Wikipedia:Village pump (technical)/Archive 78

• Village pump
  • Policy
  • Technical
  • Proposals (persistent)
  • Idea lab
  • WMF
  • Miscellaneous

Village pump (technical) archive  

This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.

---

< Older discussions · Archives: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z, AA, AB, AC, AD, AE, AF, AG, AH, AI, AJ, AK, AL, AM, AN, AO, AP, AQ, AR, AS, AT, AU, AV, AW, AX · 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215

"You have new messages (last change)." reappears after a short delay

Something is wrong with the new messages functionality for me. Maybe 30 minutes ago I started getting these notices when there were no new messages. Sometimes there's a short delay before the message stops appearing after I've checked my messages. Once it goes away, it reappears withing a few minutes. --Ronz (talk) 19:06, 2 July 2010 (UTC)

I'm getting this too. And here I thought I was just really popular. --Bongwarrior (talk) 19:11, 2 July 2010 (UTC)

This happens when the server lags. Gary King (talk · scripts) 19:35, 2 July 2010 (UTC)

Extreme server lag

I'm getting indications that articles I've been editing don't even exist, and that people I send Talk messages to are not even registered. Server lag? Everard Proudfoot (talk) 22:43, 2 July 2010 (UTC)

When I go to Recent changes, it keeps showing the same Recent changes from about half an hour ago. I have to purge my cache to see the new changes, but even then, if I go back to the Recent changes, it shows the old version yet again. Everard Proudfoot (talk) 22:45, 2 July 2010 (UTC)

I am getting the same thing. I am in the UK, if that makes any difference. The behaviour is almost random. Sometimes recently created/deleted stuff exists, sometimes it seemingly doesn't. Even history and log pages are affected. I have to keep hitting reload several times until I see what I expect. I am trying to do some new page patrol and both Twinkle and Friendly are getting very confused and messing things up when I use them. It is confusing me too. This is the first time I have ever come across anything like this. It needs to be flagged up as a serious problem.

My theory is that there are multiple cache servers and some of them are lagging and some are not. Hence it is more or less random which cache server you get served from and hence random whether you see recent data or not. The lagging servers do not show the usual red warning about the data being old. --DanielRigal (talk) 22:56, 2 July 2010 (UTC)

It looks like this needs to be exposed publicly, because I see lots of people blanking pages. Everard Proudfoot (talk) 23:17, 2 July 2010 (UTC)

Yeah, something really weird is going on with page histories. ctr-f5 helps sometimes, sometimes not. It never happened to me before, but I thought it was a cache issue on my end, but it doesn't appear to be. --Omarcheeseboro (talk) 23:19, 2 July 2010 (UTC)

If it is affecting people in different countries then it isn't some ISP running a broken transparent proxy. I think it can only be the Wikipedia infrastructure. --DanielRigal (talk) 23:23, 2 July 2010 (UTC)

Every time I go to Recent changes, I still get the page as of 22:10. I have to clear cache every time, sometimes twice. Everard Proudfoot (talk) 23:25, 2 July 2010 (UTC)

Well, I was patrolling Recent Changes (in Canada), and Gavriil Malish is viewable half the time. Sometimes, it's visible, sometimes, it says the page does not exist. The history page does not exist, yet it show up on my watchlist. Brambleclawx 23:54, 2 July 2010 (UTC) Add Refreshing helps, but that link I provided above is switching from blue to red, and it obviously hasn't been deleted... Brambleclawx 23:55, 2 July 2010 (UTC)

Does anybody know if this is being looked into? It is causing lots of screwups and making some editing tasks close to impossible! --DanielRigal (talk) 00:23, 3 July 2010 (UTC)

I think your theory makes sense Daniel; I was looking through special:logs/newusers, and when I first got there, it showed me users who had registered 2 minutes ago, then I refreshed and it showed my users who had registered 2 hours agp. I have no idea if anyone is looking into it. Brambleclawx 00:25, 3 July 2010 (UTC)

Yes, it is being looked into. With the amount of users Wikipedia has, it takes about three seconds for someone to complain about server lag to the server admins. --Deskana (talk) 00:25, 3 July 2010 (UTC)

It appears to be an issue with one of the database servers only (db40). I've alerted our sysadmins to this and it should be sorted out soon.--Eloquence* 00:25, 3 July 2010 (UTC)

The issue should be resolved.--Eloquence* 00:44, 3 July 2010 (UTC)

Appears to be ok now. Brambleclawx 00:48, 3 July 2010 (UTC)

Yes, It seems to be. Responsiveness has also improved significantly. Maybe everybody has stopped pressing F5 every couple of seconds. ;-) Thanks. --DanielRigal (talk) 00:49, 3 July 2010 (UTC)

