09, 15 February 2013 (UTC)
I could waste some time arguing that a navbox is a type of infobox, in that it has info on where to navigate to for related topics of further info. But i won't, because i think i will lose anyhow. I will point out that the use of derived templates, or not, shouldn't change the Infobox from a "Infobox chemistry" to "chembox". The "Infobox Chem" is merely an Infobox, where one would expect to find info, and in this case using the "Chem" template, as needed. I am not arguing the template or layout in this case, but the name the data may be found under on the page. If any of the projects spawned to use wikipedia go looking for an Infobox (proper) on the some random page, there may not be one, altho there may be a "Describe-o-box" or some such. The box should be labled "Infobox", followed by a generic template name. I personally don't care a huge amount over the infobox's inheritance, the template's line-by-line data arrangement, or new or missing data names from the template are/n't in it, as long as the data are easily parsed. And if i can win that arguement, then i point back to the "navbox" arguement. I think i have said all i can on this, i am in no position to do more about it, even tho i strongly believe in my position. (talk) 02:34, 17 February 2013 (UTC) aikatt

People inserting "class ktg6us78hf8vdu7" again[edit]

"Again", cause I searched and found this discussuon in the Italian Wikipedia and this one in English. It's probably some style defined in a style sheet. Latest edits inserting it: [8], [9]. --Moscow Connection (talk) 22:03, 14 February 2013 (UTC)

Google returns a multitude of hits for that class. It is inside <a>, so it is not rendered as a HTML element since <a> is not whitelisted through Sanitizer.php, but it does show the raw markup. The edit is detected by filter 345.[10] --— Gadget850 (Ed) talk 00:05, 15 February 2013 (UTC)
It does seem to be a recurring problem with about 5 apparently good faith edits a introducing it. A search of the class name [11] shows pages which contain the class.--Salix (talk): 10:37, 16 February 2013 (UTC)
I fixed [12] the 12 mainspace cases found by search yesterday. The above search currently reports 2 mainspace cases. Both were introduced after my search yesterday, and both are fixed by now (there is a delay in updating the search index). PrimeHunter (talk) 13:53, 16 February 2013 (UTC)

Watchlist recent changes count[edit]

Hi. My watchlist currently states "Below are the last 316 changes in the last 168 hours, as of 12:27, 15 February 2013". I've noticed this creeping up over the last 8hrs or so, but there's (at most) no more than 120 changes listed. Any ideas? Thanks. Lugnuts Dick Laurent is dead 12:29, 15 February 2013 (UTC)

I'm not sure but it might be related to the Wikidata deployment here. You have a link in your watchlist to show Wikidata edits. Can you click that and check if that is it? --Lydia Pintscher (WMDE) (talk) 12:41, 15 February 2013 (UTC)
Ahh, yes, toggling that shows all the hidden Wikidata edits, such as "Aballay (film) (Q4663322); 11:36 . . Sk!dbot (talk | contribs) (A Wikidata item has been linked to this page.)". Would expect the total though to show non-hidden items, inline with hiding bot edits, for example. Lugnuts Dick Laurent is dead 12:52, 15 February 2013 (UTC)
Can you file this on please? I'm having a hard time properly formulating this for you unfortunately. (Alternatively write it here and I'll put it in bugzilla.) --Lydia Pintscher (WMDE) (talk) 13:21, 15 February 2013 (UTC)
How's this: My watchlist is currently showing the total changes including Wikidata edits, despite the toggle to show Wikidata edits being turned off. This should be the same as turning off bot edits or minor edits (IE, the total isn't included when they are set to hide). Thanks. Lugnuts Dick Laurent is dead 13:32, 15 February 2013 (UTC)
Perfect. Filed as bugzilla:45041. --Lydia Pintscher (WMDE) (talk) 13:43, 15 February 2013 (UTC)
Thanks. I'm not that familar with Bugzilla. What's the usual turnaround for something like this? Lugnuts Dick Laurent is dead 14:49, 16 February 2013 (UTC)
I can't make any promises. There's a lot on the development team's plate. It doesn't seem like a difficult thing to fix. If this is super important to you then I can see if I can push its priority. If anyone who can code wants try and fix it they're welcome to too. --Lydia Pintscher (WMDE) (talk) 15:18, 16 February 2013 (UTC)
Thanks again. It is important to me, but probably not that important overall (to the project). I see someone else (below) has raised the same question, so maybe it will fall into the category of annoying a few people + easy to fix! Lugnuts Dick Laurent is dead 18:26, 16 February 2013 (UTC)

Alignments in FF vs. IE[edit]

Looking at the table here, the table below the two charts is supposed to be center aligned. It works fine in IE9, but is left aligned in FF18. Any idea on how to fix this? Rehman 07:09, 16 February 2013 (UTC)

Fixed (sort of) for those of us using desktops and laptops with this edit, though the situation on the mobile site is no better. It probably would be better to put the graphs in a more conventional gallery or thumbnail box. PleaseStand (talk) 07:37, 16 February 2013 (UTC)
Thanks, that should do for now. Rehman 11:04, 16 February 2013 (UTC)

Something wrong with the watchlist counter?[edit]


There seems to be something wrong with the watchlist page counter. With the default, 37 72 hours, it currently says 91 changes, but when I count the pages in the list of changes, it is only 60. If I select shorter intervals, the number is correct. I have also noted that the icon on my browser is not a "W" (for Wikipedia), but a small D and a small B inside an bigger H, probably's logo, as I visited their pages recently. Is it a cache problem, and if so, how do I get rid of it? I have tried ctrl+f5. My browser: IE.

HandsomeFella (talk) 14:25, 16 February 2013 (UTC)

See #Watchlist recent changes count above. --Redrose64 (talk) 14:46, 16 February 2013 (UTC)
Thanks. Ok. I think it would be better if only changes related to the hide/show selections would count. It's somewhat confusing to read that there are 91 changes, and only 60 are listed. What if the current number for each selection would appear adjacent to its hide/show link? Then the total would be made up by the sum of the numbers selected. That would make more sense. In my case "31" would appear next to the "Wikidata" selection. Think about it.
Btw, cleared the internet history, and then the correct "W" icon showed again.
HandsomeFella (talk) 22:30, 16 February 2013 (UTC)

Strange code in page Encyclopædia Britannica[edit]

See page Encyclopædia Britannica. It has code [[s:Preface to the 1st edition of the Encyclopædia Britannica|Preface]]. Question 1: what does the prefix [[s: mean? And question 2: how or why does it end up in the infobox, with label "read online" (it gives a bad & wrong output btw). -DePiep (talk) 20:35, 16 February 2013 (UTC)

Those are links to wikisource, which work, however {{Infobox book}} is broken with regards to the wikisource parameter. I really dont understand templates so I cant help solve the issue. Werieth (talk) 21:03, 16 February 2013 (UTC)
Yes, see Help:Interwiki linking. It looks as if wikisource is a new parameter, introduced after discussion here. Thincat (talk) 21:16, 16 February 2013 (UTC)
Fixed. {{Infobox book}} had |name={{lang|la|Encyclopædia Britannica}}. But that adds a hidden category link to Category:Articles containing Latin language text which breaks the link to Wikisource when the name is embedded in the s: Wikisource link (you cannot embed a link within a link). Simply specifying |name=Encyclopædia Britannica avoids this. — Richardguk (talk) 21:26, 16 February 2013 (UTC)

Missing texvc executable[edit]

The periodic "Missing texvc executable" error is back today. bugzilla:41188 is about the same problem in October–November 2012. Here is an example with simple formulas:

a = c
c = a
c = c

The formulas "a = c" and "c = a" always render for me but "c = c" sometimes gives this in bold red: "Failed to parse (Missing texvc executable; please see math/README to configure.): c = c". PrimeHunter (talk) 02:15, 16 February 2013 (UTC)

I purged this page, and now the formulas show up. Is this still a problem (for now)? PleaseStand (talk) 05:01, 16 February 2013 (UTC)
It's a periodic error, probably depending on which server processed something. I'm in Denmark, Europe if it means anything. If I preview this section repeatedly then the error is there sometimes, maybe around 20% of the times but I haven't made statistics. Whenever a formula fails, it's still always "c = c". PrimeHunter (talk) 05:45, 16 February 2013 (UTC)
I reopened bug 44653 with relevant information. PleaseStand (talk) 08:21, 16 February 2013 (UTC)
At present I am seeing this error on WP:RD/MA, but only on the final section of the page - under "February 15th" - not on the rest of the page. Gandalf61 (talk) 10:30, 16 February 2013 (UTC)
Thanks. bug 44653 now says fixed. I have previewed the above formulas 30 times without getting the error. PrimeHunter (talk) 13:41, 16 February 2013 (UTC)
Good news. WP:RD/MA looks okay to me now too. Gandalf61 (talk) 14:14, 16 February 2013 (UTC)
For the record, this problem was also seen after this edit on ptwiki. Using action=purge fixed it. Helder 22:04, 17 February 2013 (UTC)

blame visualization of whole article?[edit]

I have seen many visualizations for wikipedia. I'm wondering if there is a one which shows the final article, but then shows when each piece of text was contributed and by who. I think IBM's history flow does this, but that is not online available for use that I know of.

I'm familiar with the "blame" feature which lets you search for text, to see who added it. But I'm picturing instead having essentially all possible blames (for the current revision text only) pre-computed and available for browsing.

I have no idea how to implement this, which is why I was hoping someone has already done it. But if not, do people think its feasible or a good idea? Silas Ropac (talk) 19:45, 16 February 2013 (UTC)

WikiTrust has been down for years. Neo Poz (talk) 22:13, 16 February 2013 (UTC)
That is interesting. They conflated annotation (last-revision and author per word) with this trust thing, where I think even just annotation would be useful. But they say it required a mediawiki plugin, I wonder if can be done externally. Probably not unless you have the full history of the article, which is probably too big to fetch. Silas Ropac (talk) 01:09, 17 February 2013 (UTC)
WikiTrust shows authors and versions of pages. Who wrote what?
Have a look at WikiTrust/WikiPraise:
To activate it just insert the following code in your user script page. In English Wikipedia you will find it here. Add these lines:
Save the page and bypass your browser's cache.
Now when you visit any page and click on any text, a box opens to the left listing the text's author and the date and revision when it was added.
The script will only work on Vector and Monobook skin as expected. (Hint for users of whitelist based script blockers (like NoScript and NotScripts): allow scripts for  – otherwise nothing will happen.)
There is a similar WikiTrust/WikiPraise on Finnish Wikipedia as a gadget. It's a great tool by user:NetAction. --Atlasowa (talk) 08:15, 17 February 2013 (UTC)
When I try to execute this script on ptwiki, it returns an error Cry.png: Uncaught SyntaxError: Unexpected identifier (because the code returned is "Project pt unknown."). Helder 15:22, 17 February 2013 (UTC)
The WikiTrust/WikiPraise tool is available for German and English Wikipedia, not Portuguese. And there is a Finnish version by Zache (as mentioned above). Maybe you should have a look at Zache's description for installing on Finnish Wikipedia. Or ask at de:Benutzer_Diskussion:NetAction/WikiTrust/WikiPraise (english is OK). --Atlasowa (talk) 19:13, 17 February 2013 (UTC)
Wow great to know about. Works on english for me. Very slick. So at least on the back-end this is the same project Neo Poz said has been down for years. So I wonder what the story is, certainly not advertising itself that well. Yeah it seems like they can't do this with standard API's, they run a whole separate database with special compressed change information, and the script sucks that down for each page you visit. The Firefox Add-in looked like it was all about showing "trust" where this cross-browser userscript seems strictly to be annotation, showing who wrote what without judgement. All it needs is a quick way to toggle it on/off though, because you really don't want it on all the time, but you don't want to go and edit your common.js to toggle it. I wonder it can be inserted under Toolbox? Very cool though, lots of hard work behind this I'm sure. I think this is a great tool. I'd like to see something like this built-in to a future version of mediawiki. There is a lot more you could do with this data, if it was available natively. Silas Ropac (talk) 16:09, 17 February 2013 (UTC)
Screenshot ASQM (Article Statistics and Quality Monitor) beta without WikiTrust/WikiPraise
I think the toggle suggestion is good, maybe you can leave a note at de:Benutzer_Diskussion:NetAction/WikiTrust/WikiPraise (english is OK)? User:NetAction seems quite open to the idea. And regarding "built-in", there are indeed plans to integrate WikiTrust/WikiPraise in the ASQM-tool (Article Statistics and Quality Monitor). --Atlasowa (talk) 19:13, 17 February 2013 (UTC)
Yes I will drop him a note, thanks. I think the original WikiTrust browser plugin was meant for readers whereas this version feels like it is more for editors. In that readers care about "trust" but are less likely to want to know the usernames of all the editors who wrote the article. I'd like to see a reader-level tool again some day, built-in to the site eventually. I don't know that "trust" is the answer, if that's what people want to see. Naively I'd rather see something more neutral, that reveals what went into the article, but lets you judge for yourself how much to trust it. But I imagine they thought about this quite a bit before coming with their trust metric, so maybe that is best. Silas Ropac (talk) 21:09, 17 February 2013 (UTC)

Archive links[edit]

I recently archived some links at Botik of Peter the Great. The problem is, the main link of each goes to the archive link. Is there any way to switch the linking using the cite web template? Otherwise I'll just tack {{WebCite}} to the end of the ref. Ryan Vesey 17:21, 17 February 2013 (UTC)

Set |deadurl=no --Redrose64 (talk) 17:30, 17 February 2013 (UTC)
Thank you! I misread the instructions and thought it defaulted to no. Ryan Vesey 17:34, 17 February 2013 (UTC)

LinqToWiki: new library for accessing the API from .Net[edit]

I'd like to introduce LinqToWiki: a new library for accessing the MediaWiki API from .Net languages (e.g. C#). Its main advantage is that it knows the API and is strongly-typed, which means autocompletion works on API modules, module parameters and result properties and correctness is checked at compile time. Any comments are welcome. User<Svick>.Talk(); 17:56, 17 February 2013 (UTC)

Is this possible in a template?[edit]

{{Year nav topic}}, {{Year nav topic5}} etc will show redinked years at the lower and upper end of the range. See 2016 in the environment for example. Would there be some way of modifying the templates so the range of years is truncated so that the redlinks are prevented from showing up? I envisage other parameters would be stipulated for upper and lower limits of the range. -- Alan Liefting (talk - contribs) 02:06, 18 February 2013 (UTC)

I suppose it would be possible by using the {{#ifexist:}} parser, but wouldn't it create de-standardisation and unneeded maintenance work? (i.e., as those articles are created, the upper and lower limits will have to be adjusted) This is how {{#ifexist:}} is used: {{#ifexist:Foo|What happens if Foo exists|What happens if it doesn't}}. So for example {{#ifexist:2016 in the environment|[[2016 in the environment]] exists!|It doesn't.}} gives: 2016 in the environment exists! Intelligentsium 02:43, 18 February 2013 (UTC)

Watchlist lag but no notice[edit]


My watchlist hasn't updated since 04:05 and I know for a fact that there have been edits since then. There is no notice mentioning a lag. I've refreshed and purged my watchlist. What's up? Ryan Vesey 04:29, 18 February 2013 (UTC)

Came here for the same thing. All contributions-logs lost since around 4:00. Choyoołʼįįhí:Seb az86556 > haneʼ 04:33, 18 February 2013 (UTC)

  • Mine has been borked since 4:02. No lag notice, but no updates since then. --Jayron32 04:34, 18 February 2013 (UTC)
  • So is this a bugzilla issue? Ryan Vesey 04:35, 18 February 2013 (UTC)
Looks like a general watchlisting problem. Watchlists to an hour or so back seem to stop at 04:05 UTC. Attempts to open the general Special:Watchlist results in 504 gateway timeouts. WM status page now seems to show "performance issues" on a couple areas. Dl2000 (talk) 04:41, 18 February 2013 (UTC)
Special:RecentChanges seems to be working fine, though. Dl2000 (talk) 04:50, 18 February 2013 (UTC)
I've just tried meta, commons, and wikidata, and all seem to work, so it seems it's just us. --Rschen7754 04:44, 18 February 2013 (UTC)

They are aware of the problem, I think they are working on it now. Secret account 04:58, 18 February 2013 (UTC)

Same with me, I just edited some pages on my watchlist; the most recent changes showing are from a few hours ago. Canuck89 (what's up?) 05:13, February 18, 2013 (UTC)
Seems fixed. --Makecat 05:17, 18 February 2013 (UTC)
Yeah, seems fixed. Binksternet (talk) 05:19, 18 February 2013 (UTC)
Looks fixed. I was having the same issue. Steel1943 (talk) 05:26, 18 February 2013 (UTC)

Search function has gone haywire[edit]

I'm having major problems with the search function. Whether I'm searching for an existing article/template/page or for one that might not exist, rather than going to the article or the search page, I get sent to a blank search screen. It only happens some of the times. Ryan Vesey 01:01, 13 February 2013 (UTC)

Does this mean that nothing at all is shown in the browser, or "just" no search results? --AKlapper (WMF) (talk) 09:02, 13 February 2013 (UTC)
  • I've had the issue before too, although not recently. No search results at all, but the interface shows up. — Crisco 1492 (talk) 13:06, 13 February 2013 (UTC)
  • What Crisco said. The big problem is the search issue probably appears every 7-10 searches or so. Then I can't use the search bar for about the same number of searches. Then it will work again. Sometimes it will open some pages but not others. It's really glitchy right now. Ryan Vesey 00:11, 14 February 2013 (UTC)
This is tracked in bugzilla:42423, summary of current status in comment 49 (as every comment in Bugzilla creates bugmail I kindly ask everybody to please avoid "I have this problem too!" comments in the bug report). A script to reproduce the problem is attached in bugzilla:43544 which also describes the first step (better logging) to track this problem down in the next weeks, once Ram (new search developer) once he has settled in a bit. --AKlapper (WMF) (talk) 16:44, 18 February 2013 (UTC)

Wikidata phase 1 is live now[edit]

Heya :)

Third time's a charm, right? Wikidata is live on this Wikipedia with phase 1 now. Details are in this blog post. I am happy to answer questions here. An FAQ is here. Please also let me know about any issues. --Lydia Pintscher (WMDE) (talk) 21:01, 13 February 2013 (UTC)

Congrats! --AKlapper (WMF) (talk) 22:58, 13 February 2013 (UTC)
Seems to work great, except that the "Show Wikidata edits by default in recent changes and watchlist" option in Preferences isn't working (and yes, I disabled enhanced recent changes to test it and bypassed my cache). Even with that box checked, I have to manually click "Show" on "Show Wikidata" on my watchlist to get Wikidata edits. jcgoble3 (talk) 23:14, 13 February 2013 (UTC)
Thanks. We'll look into it. --Lydia Pintscher (WMDE) (talk) 23:17, 13 February 2013 (UTC)
I can reproduce it and have filed bugzilla:44990 for this. --Lydia Pintscher (WMDE) (talk) 10:37, 14 February 2013 (UTC)

When clicking Edit links e.g. in Template:Location map Germany I don't get linked to d:Q12205 but to d:Special:ItemByTitle/enwiki/Template:Location_map_Germany. In he:WP and hu:WP it works. NNW (talk) 10:14, 14 February 2013 (UTC)

There was a bug report filed within the last few hours. --Rschen7754 10:17, 14 February 2013 (UTC)
It's bugzilla:44985. --Lydia Pintscher (WMDE) (talk) 10:27, 14 February 2013 (UTC)

Hmm... I seem to have the opposite problem from jcgoble3 above. I have "Show Wikidata edits by default in recent changes and watchlist (does not work yet with enhanced changes)" turned off in my preferences, yet I still see Wikidata edits in my watchlist. I'm using Monobook for my skin. I am using an expanded watchlist, but not an enhanced watchlist. I've instead resorted to this one line of CSS to hide the Wikidata edits. – PartTimeGnome (talk | contribs) 23:37, 14 February 2013 (UTC)

Ok we will look into that too. --Lydia Pintscher (WMDE) (talk) 11:03, 15 February 2013 (UTC)
  • How does this affect Commons? – specifically, iwl links from Commons to the Wikipedias? Even more specifically, what about Commons categories that link to article pages? Andy Dingley (talk) 23:40, 14 February 2013 (UTC)
    • They will work as they used to for now. In the future they might also be moved to Wikidata but that's not possible at the moment. --Lydia Pintscher (WMDE) (talk) 11:03, 15 February 2013 (UTC)
  • Are Wikidata-derived iwl links commutative? Specifically if fr:WP has a topic on fr:BigTopic, but not on fr:SubTopic, we may well have articles at en:WP on both en:BigTopic and en:SubTopic. Practice has often been that en:Subtopic would include an iwl to fr:BigTopic (as the best available, and possibly quite a good overview), but we need to ensure that fr:BigTopic links only to en:BigTopic. Andy Dingley (talk) 23:43, 14 February 2013 (UTC)
    • Wikidata only handles 1-to-1 links which is the large majority. The rest will have to use links in the wikitext. --Lydia Pintscher (WMDE) (talk) 11:03, 15 February 2013 (UTC)

I started a description page at WP:Wikidata. Editors are welcome to expand it. --Izno (talk) 16:22, 15 February 2013 (UTC)

Thank you! --Lydia Pintscher (WMDE) (talk) 16:27, 15 February 2013 (UTC)
  • There is something seriously wrong with editing the interlanguage links on Wikidata for pages outside mainspace. I've started a thread at d:Wikidata:Project chat#Namespace ignored. --Redrose64 (talk) 12:23, 18 February 2013 (UTC)
    Known bug, as listed at that discussion topic. --Izno (talk) 14:22, 18 February 2013 (UTC)

Import from[edit]

With the Scribunto happening soon, I've been asked if there is a way to import some modules from to after that rollout has happened. I've looked at Special:Import and test2 is not currently listed as a source. Just wondering if it's possible to get it added? -- WOSlinker (talk) 08:30, 17 February 2013 (UTC)

After getting consensus for this, you need to make a request on bugzilla (such as this one) to change the $wgImportSources for enwiki. Helder 15:28, 17 February 2013 (UTC)
I am supporting this import. I will notify WMF staff about it.--Kozuch (talk) 08:17, 18 February 2013 (UTC)
Done. -- Tim Starling (talk) 00:41, 19 February 2013 (UTC)

Preview blanks a section[edit]

My work PC has just been replaced with a new one running Windows 7 and Internet Explorer 9. I now have an intermittent problem when editing, in that the whole section may be blanked when I Preview before saving an edit. I recently found exactly the same intermittent problem using a colleague's PC with the same spec. Sometimes I can recover some or all of the draft content by using the browser's Back button and reloading. I'm having to get into the habit of Ctrl-A Ctrl-C before I preview.

This sounds the same as Wikipedia:Village pump (technical)/Archive 96#Preview blanks a section, but that was a year ago under Chromium, and no diagnosis or resolution was noted. – Fayenatic London 15:03, 18 February 2013 (UTC)


Hi. I'm helping some folks on an off-wikipedia project called Cite Club designed to address backlogged citation needed tags. Simply, they present an article and a sentence that needs citing and guide users through the process of finding and referencing a reliable source. This all happens through the mediawiki api. They are running into some trouble with the CAPTCHA feature and want to know if it's possible to test their code against it. I assume the details of precisely when Captcha's are triggered is kept secret to deter spammers. Is there any way they can bugtest their code either on English Wikipedia or on test.wikipedia against Captcha's? (does test wp even have captcha's?). Thanks for any help. Cheers, Ocaasi t | c 18:15, 18 February 2013 (UTC)

If a remember correctly, the only way to completely avoid the captcha is for the user account to be autoconfirmed (or confirmed), which means that the API has to be run through an autoconfirmed account. This requires making about two API requests to login in, and optionally one more request to logout. (mw:API:Login and mw:API:Logout) The Anonymouse (talk | contribs) 18:23, 18 February 2013 (UTC)

Pending changes question[edit]

An pending edit was Huggle reverted by an admin back to its last accepted version, but both changes were showing on Special:PendingChanges as needing review. Seems odd. It also isn't clear to me how I was supposed to "accept" the change because the accept option was grayed out -- ended up "rejecting" and then accepting the edits. NE Ent 21:05, 18 February 2013 (UTC)

Parser function with special tags[edit]

Hello :) I'm trying to make this code working:

{{#if:{{{a|}}}{{{b|}}}{{{c|}}}|<abbr title="{{#if:{{{a|}}}|{{{a}}}{{#tag:nowiki|″}} {{#if:{{{b|}}}|{{{b}}}}}{{#if:{{{c|}}}|, {{{c}}}}}">{{{d}}}{{#if:{{{a|}}}{{{b|}}}{{{c|}}}|</abbr>}}}}

and I've got errors. I guess there's something wrong while parsing < & > and " symbols? What is the right way to write this code? The symbol in {{#tag:nowiki|″}} is an inch symbol. (talk) 22:23, 18 February 2013 (UTC)

Count your braces. I see 51 '{', but only 49 '}'. I think you forgot to close the #if that starts just after title=". – PartTimeGnome (talk | contribs) 23:18, 18 February 2013 (UTC)
Yes, my mistake. But there were still errors. I looked at the page's code:
<abbr title="">d</abbr>
Now I'm using Comment template on, it works better with parser! Thanks a lot for yout help!! (talk) 01:45, 19 February 2013 (UTC)


With the deployment of Wikidata we now have the opportunity to edit this data directly from the language links area. The problem is that there appears no way when viewing the article of knowing if the links displayed here are from Wikidata or are being overridden by an entry in the wikicode of the article. Is there some way to know which is serving each of the links, Wikedata or article code? Keith D (talk) 01:50, 19 February 2013 (UTC)

Please check Wikipedia:Wikidata. --Rschen7754 03:59, 19 February 2013 (UTC)

Searching block by "id" information[edit]

Sorry for asking for this here because I had no answer in the appropriate place. I'm sysop of pt-Wikipedia and recently included in the mailing list of unblock requests. I use to see requests with some "id" information of the blocking (perhaps blocked ips are directed to that list with that specific data). How can I search block by "id"? E. Feld talk 04:13, 19 February 2013 (UTC)

For example, for block 123, enter #123 on Special:BlockList. Prodego talk 04:15, 19 February 2013 (UTC)

::But this page has no field for id, only for user so I can't list all blocks and search it with the browser. Have I to search it mannually? E. Feld talk 04:21, 19 February 2013 (UTC)

Sorry, now I got it, I have to enter this as it were the username. Thank you very much. E. Feld talk 04:27, 19 February 2013 (UTC)

wikidata query[edit]

Having remove the iw links from an article and adding the missing entry to wikidata the new entry does not show on the page, even after refreshing. Is there some delay in the propagation for new entries? Keith D (talk) 13:09, 15 February 2013 (UTC)

I created an article (Vyritsa) earlier today without iw links, then went to Wikidata and added the article to the entry there. The links were on the article not immediately, but within five minutes. Not sure what happens for completely new entries though.--Ymblanter (talk) 13:13, 15 February 2013 (UTC)
Yes it takes a few minutes for them to show up. --Lydia Pintscher (WMDE) (talk) 13:16, 15 February 2013 (UTC)
If you perform a (null) edit on the article, the new entries will show up immediately. —Naddy (talk) 13:27, 15 February 2013 (UTC)
Thanks for the information. Now I know for future reference. Keith D (talk) 13:55, 15 February 2013 (UTC)
null edit is not even required if you add it to wikidata first and remove the IW from enwiki later. Also, note the "slurp" gadget on wikidata: you can add the enwiki article to wikidata, perform the "slurp" to ensure all the interwiki links that exist in enwiki also exist on wikidata, and verify no conflicts present, and only then remove the IW links from the enwiki articles. peace - קיפודנחש (aka kipod) (talk) 22:39, 15 February 2013 (UTC)
...if you have 'slurp' turned on in your Wikidata preferences and the en:WP page has language links. filceolaire (talk) 19:38, 19 February 2013 (UTC)

Format Change of Page View Stats[edit]

(restoring from archives) I process the raw page view statistics provided by the WMF (to make things like WP:5000). Prior to February 1, this file was always in a 4-column space-delimited format. However, beginning on that date, page titles started to appear which had spaces instead of underscores in them -- although this is not always the case (only a couple thousand in a file with several million lines). My processor started throwing errors and I've since fixed it -- but I could imagine this being a potential problem for those who haven't noticed the change or are not actively maintained (e.g., Henrik's tool). Is there any explanation why this occurred? Thanks, West.andrew.g (talk) 18:05, 3 February 2013 (UTC)

They changed the format of the source log files to be tab-delimited instead of space-delimited, to make processing easier when the field values contain spaces. They intended to not change the format of pagecounts-raw,[13] but apparently something earlier in the process used to convert spaces in the page names to underscores but doesn't anymore. Hopefully that can be fixed quickly. Anomie 20:44, 3 February 2013 (UTC)
I am just noting that spaces continue to appear in some article titles. West.andrew.g (talk) 15:39, 16 February 2013 (UTC)
Are you subscribed to the analytics mailing list? If not, it might be worth subscribing to for warnings of changes like this - see here for the discussion/notification. Andrew Gray (talk) 19:55, 16 February 2013 (UTC)
Thanks for the pointer (I subscribed purely out of personal interests). However, the link Anomie provided above suggests this was a change for internal logging and that spaces should not be appearing the published "pagecounts" files. I'll note that only a couple hundred or thousand (of millions) of lines have these spaces. Thanks, West.andrew.g (talk) 20:25, 19 February 2013 (UTC)

Two suggestions for the improvement of UI of the wikipedia pages[edit]

Hi, two things I want to mention as follows:

First, the content of some pages are so large and when I arrive at the bottom of the page, it is not convenient for I to review the things at the top. Such a botton "Back to the Top" is considered to be placed.

The other thing is that when I read a page and find something interested, I click the link and then the current page will just change into another page, and I can't review the last page I've read and cannot learn something by contrast.Columns for different things may take effort and or just create a new page for a new link.