Thanks, Eloquence. Everard Proudfoot (talk) 01:34, 3 July 2010 (UTC)

I'm still having problems. After saving an edit the page is not updated. Using purge doesn't work. Saving the page again doesn't work. I was working on User:Droll/subpages/testcases.  –droll [chat] 03:22, 3 July 2010 (UTC)

False alarm. My bad.  –droll [chat] 03:25, 3 July 2010 (UTC)

New messages bar spazzing

I'm getting the new messages bar really frequently even after I check the message. Its been about and hour since the real new message appeared. Derild4921☼ 23:26, 2 July 2010 (UTC)

Wait for it to fix itself. --Deskana (talk) 23:30, 2 July 2010 (UTC)

MediaWiki:Gadget-textareasansserif.css

This gadget is apparently redundant since there's an "edit area font style" option in preferences. vvv^t 00:32, 3 July 2010 (UTC)

If an edit conflict happens, the lower textbox shown is a textarea but it is not #wpTextbox1. The "edit area font style" affects only the main text box; the Gadget affects all textareas. The same is true for scripts such as Twinkle and closeAFD. So the two are not 100% redundant. PleaseStand ^(talk) 02:26, 3 July 2010 (UTC)

#tag

So it seems that there is a function named #tag somewhere. It's used in {{nowiki}} and in several other templates I've looked at. Does any one know if it's documented anywhere. I would expect it to be at MediaWiki somewhere but I can't find it. The syntax seems to be {{#tag: (tag name) | (string) }} —Preceding unsigned comment added by Droll (talk • contribs) 02:24, 3 July 2010 (UTC)

It is documented at Help:Magic words, fourth from the top. In essence, it allows a tag to be used (such as nowiki or imagemap) without the HTML opening and closing < >. Intelligentsock 02:30, 3 July 2010 (UTC)

Thanks! I guess I just looked in all the wrong places. –droll [chat] 02:34, 3 July 2010 (UTC)

(edit conflict)mw:Help:Magic words#Miscellaneous. It is most commonly used when you want the parser to expand template parameters before applying the tag's effect (especially nowiki's). For example, {{nowiki}} uses {{#tag:nowiki|{{{1}}}}} to accept a parameter and then nowiki it (producing code such as <nowiki>http://ezinearticles.com/</nowiki> instead of <nowiki>{{{1}}}</nowiki> from the parameter http://ezinearticles.com/). PleaseStand ^(talk) 02:39, 3 July 2010 (UTC)

Template for colors

I'm looking for a template to easily equip a template with hex no.s such as in Template:Infobox military unit. Where do I find it? E.g. {{hexcol|#######|########}}. --217.189.229.123 (talk) 23:36, 2 July 2010 (UTC)

I am not sure what you are trying to accomplish. Can you clarify what such a template would do? Intelligentsock 23:51, 2 July 2010 (UTC)

I want to put a template to the parameter "colors=HERE" in the Template:Infobox military unit (inclusion of course). The desired template (I have seen it but I forgot it) displays one or two colors by inserting the hex code. --217.189.229.123 (talk) 23:59, 2 July 2010 (UTC)

It's {{Color box}}, example {{color box|blue}} {{color box|yellow}} gives     and <code>{{color box|blue}} {{color box|yellow}}<code> gives    . --Scriberius (talk) 11:08, 3 July 2010 (UTC)

OGG/OGV Buttons Blank on Mac Firefox

I hope I'm not reposting this question.. When I access an OGG or OGV file with Firefox on my Mac, the buttons are blank -- see this image, taken from this article. It doesn't happen with Safari; that player is sleek and works fine. Perhaps there is something I need to adjust on my browser? I expect other users are having this problem, too. Cheers! Scartol • Tok 12:55, 3 July 2010 (UTC)

Wikipedia Waiting Time

I raised this issue in mid-June. The problem persists. It happens whether I'm logged in or not logged in. It seems to get worse the more I use WP in a given session, but that could be coincidental. When I click on something, Firefox 3.6.6 gives me a "waiting for . . ." message in the status bar. It can sometimes take 20-40 seconds to come back. Other times, no wait at all. If there is an excessive wait and I click on the link again, I usually get an instantaneous response. On some things, though, I'm reluctant to click twice because I'm not sure what effect it will have on the database, i.e., saving editing changes. My Internet download connection speed is above 10Mbps. I don't have this problem on other sites. I simply don't know what to do, but editing has become very frustrating because of it.--Bbb23 (talk) 15:38, 3 July 2010 (UTC)

Clicking save twice shouldn't cause any problems database-wise. At worst you might cause an edit conflict with yourself. Try doing a ping test to the Wikipedia servers to check for packet loss. Mr.Z-man 17:43, 3 July 2010 (UTC)

When I ping en.wikipedia.org, there is no packet loss. If there's another server you want me to try, let me know.--Bbb23 (talk) 22:24, 3 July 2010 (UTC)

How can I stay signed in?

I followed advice I was given here when I found out my computer still had pages I had visited even though I appeared to have no history. Now, though, I seem to have to sign in everywhere, including Wikipedia. It's not that big a deal since my computer has the software which fills in passwords automatically, though without that I was unable to access one of my email accounts because of rules they have about passwords (though I still remember that crazy password for that address). I can remember the password if the password software messes up, since I use Wikipedia at libraries, but it's just annoying.Vchimpanzee · talk · contributions · 18:25, 3 July 2010 (UTC)

Try deleting cookies, that often 'resets' the need to constantly log in (of course, you'll still have to log in once). ♫ Melodia Chaconne ♫ (talk) 20:12, 3 July 2010 (UTC)

How to feed pages such as 'My contributions' to a self-written editor?

Hi, what I'd like to do, for example, is have a little program/macro that can interface with non-editable Wikipedia pages such as 'My contributions' (or, someone else's contributions). Ideally it would read and collect (in memory or as an appendable file) a pageful of data at a time, then virtually click on 'next 100' until the end. I can write code, but I don't know how to interface with Wikipedia, or what language is required. I'm sure this is spelled out somewhere, so if you just point me to the documentation I can figure out what I might be in for. Thank you, CliffC (talk) 22:01, 3 July 2010 (UTC)

The best option would be to use the API, located (along with some documentation) at http://en.wikipedia.org/w/api.php. Several output formats are available. For example, to get your 10 newest contributions in the XML format, use the following URL: http://en.wikipedia.org/w/api.php?action=query&list=usercontribs&ucuser=CliffC&format=xml. Or you could use one of the tools or libraries that are already there. For example, AWB can get user's latest contributions and save them to a file. Svick (talk) 22:30, 3 July 2010 (UTC)

Wow.

Thank you for that very complete answer. I never thought at all about what goes on with the bots here. I looked at that documentation and got some crazy ideas. I'll dig in when I have the time and want to be frustrated/exhilarated by getting something to work. Best, CliffC (talk) 13:22, 4 July 2010 (UTC)

Wikipedia down

Anybody care to venture as to why the insecure wikipedia is down? The secure still seems to be partially up. —Preceding unsigned comment added by 69.115.30.110 (talk) 01:44, 5 July 2010 (UTC)

The rumors are that the airconditioning in the Florida center has failed, causing the machines to shutdown. It is being worked on. —TheDJ (talk • contribs) 01:51, 5 July 2010 (UTC)

Twitter has a feed of the techlog that you can follow. —TheDJ (talk • contribs) 01:54, 5 July 2010 (UTC)

For some reason, I can get on the site, but I can't log in. --71.205.208.70 (talk) 02:15, 5 July 2010 (UTC)

en.wiki is up, but skinless (bits.wikimedia is down). 124.168.19.202 (talk) 02:18, 5 July 2010 (UTC)

Could I suggest the very professional failure page that directs you to a Google search include a link to the Twitter feed and/or another Wikipedia Status page that I seem to recall but haven't looked for this time. Mark Hurd (If I could log in) 121.45.22.104 (talk) 02:43, 5 July 2010 (UTC)

Skins are back up, but I still can't log on. —Preceding unsigned comment added by 71.205.208.70 (talk) 02:46, 5 July 2010 (UTC)

Yeah, I tried to log onto my main account and my declared sock that I use on public computers, but neither one will work; I'm running Monobook, so I guess that's the problem. I'm having to use someone else's computer to type this, since mine is set to log in automatically. 71.79.87.249 (talk) 02:48, 5 July 2010 (UTC)

Special:UserLogin login now working. — Richardguk (talk) 03:26, 5 July 2010 (UTC)

The insecure Wikipedia is now working. December21st2012Freak ^{Happy Independence Day!} 03:32, 5 July 2010 (UTC)

WikiNews error page still refers to Meta

(I know this is the wrong place, but it is up enough to receive edits :-) )

The error page for wikinews.org still refers to Meta in its title, heading and icon. Mark Hurd 121.45.22.104 (talk) 02:49, 5 July 2010 (UTC)

about PWB

I would like to know which function do we use when copying a content from a wikipage or from a textfile to a wikipage or a textfile. Thanks in advance, --Jagwar - (( talk )) 10:56, 28 June 2010 (UTC)