I am looking forward to your reply. And wish wikipedia a nice future.Happy Chinese New Year.

Sincerely, XuWen — Preceding unsigned comment added by Hotdog3645 (talkcontribs) 02:35, 19 February 2013 (UTC)

Both of these things are already done by web browsers. To get back to the top of a page, press the Home key on your keyboard, or drag the scrollbar to the top with the mouse. However, a "top of page" link may be useful for mobile users, since many mobile phones don't have a quick way to get back to the top. Such a link could show on the mobile site, while not showing on the main site where it isn't needed.
To open a link in a new page, right-click it and select "Open in new tab" or "Open in new window" (or similar). You may also be able to get the same effect by holding down Ctrl and/or Shift as you click the link. Depending on your browser, you may be able to arrange the pages side-by-side after opening them.
Though opening in a new page is very useful (I do it regularly), making it the default would be very annoying, as it goes against the normal expected behaviour when clicking links. In most cases I only navigate to another page when I'm done with the current one (if not, I'll just open a new tab). – PartTimeGnome (talk | contribs) 23:45, 19 February 2013 (UTC)

Who is Billy and why is his name at the bottom of my Watchlist?[edit]

Under the last entry, on the lower left hand corner above the words privacy policy (and on my PC, a blue border and a grey line) I see the the name Billy in bold letters in line with and similar to the other dividers (what's the word for this - you know, Main, Wikipedia, Wikipedia Talk). No entries for Billy though. Dougweller (talk) 11:52, 19 February 2013 (UTC)

I don't see it ... maybe we've got some advertising for a bookshelf by a Swedish company? (✉→BWilkins←✎) 12:06, 19 February 2013 (UTC)
Lol, I've got 2 or 3 of them. And Billy's still there on my Watchlist. Dougweller (talk) 12:50, 19 February 2013 (UTC)
Just refreshed and he's gone. Odd though, he was on there for at least on refresh. Dougweller (talk) 12:53, 19 February 2013 (UTC)

Math symbols[edit]

I recently bought an HP laptop running Windows 8, using browser Internet Explorer. When I open Wickipedia's articles on mathematics, the math symbols displayed are only readable if you know what to be looking for. They are very fuzzy and seem made up of dots. Is there something I can change in my machine or in Wickipedia to have its math symbols cleanly displayed? Any advice appreciated. Henry Young — Preceding unsigned comment added by (talk) 20:26, 19 February 2013 (UTC)

Replied on IP talk page. ▫ JohnnyMrNinja 20:41, 19 February 2013 (UTC)

Split histories[edit]

Is it still impossible to split article histories? For example, can an article be deleted, the most recent half of the revisions be restored and moved to a new article name, and then the other revisions restored at the original title? I ask because I know it was impossible, but I could've sworn I saw someone do it recently in a cut-and-paste fix. ▫ JohnnyMrNinja 19:06, 19 February 2013 (UTC)

it should be possible, but a real pain to to successfully. Werieth (talk) 19:12, 19 February 2013 (UTC)
This sounds very much like WP:HISTSPLIT. If you've never done one before, I suggest you liase with Anthony Appleyard (talk · contribs), who probably has the most experience of such matters. --Redrose64 (talk) 19:35, 19 February 2013 (UTC)
Perfect, thank you. I'm a non-admin, so WP:REPAIR was my next stop (so, basically, Anthony Appleyard), once I was certain this was technically possible. It looks like it's been possible since at least 2010, maybe it always was and I was mistaken. Oh well, thanks! ▫ JohnnyMrNinja 19:47, 19 February 2013 (UTC)
So, apparently TParis watches that board and performs histmerges as well. Just writing this as an acknowledgment that my foot is in my mouth keyboard. ▫ JohnnyMrNinja 21:06, 19 February 2013 (UTC)
It's been possible to do history splits ever since [[selective deletion was introduced in December 2004. Here is a rather extreme example that I performed in 2008. Graham87 06:43, 20 February 2013 (UTC)

File:Tamil Virtual University.jpg[edit]

Tamil Virtual Academy Logo File:Tamil Virtual University.jpg. Now the Tamil virtual university is shifted as Tamil virtual acadamy. This is the updated logo of Tamil virtual acadamy as on 20 feb 2013. But the article page still showing the old image. I refresh thumbnails, to purge both the image page and the article page but still problem is there. --Srither (talk) 09:34, 20 February 2013 (UTC)

Problem Resolved. --Srither (talk) 13:06, 20 February 2013 (UTC)

Scribunto, Lua, Module: namespace and /doc subpages[edit]

As we are starting deploy Scribunto on enwp, I think it would be helpful if the scripts in the Module: namespace had /doc subpages similar to the the ones used in the Template: namespace. This would help editors with the proper usage of scripts, etc., so it would be less likely they would mess up something. All comments are welcome. (talk) 14:07, 20 February 2013 (UTC)

unlike in Template: namespace, the pages in Module: namespace behave differently, and do not support internal links embedded in the code (i.e., one can't place {{\Documentation}} in the page and expect the viewers to see the documentation). instinctively, i would expect the documentation to reside in the corresponding "Talk" page of the module - exactly one mouse-click away. there might be a better way, but i can't think of one. peace - קיפודנחש (aka kipod) (talk) 14:35, 20 February 2013 (UTC)
Thanks, kipod. I didn't know about that aspect. The talk page sounds like a good alternative to me. (talk) 14:47, 20 February 2013 (UTC)
See also mw:Extension talk:Scribunto/Documentation specification and bugzilla:39609#c3. Helder 15:53, 20 February 2013 (UTC)
I'm already working on this. Anomie 01:39, 21 February 2013 (UTC)

Wikidata scheduled downtime[edit]

Wikidata will be read-only from 19:00 UTC today through 02:00 UTC tomorrow (February 21) in order to upgrade the database schema. During that time the site will not be editable, and it will not be possible to add or remove language links. --Rschen7754 18:11, 20 February 2013 (UTC)

Are there details posted somewhere about those changes? Werieth (talk) 18:30, 20 February 2013 (UTC)
No, not exactly - it's probably to support additional features still in development. --Rschen7754 18:35, 20 February 2013 (UTC)
Yes it's changes to the database that shouldn't make a difference for editors. It's mainly to prepare for the future to improve things like search. --Lydia Pintscher (WMDE) (talk) 19:32, 20 February 2013 (UTC)
It's 21:06 UTC now. Why am I able to edit? Was this postponed or something? jcgoble3 (talk) 21:06, 20 February 2013 (UTC)

The read-only time will be starting soon and ending a few hours late due to some delays in San Francisco. --Rschen7754 21:07, 20 February 2013 (UTC)

User-script to hide Atom tool?[edit]


Is there some user-JS or -CSS you can add to hide ToolboxAtom feed for this page? It Is Me Here t / c 18:48, 20 February 2013 (UTC)

I think that the following would be best:
li#feedlinks { display: none; }
although it should also be possible by applying the same styling to either a#feed-atom or a.feedlink It would go in your Special:MyPage/common.css. --Redrose64 (talk) 20:17, 20 February 2013 (UTC)
Face-smile.svg Thank you very much! It Is Me Here t / c 23:27, 20 February 2013 (UTC)

How to edit—or request an edit—of Special:SpecialPages[edit]

I don't see a view source, and there's no button for a talk page. There is no link for Special:PermanentLink on Special:SpecialPages. Help:Special page says

(I boldfaced all for emphasis). Now Special:Permalink doesn't document how to use it (shouldn't it), e.g., the first page is Special:Permalink/1. Is there a way to permalink to a diff, i.e., can I permalink to this page? Thanks, Wbm1058 (talk) 19:50, 20 February 2013 (UTC)

Instead of &oldid=1, use &diff=539289786. Legoktm (talk) 20:22, 20 February 2013 (UTC)
(edit conflict) Essentially what you've given is the permalink, in that it's got a valid oldid in the query string. You can build that as a template if you prefer, i.e. {{diff|Wikipedia:Wikipedians|next|1|this page}} which gives this page
Legoktm - that won't work, you can't give diff twice in the same query string. --Redrose64 (talk) 20:24, 20 February 2013 (UTC)
Thanks, the template is nice, but, it's a template, and those don't work on edit summaries. I should have mentioned that there is a help:permanent link page, maybe Special:permanentlink should redirect there or transclude the help. WP:Permalink redirects there. Help:Diff doesn't mention a permalink syntax. Wbm1058 (talk) 20:50, 20 February 2013 (UTC)
You can say Special:PermaLink/484941430 to get a link to [14], for example. As for linking to a diff, try Special:ComparePages. πr2 (tc) 21:01, 20 February 2013 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Thanks, but what I'm looking for is something can create a link in edit summaries. I know I can drop a url there that can be cut&pasted, but it would be nice to make it clickable instead. We also have {{oldid}}, the template equivalent of a permalink:

two ways to do the same thing, but only the second hyperlinks in an edit summary. Interestingly enough, only the first works with WP:NAVPOPS. I also note that {{oldid}} requires specification of the page name in the first parameter, while the permalink demonstrates that the page name isn't really necessary. – Wbm1058 (talk) 22:01, 20 February 2013 (UTC)

Check out the edit summary of this edit for an example of permanent link usage. Wbm1058 (talk) 22:33, 20 February 2013 (UTC)

for convenience's sake, here is Wbm1058's aforementioned edit so you can see the edit summary EVula // talk // // 22:39, 20 February 2013 (UTC)
The link in the edit summary is a link to a particular version, which (as shown above) we can already obtain. Wbm1058 is asking for a link to a diff - one which shows the actual change above the page text. --Redrose64 (talk) 23:17, 20 February 2013 (UTC)
Asked on πr2 (tc) 03:04, 21 February 2013 (UTC)

Native/Official name parameters in Infobox settlement[edit]

Per the Manual of style for Sinosphere-related articles, bolding or italicising of Chinese characters is discouraged, yet Infobox settlement applies bold anyway. If no one objects within 36 hours, I will go on an AWB run to apply {{nobold}} in all relevant articles. See Macau present and old Macau for an example. GotR Talk 04:39, 21 February 2013 (UTC)

Lua cites run 85/sec depth=3[edit]

(tangent to thread: #Scribunto deploy).

I have tested the draft version of Template:Cite_web/lua, with Module:Citation. Even without optimizing of the Lua script, it formats cites at ~85/second (with COinS metadata), or about 6x faster than old {cite_web} which ran 14/second with COinS. Again, that is non-optimized Lua, using whatever expensive features of Lua table/string processing, and 38% of processing time was inside the Lua modules. The expansion depth for Lua cites was just 3 of the 41 possible nesting levels. Maximum capacity was 1,730 cites to hit post-expand include-size limit of 2,048,000 bytes, in 25 seconds.

For any future optimation, the limit would be Lua as 9.6x faster than markup-based cites, as 100%/62% or 1.61x faster if Lua used 0% of processing time. Optimized markup cites can have equivalent speed, as rapid Template:Cite_quick runs as fast as the Lua cites, by omitting rare parameters, and allows 2,330 cites rather than only 1,730. However, prior tests of string-scan searches have run over 180,000x faster in Lua because markup-based strings have used zillions of padleft+#ifeq parser functions. Since the cite templates are an example of typical "data processing", then expect Lua to process numerous data parameters over 6x faster than markup, although string-scans would be thousands faster.

Preprocessor stats for 500 Lua cites: The following is copied from a typical run of 500 instances of Lua-based Template:Cite_web/lua, with cite parameters shown further below, after the NewPP report.

For 500 of Empty citation (help) :
  NewPP limit report
  Preprocessor visited node count: 14501/1000000
  Preprocessor generated node count: 60037/1500000
  Post-expand include size: 591000/2048000 bytes
  Template argument size: 0/2048000 bytes
  Highest expansion depth: 3/40
  Expensive parser function count: 0/500
  Lua time usage: 2.360s
  Lua memory usage: 992 KB
 (Total time: 6.194 seconds)

The 9 cite parameters used in the various tests were:

  • {{cite web/lua | title=My Page | last1=Doe | first=John | publisher=Acme | location=London
    | url= |date=5 May 2009| page=6 |accessdate=8 February 2013}}

That use of {cite_web/lua} was repeated 500 times (or 2,500), for several timing tests.

Again, the Lua-based cite templates seem to run as fast as the optimized, markup-based Template:Cite_quick, but Lua allows hundreds of parameter names to be used faster, rather than just the major parameters in {cite_quick}. To estimate edit-preview speed of an article, then divide cite-format time by 6: as 18 seconds of {cite...} would become 3 seconds to preview with 250 Lua-cite templates. More later. -Wikid77 (talk) 13:09, 19 February 2013 (UTC)

From my own testing with our modules, Lua is limited to 10s runtime for all scripts on the page. -- WOSlinker (talk) 13:57, 19 February 2013 (UTC)
This is great to see. Thanks for laying this out. Superm401 - Talk 07:39, 21 February 2013 (UTC)
  • Lua cite speed confirmed after tests/updates: After days of modifying the Lua-based cite templates, as tested to closely match the current markup-based cite templates, the speed is still ~85 Lua-cites per second (or better), despite adding more if-statements to test other cite parameter values (and ignore dozens of blank parameters). Hence, the speed, as 85/second, is not just for experimental small citations, but rather the speed for typical citations which allow numerous other optional parameters. Also, other tests have confirmed that using fewer parameters will run even faster, so short cites (without author or publisher) could run faster than 110/second, just as markup-based cites can run faster when fewer parameters are used, among the various optional parameters. -Wikid77 (talk) 17:37, 21 February 2013 (UTC)

VisualEditor fortnightly update - 2013-02-18 (MW 1.21wmf10)[edit]

Hey all,

Here is a copy of the regular (every fortnight) update for the VisualEditor project, so that you all know what is happening (and make sure you have as much opportunity to tell us when we're wrong, as well as help guide the priorities for development and improvement).

The VisualEditor was updated as part of the wider MediaWiki 1.21wmf10 branch deployment on Monday 18 February. During this time, the team has expanded with the recruitment of Ed Sanders who will focus on the data infrastructure of the VisualEditor platform.

In the two weeks since 1.21wmf9, the team have worked on the infrastructure needed to support the new features being added for wider launch of the VisualEditor as the default way users will edit wikis. This infrastructure work has included further expansion in the capabilities of the 'document model' module that converts between the HTML+RDFa rendered by mw:Parsoid and an editable document, such as adding the ability in the system to edit "meta-data" like categories (45029); note that the user-facing editing component for this is yet to come. Other work has included supporting Microsoft Internet Explorer, adding new user interface widgets for a consistent UI, and reviewing and expanding the design of the user experience for inspectors and editors to cover template and reference editing.

Alongside this work, the team have fixed a number of bugs. These include integration improvements so that the toolbar doesn't appear multiple times if a user clicks back and forth (44838), errors thrown when users click at the end of a document (41199) or when using the link inspector in the stand-alone VE environment (44686), and making sure that inserted links get the link applied correctly (44086).

A complete list of individual code commits is available in the 1.21/wmf10 changelog, and all Bugzilla bugs closed in this period on Bugzilla's list.

Per the MediaWiki deployment roadmap, this should be deployed here on Monday, February 25.

Hope this is helpful! As always, feedback gratefully received, either here or on the enwiki-specific feedback page.

Jdforrester (WMF) (talk) 00:23, 20 February 2013 (UTC)

Please remember to edit the site notice before deployment so that those who do not want to use the VisualEditor can edit their preferences. – Allen4names (IPv6 contributions) 06:52, 20 February 2013 (UTC)
I'm not sure what you're asking; the code was originally deployed to the English Wikipedia in December. This is just an update to that deployment. Additionally, it is currently running as an opt-in tool - you have to set your preferences to ask for it for the VisualEditor to appear. Jdforrester (WMF) (talk) 01:41, 22 February 2013 (UTC)

Editing with a Samsung Galaxy Note 10.1 running Android Ice Cream Sandwich OS[edit]

Hi. It takes forever to open watchlist, history and diffs. Is there anything I can do in my settings or ..preferences to speed it up? --Anthonyhcole (talk) 07:04, 20 February 2013 (UTC)

yes. you can change you preferences (under Preferences => Recent changes) to decrease the number of changes you want to see. on galaxy tab 2 (mine is 7", not your 10"), reducing it from the default 200 to 50 improved things significantly. peace - קיפודנחש (aka kipod) (talk) 14:39, 20 February 2013 (UTC)
Awesome. Thanks. Also, the "find on this page" function doesn't work in the edit box. Any ideas? --Anthonyhcole (talk · contribs · email) 13:02, 21 February 2013 (UTC)

Commons mobile uploads testing week[edit]

Hammer - Noun project 1306.svg

Uploading pictures from mobile devices to Wikimedia Commons must be simple for everybody! The Wikimedia Mobile engineering team has fresh software and you can help testing it.



PS: this activity is part of the MediaWiki QA weekly goals. We want to do better at inviting and engaging the Wikipedia community, including technical and non-technical users. We are also open to help co-organizing testing activities for community software projects such as gadgets, bots and more. Feedback and suggestions welcome!--Qgil (talk) 07:14, 21 February 2013 (UTC)

Preventing additions of categories in certain area[edit]

Origionally raised at Wikipedia talk:WikiProject Articles for creation/2013 1#Categories - again Hi all! It has been raised multiple times to keep categories off Articles for Creation submissions. Is there any easy way to do this - The bot that normally does this is unreliable. Would a solution be an edit filter, or does the software contain a function that means we can restrict categories in the AFC area. Thanks, Mdann52 (talk) 11:08, 21 February 2013 (UTC)

The is nothing in the MediaWiki software (except edit filter) that can restrict categories in the AFC area. Ruslik_Zero 12:20, 21 February 2013 (UTC)

What do you mean by unreliable? If it sometimes works well and sometimes not then the code needs to be rewritten. If it does not run on a regular base and should be run more often, it needs a new bot operator. Once there is a bot for a task, the easy way is to use it. (IMO.) Bináris (talk) 13:10, 21 February 2013 (UTC)

Tagging data for future database use[edit]

I'm building a table of award recipients for an article. The Name column is actually a concatenation of any of title, first_name, middle_name, last_name, and suffix elements, which reads better than creating separate columns for them. I'd like to tag those individual components so if/when they are moved into a database, they can be separated back into those components. Something like:

{| class="wikitable"
! Recipient
! Department
| <name_title>Battalion Chief</name_title> <name_first>Tom</name_first> <name_mid>Q</name_mid>. <name_last>Smith</name_last>, <name_suff>III</name_suff>
| Glendale Fire Department, CA

Has this been done elsewhere on wiki? Is there a "preferred" way to do it? —[AlanM1(talk)]— 17:47, 21 February 2013 (UTC)

We're not using XHTML any more, so user-created tags like <name_title>...</name_title> are not a good idea. But what you describe is essentially metadata, which we usually mark by using particular classes. See Wikipedia:Metadata. --Redrose64 (talk) 18:13, 21 February 2013 (UTC)
actually, afaik, we *are* using html5, and the way to express this idea with html5 is to use private *attribute* rather than a private *tag*. html5 supports private attributes whose name begins with "data-", so maybe what you described will look like so:
{| class="wikitable"
! Recipient
! Department
| <span data-title>Battalion Chief</span> <span data-first-name>Tom</span> <span data-middle-name>Q</span>. <span data-last-name>Smith</span>, <span data-name-suffix>III</span>
| Glendale Fire Department, CA

or, better yet:
{| class="wikitable"
! Recipient
! Department
| <span data-title="Battalion Chief" data-first-name="Tom" data-middle-name="Q" data-last-name="Smith" data-name-suffix = "III">Battalion Chief Tom Q Smith III</span>
| Glendale Fire Department, CA

peace - קיפודנחש (aka kipod) (talk) 18:36, 21 February 2013 (UTC)

Templates with too many parser functions for lualizing[edit]

I wrote a script to search templates with many parser functions that are worth to convert to Lua now: mw:Special:Code/pywikipedia/11099. Bináris (talk) 06:26, 21 February 2013 (UTC)

You should consider creating a wiktionary entry for this new word: "lualizing". Ruslik_Zero 12:25, 21 February 2013 (UTC)
I already invented a couple of Hungarian words but this is my first in English. Feel free to spread it. :-) Bináris (talk) 13:05, 21 February 2013 (UTC)
  • Start with known slow templates: After years of analysis, several templates have been confirmed as causing the largest slow-down of edit-preview, such as the wp:CS1 cite templates, Template:Sfn, or Template:Weather_box, or similar:
  • Rewrites are in progress for wp:CS1 templates {cite_web}, {cite_news}, {cite_journal}, etc.
  • A markup-based rewrite of Template:Sfn clocks as 2x faster, but Lua could be 6x? faster (3x faster than rewrite).
  • The climate table Template:Weather_box needs to be optimized to skip whole sections when parameters are not used.
For each template, it has been good to consider, first, a rewrite of the basic functionality in markup, such as {cite_quick} which proved to run 6x faster when supporting fewer parameters. Then, after optimizing the markup-based logic, compare operation using an equivalent Lua module. -Wikid77 (talk) 16:49, 22 February 2013 (UTC)

ISBN linking[edit]

Per the discussion at Talk:International_Standard_Book_Number#Change_the_ISBN_link can the link to the acronym "ISBN" in, say, ISBN 0-14-020652-3 be removed somehow? If so, do we need a wider discussion? -- Alan Liefting (talk - contribs) 01:32, 22 February 2013 (UTC)

It was thought to be part of the MAGICWORDs but it appears to be part of the {cite} templates. -- Alan Liefting (talk - contribs) 06:25, 22 February 2013 (UTC)

Yields this link: 0-14-020652-3. ―cobaltcigs 10:22, 22 February 2013 (UTC)

Replied on that talk. --— Gadget850 (Ed) talk 14:07, 22 February 2013 (UTC)

Proposed transition plan for CS1 Lua cites[edit]

Several editors have submitted improvements for the Lua-based cite templates (of Citation Style 1, CS1), and the results are getting very close to the format of the markup-based cite templates. Lua-cite anchor links are working now, as needed to cross-link author/year links in article "India". These Lua-based cites have the potential to make articles format, or edit-preview, 2x (twice) as fast, where a 20-second edit-preview would reduce to below 10-second preview, even with non-optimized Lua script. Consequently, we are developing a transition plan to begin roll-out of Lua-based cite templates:

The strategy is to begin gaining the speed and format-checks of some Lua-based cites, even if not "perfect" yet, and also run wider tests to find more format issues to address. In the event of trouble, each transition step can be reverted to allow time to resolve related problems. Beyond the much faster speed, the Lua-based cites will provide other improvements:

  • Some parameters, such as "location=xx" will shift to more logical order.
  • A single page will show "p." not "pp." when "pages=xx" has one page.
  • An empty "separator=" will default as dot "." or, for {citation}, as comma.

Hence, even though some format alignment might need to be adjusted, the Lua-based cites will improve other format issues, while also allowing pages to edit-preview 2x faster. As techniques for optimizing Lua are analyzed, then articles might edit-preview 3x faster than with markup-based CS1 cite templates. -Wikid77 (talk) 18:46, 22 February 2013 (UTC)

Sudden change in rollback?[edit]

Basement cats on rollback page

Went to rollback via Twinkle a test-ish edit on Turtle (submersible) just now, and got this. I don't recall seeing categories on the page that popped up when rollbacking before. What's up? - The Bushranger One ping only 21:25, 22 February 2013 (UTC)

This was a change in Twinkle. I had to change to "container" used by the rollback status because it wasn't working at all in the Classic skin. I hope it doesn't bother you too much... if you like I could get Twinkle to hide the category box when rolling back, as well. — This, that and the other (talk) 05:09, 23 February 2013 (UTC)
Well, there's really no need, if it's a programming bother. Face-smile.svg I was just startled by the sudden change, primarily. - The Bushranger One ping only 16:30, 23 February 2013 (UTC)

Reference tag checking[edit]

I'm finding that AnomieBot is stomping on references created by {{singlechart}} because it doesn't recognize that it is creating a named ref. Is there anyway for me to check inside of {{singlechart}} if the reference name I have been asked to create has already been created in the article so that I can at least flag the problem?—Kww(talk) 02:50, 23 February 2013 (UTC)

I see that {{singlechart}} always generates a named ref, even when |refname= is unspecified. Does it do this for the same reason as {{sfn}}, i.e. to consolidate duplicate refs automatically? If not, is the name actually needed? I often see named refs where the name isn't used anywhere; some people have a mistaken belief that all refs must be named.
One thing that could be tried is a script written along the lines of User:Ucucha/HarvErrors. --Redrose64 (talk) 08:31, 23 February 2013 (UTC)
My recommended practice is to use the automatically generated name if the author wants to use the reference generated by {{singlechart}} in the article. If the author wants to say "It reached number 3 on the UK Singles chart" in the text of an article about the Sugababes, he can reference that to the reference named sc_uk_sugababes without doing anything. The refname is an override. Unfortunately, wikis wind up with a lot of cutting and pasting and very little best practice, so once one editor got in the habit of overriding so that he didn't have to remember the pattern, it spread.
Mutley1989 wrote me a detection script so that I can fix things after the fact. What I'm hoping for is something I can insert in the template so that I can generate a big red error message when it's done in the first place.—Kww(talk) 15:19, 23 February 2013 (UTC)

Removing Wikidata interwiki links and their reversal[edit]

There has been some great progress in removing interwiki links because of the new format used for Wikidata. However many many of these edits are being reverted by people not aware of the situation - is there a link to the topic (a talk) that can be added to the edit summaries to guide our editors to a page that explains whats going on? We cant have editwars over this - all we need is an explanation for all to see.Moxy (talk) 23:42, 14 February 2013 (UTC)

There is some information on WP:Wikidata interwiki RFC, however I agree that there should be a clear explanation posted somewhere to prevent arguments -- Marek.69 talk 00:04, 15 February 2013 (UTC)
The above proposal failed, so there's nothing preventing people from removing the links. However please check to make sure all the links have made it over to Wikidata... most have but not all of them. --Rschen7754 00:30, 15 February 2013 (UTC)
I think most interlanguage links added after December didn't make it to Wikidata. It's a mess. I think it's too early to remove them. At least, not manually. I saw people in the Hungarian Wikipedia manually remove the links without even checking if every single one was moved to Wikidata (and it wasn't.) --Moscow Connection (talk) 00:43, 15 February 2013 (UTC)
Indeed, I did around 25 articles today (with two exceptions, one's I'd authored myself), and I believe 2-3 of them needed work, the rest were just fine. Fine if you're paying attention, but not ready for automated or inattentive handling. --j⚛e deckertalk 00:57, 15 February 2013 (UTC)
Try [15] - that's supposed to help out. --Rschen7754 01:11, 15 February 2013 (UTC)
Actually I'm not sure how that script works... --Rschen7754 01:17, 15 February 2013 (UTC)
Figured it out - you add mw.loader.load('//'); to your .js file, edit a page, click the new button (on vector, it's in the drop-down menu next to the edit button), and hit save. It only removes links that are actually on Wikidata. --Rschen7754 01:21, 15 February 2013 (UTC)
I think bots are not ready yet. Maybe they can't work with Wikidata or don't have permission. For example, I created an article in the Japanese Wikipedia today, and ZéroBot linked it in all Wikipedias except the English and the Italian ones, and it didn't update Wikidata. (There's no Hungarian article, but it wouldn't add the link to the Hungarian Wikipedia too.) If you need to see an example, I'm talking about this article: ja:ワン・ポンド・フィッシュ, but don't look, it's a stub. :) There appear to be quite a lot of articles that aren't linked in the 3 above-mentioned Wikipedias because of this innovation. -Moscow Connection (talk) 01:28, 15 February 2013 (UTC)
Wikidata will be launched on all Wikipedia sites around February 27, so this will become moot. --Rschen7754 01:31, 15 February 2013 (UTC)

I started a description page at WP:Wikidata. Editors are welcome to expand it. --Izno (talk) 16:21, 15 February 2013 (UTC)

I think people should be stopped from deleting interlanguage links manually. Really, if I were an admin, I would stop it immediately and revert. I see people removing interlanguage links, obviously without checking anything (cause they have 0 edits in Wikidata). And you surely know that there are many interwiki conflicts, so there are bound to be many links that are present only in the English Wikipedia and nowhere else. I think this is going to be a major problem. (Probably solvable by a special bot assignment to go through all articles, compare their old revisions with Wikidata, and add missing links back.) --Moscow Connection (talk) 18:30, 15 February 2013 (UTC)

But some of has have enough edits on Wikidata and check the items. I am more worried about watchlist contamination.--Ymblanter (talk) 18:51, 15 February 2013 (UTC)
I have no edits on Wikidata. Yet I have verified that my edits to Hyde Park Corner tube station, Macedonia (Greece) and Bærum were correct, in that they did not remove any H:ILLs from the rendered page. Am I to be reverted and taken to WP:AIV - like Marek69, who was found innocent? --Redrose64 (talk) 19:44, 15 February 2013 (UTC)
It was just a figure of speech. Cause I said "if I were an admin", and the condition is not met. (I wasn't talking about you, I was worried about people who didn't see the thread and who simply believed that all interlanguage links had to be deleted immediately.) --Moscow Connection (talk) 20:01, 15 February 2013 (UTC)
(edit conflict) Okay. It was a figure of speech about reverting and prohibiting people from touching interlanguage links. :) Anyway, I'm worried. I think it's generally impossible to do it correctly without employing bots and scripts. (By the way, I haven't edited Wikidata myself, so I'm just a bystander.) --Moscow Connection (talk) 20:01, 15 February 2013 (UTC)
Do you know what's depressing? I've just restored 4 interwiki links on this very page that were removed without first checking they existed in Wikidata, despite this discussion specifically mentioning that this should be checked. – PartTimeGnome (talk | contribs) 21:12, 15 February 2013 (UTC)
As a double-check: If the interwiki links do not appear visible after removal is a very strong indication that they are not on Wikidata and should not have been removed.--Ymblanter (talk) 21:28, 15 February 2013 (UTC)
Actually, an even easier way to check is to look for the "Edit..." link at the bottom of the interwikies. If there isn't one, then Wikidata doesn't have them, and they shouldn't be removed. If there is, then, of course, it doesn't hurt to check if our iwikies match what's in Wikidata.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); February 15, 2013; 21:32 (UTC)
"To check if our iwikies match what's in Wikidata" - this is a very important remark. What was Wikidata populated from, the English Wikipedia or from other language? There are many interwiki conflicts. (There are many articles that are crosslinked to other articles that don't link back, but link to something completely different, cause words and terms and article scopes in different languages don't always correspond to each other. I'm sure Ezhiki and Ymblanter know what interwiki conflicts are, I'm explaining it for people who may not.) --Moscow Connection (talk) 21:42, 15 February 2013 (UTC)
May be one more thing to remark would be that out of the five last items I checked the Wikidata was lagging behind Wikipedia, one or two links were missing there. A good style in this case would be to import interwiki links on Wikidata.--Ymblanter (talk) 22:23, 15 February 2013 (UTC)

By the way, how to make the script work? I'm not sure what .js file Rschen7754 is referring to. Also, does this mean edits that removed interlanguage links will be reverted? Narutolovehinata5 tccsdnew 01:58, 16 February 2013 (UTC)

I've tried removing the links from Magic: The Gathering 3 times now. I give up. This project needs to be announced better. (talk) 14:32, 16 February 2013 (UTC)

Some people, when they see an edit by an IP address which does nothing other than remove something (even if it's a valid removal) get trigger happy and go for "rollback" without thinking about why. I suggest that you register an account and log in, because some people don't believe that IPs are human too. Anyway, I've redone your edit (and anybody who reverts me will be templated {{subst:uw-disruptive1}}). --Redrose64 (talk) 15:03, 16 February 2013 (UTC)
Those are people who shouldn't be using rollback. A rollback is a revert and when good faith IP edits are reverted, it harms the project. Jason Quinn (talk) 07:14, 24 February 2013 (UTC)

Hey, everybody, for starters, can we turn off whatever edit filter adds the "removal of interwiki link" tag? Then, for bonus points, we can add a filter that checks if {{Link GA}}s or {{Link FA}}s have been removed, tag those and have the tag link to some subsection of WP:WDATA that will explain why you shouldn't remove those templates.
Other ideas:

  • If this is possible, why not set up an automatic edit summary for edits that remove all interwiki links, saying something like ( removed interwiki links following deployment of Wikidata).
  • If the above is not possible, maybe set up another edit filter, this one a) warning users who remove ILL's without leaving a summary, and politely requesting that they do so, linking to either Wikipedia:Wikidata or a relevant page on Wikidata itself (perhaps d:WD:Introduction#Phase 1), and b) warning users who restore the ILLs, explaining the Wikidata client deployment, and saying that they should only restore links that include the {{Link GA}} or {{Link FA}} templates. In both cases, encourage the editors to check to make sure all links have been copied over to Wikidata.