Can you clarify what you are trying to do? Your question is worded vaguely. — Richardguk (talk) 02:15, 29 June 2010 (UTC)

I sould like to create a program that could generate several wiki pages from a single file in TXT format. Currently, I do not know how to use the script pagefromfile.py that generates constant errors when used (I don't know how to use it and what is the syntax used to make the file readable by the program) ; I am still not familiarized with the Python language. --Jagwar - (( talk )) 10:01, 2 July 2010 (UTC)

This probably isn't the right place to ask. But there is some information and an example at meta:Pywikipediabot/pagefromfile.py and a mailing list at http://lists.wikimedia.org/pipermail/pywikipedia-l/

You'll need to have configured your general settings in user-config.py. Then it looks like the simplest settings would require a file named dict.txt with the content encoded as UTF-8 and set out like this:

{{-start-}}
'''Title of first article'''
Wikitext of first article
goes here
{{-stop-}}
{{-start-}}
'''Title of second article'''
Wikitext of second article
goes here
{{-stop-}}

— Richardguk (talk) 15:03, 2 July 2010 (UTC)

Thanks you for the indication :). But I would like to know : sometimes the generators "says" :

X:/.../pagefromfile.py -start:{st} -end:{ed}
reading dict.txt
End of file.

Is it normal ? --Jagwar - (( talk )) 19:27, 2 July 2010 (UTC)

You probably need to post to http://lists.wikimedia.org/pipermail/pywikipedia-l/ for specific advice. — Richardguk (talk) 19:45, 2 July 2010 (UTC)

I'll see that. Thanks you and good continuation. --Jagwar - (( talk )) 12:20, 5 July 2010 (UTC)

File:Information processing system (english).svg

Resolved

This might be more appropriate on Commons, but I got a positive result here the last time I had the same problem. This file was moved to Commons a few days ago, but thumbnail generation seems to be broken: is there a way to force it to be regenerated? The SVG itself looks to be okay once it's clicked through. Chris Cunningham (not at work) - talk 09:17, 30 June 2010 (UTC)

Something's broken: Error generating thumbnail Error creating thumbnail librsvg-ERROR **: _rsvg_acquire_xlink_href_resource called for external resource: " base: (null) aborting... Appending a number to the end of the URL as described at WP:Purge#For images to regenerate the thumbnail does not seem to fix the problem. And I checked the source code and there are no external links within the SVG. PleaseStand ^(talk) 13:19, 30 June 2010 (UTC)

I suspect the brackets in the filename are the problem. Not all characters are allowed. I've tagged the image on commons with {{rename|Information processing system English.svg}}. When an admin or filemover actions that, we'll see if that clears it up. LeadSongDog come howl! 14:39, 30 June 2010 (UTC)

Fixed by commons:User:Mormegil — it was an internal link using incorrect syntax (diff). PleaseStand ^(talk) 19:54, 30 June 2010 (UTC)

This file was definitely problematic. It had transformation matrices on groups, which caused the rendered size of text to be incorrect. An ungroup/group of all groups fixed this. PleaseStand ^(talk) 20:44, 30 June 2010 (UTC)

Hmmm. Was this caused by the upload process, or just a bug in the original file? Chris Cunningham (not at work) - talk 21:19, 30 June 2010 (UTC)

Which problem are you referring to? The one that caused rsvg to crash or the one that caused incorrect text sizing? PleaseStand ^(talk) 02:43, 3 July 2010 (UTC)

The former. As I say, the file had previously been uploaded here and thumbnail generation worked fine. Chris Cunningham (not at work) - talk 22:15, 5 July 2010 (UTC)

Most likely the SVG parser/renderer has become more strict since the initial upload. Since thumbnails are cached, thumbnails that already exist will never re-render and the problem will never show. When the file was uploaded to another location, all thumbnails had to be generated again, making the problem very clear. —TheDJ (talk • contribs) 22:54, 5 July 2010 (UTC)

Image height?

I'm hoping there is a way to discover the height of an image assuming that the name of the image file and the width are known. I realize this not likely but I thought I'd ask anyway. It would be useful in a template I've been working on. –droll [chat] 04:53, 1 July 2010 (UTC)

P.S. So all I need to know is the aspect ratio. –droll [chat] 12:24, 1 July 2010 (UTC)

I don't think you can do that. Won't one of the alternative ways to specify image size (xHeightpx or WidthxHeightpx) help you? Svick (talk) 12:31, 1 July 2010 (UTC)

Perhaps {{panorama}} might be of help? ---— Gadget850 (Ed) ^talk 13:04, 1 July 2010 (UTC)

And you can of course always query the api, but not from a template, which is what I suspect you are after. —TheDJ (talk • contribs) 13:51, 1 July 2010 (UTC)

• TheDJ is right. {{Site plan}}, a.k.a {{Lageplan}} with the name transliterated, needs to know the height. I'm not sure how useful the template really is. It's like {{Superimpose}} but uses percentage coordinates instead of pixel coordinates. If the could be determined in the template it would be a good alternative for non-techies. Firebug and similar tools provide pixel detentions and the aspect ratio can be computed or the editor could just download an image, inspect the image properties and do a little math. These methods are not exactly user friendly.

  P.S. The template could use a better name. –droll [chat] 02:05, 3 July 2010 (UTC)

The MediaFunctions extension provides magic words for this (#mediaheight and #mediawidth) but is not currently installed on English Wikipedia. So the information cannot be obtained by templates on this wiki. — Richardguk (talk) 06:30, 3 July 2010 (UTC)

Thanks for the information. It's good to know it's out of my control.  –droll [chat] 07:45, 5 July 2010 (UTC)

Link doesn't work right

Maybe this is a strange way to do it, but I put links to web pages I want to go to in emails to myself so I can do it from any computer.

There was a link to [1]--at least that's what is visible in the email. [2] is what came up.

Likewise, [3] appears in the email, but clicking on it produces [4]. You'll probably send me to the Computing Reference Desk since it's not likely a Wikipedia problem, but I don't see how something like this could happen.Vchimpanzee · talk · contributions · 17:44, 1 July 2010 (UTC)

Is it an HTML format email? If so, its likely the actual HREF tag is wrong, while the text is showing what you really wanted. Do a view source. -- AnmaFinotera (talk · contribs) 18:20, 1 July 2010 (UTC)

I don't have a "View Source" button.Vchimpanzee · talk · contributions · 18:43, 1 July 2010 (UTC)

Email me if you want to establish a link outside of Wikipedia mail, then you can forward your problem to me when I respond. (Yes, that will unveil your email address) I'm suspecting "operator error" as the first port of call, but it's hard to say without the raw text. "View Source" would probably only work for web-based email clients, but a host-based client should allow you to save a message with the headers and MIME-encoding intact. For instance Outlook Express lets you save a .eml file which has all the good stuff. Franamax (talk) 20:38, 1 July 2010 (UTC)

All I've ever done is type or click (or copy and paste) on the address of the web site for the email service.Vchimpanzee · talk · contributions · 20:41, 1 July 2010 (UTC)

OK, so it sounds like you do use a web-based email client. Then you should be able to somehow or other view the HTML source of the page to see what is happening behind the scenes. Generally when I encounter a client who has a problem and I recommend a diagnostic procedure which they decide not to follow, I close the ticket as a PEBCAK, no offense meant. You could try pasting the copied link into your browser address bar to be sure you harvested the link you wanted. Franamax (talk) 22:19, 1 July 2010 (UTC)

"Somehow or other" doesn't answer the question. It doesn't seem to provide me with a way to "View Source".Vchimpanzee · talk · contributions · 22:28, 1 July 2010 (UTC)

What it sounds like to me is that instead of copying the existing source of the page when you're mailing it, you're actually sending a pointer to the current page. But without further details (what "email service" you're talking about, or the actual page source) it's difficult to be sure. To view the page source, click on View -> Page Source in your browser's menu bar. Chris Cunningham (not at work) - talk 08:46, 2 July 2010 (UTC)

Okay, we're getting closer. I don't see anything that says "View Source".Vchimpanzee · talk · contributions · 15:06, 2 July 2010 (UTC)

I figured it out. Right-click on the page, then click "View Source" ... except it doesn't seem to give me anything that's in the actual email. I recognize stuff that's displayed along with the email itself.Vchimpanzee · talk · contributions · 17:53, 2 July 2010 (UTC)

The above suggestion is, in effect, that your email is displaying the HTML equivalent of a piped link, <a href = "http://www.example.com/the_old_version" >http://www.example.com/the_new_version</a> (which in Wikitext is equivalent to http://www.example.com/the_new_version or loosely equivalent to a misleading piped link like [[the_old_version|the_new_version]]). So just delete the old stuff and paste the URL in from plain text afresh, to ensure that you are not inadvertently pasting in hidden code. (If you are copying from HTML, such as your web browser, and don't know how to make sure it is only plain text when you paste it in, first paste it into a plain text editor such as Notepad, then copy it from Notepad and paste into your email textarea.) — Richardguk (talk) 18:17, 2 July 2010 (UTC)

I can copy it here in the box I am typing in right now. The links probably won't be links I can click on, though.Vchimpanzee · talk · contributions · 18:27, 2 July 2010 (UTC)