To be clear, as the largest 'pedia, enwiki has had a significantly lower number of its articles' ILLs than some other wikis – IIRC, frwiki, nlwiki, and possibly others claim to have all articles older than a few weeks added to the database. Anyways, if I get the chance I'll work on the description side of things, but I think some technical solutions along the lines of what I'm suggesting would really do us some good. — PinkAmpers&(Je vous invite à me parler) 15:28, 16 February 2013 (UTC)

I would probably not want the filter turned off, nor would I want an automatic summary, nor the filters in your second bullet. There will be cases where the removal of the links is actually vandalism. A filter for the links FA/GA seems like a good idea, though. I would agree that technical changes would be useful, but I do not think your suggestions really help to fix the problem while maintaining the stability of the 'pedia. --Izno (talk) 19:18, 16 February 2013 (UTC)

Even if removals are made with vandalistic intent (which you'd almost never be able to be sure of), they'll only have minimal impact on the page. (An impact that will eventually be counteracted in the course of routine housekeeping at Wikidata.) At this point, I'd say the tag does more harm than it does good. — PinkAmpers&(Je vous invite à me parler) 22:43, 17 February 2013 (UTC)
I like the change to the tag that was implemented. I guess you're fine with it as well? --Izno (talk) 14:27, 19 February 2013 (UTC)
In my experience, it's not that the interwikis are deliberately removed as an act of vandalism, it's that other acts who are vandalism (blanking chunks of the text, etc) are often highlighted by the loss of interwikis. A small but subtle difference! Andrew Gray (talk) 14:36, 19 February 2013 (UTC)

Where to test Lua modules[edit]

The Scribunto documentation gives the example of {{#invoke:Bananas|hello}} for invoking a module. If I want to test out creating a module in a subpage of my user space (rather than polluting Module: with a test), what's the syntax for invoking it? {{#invoke:User:Hex/Test module}}? — Hex (❝?!❞) 15:46, 20 February 2013 (UTC)

Might work, might not; but in any case you can go to // and try out Lua there. 16:00, 20 February 2013 (UTC) — Preceding unsigned comment added by Jarry1250 (talkcontribs)
There is no way to invoke a page outside of the module namespace; {{#invoke:User:Hex/Test module}} would try to invoke Module:User:Hex/Test module. Which is a valid module page, BTW. Anomie 01:42, 21 February 2013 (UTC)
Thanks Anomie. Given that there is inevitably going to have to be a period of people learning how to use Lua, perhaps we should be establishing some guidelines on module naming so that people can experiment without causing a huge mess. — Hex (❝?!❞) 16:34, 21 February 2013 (UTC)
there is a dedicated place for experimentation. i think it's called or somesuch. as a side, i do not think we should worry too much about "causing a huge mess". if/when someone creates a page that needs to be deleted, there is an established process for deletion requests. peace - קיפודנחש (aka kipod) (talk) 17:03, 21 February 2013 (UTC)
To be honest, I think that if the only way for people to have a sandbox environment is to use another site, uptake of Lua is going to be much slower than it ought to be. (How many people do you think there are that have written template code that have never even heard of test2? I'd be willing to bet good money that the answer is "a lot".) And yes, we have a deletion process; but how many times will we get people MfD'ing apparently abandoned modules, only for their author to pop up and say hold on a moment, that's a project of mine on the back burner? We should proactively prevent confusion and people's time being wasted by coming up with standards for people to work in Lua here on the live database. — Hex (❝?!❞) 20:09, 22 February 2013 (UTC)
if you argue that some framework for sandboxing LUA will be nice, i can't argue with that. however, unlike javascript which is run by the browser, and hence can be tested and sandboxed without actually saving anything on the server, the LUA code is run by the server, and ttbomk, at present time, there is no way to test it without actually saving the code to some page in the "Module:" namespace. maybe it would be a good idea to to dediscate, say, Module:Sandbox as the root of all experimental modules, and decide on some convention: e.g., you will keep all your experiments as subpages of Module:Sandbox/Hex, and i will keep mine as subpages of Module:Sandbox/קיפודנחש. this way, even though the "Module:" namespace *will* be "contaminated", the contaminatin will be contained, and easier to control. of course, such a scheme still requires discipline - there is nothing (at least, nothing i know of) build in the system that can enforce it. peace - קיפודנחש (aka kipod) (talk) 20:43, 22 February 2013 (UTC)
  • Yes, that's exactly what I meant, a convention. In the spirit of being bold, I've created Module:Sandbox. If the community doesn't take up the convention, after a while it can be deleted. — Hex (❝?!❞) 16:15, 24 February 2013 (UTC)
Why can't we do it the same way we do templates? That is, we already have Template:Str len which is a live template and Template:Str len/sandbox which is the test version, so why not have e.g. Module:String for the live code, and Module:String/sandbox for the test version. Then, in Template:Str len/sandbox, we would put #invoke:String/sandbox instead of #invoke:String --Redrose64 (talk) 21:14, 22 February 2013 (UTC)
It seems to me that both can be done: one for personal testing of a module/dinking around, and one for dedicated template sandboxes. I think that's a primary point that is being made. --Izno (talk) 22:54, 22 February 2013 (UTC)
Note that you can also use Special:TemplateSandbox to test changes to Lua modules, just as you can with templates. Anomie 01:31, 23 February 2013 (UTC)
is this documented anywhere? can you elaborate? thanks. peace-קיפודנחש (aka kipod) (talk) 03:37, 23 February 2013 (UTC)
See mw:Extension:TemplateSandbox. PrimeHunter (talk) 16:12, 24 February 2013 (UTC)

Template {{Reflist|30em}}[edit]

This template doesn't work on Android Ice Cream Sandwich, desktop view. It adds a column to the article with no text, or just a couple of references, while reducing the font display size to about half. Compare Cancer with {{Reflist|30em}}[16] and {{Reflist}}[17]; or Android (operating system) with {{Reflist|30em}}[18] and {{Reflist}}[19]. Should something be done about this template? --Anthonyhcole (talk · contribs · email) 15:47, 21 February 2013 (UTC)

Does anybody know who is developing/maintaining en.WP's interface on Android? --Anthonyhcole (talk · contribs · email) 06:25, 23 February 2013 (UTC)

Not much can be done; the template emits standard CSS, which not all browsers may yet fully understand. I don't know what browser engine runs on Android (Webkit?). We don't have a developer specifically for android desktop either; we basically have to display modes: desktop and mobile, and devs working onn either one generally verify their work on the major platforms/borwsers. Edokter (talk) — 10:46, 24 February 2013 (UTC)

Pages with media complain if Javascript not enabled[edit]

Today, all of a sudden, any page with a media file (sound/video) on it is giving a complaint if Javascript is not enabled. Any chance of stopping this behaviour please? IE8, Vector. Thanks. --Stfg (talk) 20:03, 22 February 2013 (UTC)

What sort of complaint? Can you provide a screenshot? --brion (talk) 00:49, 23 February 2013 (UTC)
Hamster wants Java Runtime

Sorry about the too-short (and possibly inaccurate) description. Will this do? --Stfg (talk) 10:07, 23 February 2013 (UTC)

Thanks, that shows the complaint is about Java (programming language) and not JavaScript which is something different. Some browser plugins for certain media formats require Java. Which plugins do you have in IE8? See [20] for how to check it. PrimeHunter (talk) 10:55, 23 February 2013 (UTC)
I also see it in IE9 now. I guess it's not about an addon. I don't know whether it's related or new but in IE on media file pages, for example File:Dwarf hamsters eating cooked, frozen beans - 03.ogv, I see the message "For a better video playback experience we recommend a html5 video browser". At commons the link fails. commons:File:Dwarf hamsters eating cooked, frozen beans - 03.ogv displays the unlinked "For a better video playback experience we recommend a [ html5 video browser]." PrimeHunter (talk) 12:54, 23 February 2013 (UTC)
Ah, yes, <cough>. I have "Java(tm) Plug-In SSV Helper" and "Java(tm) Plug-In 2 SSV Helper", both at version, installed yesterday morning with the proffered upgrade to that version. That's when this started, and I failed to put two and two togetther <blush>. I also have various player plug-ins -- Flash, Real, DivX -- and the Adobe PDF Link Helper, but those were installed longer ago and have been on the system without this happening. By the way, with sound files, I get a message "Sorry, your browser either has JavaScript disabled or does not have any supported player" in the window, which is why I thought of Javascript, but I don't need to play those clips, so that's not a worry. The main thing is to get rid of that yellow bar at the top, as shown in the screenshot. Thanks. --Stfg (talk) 12:56, 23 February 2013 (UTC)
I'm not sure this is the whole story about the Java message. Maybe somebody who knows more will comment, but weekends often have less activity. PrimeHunter (talk) 13:49, 23 February 2013 (UTC)
I've decided to enable it for Wikipedia. Thanks very much for your help. --Stfg (talk) 19:46, 24 February 2013 (UTC)

Inability to edit[edit]

Currently and since this afternoon, I have been unable to edit Wikipedia_talk:Did_you_know#New code... what's with making it so complicated?, or the page as a whole. Is this happening to anyone else? Trying to edit the section results in a totally blank edit window and it won't save either - I had to add a new section to post my comment.--Gilderien Chat|List of good deeds 19:12, 23 February 2013 (UTC)

No issue here displaying (Chrome, XP, United States). I haven't tried saving since I'm not engaged in the discussion. Chris857 (talk) 19:27, 23 February 2013 (UTC)
Sorry, should have added doesn't work in Windows 7, Chrome and Firefox, United Kingdom.--Gilderien Chat|List of good deeds 20:00, 23 February 2013 (UTC)
Works here on Windows 7/Firefox.—Kww(talk) 21:00, 23 February 2013 (UTC)
Sounds like there might be a problem with your custom JavaScript. On that basis, have you tried commenting out your user scripts then (if that works) re-enabling them one-by-one to see which script maintainer to contact? — Richardguk (talk) 22:09, 23 February 2013 (UTC)
Correction - it is all the sections on the page from "Promotion error" to "Antedon mediterranea" but no others, and also when I attempt to edit the page as a whole. I am testing the scripts now. I just blanked my scripts, and still can't edit, although I can in IE9.--Gilderien Chat|List of good deeds 23:37, 23 February 2013 (UTC)
Can you edit it if you log out? What about at the secure PrimeHunter (talk) 01:29, 24 February 2013 (UTC)
The edit screen still blanks on the secure server, but it appears I can if I have logged out.--Gilderien Chat|List of good deeds 11:41, 24 February 2013 (UTC)
Have you tried to clear your entire cache? Do you use Vector? Does it work in other skins? Have you tried to disable Editing gadgets at Special:Preferences#mw-prefsection-gadgets? PrimeHunter (talk) 12:40, 24 February 2013 (UTC)
I already tried clearing my cache and purging several relevant pages - by disabling and enabling various gadgets, I have determined that the problem is due to the wikiEd extension. What should I do from here?--Gilderien Chat|List of good deeds 14:28, 24 February 2013 (UTC)
I get the same issue with Windows Vista, Firefox, Vector, when wikEd is enabled. The edit box has text at first for me but the text disappears after a second, and the wikEd icon in the upper right corner has a red x with mouseover text starting "Loading error". Based on experiments at the mentioned page and elsewhere, it's caused by My76Strat's signature: '''''<sub><font size="1">[[User:My76Strat|<font color="black">My</font><font color="#ff4500;">76</font><font color="black">Strat</font>]]{{*}}[[User talk:My76Strat|talk]]{{*}}[[Special:EmailUser/My76Strat|email]] </sub></font>'''''. You can report at User talk:Cacycle/wikEd. PrimeHunter (talk) 15:12, 24 February 2013 (UTC)
<sub><font size="1">...</sub></font> is in the wrong order. This trips up wikEd. The signature should be fixed but wikEd should probably also be able to handle it better. PrimeHunter (talk) 15:23, 24 February 2013 (UTC)
Ok I see you've notified those relevant ... thanks for fixing this for me.--Gilderien Chat|List of good deeds 15:44, 24 February 2013 (UTC)
I can only apologize for that, and gesture that I am finished with custom signatures. I don't mind replacing any of my signatures with the standard block, that I now use. Maybe a bot can be tasked to do it? Again, I apologize.My76Strat (talk) 01:36, 25 February 2013 (UTC)
No worries; many editors get away with much worse signatures! Mismatched tags should always be avoided, but your signature did at least display OK and we've gained the knowledge that wikEd needs to be more robust. — Richardguk (talk) 01:54, 25 February 2013 (UTC)

Wikipedia via text.[edit]


Would you please let me know when "Wikipedia via text" service is going to be availed in India and from when?

And please shed a little light on details on how the article is “delivered”.

Is the plan to send them the entire article, broken up into 160-character chunks? Pre-iPhone, I remember having motorola and nokia phones whose SMS clients made you back out of each message in your SMS inbox to read the next one, so if someone sent you a text while you were reading the current one, you had to hit back, down, center to read the next one.

So imagine doing that about a million times for each article.

Regards Bhushan — Preceding unsigned comment added by (talk) 06:32, 24 February 2013 (UTC)

I found this, and this is some mysterious-looking technical stuff related to it. It looks like it is done a section at a time; however I really don't know how it works. Maybe the user will first get an SMS of the table of contents, from which they can choose the sections they want to read. — This, that and the other (talk) 06:53, 24 February 2013 (UTC)
The text in black here is an example of what you might get. — This, that and the other (talk) 06:55, 24 February 2013 (UTC)

IPs creating articles[edit]

Jesus Chameleon was created today by an IP. Is this a change in wikipedia policies or a software bug? Materialscientist (talk) 07:46, 24 February 2013 (UTC)

The page history shows it was created at Talk:Jesus Chameleon and later moved to article space. So neither a change in policy nor a bug. -- John of Reading (talk) 07:50, 24 February 2013 (UTC)
Facepalm. I've overlooked the page move, thanks. Materialscientist (talk) 07:56, 24 February 2013 (UTC)

LUA categories, and performance[edit]

Should we ask for support of categories for module: (LUA) pages? Before it ends in some kind of chaos? and a special place for testing, e.g. module:testing/xyz? or ask for support in user space?

What is the performance gain by using LUA? and is it a good idea to have a lot of functions in the same modules, like module:math? If its is changed, a lot of pages has to be updated! (talk) 11:11, 24 February 2013 (UTC)

Regarding testing, see the discussion further up this page titled #Where to test Lua modules. There is now a Module:Sandbox. — Hex (❝?!❞) 16:04, 24 February 2013 (UTC)
  • Categories can be split/renamed and Lua performance is 4x-180,000x faster: For years, category names have been renamed or split into subcategories, happily, so it is not much of a problem. For starters, I created "Category:Lua-based templates" as just an initial index, but that could be renamed or split. As for performance, the speed depends greatly on the application, and early results show Lua can process template parameters 4x-7x times faster than markup-based templates, while scans of text strings have often run an astounding 180,000x faster than markup-based string-handling templates. The performance issues are complex, so for more experiments about the Lua benchmarking, see new essay: wp:Lua speed. -Wikid77 (talk) 16:02, 24 February 2013 (UTC)

Support for categorizing modules is coming soon. Toohool (talk) 20:30, 24 February 2013 (UTC)


Can someone do me a favor and give this template a minor re-write? right now it spews files even if the file doesn't exist. Which leads to pages like List of B8 polytopes with 216 missing files. I am not sure if LUA would help in this case or not. Werieth (talk) 12:08, 24 February 2013 (UTC)

Templates cannot check if an (non-local) image exists or not. I doubt lua can either. But besides that, every occurence of CDD seems to work as expected; none of the redlinked images uses the CDD template. Edokter (talk) — 12:47, 24 February 2013 (UTC)
Templates can check if an image exists or not, by using the Media: namespace, as in {{#ifexist:Media:imagename|code for image that exists|code for image that does not exist}}. For example, we may put:
*The image [[:File:A-36 "Apache" of the 27th Fighter Bomber Group.jpg]] {{#ifexist:Media:A-36 "Apache" of the 27th Fighter Bomber Group.jpg|exists at [[:Media:A-36 "Apache" of the 27th Fighter Bomber Group.jpg]]|does not exist}}
*The image [[:File:A-36 "Apache" of the 27th Fighter Bomber Group12345.jpg]] {{#ifexist:Media:A-36 "Apache" of the 27th Fighter Bomber Group12345.jpg|exists at [[:Media:A-36 "Apache" of the 27th Fighter Bomber Group12345.jpg]]|does not exist}}
It used to be that Media: did not work through redirects, but this seems to have been fixed. For example, File:A-36.jpg is a redirect to File:A-36 "Apache" of the 27th Fighter Bomber Group.jpg so we may also put:
*The redirected image [[:File:A-36.jpg]] {{#ifexist:Media:A-36.jpg|exists at [[:Media:A-36.jpg]]|does not exist}}
It works for non-local images too. File:BSicon leer.svg is a commons-hosted redirect to the commons-hosted image File:BSicon .svg so we may put:
So there is a technique. --Redrose64 (talk) 15:58, 24 February 2013 (UTC)

Old parser bug?[edit]

Consider the above navbox... both lists should be ordered (ol), but the second list in unordered (ul). I recall this happning way back and was attributed to a parser bug, which I believe should since have been fixed. However, I can't find any relevant bugs in Bugzilla. Edokter (talk) — 16:27, 24 February 2013 (UTC)

Looking at the HTML generated by the above wikitext at Special:ExpandTemplates (to bypass HTML Tidy), a single <ol>...</ol> is wrapping both lists, even though they are in separate rows and divs. So the <li> element in the second list is treated as having an implied <ul>...</ul> and the closing </ol> is ignored. (In non-Special pages, Tidy attempts to repair the tag soup by wrapping the orphan list item in an explicit <ul>...</ul>, as can be seen from the HTML source of this page.) So definitely a failure by the parser to recognise that a list should not straddle higher-level block elements. — Richardguk (talk) 00:27, 25 February 2013 (UTC)
Moving some of the </div> tags to new lines appears to fix it in that table (but I haven't checked whether it works when the image sections are used). The bug you're referring to may be bugzilla:40274. Peter James (talk) 01:09, 25 February 2013 (UTC)


It appears is having problems causing pages to load extremely slowly at times. - The Bushranger One ping only 00:24, 25 February 2013 (UTC)

Basic HTMLish view[edit]

Am I the only one having Wikipedia load with really basic formatting? — nerdfighter(academy) 00:55, 25 February 2013 (UTC)

bits is having trouble so no css/other stuff.Smallman12q (talk) 00:57, 25 February 2013 (UTC)
See the above. bits is on the fritz. - The Bushranger One ping only 01:05, 25 February 2013 (UTC)

Hide Article Feedback Activity Log in Special:RecentChanges?[edit]

This is worthless information to me, and it tends to occasionally clog up recent changes. Is there a switch somewhere that I'm not seeing, or is there some sort of workaround that will accomplish this? Thanks. --Bongwarrior (talk) 02:16, 25 February 2013 (UTC)

Scribunto deploy[edit]

Just a reminder, the Scribunto deploy is scheduled to start at 23:00 UTC (about 5 minutes from now). If you run into problems, feel free to join #wikimedia-tech connect. Anomie 22:55, 18 February 2013 (UTC)

And we're deployed! Remember, it's suggested we don't convert everything all at once, so not so many templates need to be reverted if for some reason it has to be undeployed. Anomie 23:53, 18 February 2013 (UTC)
Ww! -DePiep (talk) 00:06, 19 February 2013 (UTC)

I've gone ahead and converted {{str len}} to use Module:String. I've also imported Module:Citation and its helpers, but someone needs to look over them to make sure they don't need fixing (e.g. using mw.ustring instead of string) before anyone tries to deploy them. And remember, let's not go too wild right away with converting everything, Just In Case™ issues come up and it needs to be reverted. Anomie 03:51, 19 February 2013 (UTC)

One thing I've wondered about is what kind of server load would result from (say) replacing the citation templates with a module (when pages are rendered). And if that change occurred, how much impact would result from someone adjusting a comment in the Lua code, or from a minor refactor of the code. Would all pages with a citation template have their caches invalidated? I know we don't worry about performance, but it would be useful to have some general guidance in due course. Johnuniq (talk) 05:59, 19 February 2013 (UTC)
All pages that use a citation template will have their caches invalidated, but the invalidation will be stretched over a period of hours, and Squid won't be purged. For performance reasons, we definitely do want heavily-used templates to be replaced, so that the parse time will be lower in the future. But please test the module before you change the main template, for example using the TemplateSandbox extension, to minimise the number of updates. Note that updates to heavily-used Lua modules will have the same cache-invalidating effect as updates to heavily-used templates. -- Tim Starling (talk) 09:11, 19 February 2013 (UTC)
Observationally, when templates and such invalidate a very large numbers of pages on enwiki, I've observed them being rebuilt at around 10 per second, suggesting a background refresh rate of about a million pages a day. As Tim says, we are lazy about recaching after a page is invalidated. If a registered user requests the content it can get rebuilt immediately, but content for readers gets refreshed in the background and stretched over hours or days. Dragons flight (talk) 01:49, 26 February 2013 (UTC)

While we're waiting for the dust to settle, it would be nice just to gradually import modules from test2, so we can get what we want and leave behind the rubbish. Then we can hook them up to the templates in a few days' time. — This, that and the other (talk) 09:25, 19 February 2013 (UTC)

Any admin has the ability to import modules from test2wiki. Make a list of the ones that seem to be useful; there's no point in importing things like test2wiki:Module:What. Anomie 11:46, 19 February 2013 (UTC)
Basically all these, if deemed relevant: test2wiki:Wikipedia:Scribunto_experimentationThis, that and the other (talk) 00:01, 20 February 2013 (UTC)
  • See tangent: "#Lua cites run 85/sec depth=3" which shows Lua to be 6x faster for parameter processing, although thousands faster for string-scan searches. -Wikid77 13:40, 19 February 2013 (UTC)
  • I have created essay "wp:Lua in Wikipedia" (wp:Lua) to explain using Lua for various applications. Feel free to expand that essay. -Wikid77 13:40, 19 February 2013 (UTC)

I see {{Coord}} has been proposed as a good candidate for a rewrite; I'd suggest the {{Start date}} and {{Birth date}} families, also. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:57, 20 February 2013 (UTC)

Strange table tidying on unclosed wikitable[edit]

Rockwall, Texas (rev 539000930): Failure to close the table in the ==Economy== section apparently results in it jumping down into the ==Government== section and strangely concatenating to the next table of unrelated data. I would consider this behavior a bug rather than a feature. For one reason, it deceives editors as to which section must be edited to fix it. Alternatives would include just closing any table upon finding a section heading (on the basis that a table should not span multiple editable sections), or closing it at the end of the article to make the invalid table markup immediately obvious. I would further argue that any invocation of non-trivial html tidying such as this should populate some hidden cleanup category rather than silently doing strange things. ―cobaltcigs 07:13, 19 February 2013 (UTC)

It's not a bug as such, it's the way that badly-formed HTML tables work in most browsers. There are two things at work here, which might seem contradictory.
  • If a table lacks a </table> tag - or the wikicode equivalent |} - it persists until the end of the document, therefore everything after the unbalanced <table> or {| tag is treated as if it is enclosed by that table.
  • If anything other than one of the valid table-forming elements (caption colgroup col tbody thead tfoot tr td th) is placed inside a <table>...</table> element, and is also is not within either <caption>...</caption> <col>...</col> <th>...</th> or <td>...</td>, it is treated as if it should be displayed above the table. Consider the following HTML markup
<div style="background-color:#ffd">
  <table border=1>
      <td>Data properly placed in a cell</td>Data which is misplaced outside of cells
This displays as follows
Data which is misplaced outside of cells
Data properly placed in a cell
Notice how the text "Data which is misplaced outside of cells" appears above the table, even though it was entered inside the table. --Redrose64 (talk) 14:19, 19 February 2013 (UTC)
<h2> <span class="mw-headline" id="Economy">Economy</span></h2>
<p>According to the City's 2009 Comprehensive Annual Financial Report,&#160;:<sup id="cite_ref-7" class="reference"><a href="#cite_note-7"><span>[</span>7<span>]</span></a></sup> the top employers in the city are:</p>
<h2> <span class="mw-headline" id="Government">Government</span></h2>
<h3> <span class="mw-headline" id="Local_government">Local government</span></h3>
<p>According to the city’s most recent Comprehensive Annual Financial Report Fund Financial Statements, the city’s various funds had $36.6 million in Revenues, $57.2 million in expenditures, $32.3 million in total assets, $3.6 million in total liabilities, and $25.5 million in investments. <sup id="cite_ref-8" class="reference"><a href="#cite_note-8"><span>[</span>8<span>]</span></a></sup></p>
<p>The structure of the management and coordination of city services is:<sup id="cite_ref-9" class="reference"><a href="#cite_note-9"><span>[</span>9<span>]</span></a></sup></p>
<table class="wikitable">
<th># of Employees</th>
<td><a href="/wiki/Rockwall_Independent_School_District" title="Rockwall Independent School District">Rockwall Independent School District</a></td>
<td>Texas Star Express</td>
<td><a href="/wiki/Wal-Mart" title="Wal-Mart" class="mw-redirect">Wal-Mart Superstore</a></td>
<td>Presbyterian Hospital</td>
<td>City of Rockwall</td>
<td>Rockwall County</td>
<td>Special Products</td>
<td><a href="/wiki/Target_Corporation" title="Target Corporation">Target Corporation</a></td>
<td>Rockwall Nursing Home</td>
<td>Kohl's Department Store</td>
<td>City Manager</td>
<td>Rick Crowley</td>
<td>Assistant City Manager</td>
<td>Director of Finance</td>
<td>Mary Smith</td>
<td>Director of Administrative Services</td>
<td>Chery I Dunlop</td>
<td>Chief of Police</td>
<td>Mark Moeller</td>
<td>Fire Chief</td>
<td>Mark Poindexter</td>
<td>Director of Community Development</td>
<td>Robert LaCroix</td>
<td>City Engineer/Director of Public Works</td>
<td>Chuck Todd</td>
<td>Director of Building Inspections &amp; Code Enforcement</td>
<td>Jeffrey Widmer</td>
<td>Director of Parks, Recreation and Rights-of-Way</td>
<td>Brad Griggs</td>

B-b-but, I looked at “view page source” in my web browser and this appears to confirm that the two tables which are supposed to be in different sections are in fact concatenated into a single badly formed table in the ==Government== section whilst the ==Economy== section contains neither any table html nor fragments thereof. Are you saying this html is not the same html sent from the server? ―cobaltcigs 15:52, 19 February 2013 (UTC)

I generally use Firefox (18.0.1). Here is the same url of the old revision, viewed in Lynx and still showing a single malformed table in the ==Government== section: [21]. ―cobaltcigs 16:02, 19 February 2013 (UTC)

MediaWiki uses HTML Tidy to tidy up the HTML sent to the browser, hence why the HTML seen by the browser is different. For example, RedRose64's example above looks like this if you view the source of this page:
<div style="background-color:#ffd">Data which is misplaced outside of cells
<table border="1">
<td>Data properly placed in a cell</td>
It appears HTML Tidy uses the same rules as browsers to work out how to display invalid markup, then outputs valid markup that displays the same way. – PartTimeGnome (talk | contribs) 23:31, 19 February 2013 (UTC)
The "untidy" behaviour can be verified by pasting Redrose64's broken markup into the Input Text box at Special:ExpandTemplates (Special pages currently bypass HTML Tidy – Bugzilla:39617). — Richardguk (talk) 00:00, 20 February 2013 (UTC)

Error category for html-tidy abuse[edit]

[ragecomic_face_okay.gif] Thanks for clarifying that, PTG. I misunderstood and thought Redrose was implicating the browser itself. Now, as for my second question: At the software level, how feasible would it be to populate a hidden cleanup category (similar to Category:Pages with incorrect ref formatting) for systematic human review in the case of tag nesting/closure error, rather than assuming HTML Tidy will do what the average non-cautious user/editor expects?
I stumbled upon this particular error by chance just reading about Rockwall, TX for my own amusement, but I’m sure this happens in other articles as well. Examining wikitext in the ==Government== section left me confused for a few minutes, and I initially suspected some kind of cache problem before checking the higher section. And more generally, consider the possibility other wikis may reuse Wikipedia content in raw markup form without themselves enabling HTML Tidy, or without using the same configuration, etc. ―cobaltcigs 23:09, 20 February 2013 (UTC)

You might be interested in Wikipedia:WikiProject Check Wikipedia. E.g, until you fixed it, Rockwall, Texas was probably listed as Table not correct end. – PartTimeGnome (talk | contribs) 22:25, 21 February 2013 (UTC)

Thanks again for this suggestion, which I'll be sure to check out. I see it says "Dumps and new articles are run through a series of checks to produce lists for each check" which seems to imply that errant edits to existing articles are not included until the next dump (and an instant/automatic category for serious tidy-fails might still be worth considering). I don't know how many new articles/stubs are created with a table in the first revision, but I suspect the number is low and the "table not correct end" list comes almost exclusively from dumps (though it looks like said dumps succeed far more often than they used to). ―cobaltcigs 07:23, 25 February 2013 (UTC)

Is it possible to make more things transcludeable?[edit]

I have added some description of limitations I've encountered in Lua at Wikipedia:Lua - namely, the scripts can't take input from SVG files, categories, or Special: pages because when these are transcluded you don't get the goods. I understand the scripts are only run at parse time, so some of these limits are hard - for example, a Lua script can't update when something is added to a category. However, I have to ask: is anything transcludeable that isn't now?

For example, I would like {{Media:Letters SVG.svg}} to parse as

<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" "">

etc. Can this be done? Are there any present limits on transclusion that are "ideological" (someone decided, pre-Lua, that there was no use for it) and therefore could be changed, as opposed to being limitations of what can readily be implemented by the software? Wnt (talk) 17:35, 24 February 2013 (UTC)

That could get really scary real quick and introduce all kinds of issues including XSS issues. Werieth (talk) 18:15, 24 February 2013 (UTC)
How? SVG is just SVG. — This, that and the other (talk) 00:12, 25 February 2013 (UTC)
SVGs are just xml see for some XSS issues. Werieth (talk) 00:37, 25 February 2013 (UTC)
Isn't the question about transcluding an SVG into wikitext (or a Lua script)? Since SVG markup has no particular meaning in wikitext, browsers would show the SVG source rather than an image. However, now you've raised the question of XSS, I wonder if we already have an issue: If you go to a file page for an SVG and click the displayed thumbnail, the browser will load the original SVG file. A Media: link will do the same thing (e.g. Media:Information icon.svg). Do we do anything to sanitise the SVG when it is uploaded?
What concerns me more about transcluding files is that files can be quite large (larger than the maximum size for a normal wiki page). You'd probably run into the transclusion limits pretty quickly. – PartTimeGnome (talk | contribs) 01:58, 25 February 2013 (UTC)
That's right, the MediaWiki sanitiser will ignore svg tags, since they aren't on its whitelist of allowed HTML tags. — This, that and the other (talk) 02:11, 25 February 2013 (UTC)
I should point out that at test2 I was transcluding entire SVG files through a template (but I never finished debugging what was the matter with it - it was supposed to be capable of filling in a SVG map with a set of data. I just had to copy and paste the SVG file into a Wiki page to do it. But for a general template here on Wikipedia it might be desirable to be able to do it to an arbitrary SVG map file on Commons without having to copy and paste each one. Wnt (talk) 02:31, 25 February 2013 (UTC)
It would be worth filing a bug about this. — This, that and the other (talk) 03:25, 25 February 2013 (UTC)
i am probably just an idiot, but i don't see how you expect image data to be "transcluded". normally, the server just send the address of the image to the browser, and the browser pulls the image from the image server - not only is the "image server" a different machine, and not only is it a different url, but it's usually not even in the same domain (not that it would make any difference if they were on the same domain - this just illustarate the fact that the image data flows in a completely different rout than the html data). afaik, the image data never even makes it to the server that executes the Lua script, so it's never "transcluded" per se, regardless whether the image is jpg, png, svg, tiff, gif or schmif. peace - קיפודנחש (aka kipod) (talk) 05:21, 25 February 2013 (UTC)
No, you're not an idiot at all, that is a good point. I hadn't thought of the WMF server configuration. It's a pity SVG files don't live in the database instead of on some file server somewhere else... — This, that and the other (talk) 10:27, 25 February 2013 (UTC)

Converting TIF with imagemagick[edit]

How do I convert the following TIF to a jpg with the proper coloring using imagemagick? A plain "convert a.tif a.jpg" gives the wrong result.

Smallman12q (talk) 22:41, 24 February 2013 (UTC)

You can't convert anything to a JPEG and expect to keep the proper colouring. It's a nasty format that uses lossy compression. --Redrose64 (talk) 22:59, 24 February 2013 (UTC)
Yes. The TIFF is uncompressed, so that is a good start. You would probably get better results converting to PNG. --— Gadget850 (Ed) talk 23:51, 24 February 2013 (UTC)
I've looked at the full-res "JPG created with ImageMagick" again and it displays properly in infranview, and cs, but not in my browser (FF) or in the thumbnail generated by the wiki.Smallman12q (talk) 00:46, 25 February 2013 (UTC)
You're running into color space conversion issues. Some browsers, software, etc., are not colorspace aware. It's highly annoying. A full treatment of the problem is more than I'm up to writing right now, but [22] provides an overview of the problem from one photographer's point of view, in geeky, geeky detail. The best thing you can do is to convert your image to sRGB (and accept the consequences of that) before conversion, and have the image tagged as such. --j⚛e deckertalk 17:02, 25 February 2013 (UTC)

Comparison colorspace1.jpg

I've tried stripping the profile, and converting to sRGB and CMYK colorspace but it still looks like the "Original Tif" from IE. IE renders the the 'JPG created with ImageMagick' properly, but I can't get this to work in FF or Chrome. Also, it's odd that the Tif colorspace appears broken in IE and FF...that may be a bug.Smallman12q (talk) 00:47, 26 February 2013 (UTC)

Deleted Edits?[edit]

I wanted to know what actually does the Live Edits and Deleted Edits in the edit counter mean? But when seeing the Special:Preferences page, all the edits show together. Please help. Jayadevp13 (talk) 05:31, 25 February 2013 (UTC)

Live edits are edits to pages that are currently visible; while deleted edits are edits to pages that have been deleted (well actually, hidden from public view). All edits normally count as live edits, but if you edit a page and it is later deleted, all of your edits to that page are counted as deleted edits. The Anonymouse (talk | contribs) 05:47, 25 February 2013 (UTC)
Also, I don't know exactly why, but for some reason Preferences shows total edits (live edits + deleted edits), but the edit counter shows them individually. I don't think there is any particular reason why, but someone else may know. The Anonymouse (talk | contribs) 05:50, 25 February 2013 (UTC)
Thank You for telling.
Jayadevp13 (talk) 05:54, 25 February 2013 (UTC)
Face-smile.svg You're welcome! The Anonymouse (talk | contribs) 05:57, 25 February 2013 (UTC)
There are two ways to get an edit count: either by using the user_editcount field of the user table in the database, which is the count displayed in your preferences, or by counting your contributions individually, which is the method used by tools such as X!'s edit counter. The database field is just a number which combines live and deleted edits, while the tools count live and deleted contributions separately. Also, the database field does not count edits made while moving a page (among other things), while the live tools do. Graham87 07:37, 25 February 2013 (UTC)
So that's where the difference is, I guess. Thanks for the detailed info! The Anonymouse (talk | contribs) 08:22, 25 February 2013 (UTC)
Ya really. Thank you for telling. Jayadevp13 (talk) 07:17, 10 March 2013 (UTC)

Semi-protecting Lua scripts[edit]

Most of the Lua scripts I've looked at so far are unprotected, which is something of a concern. I think most of the scripts should probably be semi-protected while they're being finalized. I also think that the heavily used ones—like the citation templates—should be fully protected once they are tested, complete and we are ready to switch over to them. Any IPs that want to run tests can do so over at test2 as all those are unprotected. Any other opinions? (talk) 17:28, 25 February 2013 (UTC)

The two most heavily used at present, String and Math, are already indef protected. I assume others will be as well once they become widely used based on the same standard that we protect Wikipedia:High-risk templates, since they are essentially another approach to templating. Incidentally, it would be nice to update the "transclusion count" tool found on "what links here" for templates (e.g. [23]) so that it also appears for Modules. I presently have no idea how that tool is installed though. Dragons flight (talk) 18:00, 25 February 2013 (UTC)
there is no real difference, as far as "the right protection level" between "lua scripts" and templates, and the same logic should serve. most templates are not protected, and so should be most scripts. קיפודנחש (aka kipod) (talk) 18:22, 25 February 2013 (UTC)
In general, if a template is protected, its subtemplates are also prot to the same (or a higher) level. It's therefore logical to protect modules to a level similar to the templates that invoke them. --Redrose64 (talk) 18:34, 25 February 2013 (UTC)
Transclusion count enabled for the module namespace. - Jarry1250 [Vacation needed] 22:11, 25 February 2013 (UTC)
Not quite. The link it gives here is , which only returns zero transclusions. The form that apparently does work is where the namespace has been removed. Dragons flight (talk) 22:19, 25 February 2013 (UTC)
My bad. What about now? - Jarry1250 [Vacation needed] 23:07, 25 February 2013 (UTC)
That seems good, thanks. Dragons flight (talk) 00:13, 26 February 2013 (UTC)

Help test out new mobile upload features[edit]

Starting today, the Wikimedia mobile web and apps teams are kicking off a week-long community QA sprint. We need your help to test out the new mobile upload features we've built and make them better! Please visit our event page on Mediawiki and sign up to test out our new mobile web upload features: donating an image to Commons, and inserting an image from your phone directly into a Wikipedia (or any other local wiki) article from the article itself. You can also test out our new Commons upload apps for iOS and Android.

You're free to start testing any time, but our whole team (QA, engineering, product, and design) will make an extra effort to be available today (Monday February 25 starting at 17:30 UTC) on IRC, Bugzilla, email, etc., to answer any questions you may have. Come join us in #wikimedia-mobile connect, sign up for the mobile-l mailing list, and feel free to leave feedback on the event discussion page.

Looking forward to your bug reports :) Maryana (WMF) (talk) 17:54, 25 February 2013 (UTC)

has hacker taken over web addresses?[edit]


I wrote an article Sacred Himalayan Landscape and the web address for one of the references (Government of Nepal - Department of National Parks and Wildlife Conservation) has changed in the last day or so. Today it says:
owned by dr.m1st3r

If I look up this on the internet, e.g.owned by dr.m1st3r, this seems to be an "ethnic Albanian" who has a bunch of normal sounding companies, such as Men's Shirts| Glamour PK that all result in an organge symbol with strange music in the background. I'm not sure what's going on.

Is this a known problem, a known hacker? Thanks, Star767 (talk) 19:00, 25 February 2013 (UTC)

From the looks of a Google search for is dr.m1st3r a hacker, I'd say so. jcgoble3 (talk) 19:29, 25 February 2013 (UTC)
Thanks! Guess there's nothing to do about it. Star767 (talk) 20:19, 25 February 2013 (UTC)

densdisp template is now broken[edit]

RESOLVED: Fixes submitted. -Wikid77 01:40, 27 February 2013 (UTC)

The {{Infobox settlement/densdisp}} template is now broken, resulting in a whole bunch of infoboxes around the project showing errors. It seems like the recent migration of {{rnd}} to Lua coincides with everything falling apart, but the testcases for rnd seem to be doing fine. What's the problem here?  — daranzt ] 22:13, 22 February 2013 (UTC)

{{rnd}} has been rolled back for now. As I explained on the talk page, this is essentially a problem with the infobox template. As the result of a malformed calculation, the infobox is essentially requesting:
{{ rnd | 1234.56| Expression error: Unclosed bracket.}}
The existing {{rnd}} silently suppressed the error, while the new Lua version, available as {{rnd/sandbox}}, exposed the issue by reporting an error if fed non-numeric values. In essence {{Infobox settlement/densdisp}} was always doing something incorrect, but it wasn't till now that it was noticed. It is possible that more such issues will crop up as various templates are migrated. Dragons flight (talk) 01:02, 23 February 2013 (UTC)
  • Pinpointed error as unclosed ") )" in density calculations: I have submitted an {editprotected} request to install the fix from the "/sandbox2" version to correct the nesting of mismatched parenthesis brackets "(_(_(__)" in 5 formulas within that template. See fix:
To debug the complex, nested population-density precision calculations, I used the new run-preview "Show-preview" of a test page running with the live edit-buffer as the /sandbox2 version was being edited to insert the "))" symbols and used Template:Echox to echo/show the parameter values as the text for each sub-expression, to proofread each "(((_)))", within the nested intermediate calculation results, during the run-preview execution. -Wikid77 01:40, 27 February 2013 (UTC)

Machine learning for new articles?[edit]

Has someone looked at machine learning for identifying new articles likely to be deleted?

I know there is some ML in play for vandalism detection. New articles seem like the next-best application. Today between WP:AFC and WP:NPP a lot of human labor goes into the problem of making sure new articles are up to snuff. An ML approach could employ features like length of article, avg. sentence length, vocabulary, prevalence of non-english or garbage words, random capitalization, number of references, number of reference with working links, number of sections, length of title and headings, length of each section, number of pictures, posting history (posting many articles in a row), etc.

If ML says it's highly likely to be deleted, put it into a queue instead of posting. If ML says it's highly likely to not be deleted, post it without delay. In the middle, maybe post it with a stern warning that this article has not been reviewed and may be deleted. But the idea would be to catch the very worst and the very best, and get them to where they should be fast.

Now someone that's actually done AfC and NPP might say this is hopeless. That none of those features would discriminate well enough. But maybe all the them together could statistically be useful? The magic of ML is statistically there might be patterns when considering all the features at once. And remember we're not taking irreversible action. If a bad article gets posted or a good one queued up for review, it's not the end of the world, as long as the ML is decreasing the workload on humans, and is contributing in a positive way Silas Ropac (talk) 16:29, 24 February 2013 (UTC)

The problem is that virtually all human time goes on the ones in the middle: the new article that are not obviously terrible, but nor are clearly good. These are the same articles an ML-based system wouldn't be able to help with. - Jarry1250 [Vacation needed] 23:10, 25 February 2013 (UTC)
Yes I could see that being a case. But I wonder, could ML still help anyway. If I'm looking at the right place on WP:NPP it says the queue is 30 days long. Now suppose you could shorten that to 2 minutes for the very worst and the very best. Yes as you say you haven't lowered the overall effort much, because those were the easiest ones. But you've shortened the wait for those people by 30 days. It just intuitively seems like that would be good. I can't put my finger on exactly how, but faster seems better, and that's a lot faster. Also, for the ones in the middle, you could provide feedback even if you don't take any automatic action. Give them a rating (you are 2 out of 5; you might get deleted) then tell them a real editor will review their article, but that they can do these N things to improve it in the mean time. And re-rate them if they improve it. Again I haven't done NPP (or any ML either) so really I don't know anything about anything here, but I feel if we had a decent ML signal about new article quality, we'd figure out a way to wire it into the process in a productive manner. Silas Ropac (talk) 03:35, 26 February 2013 (UTC)
I think you're misunderstanding what the NPP "queue" is. It's not a waiting list where articles must be approved before going live; everything on the new pages list is live now whether reviewed or not. People do not have to wait for their article to be accepted for posting; it's posted as soon as they hit "Save page". The "queue" is simply a way of ensuring that all articles not created by a trusted user get looked at by someone. 30 days is simply a technical limit that dictates how long an article remains on the list before falling off—and it stays on the list even after being reviewed (though there is an option to hide reviewed items). jcgoble3 (talk) 04:56, 26 February 2013 (UTC)
Thanks. I did realize the NPP "queue" is of articles already fully visible in mainspace, where AfC is articles not yet in mainspace. But I did not realize already reviewed items for NPP stay in the queue! So that means the number of days in the NPP queue is strictly a function of the rate at which new articles are being created, and has nothing to do with rate at which NPP is reviewing those new articles? Seems like they'd want to report and graph the length of the non-reviewed items in queue, since that is what can be ahead or behind.
But all of that said, it still seems getting feedback faster is better, even if it's preliminary automated feedback. For the middle-quality articles, seems like if the creator got immediate preliminary feedback, they could iterate on their article before the human reviewer saw it. But maybe that would come off as a glorified Clippy, and everyone would ignore it. 12:30, 26 February 2013 (UTC)
Special:NewPages by default shows reviewed pages with a white background, and unreviewed pages with a yellow background (you can click on "Hide patrolled edits" to remove the white ones). Both stay on the list for one month, and then drop out whether reviewed or not. Articles still in AfC and not yet moved to mainspace are also in Special:NewPages, but are hidden because that page can only show one namespace at once. Select "Wikipedia talk" from the "Namespace" drop-down menu, and you'll get this - absolutely hundreds of the pesky things. I say "pesky" because they get in the way when I'm searching for newly-created Wikipedia talk: pages that are not AfC submissions. --Redrose64 (talk) 14:50, 26 February 2013 (UTC)
Thanks, that is very helpful. I see the clutter. In general it feels to me like AfC and NPP don't cooperate that much. I just don't understand the AfC plus NPP system.
I would have figured things worked like this: there are two statuses for articles, Live (in mainspace) or NotReady (squirreled away somewhere out of the limelight). Articles in NotReady need to be reviewed to go to Live, like AfC. Articles in Live are reviewed by NPP, like today, but the big difference is the most common action taken by NPP would be throwing an article from Live into NotReady. Isn't that what NPP is determining, is this ready or not? Right now it seems like NPP can delete, or tag, or clean it up themselves. Throwing it into NotReady seems like a much better option, it's super fast to do, it gives the article time to develop, it gives the author feedback, it keeps not-ready articles out of the encyclopedia.
And that's where ML could help. Look for example at this article the day it was posted live. Actually forget ML, even a tiny script could tell you it's not ready, it has no references at all. What's it doing in mainspace? For 30 days? A month later it still has no references, and no tag either. An automated process would have slapped a tag on there 29.9 days ago, and better yet booted it back to NotReady, since it's clearly not ready. Anyway, I realize there are all sorts of complexities and subtleties and social issues I'm missing, but by making unrealistic suggestions I am slowly learning stuff. Silas Ropac (talk) 20:18, 26 February 2013 (UTC)

Wikidata suggestion[edit]

I've noticed today that articles without interwiki links now have a "None" tag under "Languages" in the toolbar. While this makes sense - it makes it clear that there are no interwikis known to Wikidata - perhaps the "Edit links" link should be there instead? - The Bushranger One ping only 21:16, 25 February 2013 (UTC)

Second - good idea. ~ Amory (utc) 21:58, 25 February 2013 (UTC)
That would require a designated page to already exist on Wikidata, so not as trivial as one might think. Someguy1221 (talk) 05:47, 26 February 2013 (UTC)

I think it should show "Edit links" if an article has an item on Wikidata regardless of how many interlanguage links it has. Why should an article with two interlanguage links (English + another language) get the "Edit links" link but an article with only one interlanguage link (only English) get "(none)", if they both have an item on Wikidata? Just my opinion... The Anonymouse (talk | contribs) 08:08, 26 February 2013 (UTC)

We're working on this. You can give it a try on the test system. It'll unfortunately still take a bit until it's going to be usable on the Wikipedias. --Lydia Pintscher (WMDE) (talk) 10:18, 26 February 2013 (UTC)

Article Feedback Tool bug[edit]

The Article Feedback Tool (aka page ratings) seems to have some problems; when you rate a page the reply is "An error has occurred. Please try again later." See also User talk:Alkollar. Do we know whether the ratings are still saved? Huon (talk) 06:11, 26 February 2013 (UTC)

Eek, that's probably related to the 1.21wmf10 deployment today. We'll look into it ASAP.--Eloquence* 07:49, 26 February 2013 (UTC)
Should be fixed now.--Eloquence* 16:43, 26 February 2013 (UTC)


I was just curious if anyone is using non-standard fonts on Wikipedia. For example, Distributed Proofreaders has DPCustomMono2 at, which is supposed to make it easier to recognize look-alike characters. Personally, I tend more towards Consolas myself, but haven't actually changed any settings to enforce that. --SarekOfVulcan (talk) 13:20, 26 February 2013 (UTC)

I have 746 fonts installed, but not that one. My browser is set to use Consolas for monospaced text, and Arial/Times as the other default fonts. Edokter (talk) — 18:21, 26 February 2013 (UTC)

Lua strings[edit]

Just a note, almost all of the string templates have now been converted to use Lua Module:String. This should give hugely better performance and remove the limitations associated with the old string templates (i.e. limits on string length, character set, etc.). Hopefully this will better serve everyone's string manipulation needs. During the conversion process, a few bugs cropped up that should now have been squashed, but if people notice any string manipulations behaving strangely feel free to report them. Dragons flight (talk) 16:11, 26 February 2013 (UTC)

There's probably also templates out there that use multiple string templates that could now be modified to do things better with different templates now or a direct lua call. -- WOSlinker (talk) 18:36, 26 February 2013 (UTC)
*cough* Template:Asbox/stubtree/2 *cough* Dragons flight (talk) 21:35, 26 February 2013 (UTC)
Yes, I'm working on it but it's not quite ready yet. -- WOSlinker (talk) 21:37, 26 February 2013 (UTC)
  • Created {str_left_long} because old {str_left} 6x faster than Lua: Also, I have created new Template:Str_left_long to extract left-hand text from strings longer than 500 characters. After running extensive timing tests, for months from test2 wiki, I have again confirmed, on English Wikipedia, how the wp:parser function {padleft:} is over 6x faster than the Lua Module:String function 'sub'. The extreme speed of the parser functions, when used inside the markup-based templates, is what many of us had come to realize, so to keep the speed of {str_left}, I created the slower, Lua-based {str_left_long} to handle strings 7x slower but allowing 500 to thousands of characters long. See more: #Lua faster/slower but a helper. -Wikid77 (talk) 22:07, 26 February 2013 (UTC)

Citation tool issue[edit]

Are there any reports of the citation tool missing from the editing toolbar today? I'm not seeing it appear on the mainspace, but I am seeing it here. If anyone could help with figuring what's up, that would be awesome. Thanks! Kevin Rutherford (talk) 19:12, 26 February 2013 (UTC)

Fundraising proposal for modeling user behaviour on WMF properties[edit]

All, the Fundraising team is looking for methodologies for modeling how users use our sites so that we can attempt to devise new fundraising strategies. I've written an RfC describing one such experiment and welcome your feedback. Mwalker (WMF) (talk) 01:34, 27 February 2013 (UTC)

Can category pages link to the actual articles?[edit]

I posted this to ideas, but was told maybe to try here. For a category page like Category:GA-Class Book articles why does it have links only to the talk pages? So today each link on the category page is like:

it seems like this would be more useful:

Then you can jump to the talk or the article page, whichever you want. And note that it's identical to the default signature for users. And it doesn't take up much more space. When I'm browsing a category generally it's because I want to view articles in that category, not their talk pages. But this way would let you do either. Silas Ropac (talk) 02:29, 18 February 2013 (UTC)

I don't see the technical possibility of your proposal within the boundaries of MediaWiki. Once category is inserted into talk page, talk page will be categorized. That's all. And these kind of categories that have local interest for Wikipedians and tell something about the life of wikicommunity, are inserted into talk pages because they are not interesting for readers. A possible alternative is to use hidden categories instead, and they may be used on content pages. Bináris (talk) 14:40, 21 February 2013 (UTC)
(edit conflict) Category pages contain links to the page where the category is set; they do not contain links to pages where the category is not set. The category Category:GA-Class Book articles is set by using the parameter |class=GA within {{WikiProject Books}} (or in the specific case of Talk:A Moral Reckoning, the shortcut {{WPBooks}}). WikiProject banner templates must be used on talk pages, not on article pages, therefore it is always the talk page that appears in Category:GA-Class Book articles. --Redrose64 (talk) 15:51, 21 February 2013 (UTC)
The script User:Equazcion/CatListMainTalkLinks seems to do what you want. -- John of Reading (talk) 15:53, 21 February 2013 (UTC)
Wow cool, that is really close. For a page of talk links it does:
when this would be better:
But still cool someone wrote that, shows they had the same itch. I will use it, thanks. Ideally though, mediawiki would do it natively, without a userscript. I'm sure there might be some technical obstacles, but if we can put a car on mars, we should be able put dual-links on our category pages. Silas Ropac (talk) 20:16, 21 February 2013 (UTC)
Thanks for posting that John. I have been mulling over whether that was feasible. I am off to install it right now! -- Alan Liefting (talk - contribs) 19:59, 21 February 2013 (UTC)
In the absence of native MediaWiki support, what about implementing it ourselves like this: Create a template (let's say {{SUBJECTLINKS}}) to be used on categories like this where talk pages are categorized and links to articles would be helpful. It could be a banner like the others used in the category namespace or an "invisible" one consisting of simply an empty <span>; the important thing is to have an id= attribute set. Then put something in MediaWiki:Common.js that identifies that id, using document.getElementById("id used in template"), and when the id is present, invokes User:Equazcion/CatListMainTalkLinks or a variation of it. Essentially, {{SUBJECTLINKS}} would be used like a magic word. Would that work? jcgoble3 (talk) 01:24, 22 February 2013 (UTC)
Do we do this kind of re-writing in MediaWiki:Common.js already? I wouldn't pave new ground for this, far too minor. I did read through the titles of 329 mediawiki bugs related to Categories, I don't think this is in there. They did mention CategoryPageView (called before viewing a categorypage in CategoryPage::View) I don't know if that can be used. I would file a new mediawiki bug but not sure what to ask for. They can't magically pair up article and talk pages I don't think, they don't know WP's namespaces, instead it would have to be a hook or something which allows logic similar to User:Equazcion/CatListMainTalkLinks which spells out which articles pages go with which talk pages. Silas Ropac (talk) 11:48, 26 February 2013 (UTC)
────────────────────────────────────────────────────────────────────────────────────────────────────MediaWiki is aware of the link between subject and talk space pages. If you think about it, it needs to be for the "Article" and "Talk" tabs at the top of each page to go to the right namespace. The {{SUBJECTSPACE}} and {{TALKSPACE}} magic words also need to know which subject space corresponds to which talk space. – PartTimeGnome (talk | contribs) 21:51, 26 February 2013 (UTC)
Of course, so we can write "see talk page" or "see article page" today, yeah so they must know. What is this CategoryPageView thing that sounded promising? Anyway perhaps I will just file a bug/suggestion with mediawiki and see they say. If the answer is "bof, you can already do that" then so much the better if they tell me how. I don't think this all that important, especially with the user script, but it's an excuse to learn. Silas Ropac (talk) 18:08, 27 February 2013 (UTC)

Cannot edit watchlist[edit]

Is there a way to just delete it and start anew? Carlossuarez46 (talk) 04:48, 22 February 2013 (UTC)

You can go to Special:EditWatchlist/raw and delete some or all of it. Dragons flight (talk) 05:09, 22 February 2013 (UTC)
I tried that; I keep getting "HTTP Error 500 (Internal Server Error): An unexpected condition was encountered while the server was attempting to fulfill the request". Carlossuarez46 (talk) 17:51, 22 February 2013 (UTC)
How many items do you have on your watchlist? Werieth (talk) 17:58, 22 February 2013 (UTC)
I doubt it's size: I'm at 16870, and I can still edit mine.—Kww(talk) 21:30, 22 February 2013 (UTC)
I have hear reports about cases where the watchlist is between 30,000 and 35,000 pages and becomes un-editable via the UI. Werieth (talk) 21:43, 22 February 2013 (UTC)
How does one edit it then? Carlossuarez46 (talk) 00:22, 23 February 2013 (UTC)
If it can't be done through the UI, I might suggest a bot script to clear it out. I believe my bot framework is capable of such a task. If that doesn't work, the it'll have to be done by the tech people.—cyberpower ChatOnline 00:48, 23 February 2013 (UTC)
Does the "unwatch" tab (Monobook) or blue star (Vector) still work? If so, there is a way but it will take some time. First go to Preferences → Watchlist and turn off all the options from "Add pages and files I edit to my watchlist" downwards; this will prevent more being unconsciously added. Then go through Special:Contributions/Carlossuarez46, check each page in turn, and "unwatch" all unless you really want to keep watching. Don't try to do it all at once; do some each day, noting how far back you've been each time so you can restart at the same point. Periodically try Special:EditWatchlist/raw and when this becomes editable, you can remove pages much more quickly. --Redrose64 (talk) 08:02, 23 February 2013 (UTC)
I've filed a bug at bugzilla:45380 asking a sysadmin to clear it. MBisanz talk 20:59, 25 February 2013 (UTC)
It's been cleared today. User had 221,044 items on it.—cyberpower ChatOffline 02:37, 28 February 2013 (UTC)