Oh, wait, I can click on them. And they work. It's the OTHER email address that's the problem because I have to send the links BACK, and I can't click on plain text emails FROM that address. How this email ended up not being plain text, I have no idea.Vchimpanzee · talk · contributions · 18:27, 2 July 2010 (UTC)

For the umpteenth time - this is not your personal computer help desk. This is for discussing technical issues regarding English Wikipedia. OrangeDog (τ • ε) 13:03, 5 July 2010 (UTC)

Secure problems

I notice that https://secure.wikimedia.org/ throws an error about containing unencrypted content in Firefox (latest update). Can this be fixed? 76.117.247.55 (talk) 05:00, 4 July 2010 (UTC)

That is because the images are served as non-secure. Peachey88 ^{(Talk Page · Contribs)} 05:06, 4 July 2010 (UTC)

There is some previous discussion here. And gives a rather ugly "fix" that blocks the images. HumphreyW (talk) 06:13, 4 July 2010 (UTC)

Currently, the Wikimedia foundation does not have the financial resources to deliver media in an encrypted manner...this is why you are getting a msg about unencrypted content.Smallman12q (talk) 12:12, 4 July 2010 (UTC)

Well not only that, it is probably more of a time and manpower issue. That could be solved with throwing money at it, but apparently it isn't that high up on the priority list. Using the financial resources to fix problems like this really requires them to be of top priority. —TheDJ (talk • contribs) 12:29, 5 July 2010 (UTC)

mwEmbed + Twinkle = video not loading

When I enable the mwEmbed and Twinkle gadgets, videos (such as the one on today's featured article) don't load. Firefox's error console says:

Error: _this.moduleLoadQueue[moduleName] is undefined
Source File: http://prototype.wikimedia.org/mwe-gadget/mwEmbed/ResourceLoader.php?class=mwEmbed,mw.style.mwCommon,mw.EmbedPlayer,mw.style.EmbedPlayer,$j.ui,mw.PlayerControlBuilder,$j.fn.hoverIntent,$j.ui.slider,mw.PlayerSkinKskin,mw.style.PlayerSkinKskin,$j.fn.menu,mw.style.jquerymenu,mw.TimedText,mw.style.TimedText,mw.SwarmTransport,mw.EmbedPlayerNative&uselang=en&urid=r136
Line: 55

Chrome says:

ControlBuilder,$j.fn.hoverIntent,$j.ui.slider,mw.PlayerSkinKskin,mw.style.PlayerSkinKskin,$j.fn.menu,mw.style.jquerymenu,mw.TimedText,mw.style.TimedText,mw.SwarmTransport&uselang=en&urid=r136:55Uncaught TypeError: Cannot set property 'loaded' of undefined

Without Twinkle, mwEmbed works. -- Nx / talk 13:42, 5 July 2010 (UTC)

Probably clashing global variables or something. I'll ask mdale to look into it, though with Wikimania arriving, it might take a while before it is fixed. —TheDJ (talk • contribs) 14:06, 5 July 2010 (UTC)

User:AzaToth/morebits.js messes with the Array type's prototype. Here are the culprits. -- Nx / talk 15:30, 5 July 2010 (UTC)

That means mwEmbed is broken! _→AzaToth 15:35, 5 July 2010 (UTC)

Yeah, it's trying to use an object as a php-style associative array: for(var moduleName in moduleRequestSet) -- Nx / talk 15:38, 5 July 2010 (UTC)

Don't you mean an "Array ([])" as a "Object ({})"? _→AzaToth 15:40, 5 July 2010 (UTC)

Array() is an object :) But yeah, Javascript's confusing for...in / foreach implementation is to blame here. -- Nx / talk 15:41, 5 July 2010 (UTC)

This was fixed in r69064, but it has not yet been deployed. —TheDJ (talk • contribs) 23:05, 5 July 2010 (UTC)

Image not displaying, as if deleted

I noticed today that at Crucifixion and Crucifixion in the arts, the image file File:Dionysus Crucifixion.gif displays as though the file has been deleted (a red link), as it does here, and yet the file does indeed still exist at Commons commons:File:Dionysus Crucifixion.gif. I cannot find anything wrong with how the file name is entered on these pages. Am I missing something, or is there a bug? Thanks. --Tryptofish (talk) 19:01, 5 July 2010 (UTC)

We had the same issue for the WPVG ad, see Wikipedia talk:VG#WikiProject ad redlinked. Anomie was able to fix it with some trickery, apparently using a full url with &action=purge: http://commons.wikimedia.org/w/index.php?title=File:Dionysus_Crucifixion.gif&action=purge. I would assume this is related to the recent outage. –xeno^talk 19:03, 5 July 2010 (UTC)

Thanks. I see that fixes it at this page, but not at the two articles named above. Sorry for my relative lack of technical fluency, but how do I do that purge at the Commons page? --Tryptofish (talk) 19:20, 5 July 2010 (UTC)

They just needed a purge as well. The easiest way is to enable the Gadget called UTCLiveClock which is available both here and at commons, and then click the clock that appears in the top right corner which acts as a purge link. –xeno^talk 19:23, 5 July 2010 (UTC)

Thanks for both the help and the gadget advice. I've just added the gadget. --Tryptofish (talk) 19:29, 5 July 2010 (UTC)

A bug (?)