Have not been able to view watchlist...[edit]

I haven't been able to view my watchlist for a couple of had about 10,000 items on this just me or are others having problems? Casliber (talk · contribs) 19:44, 27 February 2013 (UTC)

Can you tell me if the problems have been since Monday? or before that? There was a code update on Monday, although I am not sure exactly what watchlist-related changes (if any) there were. For Wikidata stuff, we're looking at making that part faster, but no idea if there's more going on or what. --Aude (talk) 19:58, 27 February 2013 (UTC)
Have you tried limiting the time frame that your watchlist displays? (1 hour, 1 day ect) Werieth (talk) 20:03, 27 February 2013 (UTC)
I can see mine, and it has over 20000 items on it and shows the last 72 hours. Ucucha (talk) 20:11, 27 February 2013 (UTC)
It was since Monday that I hadn't been able to see it, but before that it had been slowing over some time. It times out on my smartphone most of the time - it was over a long period (7 days+?), but it is all academic now as I wiped the slate clean. Casliber (talk · contribs) 02:02, 28 February 2013 (UTC)

Lua: module names back to CamelCase?[edit]

I see that new module names are using CamelCase: Special:PrefixIndex/Module:. Earlier on the test2 site, spaces could be used [24]. I'd say natural wording names is preferable. Is this a (new) technical advised restriction, or maybe a "prevent future problems" choice? -DePiep (talk) 10:59, 25 February 2013 (UTC)

I think it is just for consistency with other MediaWiki related code. See e.g. mw:Manual:Coding conventions/JavaScript#Naming. Helder 11:25, 25 February 2013 (UTC)
Clear. Thanks. -DePiep (talk) 17:10, 25 February 2013 (UTC)
See also: Wikipedia:Lua style guide. Helder 13:11, 27 February 2013 (UTC)

"Stop running this script?" message[edit]

Whenever I log into wikipedia on my home PC, I keep getting a message, as shown above, which tells me that a script running on this page is causing Internet Explorer to run slowly and could bring my system to a halt. Although I use Explorer on other PCs, I don't get this message there. Has anyone else seen this and knows what the cause is? Deb (talk) 19:07, 25 February 2013 (UTC)

Nope. Could be an issue with the javascript engine on your computer.—cyberpower ChatAbsent 19:36, 25 February 2013 (UTC)
I see that message sometimes on my work computer (which runs an "earlier" version of Internet Explorer ... possibly IE8? Can't remember offhand) when I go to long articles—or more specifically, articles with a lot of {{coord}} templates. Grade II listed buildings in Brighton and Hove: M is an example. It seems to struggle to load the little "globe" icons next to the coordinates, and brings up this message. My IE9-equipped home PC is unaffected. Hassocks5489 (Floreat Hova!) 19:55, 25 February 2013 (UTC)
The "bits" server - where the site .css and .js files are held - has been reported as having problems (see #Slowness... above). If a .js page isn't getting through to your browser, you may well see messages like this. See also Wikipedia:Village pump (technical)/Archive 108#Page loading issues. --Redrose64 (talk) 20:15, 25 February 2013 (UTC)
Thanks. I think Hassocks5489's analysis is correct, as I have now realised it happens with the "Recent Deaths" page but not noticeably on any of the others. Deb (talk) 19:00, 27 February 2013 (UTC)

Watchlist related tools?[edit]

Are there any good tools to visualize or browse changes on your watchlist, I see LiveRC but it's unclear if it works for English. The stuff on toolserver I mostly see people link to is edit count related, I saw one watchlist related tool but it wasn't clear to me what it really did. Just wondering what people actually use regularly, if anything. Thanks. Silas Ropac (talk) 03:56, 26 February 2013 (UTC)

I'm not sure if any exist, but I know there is discussion at Wikipedia:Village pump (proposals)/Persistent proposals. Does anyone know if anything the WMF developers are working on will help out with watchlists? Biosthmors (talk) 16:42, 26 February 2013 (UTC)
Sounds like multiple watchlists are the big idea there. Which sounds useful, you could group by how much you care about stuff. One to watch closely, others to watch only if you have time to kill. I guess the "Group changes by page in recent changes and watchlist" is a watchlist tool. Do people use that? I tried it and turned it off initially. But now I think it might be useful. With "Expand watchlist to show all changes" on of course. It only groups within one day, which I guess is right. I don't know what exactly to improve about watchlists, just was wondering what was out there. Silas Ropac (talk) 00:58, 27 February 2013 (UTC)
Oh I think I remember why I turned it off. With it off you get (diff | hist) by every change. With it on you get (cur | prev) by every change. no diff. I think the diff is the only reasonable way to read a lot of changes, seeing the full article (before or after) usually gives me no idea what changed So this seems like a deal breaker for "group changes" mode. What I'd really like is a diff of the rendered page. I know a lot of time you need to see diff of the wikitext, if references or other markup changed. But often I don't care at first, I just want to know what visible text changed. Seeing the actual real page with just a highlight of what's new seems like it would useful. Does any tool do that? Silas Ropac (talk) 01:11, 27 February 2013 (UTC)
I use the grouped changes and always have. "cur" and "prev" are actually both diff links; "prev", the equivalent to "diff" in the non-grouped list, is the diff of that edit/revision (compared to the previous revision, hence the name of the link), and "cur" is the combined diff of all subsequent edits to the page, i.e. the diff between that revision on the left and the current revision of the page on the right (useful when, for example, you haven't viewed the last five changes to a page out of eight today; simply click "cur" on the sixth revision to view the combined diff of all five edits at once). jcgoble3 (talk) 01:41, 27 February 2013 (UTC)
Oh, duh, thanks. That is simple enough once you get it. Also the summary line where it says (6 changes | hist), the "6 changes" part is a diff also. Yes seems nice. Silas Ropac (talk) 02:40, 27 February 2013 (UTC)

For the grouped watchlist, for pages with tons of activity like this page, would it make sense to have a 2nd level of grouping by section title? So right now my Village pump (technical) group has 71 entries, but 23 of them are either (→‎ User talk page is in error) or (→‎ Another slowness thread). If I'm not following either of those, seems like having them folded up into one line for each would be nice. But maybe that would have other drawbacks? Silas Ropac (talk) 01:14, 28 February 2013 (UTC)

Another "slowness" thread[edit]

The Watchlist seems to be taking longer and longer to load as they days go by - it used to be it would take <2-3 sec, now it takes 10+ sec sometimes. What gives? Is it just the watchlist temporarily growing longer due to all the bot edits converting interwikis to Wikidata? - The Bushranger One ping only 17:04, 26 February 2013 (UTC)

Having the same problem. Elockid (Talk) 01:14, 27 February 2013 (UTC)
I am experiencing serious watchlist slowness as well. --Saddhiyama (talk) 09:15, 27 February 2013 (UTC)
Thanks for the reports. I'll ask someone from the Wikidata team to look if this is caused by Wikidata and if so what we can do something about it. --Lydia Pintscher (WMDE) (talk) 10:48, 27 February 2013 (UTC)

Can you do me a favor and try the enhanced changes JS preference (under recent changes), and see if that helps or not? That completely omits Wikidata in the query. --Lydia Pintscher (WMDE) (talk) 12:49, 27 February 2013 (UTC)

It is to do with WikiData. See this that I raised sometime ago. Currently, my watchlist says it has 1,338 changes on it, but only 120 (approx) listed. The others are Wikidata changes that aren't been hidden in the same way you can hide bot or minor edits. This really needs urgent attention. Lugnuts Dick Laurent is dead 12:51, 27 February 2013 (UTC)

That doesn't necessarily mean it is causing the slowness. I understand that watchlist slowness is bad. Which is why we need to investigate it. --Lydia Pintscher (WMDE) (talk) 12:59, 27 February 2013 (UTC)
Odd that is started happening as soon as WD was deployed and that it's continuing while bots are adding thousands of links in WD. Sounds like it's WD to me. Lugnuts Dick Laurent is dead 15:00, 27 February 2013 (UTC) should help make the watchlist faster. It's a different issue than the counter, which we are also working to fix asap. Thanks for reporting the issue here. Cheers. Katie Filbert (WMDE) (talk) 19:43, 27 February 2013 (UTC)
Just as a note, I wasn't saying it had anything to do with WikiData itsself, but simply the trememdous number of bot edits on itself removing interwiki links because of WikiData, which runs up the number of recently-changed pages that the watchlist has to display. - The Bushranger One ping only 23:50, 27 February 2013 (UTC)

Useless edit filter[edit]

An article I dealt with in early 2011 was undergoing high levels of persistent vandalism and I repeatedly requested and gained semi-protection for the article. After several months another editor took notice of the situation and created a filter for the page. Because the vandalism took a rather specific form, the filter worked like a charm for years until the page was moved to a different title. The new title reflects the proper name of the article's subject, but sadly this means that the old filter is now protecting a redirect and the vandalism has resumed. In December 2012 I made a request that the edit filter be moved to the article's new name, but either nobody monitors that page or nobody there cares about this issue. I'm wondering where to go from here. This problem is going into its third year now, and there's no end in sight. Is it possible to move a filter? If not, is it easily possible to create a new one to do the same thing? Is Wikipedia:Edit filter the only place to go to make requests in cases like this? Thanks in advance for any help you can offer me. -Thibbs (talk) 05:27, 27 February 2013 (UTC)

What is the number of the filter? Ruslik_Zero 06:29, 27 February 2013 (UTC)
I don't know. It's associated with Overblood, and I think it was created by User:Master of Puppets on 26 September 2011. I contacted him about it in December 2012 but he didn't respond. -Thibbs (talk) 06:35, 27 February 2013 (UTC)
Here's some evidence that it did a lot of good prior to the article's move to a new name. Sadly the log just calls it "an edit filter" instead of giving it a number. -Thibbs (talk) 08:10, 27 February 2013 (UTC)
A search for "Sonichu" in this list finds Filter 257. -- John of Reading (talk) 08:23, 27 February 2013 (UTC)
Filter 257 protects all articles, so is still working - however I think the problem is that one of the things it was protecting against was removed as it was causing too many false positives, and so is therefore not being stopped. The users who have been maintaining it are User:MuzeMike and User:Scottywong - you may wish to contact them. Black Kite (talk) 13:02, 27 February 2013 (UTC)
Ah, OK. Thanks for the lead. As a non-admin, I find the filter system to be rather extremely opaque, but that's probably intentional so vandals aren't able to easily circumvent the filters. Anyway thanks for the help from everyone. I'll contact those two editors as my next step. -Thibbs (talk) 14:18, 27 February 2013 (UTC)

YesY OK, problem resolved. Thanks again for all the help. -Thibbs (talk) 19:53, 27 February 2013 (UTC)

Template parameters (was: Pending changes bug)[edit]

Hi! Whenever I review an article for good article reviews and use the template

, my edits sometimes show up as pending, and the only way anyone can read them is by going onto the edit page and reading the raw wikicode. Examples can be seen here:

Any help would be appreciated, thanks! RetroLord 07:06, 27 February 2013 (UTC)

Unnamed template parameters must be explicitly named when the value of the parameter contains an = symbol as your signature does; otherwise the template parser thinks that everything preceding the = is the name of the parameter. I've fixed the two cases you linked above. Also, since this has nothing to do with WP:PC, I've changed the section title to avoid confusion. jcgoble3 (talk) 07:34, 27 February 2013 (UTC)

Sorting at List of national independence days[edit]

Folks we have received an e-mail at OTRS (ticket #2013022710006899) alerting us that the sorting in the table in this article appears to be broken. I just checked it too and I am seeing the same problem. Would someone please take a look at it? Thanks.--ukexpat (talk) 16:27, 27 February 2013 (UTC)

sortable tables have this kink, where all rows must have the same number of cells in them. 3 rows in this table were missing the last column. adding one more "|" to these 3 rows solved the problem. peace - קיפודנחש (aka kipod) (talk) 17:20, 27 February 2013 (UTC)
Thanks for the explanation and for fixing.--ukexpat (talk) 17:52, 27 February 2013 (UTC)

Secure login not working[edit]

Anyone else use secure log in? Right now it is not working. I log in, it says I am logged in, but click on anything and boom I am logged out. This only applies to secure log in. Did someone just break something? Apteva (talk) 18:51, 27 February 2013 (UTC)

And now it is working. I cleared all of my cookies and tried again. It was working with Chrome and not with Firefox. Apteva (talk) 19:11, 27 February 2013 (UTC)

Database issues[edit]

Database keeps locking. This is making new accounts for new users an issue. What's going on?—cyberpower ChatOffline 00:19, 28 February 2013 (UTC)

Gone now.—cyberpower ChatOffline 02:38, 28 February 2013 (UTC)

New Lua requests page[edit]

I've created a new request page for Lua scripts at Wikipedia:Lua requests. If you're a Lua developer, please watchlist it. If you'd like to make a request, please do that too. I'd also appreciate any help in improving the page itself, as it's very basic right now. Dcoetzee 23:09, 23 February 2013 (UTC)

A five days old post copied from #Scribunto deploy:
And we're deployed! Remember, it's suggested we don't convert everything all at once, so not so many templates need to be reverted if for some reason it has to be undeployed. Anomie 23:53, 18 February 2013 (UTC)
Is the risk of undeploying over? Many Lua requests would be for things which cannot be converted to templates. PrimeHunter (talk) 01:48, 24 February 2013 (UTC)
Yes. BJorsch (WMF) (talk) 19:08, 26 February 2013 (UTC)
Thanks. PrimeHunter (talk) 12:41, 28 February 2013 (UTC)

And I added it to Wikidata as we in Hungarian Wikipedia have had such a page for a week. :-) Bináris (talk) 09:20, 24 February 2013 (UTC)

  • Simplifying templates versus Lua-based rewrites: There are a lot of hunches as to what to do next, but we need more emphasis on where Lua script is actually needed. Many templates are already extremely fast, but the problem is rampant wp:Data hoarding with excessive clutter, using clutter-format templates to format the untrimmed clutter, which should have been deleted or put into subarticle "Climate and temperature records of Anytown". Even Lua becomes slow when buried in mountains of clutter. Hence, consider how the Template:Cite_quick (markup-based) can format citations at similar 6x speed like Template:Cite_web/lua, because {cite_quick} only handles 50 major cite parameters, while {cite_web/lua} checks the names of over 430 rare parameters and alternate alias spellings in Lua tables. In many cases, a template has become artificially complicated to handle rare options which could be formatted either by hand-coded text or by add-on templates, to mix and match with the original simple template, not to swell the template to have "everything including the kitchen sink". We need to explain to users the YAGNI principle ("You Aren't Gonna Need It") so they do not imagine every fast template must be tediously rewritten in Lua to run in 1/200th second, rather than 1/50th second during a 7-second edit-preview. The real danger is not from slow templates, but from creating a chaos of rewriting quick templates as slightly faster Lua, to hasten rather than delete the excess clutter and clutter-format templates from slow articles. Meanwhile, as Lua formats clutter 4x-7x times faster, perhaps people will think more clutter can be allowed since the page can be edited faster containing Lua-based templates. In many cases, so-called "slow templates" are actually an indication that an article has become a data hog, with too many rampant details clogging the page. When people request a feature with Lua, consider if other templates could be split, or simplified, to provide similar capabilities by reducing all the massive clutter, rather than processing clutter 4x-7x times faster. The best news is that Lua can be used to more-quickly format a page to proofread while clutter is being removed. -Wikid77 (talk) 10:32, 24 February 2013 (UTC)

Lua faster/slower but a helper[edit]

As we document more about when Lua script is much faster, or much slower compared to running wp:parser functions, it is important to remember Lua is a potential helper, even when much slower in some rare cases. Also, the speed comparisons confirm the excellent job done by the people who wrote the quick parser functions, such as {padleft:} which runs over 6x faster than Lua Module:String function 'sub' to extract the left-hand portion of a string. While there are many plans to write amazing new Lua functions, we might also encourage others to write more parser functions for rapid use inside the markup-based templates, and perhaps other templates would run faster than Lua for simple processing.

For months, we had suspected some templates would be much faster using rapid parser functions. Based on running extensive timing tests, during months on test2 wiki, I have confirmed the wp:parser function {padleft:} is over 6x faster than the Lua Module:String function 'sub'. The rapid speed of the parser functions, called inside markup-based templates, is what many of us had come to realize, so to keep the speed of {str_left}, I created the slower, Lua-based {str_left_long} to handle strings 6x slower but allowing 500 to thousands of characters long. Anyway, it is important to note when templates are faster, or easier, and beware when using Lua might make some templates run several times slower when using non-optimized Lua script. -Wikid77 (talk) 22:03, 26 February 2013 (UTC)

i doubt that what you measure is what you think you measure. i believe that when you time {{str_left_long}}, what you measure has much more to do with the overhead of #invoke: than with the performance of string manipulation in Lua.
there is little point replacing parser functions directly with a Lua call that does exactly the same: after all, the parser functions are written in native php, and the overhead of calling them is much lower than invoking a Lua module.
afaik, none of the replacements we did just replaced a parser function with a Lua function in a one to one replcement, and you are correct noting that it would be wrong doing so. peace - קיפודנחש (aka kipod) (talk) 23:23, 26 February 2013 (UTC)
  • Lua can be fast but #invoke is slow: It seems I should have said "invoking Lua modules" is 6x slower than parser functions. But now, the good news is that, if we find some ways to optimize the #invoke interface as 2x faster, than the small Lua-based templates, such as for string length or precision of numbers, could run about 2x faster. I have already timed the shell templates, which contain the #invoke markup, as running 20%-50% slower than invoking a Lua function directly as "{{#invoke:Module|myfunction|...}}". Otherwise, I am thinking:
  • "If a template runs 200 per second, invoking Lua cannot run faster."
So, I advise Lua modules should only be used when an entire set of templates runs slower than about "70 per second" such as the old markup-based wp:CS1 cite templates which ran at 14/second (with COinS metadata), while the Lua-based cite templates clock at 85/second. If people change small templates into Lua, then they might run 2x-5x slower than short markup-based templates. The other extreme would be "wp:Lua articles" as very fast to reformat, but people should consider the drawbacks of article pages written entirely in Lua. -Wikid77 (talk) 13:50, 28 February 2013 (UTC)

Redlinks that still point to other languages[edit]

Would it be possible to create a page with nothing but links to other language Wikipedias and have it still appear as a red link? Or, at the very least, have some text saying "English Wikipedia does not have an article with this title; however, there is an article in Portuguese"? — Preceding unsigned comment added by Ego White Tray (talkcontribs) 13:31, 27 February 2013 (UTC)

Not sure if this is relevant to your question, but if you go to a non-existing page in the Swedish Wikipedia, such as sv:Sghrgajr, the instructions in the box will give you the option to check a select number of other wikipedias for the same word (English, German, Danish, Finnish and both versions of Norwegian). As it only works for articles with the exact same title, it would often work for proper nouns (people, places, businesses and so on) but wouldn't work for cases where the equivalent article would have a completely different title. --Hegvald (talk) 14:53, 27 February 2013 (UTC)
More likely useful that the initial proposal would be to include the interlanguage links from Wikidata. It might need a flag there to say that en.wp doesn't have this topic yet, but would like it at $title (I'll have a hunt round Wikidata to find where to propose this). Then when someone looks for $title they would see the links to the other Wikipedias' articles on that topic; for example we'd connect Lăpuș Mountains to ro:Munții Lăpușului and hu:Lápos-hegység. Failing that the Swedish system would seem like a good one, although we'd probably choose a different selection of languages (French, Spanish, German, Mandarin, Japanese and Hindi I think are the most common second languages of English speakers?) Thryduulf (talk) 16:41, 27 February 2013 (UTC)
See d:Wikidata:Project chat#Interwiki links for non-existent articles for the question I've asked of the folks at Wikidata. Thryduulf (talk) 17:14, 27 February 2013 (UTC)
  • Redlinks are your deleted article but create pre-stubs: I would want to keep a redlink to "Mein Garage Band" as a redlink, in any language, and still read when the page was deleted. Perhaps a better plan is to accept "pre-stub pages" (called: wp:prestub) which contain a smart template to link advanced search-engine searches (such as person "born on" or organization "established") for when today's search-engine parameters will some day have better "semantic search" to find pages about someone's birthdate (or education) or when an organization was established. Remember today's search engines are so primitive, linguistically, they cannot find "2+2=4" although they can index the contents of 50(?) billion webpages but unable to switch to "math mode" to know "+...=" means match an equation, Sherlock. Anyway, if we started a concept of having pre-stub pages, with a box of formal search requests for "born/birthdate" or "origin of word" then perhaps the major search-engines would go "Eureka" to see what answers we need when researching a topic. Meanwhile, a {prestub} template could also link to other-language Wikipedias, perhaps by guessing likely "recommended other language" from syllables/alphabet in the article name (such as a huge, fast Lua script module to split words and guess related languages). However, long-term, it seems plain wrong that song and 1952 film "Einmal am Rhein" (":de", "Once on the Rhine", see googled) has not existed in any other-language Wikipedia, so that is why pre-stub articles should be quickly created, to begin to auto-inform readers how the world knows a lot about a particular title, and no longer a redlink but rather a blue prestub page which does not meet stub format. -Wikid77 (talk) 11:34, 28 February 2013 (UTC)
    • Other than using lua templates and a pseudonamespace rather than transcluding wikidata information onto the mediawiki message displayed on a non-existent page is there any other way that your proposal differs from mine? This is not meant to be denigrating of your idea, I'm just trying to understand it (I've not even attempted to understand lua yet). Thryduulf (talk) 14:55, 28 February 2013 (UTC)

User talk page js error[edit]

Hi again, I have been encountering a Javascript-related error for months now. I think I first noticed it in late December. The problem is that very frequently when I try to access an editor's talk page, my browser will hang. I notice this most often when trying to give IP editors vandalism warnings, but it happens quite frequently for other user talk pages as well. I posted this question at the Help Desk and after suggesting things like clearing my cache, setting my cache to zero, and turning off javascript, they directed me here. The did provide an important clue, though, because when I turn off javascript then the problem vanishes immediately. So there is clearly some javascript problem that is related only to user talk pages. No article makes my browser hang, no user page makes it hang (as far as I know), and no other area of Wikipedia makes it hang. It's also fine in the rest of the web including talk pages at other wikis like Wikia. It's just the user talk pages at Wikipedia that cause the problem. For reference, my browser is Firefox v.3.6.28, I'm running it on a Mac OSX v.10.4.11, and I don't have any gadgets or special skins on my Wikipedia account (everything is the normal default). Can anyone tell me what might be causing this? -Thibbs (talk) 21:03, 27 February 2013 (UTC)

I would suggest updating your Firefox. Werieth (talk) 21:15, 27 February 2013 (UTC)
Yes, the current version is 18.0.2. Versions 3.6 and earlier only checked for minor updates automatically, not major ones. 3.6.x ceased being supported almost a year ago; see Firefox release history. jcgoble3 (talk) 21:23, 27 February 2013 (UTC)
The current release is actually Firefox 19.0 - I was given the upgrade automatically at about 11:45 UTC today - just over twelve hours ago. --Redrose64 (talk) 00:11, 28 February 2013 (UTC)
Well thanks, but that's really just treating the symptom and not the disease. Firefox v.3.6.28 is the last version of the browser that is compatible with Mac OSX v.10.4.11, and that Mac OS is the last free upgrade I'm entitled to through Apple. So what this suggestion equates to is "buy a new computer". I actually do have a new computer that I just bought (a Win 8 machine), so this isn't really a big problem for me, but I'd like to understand why it's happening in the first place. Does anyone here have any ideas about why? -Thibbs (talk) 21:28, 27 February 2013 (UTC)
Well, if you're not using any non-standard Javascript, and nobody else is experiencing a similar problem (which you would expect if it's some sort of issue in vanilla Wikipedia), not really. Anyway, the old browser might in fact be the disease and not a symptom; they do get updated for a reason. :P Writ Keeper 21:44, 27 February 2013 (UTC)
While I agree that it could be useful to find out why it's started happening, yes, it is reasonable for the solution to be "get a new computer". You're running an end-of-lifed browser on an end-of-lifed OS. Now, it's not exactly like you're running Netscape on Windows 95, but the general principle of "code to what is being supported rather than obsolete software" still applies. --Golbez (talk) 21:49, 27 February 2013 (UTC)
(edit conflict) I guess I'm not really concerned with how to fix it any more since my new computer works fine, but I'm just curious what sort of user-talk-page-specific javascript calls exist. How is a user talk page handled any differently from a client-server perspective than an article page? If all Wikipedia pages caused the browser to hang then I'd assume it was something related to Wikipedia, but it intrigues me how specific the situation is. This is clearly something related specifically to the user talk pages. What sort of javascript does the user talk page call? -Thibbs (talk) 22:06, 27 February 2013 (UTC)
Can you try an alternate browser and let us know if the problem persists? Werieth (talk) 22:00, 27 February 2013 (UTC)
I guess that would be my next step. Right now the computer is choked full of personal projects and other mess so I'll need to offload some of that to an external drive to clear up the room for another browser. If I get inspired I might make an experiment like this, but otherwise it's no big deal. I'm mainly just curious. -Thibbs (talk) 22:06, 27 February 2013 (UTC)
i wonder if this has anything to do with any gadget you are using. presuming the problem is reproducible (i.e., once you found a specific talkpage that cause it, it happens on this page with some consistency), can you try to see if it happens when you are logged out? (or better yet, if you have another account for testing, like many people do, when you are logged into the other account)?
(the rest of the response assumes that the problem disappears when logged out:)
if you can't make the problem appear when logged out, good chances are that a specific gadget you use causes it (i don't think you have custom JS, so it has to be a gadget). the 3 most suspicious ones are wikEd, navpopups and twinkle, but all the rest are not automatically innocent. try to turn gadgets off one by one, and see which of them is the culprit.
peace - קיפודנחש (aka kipod) (talk) 22:26, 27 February 2013 (UTC)
It doesn't happen when I'm logged out so it does seem to be specific to my account, but the only gadgets I have running are the defaults: 1) "Ask a question" feature, 2) Reference Tooltips, 3)Hotcat, 4)Form for filing disputes at the dispute resolution noticeboard, 5) CharInsert, and 6)Add a "Sandbox" link to the personal toolbar area. I still get the error when I turn them all off. -Thibbs (talk) 23:19, 27 February 2013 (UTC)
Try changing your skin, I suggest monobook. Werieth (talk) 23:30, 27 February 2013 (UTC)
I usually use Vector, but I switched to monobook and the problem still crops up. -Thibbs (talk) 23:39, 27 February 2013 (UTC)
I don't want to be a bother, by the way. If this problem is difficult to diagnose then it's probably not worth it. I just figured it would be good to have a record of the issue somewhere online. For the record, the easiest work-around I've found is simply to turn off javascript, hit refresh on the user talk page, and then turn it back on after you're done. -Thibbs (talk) 23:42, 27 February 2013 (UTC)
  • Compare the view-source of the generated HTML: There might be something obvious in the generated HTML (when logged out), or a different CSS sub-sub-sub-class name, or such, which interferes with JavaScript. I am also trying to find ways to support older browsers, in this case, IE7 (2009) where the page title disappears, or IE8 or IE9 (with Windows 7), after I saw recent stats that the "world at large" still uses MSIE in 43% of usage (November 2012), but WP lockups have driven away over half of IE browser users in recent years, which would explain why it seems Wikipedia editor activity is declining by 2% a year. If we can get Wikipedia to again support the MSIE browsers, perhaps as a dedicated "Explorer skin", then I think worldwide usage of Wikipedia would definitely increase. Anyway, find all the bugs and lockups in the current Wikipedia interface, and we might be able to get many other browsers to work as well. It is amazing that in 2009, Wikipedia supported "every browser in the world" and now it just craters, spits garbage, and locks up browsers. Any insights people might offer about the lockups, could help to salvage Wikipedia's poor user-support fiasco. -Wikid77 (talk) 23:49, 27 February 2013 (UTC)
    • I didn't realize that. If your idea regarding declining activity is correct (and from the usage stats it sounds like it very well might be), then this is an important issue indeed. I'm not very familiar with html, sadly, but I'll conduct experiments and alert you and report back here if I find anything. -Thibbs (talk) 15:09, 28 February 2013 (UTC)

Feature articles and interwiki[edit]

Hello. Is there a way to find all feature articles of English Wikipedia that have an article in French wikipedia? Xaris333 (talk) 14:04, 28 February 2013 (UTC)

Hi, and thanks for your question. I am not sure what you mean, do you mean the French versions of all featured articles in the English WP? Nearly all of the English FAs have French versions (they might not be FAs there, though)
Or do you mean articles that are FAs in both English and French WP? Zaminamina Eh Eh Waka Waka Eh Eh 19:05, 28 February 2013 (UTC)

The first. If the English FAs have French versions. How can i find those which has french version and those shich not have? Is there a simply way? Xaris333 (talk) 21:14, 28 February 2013 (UTC)

Assuming all English FAs are properly tagged as such on the French versions, you could do a search of all articles on the French wiki for any that use {{Lien AdQ|en}}, which is the template used on the French wiki to generate the featured star next to the English interwiki link. However, I don't know how reliable that will be; if that template is handled automatically by interwiki bots, then it should work without a problem, but if it has to be applied to articles manually, then it's likely to be hit or miss. jcgoble3 (talk) 21:53, 28 February 2013 (UTC)

Thxs! Xaris333 (talk) 00:37, 1 March 2013 (UTC)

Watchlist and bots[edit]

There is a preference that allows to see or hide edits from bot in the watchlist. In either case, it applies to everything. Is it possible to make an exception, hide bot edits from all pages except for a certain one, where I may be interested to know of new bot edits? Cambalachero (talk) 14:49, 28 February 2013 (UTC)

Wikidata links broken: MediaWiki regression?[edit]

Has anybody else noticed that internal links to Wikidata broke in the last couple of hours? d:Q5561477 should go to this page but instead it shows as a redlink to this page. Similarly d:Special:ItemByTitle/enwiki/Logopolis, which should also go to this page, instead shows as a redlink to this page. All other d: links also broken, so I'm guessing that there has been a MediaWiki regression. --Redrose64 (talk) 18:32, 28 February 2013 (UTC)

Of course, d: ought to come back a.s.a.p, but just notice that wikidata:Wikidata:Introduction still works. BTW, I can’t access interwiki on m:Toolserver – the table exists and is described, but is empty. Incnis Mrsi (talk) 19:00, 28 February 2013 (UTC)
Now working again. --Redrose64 (talk) 19:14, 28 February 2013 (UTC)

Deletion statistics[edit]

I am asking this question here because I get good results and there is no other approp forum.

I want to get statistics on article deletions such as:

  • Daily or weekly number of speedy deletions
  • Daily or weekly number of PRODS
  • Daily or weekly number of AfDs results, i.e. delete, keep, merge, no consensus (Scotty Wong does then for individual users)
  • Daily or weekly number of new articles (this may already be done.)

Weekly may be often enough. This info will help with policy development and is useful for Wikipedia:WikiProject Statistics (a bit inactive though). Thanks. 23:46, 28 February 2013 (UTC) — Preceding unsigned comment added by Alan Liefting (talkcontribs)

You could ask at Wikipedia talk:Database reports.
Wavelength (talk) 00:05, 1 March 2013 (UTC)

Strange-looking infobox[edit]

Yes check.svg Done

Have a look here, between the flag and the picture. Is this a template issue? pablo 16:35, 28 February 2013 (UTC)

No, it's a bad edit. |pushpin_map=Malaysia peninsula was altered to |pushpin_map=Malay peninsula - {{location map Malaysia peninsula}} exists, but {{location map Malay peninsula}} does not. --Redrose64 (talk) 16:47, 28 February 2013 (UTC)
Cool, thanks. pablo 10:46, 1 March 2013 (UTC)

Watchlist email options[edit]

Is it possible to receive email about only a subset of our watchlist? I have a small group of pages in my watchlist I care enough about to want to receive emails, but most pages I'd rather not be bothered. Is there a way to that that up? Do people even use email notification? Seems like it would get too chatty with even moderately big watchlists. I thought about trying to use email filters to accomplish this, but it seems like kind of pain considering the set of interested pages will change over time. Thanks for any ideas. Silas Ropac (talk) 00:54, 1 March 2013 (UTC)

A workaround would be to link to all the pages on a subpage of yours, like User:Silas Ropac/watch, which then appears watchlist-like via the "Related changes". Now, on RecentChangesLinked, if the page exists, you will see a link like "Feed-icon.svg Atom", which is a type of feed (see Atom). You can then put that link into a feed digest or realtime feed service that will email you any changes. This would require your sublist to be public, though. A more aggressive solution is to put your whole watchlist on a feed service that would let you filter for specific changes for email notification, but I don't know of any free services with that much granularity. --Splarka (rant) 09:03, 1 March 2013 (UTC)
Thanks, so even without feeds you could put your most important pages in your real watchlist, and set that to send emails, and then put your less important pages on a subpage, and use related changes for yet. I will try it. Silas Ropac (talk) 13:13, 1 March 2013 (UTC)
  • Yes. Note that watchlists always add talk pages and nontalk pages in pairs, regardless of whether you watch one or the other. But if you use the above Special:RecentChangesLinked method, only the actual links are monitored. So if you use that method, you should probably take care to add separate links to both talk and nontalk pages, to be sure that you don't miss significant talkpage posts. Also, you would have to manually add or amend links if a page is moved, whereas watchlists add new pages automatically after a move. — Richardguk (talk) 14:23, 1 March 2013 (UTC)
Can an editor not just set up a second account, and have the short selected watchlist on that account, set to e-mail? or is that against some rule? - I'm asking because I was just about to do that. - Arjayay (talk) 13:20, 1 March 2013 (UTC)
A second account may be legitimate; Redrose64a (talk · contribs) is an example. Such accounts have their own watchlists. --Redrose64 (talk) 14:47, 1 March 2013 (UTC)
Thanks - it seems easier and overcomes all the problems explained above - Arjayay (talk) 15:20, 1 March 2013 (UTC)
Just found this which says: "... one may be tempted to log in under a different username just to specify a different (typically smaller) watchlist for e-mail notification. However, after viewing a page under one username, one would have to clear the notification flag for that page for the other username too." That sounds like a pain, because today when I get notification email and click on it, my main account is already logged on. So I'd have to log out or open up another browser just to clear the flag. In addition to the "talk" and "move" problems Richardguk mentioned, subpages don't work with the little star icon and the "watch this page" checkbox. So yeah I'm down on both workarounds at the moment. Maybe Echo will help, although at first I think they are targeting new and casual users, power users comes later. Silas Ropac (talk) 15:56, 1 March 2013 (UTC)
See also gerrit:16419 (Watchlist grouping). Helder 16:31, 1 March 2013 (UTC)
That's really cool. Not obvious it includes per-group email settings, but that's probably an easy addition once the infrastructure is in place. And I noticed Echo includes email options of instant, daily digest, and weekly digest. Doesn't sound like Echo knows about watchlists yet, but eventually you could imagine being able to get instant updates from one watchlist group, while getting daily digests from the rest. That sounds ideal really. Silas Ropac (talk) 17:58, 1 March 2013 (UTC)


Toolserver appears to be having issues at the moment. ​—DoRD (talk)​ 12:54, 1 March 2013 (UTC)

Geohack down?[edit]

since last night I have been trying to access coordinates of places but either the server takes too long to respond or there is an internal error. Anyone else having the same problem? Simply south...... catching SNOWballs for just 6 years 12:27, 1 March 2013 (UTC)

Edit count at Special:Contributions[edit]

The edit count link at Special:Contributions that links to the Toolserver is down. I suggest it is replaced, unless this is only temporary. --Eleassar my talk 12:48, 1 March 2013 (UTC)

Templates for special pages[edit]


Where are the templates that produce a specified link to a WP:SPECIAL page? -DePiep (talk) 12:50, 1 March 2013 (UTC)

found: it's like [[Special:PrefixIndex/HMS|label]]. -DePiep (talk) 13:31, 1 March 2013 (UTC)

LaTeX Export[edit]

[25] — Preceding unsigned comment added by (talk) 15:53, 1 March 2013 (UTC)

Template:Course assignment[edit]

What did I do wrong with {{course assignment}} so that the word "in" is added onto and linked with "course assignment" as currently shown here when one specifies a |term= parameter? Biosthmors (talk) 19:25, 1 March 2013 (UTC)

I fixed it for you. You didn't have a space after the wikilink. Ryan Vesey 19:42, 1 March 2013 (UTC)
Thanks. But might there be an easy way to fix it without having a period before the space in the default {{course assignment}} version? Biosthmors (talk) 19:59, 1 March 2013 (UTC)
Oh sorry, I didn't even realize I did that. I've made another fix. Ryan Vesey 20:03, 1 March 2013 (UTC)
Awesome. Both versions look good now. Thanks! Biosthmors (talk) 20:05, 1 March 2013 (UTC)

Conditional table in template is adding empty lines[edit]

Hi there,

I have a markup question about templates. I created a template called {{Contact Information}} which I'd like to feed subsets of data in order to create variable tables from row-length 0 to maximum.

The template markup looks like this:

{| border="1" cellspacing="0" cellpadding="3"
{{#if:{{{FirstName|}}}| {{!}}-
{{!}}{{{FirstName}}} {{{LastName}}} }}
{{#if:{{{JobTitle|}}}| {{!}}-
{{!}}{{{JobTitle}}} }}
{{#if:{{{Department|}}}| {{!}}-
{{!}}{{{Department}}} }}
{{#if:{{{Address|}}}| {{!}}-
{{!}}{{{Address}}} }}
{{#if:{{{OfficePhone|}}}| {{!}}-
{{!}}{{{OfficePhone}}} }}
{{#if:{{{CellPhone|}}}| {{!}}-
{{!}}{{{CellPhone}}} }}
{{#if:{{{Fax|}}}| {{!}}-
{{!}}{{{Fax}}} }}
{{#if:{{{Nextel|}}}| {{!}}-
{{!}}{{{Nextel}}} }}
{{#if:{{{Email|}}}| {{!}}-
{{!}}[mailto:{{{Email}}} {{{Email}}}] }}
{{#if:{{{Website|}}}| {{!}}-
{{!}}{{{Website}}} }}

I pared it down to its simplest elements because I am having this issue. When the template is fed parameters, any parameter that does not exist (results not true) returns <p><BR></p> which causes a blank line to appear. For example, if I call the template like {{Contact Information|CellPhone=(494) 234-2039}} I'd get.

<table border="1" cellspacing="0" cellpadding="3">
<td>(494) 234-2039
<p><br />
</p><p><br />
<p><br />

I put in a default for the #if statement to see what it would do. I get the following behavior:

{| border="1" cellspacing="0" cellpadding="3"
{{#if:{{{FirstName|}}}| {{!}}-
{{!}}{{{FirstName}}} {{{LastName}}}| default }}
{{#if:{{{JobTitle|}}}| {{!}}-
{{!}}{{{JobTitle}}}| default }}
{{#if:{{{Department|}}}| {{!}}-
{{!}}{{{Department}}}| default }}
{{#if:{{{Address|}}}| {{!}}-
{{!}}{{{Address}}}| default }}
{{#if:{{{OfficePhone|}}}| {{!}}-
{{!}}{{{OfficePhone}}}| default }}
{{#if:{{{CellPhone|}}}| {{!}}-
{{!}}{{{CellPhone}}}| default }}
{{#if:{{{Fax|}}}| {{!}}-
{{!}}{{{Fax}}}| default }}
{{#if:{{{Nextel|}}}| {{!}}-
{{!}}{{{Nextel}}}| default }}
{{#if:{{{Email|}}}| {{!}}-
{{!}}[mailto:{{{Email}}} {{{Email}}}]| default }}
{{#if:{{{Website|}}}| {{!}}-
{{!}}{{{Website}}}| default }}

Results in:

<table border="1" cellspacing="0" cellpadding="3">
default default default default default
<td>(494) 234-2039
<p>default default default default
<p><br />

Any helpful advice on what I'm doing wrong is appreciated! — Preceding unsigned comment added by (talk) 23:33, 1 March 2013 (UTC)

This is the same problem described at Wikipedia:Village_pump_(technical)/Archive_108#Extra_breaks_in_template; the issue is that you're leaving newlines outside of the #if parser functions, so those newlines are always transcluded, and those newlines convert into <p>...</p> and <br /> tags in the final HTML. The solution is to move the newlines inside the parser functions such that the end of one PF and the beginning of the next butt up against each other on the same line like this: }}{{#if:. However, parser functions strip all leading and trailing whitespace from each parameter, so you must trick it into including the newline; one method of doing so is described by PartTimeGnome at the bottom of the archived discussion. jcgoble3 (talk) 23:49, 1 March 2013 (UTC)

Database error[edit]

The first page I visit is often the watchlist and the first time log in produces this error (happening for a couple of days now) which vanishes on repeat attempts. Posting this in any case. Shyamal (talk) 01:35, 2 March 2013 (UTC)

A database error has occurred. Did you forget to run maintenance/update.php after upgrading? See:
Query: UPDATE `user` SET user_touched = '20130302012542' WHERE user_id = 'XXXX' AND (user_touched < '20130302012542')
Function: User::invalidateCache
Error: 1205 Lock wait timeout exceeded; try restarting transaction (

wikidata and non-symmetrical interwiki links[edit]

Hi all,

How does wikidata deal with non-symmetrical interwiki links?

For example, look at these 3 cases:

1) 2 different articles on should go to the same article on articleA ---> articleX articleB ---> articleX

2) An article on is represented by 2 different articles on articleA ---> articleX AND articleY

3) A on should link to B on, which in turn should link to C on articleA ---> articleB articleB ---> articleC

All these cases can happen under certain circumstances, and I'm wondering how wikidata deals with them.

Thanks! Azylber (talk) 07:11, 26 February 2013 (UTC)

The same way that the interwiki bots have run for ages, namely by treating all interwikis as if they are one-to-one and symmetric. One can still maintain non-symmetric situations by locally including interwiki links, but they can't be included through wikidata at present. Dragons flight (talk) 09:07, 26 February 2013 (UTC)
Have a look at Doctor (Doctor Who)#External links - that has one H:ILL, to de:Doctor Who#Der Doktor, but the link in the opposite direction goes to en:Doctor Who. Compare d:Special:ItemByTitle/enwiki/Doctor (Doctor Who) with d:Special:ItemByTitle/enwiki/Doctor Who - only one has a dewiki entry. --Redrose64 (talk) 14:04, 26 February 2013 (UTC)
DE is typically the worst offender for conflating two different topics (a fictional person and the show in the above; a video game and a video game series, etc.). I found it was common across many separate topic groupings. In almost all cases, most links don't need to be anything more than one-to-one linkage, and I would personally strongly advise against introducing anchor links locally or otherwise, as that defeats the purpose of Wikidata in multiple ways. --Izno (talk) 14:16, 26 February 2013 (UTC)
  • Also some should link to a redirect, or disambiguation or crosslink page: In many cases, a word in one language should map to several different words in another language, and similarly, a film title might be translated to a different common phrase in another language, which often already has been used for an unrelated topic. This is a somewhat complex (or confusing) subject, which many editors would not fully follow, but part of the problem can be resolved by crosslink pages which "distinguish" related concepts, rather than "disambiguate" the same title names. Even some examples can be difficult to explain, so perhaps consider a famous book where several films were made under different titles, but the central concept is they are film versions of the non-matching book title. A related example is where separate articles for each major character from Gone with the Wind would all map to a single other-language article as "List of characters in <book>" although it might be preferable to use one-to-one mapping by redirect titles, so have each character's article map to a separate character-name redirect among the other-language titles (and those redirects would then each point into "List of characters..." in that language). I am afraid that I am not clarifying the issues, but perhaps I have emphasized how a title in one language could refer to a single article, while in another language it would be a redirect into a list article, or perhaps to a disambiguation or wp:crosslink page where a word/title in one language is covered by multiple words/titles to describe the same concept in another language. I wish I could explain these issues better. Perhaps here's the thing: a title not only designates a particular item, but sometimes a title names a concept which is covered by multiple articles in another language. To point to multiple titles, then have an interwiki link to a crosslink-page which lists all those different titles which explain the related concept from the other language (which has only one title there). Or, perhaps I should take a nap, then re-read this paragraph 3 times. :) -Wikid77 (talk) 09:55, 26 February 2013 (UTC)
Ok. Thanks for the answers. Azylber (talk) 15:42, 2 March 2013 (UTC)

.js file not loading[edit]


This just started happening - none of the scripts in my .js file and none of the gadgets I have installed are running. I have PURGED and BYPASSed multiple times to no avail.--ukexpat (talk) 21:04, 26 February 2013 (UTC)

most probably this is related to the difficulties is experiencing. you might also occasionally see WP in "plain text" mode, due to failure(s) to pull CSS data, and similar such phenomena. peace - קיפודנחש (aka kipod) (talk) 21:19, 26 February 2013 (UTC)
OK thanks - wasn't aware that bits was in difficulties.--ukexpat (talk) 21:28, 26 February 2013 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────Is bits still having problems? --ukexpat (talk) 02:17, 27 February 2013 (UTC)

Anyone else having this problem? My .js file is still not loading. Thanks.--ukexpat (talk) 13:10, 27 February 2013 (UTC)
I'm also having this issue. I'm no tech and I have no idea what's going on. Basalisk inspect damageberate 14:36, 27 February 2013 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────Anyone? Bueller?--ukexpat (talk) 20:15, 27 February 2013 (UTC)

Hello, I found several of my scripts stopped working, it turned out it was because I had a <!-- --> style comment. Try replacing any comments in that fashion with // at the start of the line. --Closedmouth (talk) 14:13, 28 February 2013 (UTC)
Thanks for the suggestion but it's not that. None of my .js scripts are running.--ukexpat (talk) 18:45, 28 February 2013 (UTC)
Do you have a javascript debuger in your browser? Werieth (talk) 19:04, 28 February 2013 (UTC)
wild guess: is it possible JS may be disabled on your browser? peace - קיפודנחש (aka kipod) (talk) 19:52, 28 February 2013 (UTC)
I've had the identical issue for at least 2 days now. I have the same problems on a different browser and on my phone. Basalisk inspect damageberate 20:08, 28 February 2013 (UTC)
Nope, javascript is enabled. I see the same problem in FF v19.0 and Chrome v25.0.1364. Not familiar with debugging, how do I do that please? Thanks for the help.--ukexpat (talk) 20:32, 28 February 2013 (UTC)
In Firefox it is Tools > Web developer > Error Console Werieth (talk) 20:40, 28 February 2013 (UTC)

Fixed - there was indeed a stray <!-- --> which I corrected to // - so all is well. Question is, the <!-- --> had been there for a long time, so why would it break the script loading just a few days ago? In any event, thanks for the help. We now return you to our scheduled programming (no pun intended, well maybe a bit). --ukexpat (talk) 16:42, 1 March 2013 (UTC)

This fix also worked for me. Happy days. Basalisk inspect damageberate 20:12, 2 March 2013 (UTC)


If a logged-in system is left unattended, does it get automatically logged out after a time? If so, how long? JohnCD (talk) 09:59, 1 March 2013 (UTC)

What are you logged in to? That is, do you mean Unix login, a Wikipedia login, or something else? --Redrose64 (talk) 13:35, 1 March 2013 (UTC)
Wikipedia login. I have someone on my talk page complaining that "I left my computer idling for quite a bit of time when I came back, I was unable to edit, because it says I did not have permissions. Why has this happened?" and it occurred to me that he had perhaps been automatically logged out. JohnCD (talk) 16:11, 1 March 2013 (UTC)
I have noticed that it varies. There is a box that you can click to stay logged in longer, which I never use, but sometimes I get logged out much sooner than other times. This is a SUL account, but sometimes I get logged out on other wiki's and sometimes not, even though I am still logged in here on this wiki. Apteva (talk) 03:14, 2 March 2013 (UTC)
Such things have come up before (see for example Wikipedia:Village pump (technical)/Archive 96#Being logged out unexpectedly, there are several others). Usually it's not that the user has actually been logged out, but the edit session ID has expired - I believe this happens if you do nothing for 30 minutes. When you try for Save page on an expired session, you get a red message like "Sorry! We could not process your edit due to a loss of session data. Please try again. If it still does not work, try logging out and logging back in."
You can tell if you really are logged out by going for any link and checking at upper right; if it only has the "Create account" and "Log in" links, you are logged out - if there are links to your user page, talk, sandbox, prefs etc. you're still logged in. --Redrose64 (talk) 22:50, 2 March 2013 (UTC)

DDA template not working on a page[edit]

James George Scott (edit | talk | history | protect | delete | links | watch | logs | views)

I've examined and examined again and I'm 'sure that the {{dda}} template is right, but for some reason nothing at all shows up in either box. Can someone diagnose why? hbdragon88 (talk) 23:52, 2 March 2013 (UTC)

Fixed by [26], but I don't know why. I deleted two spaces which were normal spaces according to my tools which usually reveal stray Unicode characters. I "replaced" each deleted space by a space. The page became 3 bytes smaller according to the page history (one of the bytes was because I also changed a number). Can somebody tell what really happened by comparing the revisions? PrimeHunter (talk) 01:02, 3 March 2013 (UTC)
There were non-breaking spaces (or something similar) after some of the infobox parameter names. Peter James (talk) 01:11, 3 March 2013 (UTC)
Thanks. In Google Chrome I can see it was non-breaking spaces. Apparently Firefox copy-pastes a non-breaking space as a normal space. I would call that a bug. PrimeHunter (talk) 01:31, 3 March 2013 (UTC) shows this has been a Firefox issue since at least 2006. PrimeHunter (talk) 01:39, 3 March 2013 (UTC)
Special:Export keeps them as non-breaking spaces. Peter James (talk) 01:49, 3 March 2013 (UTC)
OK. Just to clarify if there is any doubt, I mean this is a Firefox bug and not a MediaWiki bug. PrimeHunter (talk) 02:13, 3 March 2013 (UTC)

Vanishing blocks[edit]

As you can see in this bot edit at AIV, the IP I reported was supposed to have been blocked for one month. When you check the block log, though, nothing shows up. I saw this about a week ago with another user (registered), though I can't recall the name. Has anyone else been noticing this? Evanh2008 (talk|contribs) 13:08, 24 February 2013 (UTC)

The other user I experienced this with was also an anon, I even left a message at the blocking admin's talk page within five minutes of it happening. Evanh2008 (talk|contribs) 13:12, 24 February 2013 (UTC)
that would be due to range blocks. Werieth (talk) 13:37, 24 February 2013 (UTC)
Ah, yes. That would explain it. I am dumb! Carry on. :) Evanh2008 (talk|contribs) 13:48, 24 February 2013 (UTC)
Actually, no. That does explain the more recent situation, but I most definitely saw a block listed on the contributions page for the earlier IP, which vanished after I reloaded the page. Would a range block explain that? Evanh2008 (talk|contribs) 08:56, 26 February 2013 (UTC)
The block for is visible to me, both in the block log and at Special:Contributions/ - David Biddulph (talk) 09:07, 26 February 2013 (UTC)
Yeah, I should have clarified -- EdJohnston blocked the IP after the block by Timotheus Canens disappeared. Evanh2008 (talk|contribs) 02:44, 27 February 2013 (UTC)
Gonna bump this so it doesn't get archived. Evanh2008 (talk|contribs) 16:03, 3 March 2013 (UTC)


When I cut and paste the date from the history to sign something, I get a date that looks like 2013, but our RMCD bot can not recognize. So I type in 2013 and it works. Can anyone tell what is wrong with the date in history that it can not be used? Is it a special font, or are there hidden characters, or are they some kind of funny unicode that is somehow different from what the bot is looking for but it still looks like 2013?[27] Basically, why are we using unicode numbers in history but ascii characters when using normal tilda signatures? Or actually, why are we using unicode characters in history? Apteva (talk) 02:50, 2 March 2013 (UTC)

  • Yes, I got stumped by that: by playing around with it, I found it was the space character after the [year] that isn't a traditional space. It would be useful to know how to treat it with using regexes. -- Ohconfucius ping / poke 03:19, 2 March 2013 (UTC)
    • Interesting. I did not think that I changed the space, only 2013 from 2013. You are right. History has two spaces after 2013 and before the Username. The first is unicode, the second ascii. You know what? It is not the date that is the problem, it is the invisible space after the date that is unicode. Like this one: 2 March 2013 ← Apteva (talk) 04:05, 2 March 2013 (UTC)
      • The actual issue is that there is an invisible left-to-right mark at the end of the date. See this breakdown of a copy-and-paste from the page history. jcgoble3 (talk) 04:42, 2 March 2013 (UTC)
      • So is this use of an invisible left-to-right mark a bug or a feature? This is pure English, there is no Russian or Arabic to be found here. So why is the character lurking there? If it's a bug, is there a bugzilla entry for it yet? Wbm1058 (talk) 13:45, 3 March 2013 (UTC)
Though our articles are in English, there are a few users on the English Wikipedia with Arabic or other RTL usernames. Usernames appear next to the date on history pages. For something so trivial, the developers probably did not think it worth checking if the left-to-right mark is necessary for each instance. (And not being there consistently could cause more confusion.) – PartTimeGnome (talk | contribs) 16:25, 3 March 2013 (UTC)
Oh, I see – WP:NLS. – Wbm1058 (talk) 17:40, 3 March 2013 (UTC)
  • Strange space characters new in history-log output: Some weeks ago, I noticed those new, strange-space separator characters, between columns, in the output from running a history-log display. Formerly, the history-log only generated simple ASCII spaces, but evidentally, the peculiar Unicode separator character (this: '←') was added into the MediaWiki Wikipedia interface (surprise!), along with the other gradual, peculiar changes which have killed the support for the world's older Internet Explorer browsers, locking upon incompatible CSS class contents, which formerly worked fine in 2009. The entire interface to Wikipedia seems to be in a state of perpetual chaos, so always expect new problems whenever displaying any screen in Wikipedia. Remember "forewarned is forearmed" and perhaps try to use offline text editors to edit articles, then hold your breath, and copy the page into Wikipedia to see the newest surprises. Even the best of personnel cannot perform well in a chaotic environment, and someone needs to grab the reins. -Wikid77 (talk) 07:11, 2 March 2013 (UTC)

Module display problem[edit]

I just updated a big module. The wikitext is fine (I downloaded it again, and it matches my local file; also the module runs fine). But looking at Module:Convertdata shows corruption. The raw wikitext has lines 18 to 21 inclusive:

local SIprefixes = {
    -- The prefix field is what the prefix should be.
    ['Y'] = { exponent = 24, prefix = 'Y' , name = 'yotta', name_us = 'yotta' },
    ['Z'] = { exponent = 21, prefix = 'Z' , name = 'zetta', name_us = 'zetta' },

But the module page shows these four lines (the 'GREEK SMALL LETTER MU' is from line 33):

local SIprefixes = {
    -- The prefix field is what the prefix should be.
-- key = 'GREEK SMALL LETTER MU' (U+03BC) utf-8 CE BC
    ['µ'] = { exponent = -6, prefix = 'µ' , name = 'micro', name_us = 'micro' },  -- key = 'MICRO SIGN' (U+00B5) utf-8 C2 B5

There are other corruptions, and the total display is only 72 lines (the wikitext has 8869 lines<groan>). It's probably an ephemeral issue and will go away in due course, but I thought I should mention it to alert other Lua coders. The usual refresh/purge did not achieve anything. (See Module talk:Convert for background if interested.) Johnuniq (talk) 10:23, 2 March 2013 (UTC)

Reported as bugzilla:45669 Dragons flight (talk) 18:39, 3 March 2013 (UTC)
i changed the classification of this bug report from "extensions:scribunto" to "extensions:syntaxhighlight". the same problem appears on wikis that did not yet enable scribunto (e.g., copy the content of Module:Convertdata to your user page on dewiki, and put it inside <source lang="lua">....</source> tags, and you'll see exactly the same truncation that happens here. peace - קיפודנחש (aka kipod) (talk) 21:34, 3 March 2013 (UTC)
Ah, that sounds reasonable because trying to syntax highlight all that text is an over-use of resources. I think I saw a mention on wikitech-l that someone had a problem of very slow loading when syntax highlighting a large page, and the suggestion was that a change be made so highlighting is automatically disabled if the block of text exceeds a certain length. Perhaps that fix broke something. Johnuniq (talk) 02:24, 4 March 2013 (UTC)

Watchlist slow[edit]

The display of the watchlist has been consistently slow lately. It's been taking about 10 seconds instead of 2 or 3 seconds. I see that there's already a section here about this, #Another "slowness" thread, but I thought I'd go ahead and start a new section, to ask if the issue has been identified, and if there's an estimate for when this will be fixed. Thanks. Mudwater (Talk) 12:33, 2 March 2013 (UTC)

Yes Katie has been working on it. You can see this here: It still needs to be reviewed and go live here though. I hope this will happen soon. --Lydia Pintscher (WMDE) (talk) 12:58, 2 March 2013 (UTC)
Thanks! Mudwater (Talk) 14:38, 2 March 2013 (UTC)
Thanks for the update. Lugnuts Dick Laurent is dead 09:46, 3 March 2013 (UTC)

Odd happenings with Template:Japanese episode list[edit]

I've recently discovered that an empty entry for the "ShortSummary" parameter at {{Japanese episode list}} does not act the same as it does at {{Episode list}}.