Copied from misc page:

In my post at the talk page of WikiProject Chemistry i tried to bring to notice an issue which in my view affects every WikiProject and is thus a serious bug. Please see the post for details. I hope someone has a solution. --Siddhant (talk) 20:41, 5 July 2010 (UTC)

It is not a bug, it is normal behavior. A category is for performance reasons only really aware of it's own level and not of the pages in subcategories. —TheDJ (talk • contribs) 23:08, 5 July 2010 (UTC)

Vector Search Problem

I initially posted this at the Help Desk, although I kind of had it in the back of my head that I should post this here. When I search for certain items, it seems that the search bar brings up strange results. For instance, searching "Doctor Who (" turns the "(" into an "a" (an example here in this segment of a screen cap) and a search for "Doctor Who h" turns the "h" into an ":" (i'm using Doctor Who examples because that's what I was indeed looking for when I discovered this). I get the bug with other searches (like, "Star Trek o") or when I search for a single word and a space (like "Star " turns "Starting Pitcher" into "Star ing Pitcher" and "Starbucks" into "Star ucks" and "Life " turns "Life Imprisonment" into "Lifebimprisonment"). In fact, I just noticed this now, but searching for "life" then two spaces turns the searches into "Lifebo+(rest of search result phrase beginning with "Life"), three spaces turns the search to "Lifeboa+(rest of search result phrase beginning with "Life"), eg "Lifeboaimprisonment", and so on until "Life+17 spaces" full reveals "Lifeboat (shipboard)" over the search results. I get the same thing if I search words like "time", "star", "rose" and "royal" and a bunch of spaces or any spaces (even sometimes one). Bizarre! Doc Strange^MailboxLogbook 16:42, 3 July 2010 (UTC)

By the way, I have indeed tested this on another computer just to check if the problem was with my computer and/or keyboard and this bug indeed shows up on that computer when the same search queries (ie "Doctor Who (") are typed into the search bar. Others who replied to my message at the help desk noticed this bug and they mentioned that it only appears in the Vector skin. Doc Strange^MailboxLogbook 16:42, 3 July 2010 (UTC)

Yes, I've just noticed this too. Does anyone know what's going on?--Kotniski (talk) 09:36, 5 July 2010 (UTC)

Which browser ? No such problems using Safari here. —TheDJ (talk • contribs) 12:24, 5 July 2010 (UTC)

It's happening for me in IE.--Kotniski (talk) 12:38, 5 July 2010 (UTC)

Version ? Because they kinda differ a LOT. —TheDJ (talk • contribs) 22:59, 5 July 2010 (UTC)

The problem occurred for me in Firefox 3.6.6, IE6 and IE7 on multiple computers. Maybe this doesn't affect Safari. I wouldn't know. Doc Strange^MailboxLogbook 01:42, 6 July 2010 (UTC)

Filed as bugzilla:24282 —TheDJ (talk • contribs) 10:52, 6 July 2010 (UTC)

P.S. Everyone could have been saved a 3 day 'discussion'. If people had started by reporting their browser setup. Perhaps it is time to start new pages with a preloaded form, so that it becomes easier for people to properly report a problem. —TheDJ (talk • contribs) 10:52, 6 July 2010 (UTC)

That wasn't a personal criticism btw, it is quite a common problem in bugreporting. I have now added an editnotice in hopes that people will be reminded of these things. —TheDJ (talk • contribs) 11:11, 6 July 2010 (UTC)

New problem when editing with Chrome

I have no problems reading pages, they look normal. But when I try to edit, the edit page opens normally but within a second changes to one in which the screen is almost completely taken up with the edit field. I can see the edit toolbars (wikiedit, etc), the edit summary field, and just under that the top half of the Save Page, Show Preview, etc. buttons (I can't even remember, should the 'content that violates copyright be above these as they are in the IE browser I'm using right now?). Nothing on the left or right side of the edit field at all, and no scroll bars. It suddenly happened a few minutes ago while I was using Chrome, and all other pages look normal, it's only when editing that I have the problem. Dougweller (talk) 10:45, 6 July 2010 (UTC)