No. Title Original air date
# "Filled summary" Airdate
# "Blank summary" Airdate
# "No summary" Airdate


No. Title Original air date
# "Filled summary"
"romaji" (kanji) 
Short summary 
# "Blank summary"
"romaji" (kanji) 
# "No summary"
"romaji" (kanji) 

No. Title Original air date
# "Filled summary (sandbox)"
(kanji romaji
Short summary 
# "Blank summary (sandbox)"
(kanji romaji
# "No summary (sandbox)"
(kanji romaji

What's going on at one that's not working on the other?—Ryulong (琉竜) 13:18, 2 March 2013 (UTC)

I think it's the difference between {{#if:{{{ShortSummary|}}}| and {{#ifeq:{{{ShortSummary|ʁ}}}|ʁ| |. According to Help:Parameter default, the first example will be "true" if the content of ShortSummary is defined and non-empty, and "false" is empty or not defined - ShortSummary is used if specified, even if empty. In the second, if ShortSummary has content, including empty content, it will be used; the result is only "ʁ" if the parameter matches "ʁ" or has not been specified; the content to be used if they don't match is after the final pipe. Peter James (talk) 14:14, 2 March 2013 (UTC)
Some of the parameters have been done this way deliberately (see comments by Ned Scott at Template talk:Episode list/Archive 1#Limits part 2); this one was changed in Episode list but not in Japanese episode list. Peter James (talk) 14:23, 2 March 2013 (UTC)
So can Japanese episode list be modified to match episode list? I attempted to ask this on the talk page but because no one bothered to see why one acted one way and not the other I got told to come here because I was clogging up edit requests or something.—Ryulong (琉竜) 17:34, 3 March 2013 (UTC)
I've edited Template:Japanese episode list/sandbox and Template:Japanese episode list/sublist/sandbox, and expanded your examples to include "no summary" and the sandbox version. I've also made similar changes to Template:Japanese episode list multi-part/sandbox, an alternative template; I don't know why it exists, or whether the same should be done there. If there are no problems in the testcases pages (for Japanese episode list, /sublist and /multi-part) you can request these changes to the templates. It looks like the changes to the main and sublist templates would have to be requested together - editing the main template before the sublist would make empty summaries in sublist appear as "ʁ". Peter James (talk) 19:14, 3 March 2013 (UTC)
It's only Template:Japanese episode list that's protected, so the changes can be requested for that and its /sublist on its talk page. The "multi-part" can just be edited, and I'd make the changes myself if I was certain that they were wanted there; I'll leave it as you hadn't mentioned that template. Peter James (talk) 19:39, 3 March 2013 (UTC)

Article stats[edit]

Hi, does anyone have any idea roughly how many of the daily article views listed at the stats pages at are likely to be non-human accesses like search engine web crawlers? (talk) 03:24, 3 March 2013 (UTC)

"Including all of its different search crawlers and services hosted on its servers, Google itself requested another 31,551,740 page pages per day, representing 6.1% of our external page requests.". [28]
"In total 78,991,900 page requests (mime type text/html only!) per day are considered crawler requests, out of 518,048,900 external requests, which is 15.2%" [29]
Calvin (t·c) 03:51, 3 March 2013 (UTC)
Good heavens. And they would all be included on those stats pages, right? I mean, the stats tool would not try to identify and eliminate them? (talk) 04:00, 3 March 2013 (UTC)
So can you explain what that all means?!Curb Chain (talk) 04:13, 3 March 2013 (UTC)
As a rough estimate, 15.2% of requests for Wikipedia pages come from web crawlers (approximately 2 in 13 requests). 6.1% of page requests are from Google's crawler (approximately 2 in 33 requests). – PartTimeGnome (talk | contribs) 16:07, 3 March 2013 (UTC)
  • Beware sudden mega-pageviews: There have been numerous cases of wild, unexpected "mega-pageviews" (in fact, actually over 1 million per day) for some articles which rarely get 5,000 pageviews per day. For example, in January 2013, the article "Alabama" was viewed over 1.074 million times on 21 January (to enter the Top 1000 most-viewed pages of 2013) but then had only regular 3,500-pageview level on the next day:
For human readers to cause such a huge daily spike, from 3500 to almost 1.1 million, would require "National Day to read English article Alabama" in China or such, or else a top news story, "U.S. State of Alabama dies in plane crash". In fact, human readership typically shows elevated pageviews on 2 or 3 days around a spike as an "exponential curve" (or actually a negative binomial distribution) which tapers off for a few days later, where perhaps the next day, the pageviews would linger at 20%-50% of the prior spike day, such as 200,000 pageviews following a 1-million pageview newsflash. Compare pageviews around the 11 February 2012 death of singer/actress Whitney Houston:
Two days after her death, pageviews rocketed to over 5.97 million on 13 February, and remained at elevated levels for over 6 months. For computer-generated pageviews, instead, there will be a sudden drop where the next day returns to typical pageview levels. Anyway, mega-pageviews are not reflected in webcrawler percentages. -Wikid77 (talk) 17:23, 3 March 2013 (UTC)
Is there any proposed explanation for the Alabama-style spikes? (talk) 02:35, 4 March 2013 (UTC)

Tulip chair "infobox"[edit]

The article's "infobox" is actually formatted using a table. Is this desirable? Or should a infobox template be filled out? Can someone help with that isif so?Curb Chain (talk) —Preceding undated comment added 04:10, 3 March 2013 (UTC)

Yes check.svg Done I've changed the article to use {{Infobox furniture}}. That template doesn't have a field for colours, so I placed the colour description in the caption field instead. – PartTimeGnome (talk | contribs) 15:57, 3 March 2013 (UTC)
Thanks! I thought there was a correct infobox for this!Curb Chain (talk) 01:08, 4 March 2013 (UTC)

Page move reversal[edit]

I recently created the page Wikipedia:Help index this was boldly moved to an obscure title of Help:Outline of help resources. As a result of the move the page views has gone from 250+ a day to under 90 a day - let alone the 300+ redirects that need to be fixed. I have asked for a Requested move - but am here to find out if the bold edit can be reverted and then a talk take place - not the other way around. I want to revert the bold edit but cant because of how the redirects are set up. Any recommendations on what to do? Moxy (talk) 16:37, 3 March 2013 (UTC)

I can't see any technical reason that prevents you from reverting the move. There is no history to the redirect at Wikipedia:Help index, so you should be able to move over the redirect. (See Wikipedia:Moving a page#Undoing a move for a longer explanation.) Whether or not you should undo the move is a matter for discussion on its talk page. – PartTimeGnome (talk | contribs) 16:55, 3 March 2013 (UTC)
Ok have reverted the bold move and will see what others have to say about a new title Help talk:Outline of help resources#Requested move. Thank you for that info - some odd reason the "move" button was not working.Moxy (talk) 17:06, 3 March 2013 (UTC)

Proposal for using the overflow property on tables and other features in the mobile version[edit]

Working a bit on a wide table, I found out about the overflow property of CSS. And, I finally decided to test how "overflow:auto" would make the table look like in my smartphone (a Windows phone). To my surprise, it worked pretty fine, making otherwise inaccessible areas of the table, scrollable and readable (whereas at present, almost any table appears clipped in the mobile version on my phone). As even Microsoft has implemented support for such a feature, wouldn't it be a good idea to add (conditional) code on everything other than text so that it would appear scrollable in the mobile version if overflow is clipped? Or at least incorporate such a change in the wikitable template? The "auto" value means that the table will only become scrollable if it doesn't fit on the screen, thus not affecting at all users who access the mobile version through systems with screens wider than a phone's one.

Example to be viewed with cell phone:

  • without overflow tag:
Signatory Conclusion date Institution Majority needed In favour Against Abstentions Deposited References    Implementation law Balanced budget amendment
  • with overflow tag (can be touched and slided although no bar appears on my cell phone):
Signatory Conclusion date Institution Majority needed In favour Against Abstentions Deposited References    Implementation law Balanced budget amendment

Heracletus (talk) 23:51, 3 March 2013 (UTC)

Suppress Your edit was saved message?[edit]

Is there a way? I understand its usefulness, but don't want this appearing every time I save an edit. MeekSaffron (talk) 23:07, 2 March 2013 (UTC)

Add .postedit { display: none; } to your personal CSS. PrimeHunter (talk) 01:11, 3 March 2013 (UTC)
Thanks! MeekSaffron (talk) 10:57, 3 March 2013 (UTC)
I'll try it too, thank you. It is a pity that we have to waste time and resources to avoid this - mind my french - idiotic stuff :-| - Nabla (talk) 20:31, 3 March 2013 (UTC)
Agreed. Shame it can't be turned off globally across all wikis. Lugnuts Dick Laurent is dead 16:04, 4 March 2013 (UTC)

Extra space List of Viking metal bands[edit]

There's an extra space after the lede. I know there is a hidden comment after it, but shouldn't it not appear even if it is ontop of the opening {| of the table? I've confirmed that }}<noinclude> exists in {{Multiple image}} and {{Dynamic list}}. What am I missing, and can someone fix this?Curb Chain (talk) 01:21, 4 March 2013 (UTC)

I've removed a newline from the article. Is that a good enough fix? -- John of Reading (talk) 07:47, 4 March 2013 (UTC)
Sure! But is there something in the templates that can prevent this from happening?Curb Chain (talk) 04:25, 5 March 2013 (UTC)

Navigation popups - JS work requested[edit]

Could a JavaScript expert please look at the top of WT:POPUPS, where there is a change to the script's behavior (not overriding Reference Tooltips) that has been requested by several users since last August. Thank you. — Hex (❝?!❞) 15:42, 5 March 2013 (UTC)

VisualEditor fortnightly update - 2013-03-04 (MW 1.21wmf11)[edit]

Hey all,

Here is a copy of the regular (every fortnight) update for the VisualEditor project, so that you all know what is happening (and make sure you have as much opportunity to tell us when we're wrong, as well as help guide the priorities for development and improvement).

The VisualEditor was updated as part of the wider MediaWiki 1.21wmf11 branch deployment on Monday 4 March.

In the two weeks since 1.21wmf10, the team have worked on the new features being added for wider launch of the VisualEditor as the default way users will edit wikis. Part of this work has been some of the last infrastructure changes expanding the capabilities of the 'document model' module, including landing the last of the major re-write there (45062). This has also covered letting the converter work with full HTML documents rather than just fragments and be more robust, significant fixes to bugs around list items being merged and un-merged wrongly, better support for 'meta' nodes, fixed handling of text that in normal MediaWiki would be auto-linked (37940), and better documentation (45264 amongst others). The user interface through a major refactor and refresh now has access to a much more consistent set of 'widgets' for the main interface and for items like the link inspector as well as the to-come category, template and references dialogs which are being designed.

The 'content editable' module has improved keyboard support, fixed the ability to add new items in a list (due to an unfortunate typo) and add new lines in pre-formatted blocks using Shift-Enter (45344), added some changes to improve issues with selection of content in the browser (especially "backwards" selection), will now hide the 'phantom' overlays for non-editable content blocks if the mouse leaves the page (43461). Additionally, the dependency on a key library called 'rangy' was refactored, and the broken copy-and-paste support between VisualEditor instances was removed, in anticipation of better copy-and-paste to work with non-VisualEditor surfaces as well.

A complete list of individual code commits is available in the 1.21/wmf11 changelog, and all Bugzilla bugs closed in this period on Bugzilla's list.

Per the MediaWiki deployment roadmap, this should be deployed here on Monday 11 March.

Hope this is helpful! As always, feedback gratefully received, either here or on the enwiki-specific feedback page.

Jdforrester (WMF) (talk) 17:23, 5 March 2013 (UTC)

Article feedback tool gone?[edit]

Just curious: has the (old) article feedback tool been switched off? It no longer seems to be shown on articles I view, where I think it was still around yesterday or thereabouts. I'm talking about the old one, where people just graded articles but didn't leave comments. Will the new tool be deployed to all these pages now? Fut.Perf. 19:26, 5 March 2013 (UTC)

Yes, it was undeployed this morning. The new tool is deployed but is done so on an "opt-in" basis.--Jorm (WMF) (talk) 19:32, 5 March 2013 (UTC)
See more info in this update from Fabrice Florin.--Eloquence* 00:01, 6 March 2013 (UTC)
More at Wikipedia talk:Requests for comment/Article feedback#Discussion closed; bug filed --Redrose64 (talk) 13:02, 6 March 2013 (UTC)

Table sort question[edit]

The sorting in List of dog breeds doesn't seem to be working.

I see some recent discussions about sort issues, but at least one said there were problems that went away.

Can anyone see the problem?--SPhilbrick(Talk) 21:26, 5 March 2013 (UTC)

I think that it's the refs in the column headings. --Redrose64 (talk) 21:53, 5 March 2013 (UTC)
no, it's not. sortable tables have this quirk, where every single row must have exactly the same # of columns. in this case, 5 rows missed the last column. adding "||" to each of them solved the problem. however, i'm not sure how valuable it is to be able to sort on a column, if so many of the cells in this table are empty. peace - קיפודנחש (aka kipod) (talk) 22:25, 5 March 2013 (UTC)
Thanks! I work with sortable tables a bit, but missed that. It makes sense, but it didn't occur to me to look for that. I'm guessing it was working, then someone added an entry, and assumed that they did not to fill out the rows when some cells were blank, so it may happen again. Good to know.--SPhilbrick(Talk) 14:06, 6 March 2013 (UTC)
i doubt anyone did it this way intentionally: regardless of disabling the sortability, it's makes the table ugly (the right border of the table appears broken), and if it was intentional, why create the other empty cells? these rows ended with |||||||||| instead of |||||||||||| or somesuch (i bet anyone can spot the difference instantly). i believe it was just that whoever added these rows (prolly not the same person added them all) just did not count the columns carefully enough. wiki-code for tables is far from being friendly... (not that html table code is any better - counting <tr> and <td> is not easier than counting |- and ||). peace - קיפודנחש (aka kipod) (talk) 15:39, 6 March 2013 (UTC)
I assumed it was accidental, rather than deliberate, but I haven't actually looked at the edit that introduced the problem.--SPhilbrick(Talk) 19:23, 6 March 2013 (UTC)

Long file edits in PHP[edit]

I was all excited about finally getting time today to get Chartbot to work, but failed. It scans the files, performs the URL substitutions, verifies our links against Billboard, everything, but dies when it tries to write the file. Stripping my broken code down to its simplest form yields User:Chartbot/simplified, which works when I set the filelength to 634 bytes, but crashes at 635 with an HTTP 417 error. This seems to be because our servers want me to use a different content type. Try as I might, I can't figure out a way to control the content type that http_post_fields uses when it posts the data. Anyone got a clue? I really hate having a bot that can do all the complicated logic but can't store the final result, and I've struggled with this all day.—Kww(talk) 01:55, 6 March 2013 (UTC)

I have a feeling this is to do with the Expect: 100-continue header. Certainly that's been an issue for me in .NET in the past (in .NET it's very easy to fix too, simply set System.Net.ServicePointManager.Expect100Continue to false). I believe the correct way to fix this in PHP is to add 'Expect' with a blank value to your headers array. Not really looked at your code closely, or how you are handling headers, but it looks like that would be a case of adding $http_options['Expect']=''; after setting $http_options['cookies']. - Kingpin13 (talk) 02:47, 6 March 2013 (UTC)
Got a version working with Curl, which will let me suppress the "Expect" header. It also handles the long file encoding more reliably, so it wasn't just busywork.—Kww(talk) 16:22, 6 March 2013 (UTC)
Good to hear. Yeah, curl will probably be better in the long run. - Kingpin13 (talk) 21:51, 6 March 2013 (UTC)

Lua migration of Coord[edit]

I just migrated {{coord}} to use Lua Module:Coordinates. This provides a great improvement in rendering speed for pages with many coordinate templates. Under normal circumstances the output of the Lua version should be exactly identical to the output of the old template, except for a few edge cases. Specifically, the old template would drop trailing zeros and make some anomalous precision choices in displaying decimal formats. This has now been standardized. The Lua version is also displays more verbose error messages when given malformed content.

Dozens of test cases were run and validated before making the change. That said, there may still be some unexpected edge cases, so please look out for any problems with coordinate tags. In particular, Category:Pages with malformed coordinate tags may catch most errors. Dragons flight (talk) 05:53, 6 March 2013 (UTC)

not sure if this question is related to the lua conversion, or if this was the same even before the conversion, but i find the decimal display of coordinates (say, in Bays of the Philippines a bit strange. shouldn't it be displayed as DMS rather than a decimal fraction? how did it look before the conversion? regarding performance improvement, this is very impressive. i would not guess that innocent looking {{coord}} was be so bad, performance-wise. i sure hope we'll see similar improvements with the "cite" family of templates. peace - קיפודנחש (aka kipod) (talk) 20:12, 6 March 2013 (UTC)
The display characteristics are unchanged. The template gives a decimal display by default when fed decimal parameters {{coord|1.124|34.567}} = 1°07′26″N 34°34′01″E / 1.124°N 34.567°E / 1.124; 34.567, and a DMS display when fed DMS data {{coord|1|14|N|34|56|E}} = 1°14′N 34°56′E / 1.233°N 34.933°E / 1.233; 34.933. {{coord}} also has a format= parameter to force a DMS or decimal display when fed the other kind of data. Lastly, Template:Coord#Per-user_display_customization also gives instructions for changing the appearance of coord data on a personal basis using CSS. Dragons flight (talk) 20:32, 6 March 2013 (UTC)

Unresponsive script[edit]

Hello, I'm getting a script error whenever I open any article, pointing out to Script:

How to fix it? — Preceding unsigned comment added by (talk) 10:37, 6 March 2013 (UTC)

Remove redirects from search bar[edit]

Is there any way to tag a non-neutral or factually incorrect, but otherwise acceptable, redirect so that it doesn't autofill as if it is an article title in the search bar? Ryan Vesey 15:54, 6 March 2013 (UTC)

I believe that if the redirect page is in Category:Unprintworthy redirects, it will be excluded from the search box. It's best to use a WP:TMR - {{R unprintworthy}} will do this, but some others will too. --Redrose64 (talk) 17:47, 6 March 2013 (UTC)
selected at random one "unprintable redirect" from the category, and started typing it in the box. sure enough, it was autocompleted. maybe this is not so for all of them, but i'd be surprised.
afaik, the autocomplete is fed from the same database that powers the regular search, so something like the "noindex" magic word would do it, if there is such a trick that affects internal search. ttbomk, "noindex" only affects external search engines, and i'm not familiar with a similar magic word that removes a page from internal search, but if there is one, i'll bet it would prevent autocomplete also. (if you find something and want to try it, you should wait a day or two before it actually takes effect, just like it takes a day or two for a new article to appear in the autocomplete). peace - קיפודנחש (aka kipod) (talk) 18:04, 6 March 2013 (UTC)
This has been brought up before at Wikipedia:Village pump (technical)/Archive 105#Unprintworthy cross-namespace redirects are not being excluded from the search dropdown as expected. AFAIK, no bug has been filed for this. – PartTimeGnome (talk | contribs) 23:27, 6 March 2013 (UTC)

Ghost bot?[edit]

My watchlist shows a raft-full of recent edits by User:Sk!dbot, but these edits don't appear in the page history and this user has no contributions (including no deleted contributions). Is this a ghost bot? --Orlady (talk) 22:18, 6 March 2013 (UTC)

I expect these are wikidata edits (the new system for interwiki links on pages). You can see them over at Click "Hide Wikidata" at the top of your watchlist to get rid of them. - Kingpin13 (talk) 22:26, 6 March 2013 (UTC)

Link to AfD discussion in template incorrectly displaying as redlink[edit]

I've noticed a very weird and specific problem with the AfD template, or at least one that is occurring at Periphery (BattleTech). If I view the page while I am not logged in, the link to the AfD discussion is displayed as a redlink with alt-text indicating the page doesn't exist. However, if I click on the link, it takes me to the AfD discussion (which does in fact exist). The link displays correctly (as blue, not red) if I'm logged in.

I'm using Chrome and Windows 7. Does anyone else see this occurring? Some guy (talk) 23:56, 4 March 2013 (UTC)

Interesting. Same here, with Opera 11.61 / Linux. - Nabla (talk) 00:02, 5 March 2013 (UTC)
This often happens with AFDs and is browser/OS-independent. It can happen if the AFD page is created after the AFD template is added to the article; what has happened is that the MediaWiki software hasn't yet got around to changing the link colour from red to blue. All you need do is WP:PURGE the article page. I've switched on 'Add a "Purge" option to the top of the page, which purges the page's cache when followed.' at Preferences → Gadgets to allow this to be done quickly.
I've seen it happen with other types of XFD, like CFD, but it's rarer because they often have per-day discussion pages not per-article. --Redrose64 (talk) 00:13, 5 March 2013 (UTC)
This is frequent when Twinkle is used for XfD nominations as in Periphery (BattleTech). Twinkle edits in the best order by creating the discussion page first, but the deletion tag is added to the nominated page a moment later, too fast for the servers to register that the discussion has been created. I have seen at least 10 reports of this. Twinkle changes with purging or a delay after page creation have been suggested, for example at Wikipedia talk:Twinkle/Archive 28#AfD template red link, but no changes have been made. PrimeHunter (talk) 02:52, 5 March 2013 (UTC)
Ah, interesting. Thanks for the info! Some guy (talk) 03:52, 5 March 2013 (UTC)
Ditto - Nabla (talk) 20:43, 7 March 2013 (UTC)

PDF rendering error reported to OTRS[edit]

Folks, we have received an e-mail reporting a PDF rendering error. I am pasting the text of the issue below.

I've tried to get the PDF of 'Stabat Mater' over the past week. It does not form into a proper PDF with the proper layout - this is the first time I have come across this issue.
The message at the top of the PDF reads: "WARNING: Article could not be rendered - ouputting plain text. Potential causes of the problem are: (a) a bug in the pdf-writer software (b) problematic Mediawiki markup (c) table is too wide" --- this one on attempt dated 04Mar13

Would someone please take a look and see if there really is a problem and, if so, what the cause/solution may be. I assume the article being referred to is Stabat Mater. Thanks in advance.--ukexpat (talk) 15:34, 6 March 2013 (UTC)

A quick test in my sandbox shows that the problem is with the section Stabat Mater#Text and translation. I guess the PDF renderer doesn't like {{multicol}}. -- Toshio Yamaguchi 16:02, 6 March 2013 (UTC)
Thanks for troubleshooting. I have, temporarily, put a copy of the article in user subpage, removed the multicol formatting and asked the user to try and create a PDF from that version.--ukexpat (talk) 18:18, 6 March 2013 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────The user was able to create a PDF from the version of the article that I saved in my userspace without the multicol formatting. Would someone please report this at Bugzilla for me - the bug being that multicol formatting as in the parallel text section of Stabat Mater breaks the creation of a PDF. Thanks.--ukexpat (talk) 15:17, 7 March 2013 (UTC)

Filed a bug report myself.--ukexpat (talk) 18:38, 7 March 2013 (UTC)

Bug Day focusing on triaging MediaWiki - General/Unknown reports[edit]

Bug icon - Noun project 198.svg

You are invited to join the next Bug Day tomorrow Mar 07. We will go through fresh open reports filed against MediaWiki in the General/Unknown component. It is a good chance to learn about bug management and meet the people working on a daily basis with the bug database (which includes Wikipedia reports, yes). No MediaWiki expertise is actually required. There is work for everybody.  :) --Qgil (talk) 18:57, 6 March 2013 (UTC)

Reminder: This is in 10 minutes and everybody is welcome to join and say hello! --AKlapper (WMF) (talk) 14:52, 7 March 2013 (UTC)

Filter log and API[edit]

I don't know if this is indicative of a bug somewhere, but the filter log associated with the redirect API is surprisingly extensive (link). Many apparently innocent users get reported by the bot at Wikipedia:Administrator intervention against vandalism/TB2 because of this. Just posting here in case anyone is interested in investigating ... -- Ed (Edgar181) 14:29, 7 March 2013 (UTC)

This is mainly Wikilove tracking filter. Ruslik_Zero 19:18, 7 March 2013 (UTC)
This is bizarre, as the API redirect has been fully protected since 2010, so users shouldn't even be able to reach the point that would trigger the filter to begin with. My best guess is some kind of bug in the MediaWiki API. jcgoble3 (talk) 19:20, 7 March 2013 (UTC)
Could it be that filter hits triggered by an editing request through api.php (such as those made by the WikiLove extension) are erroneously logged as if the edit was done to the API page?—Emil J. 19:48, 7 March 2013 (UTC)
That's very likely, given that all hits on the Wikilove filter have logged on the API page since February 11. Something that was changed on February 11 broke it. Could someone file a bugzilla report? jcgoble3 (talk) 20:30, 7 March 2013 (UTC)
I believe this is already fixed by this changeset, which should be deployed here on Monday (see mw:MediaWiki 1.21/Roadmap for the schedule). Note that Feb 11 was the date 1.21wmf9 was deployed here, so rather than "something that was changed on February 11" it's more likely "some change that was included along with 1.21wmf9". BJorsch (WMF) (talk) 03:12, 8 March 2013 (UTC)

Lua templates not compatible with PDF Export[edit]

Earlier today it was noticed that the "Download as PDF" function and associated tools are completely ignoring all of the new Lua based templates. The issue was referred to Mediawiki developers, but as we have already migrated a number of templates to Lua (e.g. Category:Lua-based templates), I wanted to make sure the larger community was aware that there may be missing information or other errors in PDF exports until the problem is resolved. Dragons flight (talk) 20:41, 7 March 2013 (UTC)

New version of Special:GettingStarted[edit]

Hey all,

I wanted to update you about two developments on the projects by the experiments team at the WMF:

  1. We just finished A/B testing guided tours, and so far results are quite promising. With the tour, we see users try to edit at a greater rate (+13%) and save their first article edit in 24 hours more (+4%). Both of these are statistically significant results, so we're going keep trying to find ways to use tours effectively.
  2. Today we launched a new version of Special:GettingStarted which is delivered to all newly-registered editors (more docs, including a screenshot of the previous version). The big change is that we've added two additional task types from the backlog -- articles with too few wikilinks and articles tagged as confusing, vague or unclear. In addition to the copyediting task, we hope that we've picked things to do that don't require an expansive knowledge of policy or subject matter expertise.

Let me know if you have any questions. As as always, you can keep tabs on the edits made by new users via these interfaces by watching the gettinstarted RecentChanges tag. Steven Walling (WMF) • talk 00:51, 8 March 2013 (UTC)

CSD transclusion chaos[edit]

Category:Candidates for speedy deletion is looking pretty busy this morning - it seems that every page which transcludes the template {{Independent Production}} is now up for deletion, even though the category doesn't appear on any of the individual articles (that I've checked so far) or the template itself. I've asked CsDix, who was the last person to edit the template, to take a look, but if other technical bods wouldn't mind giving it the once over, I'd be grateful. Yunshui  09:09, 6 March 2013 (UTC)

Found the source of the problem - all the affected pages transclude the redirect {{Independent production}} (lowercase "p"), which was recently tagged with {{db-g6}} (declined). I would have thought removing the G6 tag ought to have fixed the problem, though. Yunshui  09:21, 6 March 2013 (UTC)
The page links table has had issues with updating lately, technically a purge on each individual page might be needed. Legoktm (talk) 10:27, 6 March 2013 (UTC)
I've tried purging several of the affected pages (and CAT:CSD as well, for good measure), but it doesn't seem to be having an effect. Yunshui  10:37, 6 March 2013 (UTC)
Purging only affects the purged page itself. It doesn't affect categories where the page is displayed. That requires a null edit. I haven't found an article with the problem so it seems to be gone. Please post an example if you still see it, or if you report a problem another time. PrimeHunter (talk) 12:42, 6 March 2013 (UTC)
Seems to be resolved now, yes. Clearly the technical solution was for me to go away and have a sandwich; I shall have to remember that fix in future. Yunshui  13:04, 6 March 2013 (UTC)
The "cup of tea and a cookie/biscuit" fix also works in my experience!--ukexpat (talk) 15:37, 6 March 2013 (UTC)
  • To clear them from CSD, do a null edit of each article. — RHaworth (talk · contribs) 15:43, 8 March 2013 (UTC)

Edit mode problem[edit]

I've repeatedly run across this problem: When I edit a redirect that has absolutely nothing but the #REDIRECT statement in it and add one or more "R" templates and/or one or more categories then do a Preview, there is no list of included templates and categories provided. But when I save my modifications and re-edit the redirect everything is as it should be and further added templates and categories are added to the lists when previewed. Cbbkr (talk) 00:46, 7 March 2013 (UTC)

I think a change was made at some point to not process any content on a page past the REDIRECT tag, but that it still will process on a diff. Someone else would know more of the background on that. Chris857 (talk) 01:21, 7 March 2013 (UTC)
For some years it has been the case that the content is processed on the rendered page in order to extract the categories, as with Plympton railway station, but the behaviour has been changed, at least twice: the most recent was a few weeks ago, when diffs stopped showing the content after the #REDIRECT [[]], although the categories are still shown: see here for example. This change in behaviour was noted at Wikipedia:Village pump (technical)/Archive 107#Redirect from misspelling. --Redrose64 (talk) 11:41, 7 March 2013 (UTC)
Chris857, I've been trying to use parser functions to redirect my talk page on my home wiki to no avail; are you saying that if I put the parser functions in front of the redirect and save it to a variable, then call the variable from inside the redirect it should work? In other words, this doesn't work:
#REDIRECT [[User_talk:Technical_13/{{#time:Y|now -5 hours}}/{{#expr:ceil({{#time:n|now -5 hours}}/3)}}]]
BUT my understanding of what you are saying is that this would:
{{#vardefine:Year|{{#time:Y|now -5 hours}}}}{{#vardefine:Quarter|{{#expr:ceil({{#time:n|now -5 hours}}/3)}}}}#REDIRECT [[User_talk:Technical_13/{{#var:Year}}/{{#var:Quarter}}]]
Is that correct? T13   ( C • M • Click to learn how to view this signature as intended ) 21:56, 8 March 2013 (UTC)
No. #REDIRECT must be the first thing on the page; there can't be anything before it. I'm not certain, but I don't think parser functions can be used in the redirect target either. (I'm guessing that your home wiki isn't a WMF one, since Extension:Variables isn't installed here.) – PartTimeGnome (talk | contribs) 22:49, 8 March 2013 (UTC)

Pending changes glitch?[edit]

An edit on Eleanor Roosevelt showed up as a pending revision, but when I looked at the revision it was listed as "accepted" and was also listed as "automatically accepted" in the review log, even though it was still showing up in the queue of pending revisions in the page history and on the pending changes page after I refreshed those pages several times. I had to unaccept the revision and then accept it to fix that issue. The edit was made with Huggle so perhaps there was some weird conflict.--The Devil's Advocate tlk. cntrb. 17:58, 8 March 2013 (UTC)

Did you try purging your cache first? Sounds like your cache was a bit outdated.—cyberpower ChatOnline 19:04, 8 March 2013 (UTC)
I don't think that was it. The edit showed as accepted in the diff, but not when I looked at the revision history. Also, the log says it was automatically accepted, but it still showed up as a pending revision on the page for pending changes.--The Devil's Advocate tlk. cntrb. 20:09, 8 March 2013 (UTC)

Is there a way to get this page information?[edit]

Can one find for a given article

  1. whether as a newly created article it's ever been patrolled
  2. (if it was patrolled) when it was patrolled?
  3. (if it was patrolled) what user patrolled it?

Signed: Basemetal (write to me here) 20:21, 8 March 2013 (UTC)

I don't know the answer to this. What I find strange is, I checked Special:NewPages and found the article Hydrothermal Liquefaction. I checked and it had the [Mark this page as patrolled] at the bottom. I viewed the page in edit mode and couldn't find the tag. I then returned to read mode and the tag was gone, so it seems to view it once in edit mode is enough for the tag to vanish. I wonder what the point of this is, though. -- Toshio Yamaguchi 20:42, 8 March 2013 (UTC)
Toshio, see Bugzilla: 35810 and upvote it! T13   ( C • M • Click to learn how to view this signature as intended ) 20:50, 8 March 2013 (UTC)
Somehow I can't log into Bugzilla with my account, so it seems my unified account doesn't work on Bugzilla. Is there a reason why Bugzilla is excluded from SUL? I guess I will almost never use it, so my desire to create a separate account over there is pretty low. -- Toshio Yamaguchi 21:07, 8 March 2013 (UTC)
IIRC, Bugzilla is a completely different software than MediaWiki, and so cannot have the Module:UnifiedLogin loaded to it. Just create an account, I never use mine but I still have it :) It doesn't spam your mail, promise! gwickwiretalkediting 21:30, 8 March 2013 (UTC)
Just to clarify something - bugzilla does indeed send you emails whenever a bug you commented on changes, which some people might consider spam (This can be controlled in preferences). It also publicly releases your email to any logged in user. Bawolff (talk) 11:03, 9 March 2013 (UTC)
Go to Special:Log/patrol, fill in the page name after "Target (title or user)" and hit "Go". It'll come back with "No matching items in log." if the page is unpatrolled. --Redrose64 (talk) 22:05, 8 March 2013 (UTC)
Ok, I did that for Dobroslav_Jevđević, today's featured article and it comes back with "No matching items in log". Do you mean to say that this is page was never patrolled? Signed: Basemetal (write to me here) 03:40, 9 March 2013 (UTC)
The page was created in 2007, long before new-page patrol existed. Bielle (talk) 03:50, 9 March 2013 (UTC)
You are right! And it works with newer new pages. Great! Signed: Basemetal (write to me here) 10:15, 9 March 2013 (UTC)

One more question on this topic: are unreviewed (as marked in, say, the new pages feed) and unpatrolled the same thing? Signed: Basemetal (write to me here) 10:32, 9 March 2013 (UTC)

Proposal to rename Catalogue of CSS classes[edit]

Please comment at Wikipedia talk:Catalogue of CSS classes#Suggested move. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:32, 8 March 2013 (UTC)

Edit count question[edit]

Is there a way to track who has made the most edits per namespace? I.e., who has the most for User or User talk or File or Wikipedia... etc. AutomaticStrikeout (TC) 23:44, 8 March 2013 (UTC)

You may want to check out Extension:MW-EditCount T13   ( C • M • Click to learn how to view this signature as intended ) 23:55, 8 March 2013 (UTC)
Yes, but does that show a list of the editors with the most edits per namespace, or is it just an edit count per namespace for an individual editor? AutomaticStrikeout (TC) 23:59, 8 March 2013 (UTC)
I don't know, but it isn't installed on Wikipedia in any case. See Special:Version#Installed extensions for a list of which extensions are installed here. – PartTimeGnome (talk | contribs) 00:04, 9 March 2013 (UTC)
It is currently setup to show edit count per namespace for an individual editor, but since it is not maintained and requires repair to be used on any MW installation 1.18+, it wouldn't be all that hard to cannibalize it with Extension:CountEdits T13   ( C • M • Click to learn how to view this signature as intended ) 00:10, 9 March 2013 (UTC)

Internal Error after Save Page[edit]

I was editing Audie Murphy. After Save Page, it flipped to a page headed "Internal Error" with this message:[ba13be97] 2013-03-09 00:40:45: Fatal exception of type MWException. I think I saw a similar post here not too many weeks ago. My edit actually was successful, so I don't know why this MW Exception displayed, but thought I'd mention it here if it's important. — Maile (talk) 00:48, 9 March 2013 (UTC)

I got a similar message when trying to view an image on Commons yesterday evening. Keith D (talk) 01:58, 9 March 2013 (UTC)
I got it earlier tonight after logging in; fatal error. Truthkeeper (talk) 02:04, 9 March 2013 (UTC)

JavaScript help please[edit]

At my Common.js page (on Commons), I took someone's script, but instead of having it go to "Special:Upload", I actually want it to go to the specific URL: Special:Upload&uselang=experienced. How can I do this so I can have a quick one-click uploading experience? – Kerαunoςcopiagalaxies 04:46, 9 March 2013 (UTC)

I take it you saw that your most recent version works? ;) Writ Keeper (t + c) 05:18, 9 March 2013 (UTC)
Absolutely! Someone very kindly helped me at MediaWiki and I was rushing around trying to solve other problems before coming here to cancel this query. Thanks for checking in. I'll have to keep this board in mind in the future. – Kerαunoςcopiagalaxies 05:32, 9 March 2013 (UTC)


Hello from Latvian Wikipedia sysop! I have a problem of finding the Mediawiki interface message for the "Edit links" line from the iw links list. Can somebody help? --Edgars2007 (talk/contribs) 08:45, 9 March 2013 (UTC)

It's at Mediawiki:wikibase-editlinks - at Special:Preferences, Gadgets tab, there is a checkbox to "Create a toolbox link to show the page with messages from the user interface substituted with their names", which solves problems like this one. -- John of Reading (talk) 08:59, 9 March 2013 (UTC)
Or just add uselang=qqx to the url to help find the name of the page -- WOSlinker (talk) 09:00, 9 March 2013 (UTC)
Really big thanks :) Unfortunately @lv wiki we don't have the gadgets tab, but the uselang=qqx option works good. --Edgars2007 (talk/contribs) 09:14, 9 March 2013 (UTC)
The code is at MediaWiki:Gadget-ShowMessageNames.js. It works in your personal js. PrimeHunter (talk) 12:34, 9 March 2013 (UTC)

Slow wiki[edit]

Why is Wikipedia so painfully slow?

Served by mw1181 in 19.987 secs.

I get similar times on different pages.—cyberpower ChatOnline 18:10, 9 March 2013 (UTC)

rather difficult to say without knowing which pages, but generally because people like complex templates. Are you getting times like that for cached hits too? (if so, that is rather slow. if pages are significantly faster when logged out, can be caused by certain prefs like stub threshold). Bawolff (talk) 21:21, 9 March 2013 (UTC)
Its not my preferences as I haven't changed that for months. It took 15 minutes to push the above statement.—cyberpower ChatOnline 21:50, 9 March 2013 (UTC)

Examples of convolution[edit]

I saw the wiki page, but I couldn't find any examples using actual numbers evaluating the formula. Could you give some examples of convolution, please? Mathijs Krijzer (talk) 22:41, 9 March 2013 (UTC)

Please ask at the maths reference desk. Most people here are not maths experts; you'll find people more knowledgeable about maths there. Wikipedia:Village pump (technical) is primarily for technical discussion relating to Wikipedia itself; knowledge questions are more suited to the reference desks.
(I was also going to recommend posting to the article's talk page to suggest that examples would improve the article, but I see you have already done so.) – PartTimeGnome (talk | contribs) 23:15, 9 March 2013 (UTC)

Portland, Oregon ref problems[edit]

I tried asking on the Freenode chat channel about this. I just added three citations to Portland, Oregon [30] in the "Economy" section, and noticed they weren't displaying in the ref section. They're citation numbers 81, 82, and 83 in the text, and when I click on those numbers it leads nowhere. The refs currently shown as 81, 82, and 83 in the ref section lead to the "Demographics" section when clicked on.

And the number of unique citations in the text exceeds the number displayed in the ref section, which proves that there are other citations written in the text (not just 81, 82, and 83) that aren't being displayed below. First observation: #174 in the "Sister cities" section leads to #93 in the ref section. 159, e.g., leads to 78. But citations 9, 18, 35, 37, 40, 50, 59, 77, 89 in the text (for example) lead nowhere. 1 through 8 do, but starting with 9 it appears they don't, until you get to 90.

I'm on Safari 5.1.7. I don't know what browser User:Steven Zhang was using when he tried to help me on the chat channel, but he saw these same problems. Jsayre64 (talk) 01:12, 10 March 2013 (UTC)

I had the same issues, traced it back to the 8th reference [31] incorrect formatting screwed things up. Werieth (talk) 01:22, 10 March 2013 (UTC)
Wow, nice catch. It seems to be working properly now. Thanks! Jsayre64 (talk) 01:28, 10 March 2013 (UTC)

Lua 10-second timeout unusable for articles[edit]

We cannot use big-scale Lua in articles at this point, such as the Lua-based wp:CS1 cite templates, due to Lua crashing pages at a 10-second timeout. I have been very worried about trying to use Lua in actual large articles, because during months of testing on the test2 wiki, I noticed how real articles would occasionally get "Script error" when formatting hundreds of Lua-based cite templates. The appearance was:

  • "Gee, sometimes Lua gets tired of running and cries 'Script error' for fun".

The problem was so rare, I just wondered huh, and who really has time for this, but months later finally, someone noted "Lua seems to have a 10-second timeout" which is severely small compared to markup-based templates which could reformat text for 60-second timeout. Of course, server load varies greatly, by the minute (rarely dragging to triple-slow) but often doubling a 6-second Lua run, to want 12-second Lua, and bingo, the whole page exceeds the 10-second Lua timeout, and further Lua-based templates start storing "Script error Script error Script error" repeatedly into the formatted cache-copy of a page, for thousands of readers to view. Fortunately, we have not deployed any large, major Lua-based templates (I think), so there are unlikely to be pages studded with "Script error" which readers can view daily. However, at this point, I cannot recommend using big-scale Lua until the 10-second timeout is raised, to perhaps 30-second timeout, pending other reasons why Lua modules must be limited by total run-time duration. I think Lua provides a potential net benefit to Wikipedia, but we must find a way to allow Lua modules to run longer than a 10-second timeout on busy servers. Meanwhile, we can continue to rewrite the markup-based templates, as several times faster (such as {cite_quick} ), until the Lua technology is configured to handle the "big league" world of major articles, with templates used in over 1.5 million pages. -Wikid77 (talk) 07:37, 2 March 2013 (UTC)

Can you give some example test articles where Lua cites time out or come close to timing out? It would be good to see how much wiggle room we actually have and if there are ways to make the scripts faster. Dragons flight (talk) 08:16, 2 March 2013 (UTC)
Tests show that frame:preprocess is very slow (I think around 150 of them hit the 10 second limit). The manual says that frame:expandTemplate should be used if expanding a template, but it is slow as well. Is the citation module using either of those? I don't feel like finding it, so perhaps a link? Johnuniq (talk) 10:37, 2 March 2013 (UTC)
i might be wrong here, and if so i'd love it if someone "in the know" will put things straight, but i think the above is based on misunderstanding. the 10 seconds budget is not for wall-clock time, but rather for CPU time, and as such is independent from server load. when the server is loaded, it might take more time (wall-clock) to allocate the same 4 seconds of CPU time because the CPUs are busy elsewhere, but it will still show the same "4 seconds", and is not more likely to hit the 10-seconds budget limit than when the server is idle.
i second the request to point to a page with high (let's say, above 3 seconds) Lua-engine time, as reported in the "newPP" comment inside the html. it is very possible such a page exists, and if you can find one it will be very interesting to try and understand the issue. if you have an experimental Lua-based template that can replace an existing template, why not create a sandbox with a copy of some large article where the template is switched, and see if you can push Lua time to 4-5 seconds? peace - קיפודנחש (aka kipod) (talk) 15:04, 2 March 2013 (UTC)
I would think it is CPU execution time (when executing a Lua module), and not wall clock time. I did not mean to suggest otherwise. BTW, when I said "around 150" I forgot that 150 was the approximate number of operations, but each operation actually invoked preprocess three times. Johnuniq (talk) 01:08, 3 March 2013 (UTC)
  • Imagining the fox in the Lua henhouse: When people cannot believe that Lua script in the henhouse could ever reach the 10-second timeout, and there is no such problem as a "sly fox" well, here is what dead chickens look like in the Lua henhouse; they state:
  • "The time allocated for running scripts has expired."
  • "The time allocated for running scripts has expired."
  • "The time allocated for running scripts has expired."
The way this Lua fox works in the Lua henhouse is he, very slyly, causes a Lua module which normally runs just a few seconds to expand the Lua CPU time to over 10 seconds, such as running "11.23" seconds, and then all subsequent Lua functions will be rejected due to time "expired". The problem is rare, and might take dozens of trials to reappear, so perhaps we should create an artificially large testcase, to force the rare timeouts as more likely to occur. It does not help to claim a "misunderstanding" about the dead chickens in the Lua henhouse, as if the 10-second timeout could never, ever happen. Hence, step 1, let's start thinking about why Lua must be limited to 10-second timeout, and consider if a 30-second timeout would be ok. If the Lua timeout limit cannot be raised beyond 10 seconds, then Lua cannot be used in large-scale templates, and there would be no need to "prove" anything. Then, instead, we can get back to adding rapid wp:parser functions, and restructuring markup logic, so that templates can run much faster. In future years, as technology improves to allow Lua-based templates to run over 10-second duration, then we could reconsider using Lua to help rather than crash articles with "Script error Script error Script error". Fortunately, I already found a way to make the wp:CS1 cite templates run about 4x faster without using Lua. Also, Lua can be used for other purposes, rather than rewriting large templates from markup into Lua. -Wikid77 (talk) 16:06, 2 March 2013 (UTC)
i guess i am just stupid, but i could not understand your last message at all.
i did not see anyone claiming that it's physically impossible for lua script to exceed its 10 seconds budget - such a claim would be plain silly. but can you please point to a page (can be snadbox in user space) where lua script executes in, say, 5 seconds? i do not mean a test page that does something with the sole intention of stretching the limits, but rather a copy of a "real" page from article space, where some templates are replaced with lua counterparts. this may help better understand the issue, and may even expose poorly written or misused scripts. thx, peace, קיפודנחש (aka kipod) (talk) 16:32, 2 March 2013 (UTC)
Here's one: Bays of the Philippines -- WOSlinker (talk) 17:20, 2 March 2013 (UTC)
We also have Category:Pages with script errors, which contains a few things. Most of the ones that are timeout based are coming from having a very large number of {{coord}} or a very large number of {{convert}}. Neither of which are great examples, since both families of template are still planning to be overhauled. However, it is noteworthy that the timing list shows "Scribunto_LuaSandboxCallback::getExpandedArgument" as the slowest step (96% of the time for Bays of the Philippines). If I understand correctly "Scribunto_LuaSandboxCallback::getExpandedArgument" is actually a part of the process that Mediawiki uses during the #invoke step and not something that the guts of the script is doing. That suggests that the developers may need to better optimize that step, or perhaps not count the #invoke setup time against the Lua execution time limit. Dragons flight (talk) 18:12, 2 March 2013 (UTC)
I migrated {{coord}} entirely to Lua, and the Lua execution time for the Bays article is now 0.8 seconds. Dragons flight (talk) 06:00, 6 March 2013 (UTC)
Wikid, obviously Lua can time out. Any idiot could write a script that would time out simply by designing a very bad script. That's not really the point. In creating Lua, the developers had several design goals, which included replacing some of the more complex templates such as the cite family. The best way of making a case that 10 seconds is too low is to show examples of pages in actual use that can't be converted to Lua cites (or similar Lua conversions) because the timeout limit prevents it. Developers want there to be a limit to help prevent against denial of service attacks, but they also want to improve the functionality and performance of Wikipedia. If 10 seconds is inadequate for dealing with cite and coord and infoboxes, etc., then we can make the case for higher limits, but it is better to make that case using clear examples. For the record, I tried converting Barack Obama to Lua cites (sandbox) and those 336 references formatted in 2.7 seconds of Lua execution. That's a a little long, but still has a lot of headroom, so I'd be interested in finding examples that come closer to the limit. Dragons flight (talk) 17:51, 2 March 2013 (UTC)

regarding Bays of the Philippines: this is an excellent example why we should look at those things. the problem here is that the "Math" module tries to be be smarter than what's good for it, and be able to accept arguments which are expressions as well as numbers. IMO this is a wrong way to use the tools we have. IOW, i would yank the call to "preprocess" out of Math._clean_number(), and when a template wants to be able to accept expressions as well as numbers as its arguments, the template should pass this parameter to #expr before passing it to the module. so, if you want to allow someting like {{order of magnitude|11 * 12}}, you should not force Math.order to eveluate its input and understand that 11 * 12 is really 132, but rather make {{order of magnitude}} look like so: {{#invoke:Math|order|{{#expr|{{{1}}}}}}} or somesuch. i.e., use #expr *in the template that invokes math*, instead of "preprocess" #expr from the module. doing so will bring Bays of the Philippines Lua time consumprion to the sub-second region - prolly 100ms or so.

so yes, there can definitely be pages that consume more than 10 sec on their Lua eveluation, but the remedy isn't necessarily to bump up the Lua budget, but rather to do things the right way instead of the wrong way.

peace - קיפודנחש (aka kipod) (talk) 19:27, 2 March 2013 (UTC)

You're actually mistaken, Math only calls preprocess if the input can't be understood as a number. The problem here is actually with the construction in {{coord/prec dec}}:
Max is a Lua function, but the Lua timer is being charged for the time to expand and evaluate {{precision1|{{{1}}}}} and {{precision1|{{{2}}}}}. Further, the expansion isn't in Math._clean_number but rather in Mediawiki since getExpandedArgument has to be called to fully expand all templates (in this case {{precision1}}) before the Lua function can be called. Essentially, {{#invoke:Math|max}} is being charged for the time required to evaluated {{precision1}} which isn't actually a Lua issue, but rather a Mediawiki template issue. Dragons flight (talk) 19:45, 2 March 2013 (UTC)
PS. Just to be clear, removing preprocess from Math._clean_number will do absolutely nothing here, since it isn't being called in this case. Dragons flight (talk) 20:03, 2 March 2013 (UTC)
i verified that what you wrote above is correct, by copying the Math module to a sandbox, removing the call to preprocess, and using Special:TemplateSandbox to evaluate Bays of the Philippines. i still do not think that Math should call preprocess like this (i.e., i believe that the template should bear the burden of calling #expr if it wants to allow expressions as an arguments), but this is beside the point. so the question is: is the time required to calculate/generate the parameters to max really deducted from the Lua 10 seconds budget? if this is the case, then i think it's wrong, no? so the remedy should not be to increase Lua's time budget, but rather to calculate Lua's time consumption more correctly, no? peace- קיפודנחש (aka kipod) (talk) 20:39, 2 March 2013 (UTC)
I edited {{coord/prec dec}} to make the calls more directly Lua and remove the references to {{precision1}}, and now Bays of the Philippines completes with 6.9s of Lua runtime, so we can take that one of the list (and it will be even faster once coord is entirely a Lua module). Regarding the of whether the #expr calls should be in the template or in the Math module, the main reason it is in Math is to facilitate cases like {{max}} whether the arguments are passed implicitly and wrapping each term is impossible. Also, I'm not sure it is necessarily better to have a template that wraps all arguments in #expr on every call or a Lua script that checks if the input is a number and then only calls #expr if it isn't. The latter case might be faster on average, but I'm not entirely sure. To the last point, I generally agree that it is strange to have a situation where Lua is being charged for the time it takes Mediawiki to evaluate the arguments being sent to Lua. Dragons flight (talk) 00:02, 3 March 2013 (UTC)
Possibly that was done on the cautionary principle: when creating a new system (Scribunto), make sure that denial-of-service attacks are avoided by checking CPU time in a simple and reliable manner. That might be adjusted after some months of experience. However, given the good fix that you implemented, and the scope for enormous improvements when {{coord}} is replaced by a module, I don't think any change to the way Lua execution time is charged is warranted. We can think of it as a hint that all time-intensive procedures should be updated. My brain has no spare capacity at the moment so I can't investigate, but I have implemented some precision and rounding code in Module:Convert. It was hideously tricky, and would benefit from checking, but anyone wanting to work on similar things might like a look. Johnuniq (talk) 02:40, 3 March 2013 (UTC)
  • Charging Lua for markup parameters risks 10-second limit: I have an example to confirm that a single use of a Lua function, passing templates as parameters can kill the Lua operation (due to 10-second timeout) and show "Script error" rather than store formatted text into a page. For that reason, the Lua timeout limit should be greatly increased from 10-second timeout, to perhaps 30-second timeout or more. Note: we do not need to change Lua's CPU time counter, just allow a higher limit. To run a fatal Lua-timeout test, try the following use of Lua-based {hexadecimal} in edit-preview:
# {{hexadecimal|{{#expr: 10 +
+{{convert|1|m|ft|0|disp=number}} +{{convert|1|m|ft|0|disp=number}}
+{{convert|1|m|ft|0|disp=number}} +{{convert|1|m|ft|0|disp=number}}
+ ...repeat to have 300 lines of {convert}...
}} }}
Because rounded {Convert} runs at 55 per second, then 600 {convert} typically runs over 11 seconds, so the Lua CPU time will clock nearly 12 seconds then decide the 10-second limit has expired, and not perform the hexadecimal formatting of the result after totalling the 600 conversions. Although this is an artificial example, it shows the danger of 10-second timeout in Lua-based templates, where the time needed to evaluate markup-based parameters is added to the Lua CPU clock, and then every subsequent Lua-based template (on the page) will crater and store the words "Script error" into the formatted page, for thousands of readers to view. -Wikid77 (talk) 14:51, 3 March 2013 (UTC)
Er... that's insane, and just the kind of thing I'd hope would be blocked by limits. Why would someone use 600 {{convert}} templates as arguments to a single Lua-based template? The only reason I can think of would be a deliberate denial-of-service attack trying to tie up Wikimedia's servers. If there is a genuine reason to do this, then there is most certainly a better way to do it.
In any case, I think it's a bug that the time taken to expand templates used as arguments to Lua-based templates is counted as time used by Lua. If this is the most significant cause of Lua timeouts, the solution is to stop counting template expansion as time used by Lua. There are already other limits that restrict regular template expansion. – PartTimeGnome (talk | contribs) 16:39, 3 March 2013 (UTC)

Clearing up some things here:

  1. The 10-second limit is CPU time, not wall clock time, as surmised by קיפודנחש and others. Specifically, in the LuaSandbox PHP extension used on WMF wikis, it uses POSIX timers with CLOCK_PROCESS_CPUTIME_ID to implement the timeout.
  2. "Scribunto_LuaSandboxCallback::getExpandedArgument" is not used during #invoke. It is used to handle accesses to frame.args: the first time any particular key is accessed, that function is called to get the value for the argument, and getting that value includes expanding the passed wikitext. Which means that it's actually worse to pull all possible arguments into local variables even if they won't be needed. There's nothing to be done to speed the function up, the code is very simple:
    1.     function getExpandedArgument( $frameId, $name ) {
    2.         $args = func_get_args();
    3.         $this->checkString( 'getExpandedArgument', $args, 0 );
    5.         $frame = $this->getFrameById( $frameId );
    6.         $result = $frame->getArgument( $name );
    7.         if ( $result === false ) {
    8.             return array();
    9.         } else {
    10.             return array( $result );
    11.         }
    12.     }
    The slow bit is the call to PPFrame::getArgument() on line 398, which is a part of the core parser. So anything that might speed that up would speed up non-Scribunto wikitext processing, too.
  3. Yes, frame:preprocess() and frame:expandTemplate() can be slow. They're not quite as simple as getExpandedArgument() quoted above, but they're close.
  4. frame:expandTemplate() is slightly faster than passing an equivalent string to frame:preprocess(), but the real advantages are in not having to construct the equivalent string in the first place and in not having to worry about trying to escape the argument values when doing so. frame:callParserFunction() is on my TODO list, BTW.
  5. Yes, the time spent to parse an argument is counted against the Lua time limit. The counting of calls into PHP against the time limit in general is intentional, but there is a good case for making an exception for getExpandedArguments().

Hope that helps. BJorsch (WMF) (talk) 21:48, 3 March 2013 (UTC)

Thank you for describing those issues. We certainly wish there were some clever way to speed the #invoke interface, which is showing a limit of ~180/second, while simple markup-based templates can run over 350x-1,200x per second. However, for now, just raising the 10-second timeout to about 30-second timeout would help. I have an example, below {{Weather_box/cols/sandbox}}, which shows how when Lua quits, the "scribunto-error" tags are stored in the page-cache version, where readers can see the "permanent" stoppage of Lua formatting, until the page is reformatted to reset the cache-page. -Wikid77 (talk) 22:28, 3 March 2013 (UTC)
  • Example of Lua 10-second timeout in template doc page: Based on the above conversation, questioning how could hundreds of calculations be used in an actual page, I remembered that most template limitations are first noticed in template /doc pages, and that is the case to demonstrate a typical case where Lua hits the 10-second timeout: {{Weather_box/cols/sandbox}}. In supporting Template:Weather_box, many cell-color helper templates have used the Lua-based Template:Hexadecimal over 200 times per article, but the /doc text of the helper subtemplates uses {hexadecimal} many hundreds of times. For the total Lua time used, each use of {hexadecimal} also runs several #ifexpr or #expr parser functions, and the total time often exceeds the 10-second timeout limit. At that point, the final table, of "rain days in a month" stops getting the hexadecimal values for blue table-cell colors and all table cells show: <td class="scribunto-error"> or such. Look inside the bottom of the HTML version of that /cols/sandbox page. More later. -Wikid77 (talk) 22:28, 3 March, 13:12, 4 March 2013 (UTC)
    Lua can perform an amazingly large amount of work in well under 10 seconds, so there is some other factor responsible for time outs. In my sandbox are 1689 template calls, and each call invokes Module:Convert, and each of those invokes causes Module:Convertdata to be "required". That is a stupendous amount of work, yet when I just looked the Lua time usage was 4.306s. See Module talk:Convert for info, including mention of how mw.loadData is likely to at least double that speed. Johnuniq (talk) 02:37, 4 March 2013 (UTC)
    Yes, there are other factors responsible. The other factor is that Lua is getting charged for the time that the parser takes to compute the parameters being sent to Lua. For the weatherbox box example, 85% of the time is being used to compute the parameters that are being sent to Lua, while only 15% of the time is actually being used to do computations within the 100s of Lua calls. If we didn't charge the Lua clock for the template expansions and parser functions that are being executed to generate the inputs for Lua, then the present issue would be largely avoided. This is essentially just another case of the parser being slow, but because of the implementation, that slow parser can also cause Lua to run out of time through no fault of the actual Lua code. Dragons flight (talk) 06:42, 4 March 2013 (UTC)

Counting the preprocessing of #invoke parameters against the Lua time limit certainly seems problematic. I've taken the liberty of opening bug 45684. More generally, I don't understand the need for a Lua time limit at all. As long as Lua counts against the overall limit on page render time (60 seconds, if I understand correctly), doesn't that suffice? To take an extreme example, if a page with 45 seconds of markup-based templates could be converted into 15 seconds of Lua-based templates, wouldn't that be a good thing? Toohool (talk) 06:05, 4 March 2013 (UTC)

  • Meanwhile focus on faster markup templates: Thank you for filing that Bugzilla #45684. I agree with your logic about the 60-second limit, and I have noted another danger of the 10-second timeout, here: if the timeout is not fixed (somehow), then rather than recode 45-seconds of markup into 19 seconds, as 8 markup and 11 Lua seconds, then the page will continue to use the 45-second markup-based templates, to avoid risk of Lua timeout above the 10-second limit. However, the near future is not totally grim; we can optimize the markup-based templates, without Lua, to run faster. In fact, I have written a streamlined Template:Hexdigits2 which is 3x faster than the Lua-based Template:Hexadecimal but only handles 2 digits, not full hexadecimal numbers. Also, some Lua-based templates might be used only a few dozen times per page (not 2 thousand {hexadecimal} ), and those could be used to run within 2 seconds of Lua, far below the 10-second Lua timeout limit. -Wikid77 (talk) 13:12, 4 March 2013 (UTC)
  • Example of March 10 Lua timeout at 10-second limit: I have re-created the example of a Lua timeout, in Template:Weather_box/cols/sandbox2, after the prior /sandbox timeout last week was rewritten away. Running several edit-preview sessions has confirmed that the timeout has varied slightly at 10.030-10.045 seconds, each time, after ~2,000 Lua function calls with Template:Hexadecimal, varying by 500 or more Lua functions run. The timing is generally consistent, but not exact to within 1 or 2 functions, but rather varies by 500 or more function calls each time. So, for example, the timeout might re-occur at 3 different wikitables formatted on a page. Also, other timeouts have logged "11.3" Lua seconds, to finish whatever steps before checking to stop after 10-second timeout. Because of the variation, the timeouts might occur sporadically, where a large page could display fine many times, and then, "Bam! Lua kills the page". Overall, the severe 10-second limit seems to be a rare problem, mainly just causing additional fears of using Lua, but should not stop progress in transitioning to use some Lua-based templates 5x-7x times faster than markup templates. -Wikid77 (talk) 11:48, 10 March 2013 (UTC)

Archiving help[edit]

I've been prepping Gone with the Wind (film) for a GA review, but Variety magazine has gone and deleted a whole bunch of articles, including one I've used as reference. Unfortunately it was deleted before I had a chance to run a Webcite comb, and it isn't archived at Wayback, but I have found a copy in Google cache at [32]. Now, Google cache isn't permanent, and Webcite won't archive the Google cache page, so is there any way I can get this page permanently archived somewhere. It's a 70 year old article so this is probably my only way of getting a permament copy. Assistance would be appreciated. Betty Logan (talk) 17:33, 9 March 2013 (UTC)

Its available at [33] but it is behind a paywall Werieth (talk) 17:52, 9 March 2013 (UTC)
PS I created a PDF copy of the cache if you can use that somehow. Werieth (talk) 17:57, 9 March 2013 (UTC)
I downloaded the cache copy so I can refer to it, but thanks anyway. I will cite the paywalled copy and provide the Goggle cache copy through the archive link, and hopefully it will last long enough to go through its GA assessment. Betty Logan (talk) 08:09, 10 March 2013 (UTC)