Sounds like the fullscreen mode of WikEd. In the top-right bar of WikEd click the bottom-right icon, it looks a bit like the maximize window icon of MS Windows. —TheDJ (talk • contribs) 11:00, 6 July 2010 (UTC)

Thanks, that's why, as I just discovered, clicking on preview fixes it. I clicked on that be accident, trying to click on the icon above to get down to the save page, etc buttons. Thanks again. Dougweller (talk) 11:04, 6 July 2010 (UTC)

Template sandbox and testcases templates

In my opinion the page headers provided by {{template sandbox notice}} and {{template test cases notice}} are very useful. Is there anyway that the information can be displayed on such pages system wide and without user intervention. If such a thing were to be implemented, maybe there should be a way to disable it in preferences as someone would probably object. Maybe not.  –droll [chat] 08:04, 5 July 2010 (UTC)

You mean like edit notices, but then without editing? In that case, no such a thing does not exist. These boxes are auto-added by the {{documentation}} template however. —TheDJ (talk • contribs) 23:02, 5 July 2010 (UTC)

You have lost me. Are you saying the the documentation template on the template page causes some message to appear on a different sub-page. I don't think so. So lets start over. The {{template sandbox notice}} template displayed a message on the sandbox page. At least I think it does. What I am asking is is it possible to display the message without the template. All kinds of messages appear on some of my pages. So it seems there is a mechanism that in place to the kind of work I'm talking about. It's sleepy time. talk more tomorrow.  –droll [chat] 11:17, 7 July 2010 (UTC)

anon-only options showing up on named users?

Resolved

 – Local issue only. –xeno^talk 12:45, 7 July 2010 (UTC)

Strange, I have the option to

• Block anonymous users only, allowing registered users to edit as normal.

when looking at the block form for a named user? –xeno^talk 23:34, 6 July 2010 (UTC)

I'm not seeing it, myself. All accounts, or just one user in particular? —DoRD (talk) 00:03, 7 July 2010 (UTC)

This option is always there, but a script hides it when you enter a username. Is javascript enabled? Is it throwing any error messages? This is the script that does that: http://bits.wikimedia.org/skins-1.5/common/block.js -- Nx / talk 00:15, 7 July 2010 (UTC)

Thanks, this explains it. –xeno^talk 12:46, 7 July 2010 (UTC)

Works for me (Safari 5.0 under Mac OS X 10.6); the problem's certainly not universal. {{Nihiltres|talk|edits|⚡}} 00:32, 7 July 2010 (UTC)

Making Switch parser active only when there is a certain entry in an template

Resolved

 – Fixed using #ifeq:. —MC10 (T•C•GB•L) 19:56, 7 July 2010 (UTC)

The m:Template:Extension changes color depending on what an editor enters as the property of {{{switch}}}

Unfortunately, the parser instructions at m:Help:Extension:ParserFunctions are not very helpful.

So if I wanted to make a section of an infobox template appear or not appear dependent upon what the editor enters in, how would I do this?

Say I have this template:

{{Infobox
|header1 = {{#switch: {{{{{found|}}}|Idaho={{{found}}}|}} }}
| label1 =  Location
| data1 =  {{#if:{{{found|}}}|{{{found}}} }}
}}

I know this coding is incorrect....

How would I make it that, ONLY if {{{found}}} = Idaho then header1 will be visible? But if {{{found}}} is something else, it will not be visible?

Thank you so much in advance for your help, I am so confused and don't know where else to ask on wikipedia. God bless. Adamtheclown (talk) 05:40, 7 July 2010 (UTC)

maybe try:

{{Infobox
|header1={{#ifeq:{{{found}}}|Idaho|{{{found}}}}}
| label1 =  Location
| data1 =  {{{found|}}}
}}

Bawolff (talk) 05:57, 7 July 2010 (UTC)

(ec) Bawolff on IRC was kind enough (again) to write:

after reading your post on the vp, sounds like your more after #ifeq than #switch:

{{#ifeq:{{{param}}}|some value|what to print if some value = param| what to print if not}}

Adamtheclown (talk) 06:00, 7 July 2010 (UTC)

Thanks Bawolff! you are so wonderful. Adamtheclown (talk) 06:00, 7 July 2010 (UTC)

Content Encoding Error

Any time I try to view a page that does not exist on Wikipedia (a redlink), I have gotten this message. This just recently started happening. Anyone got answers?  A p3rson ‽ 20:32, 7 July 2010 (UTC)

P.S.: Firefox 3.6.4, running on Vista Home Premium x64

Seems to be a known problem with Firefox, with somewhat vague advice for fixing your PC config given here: http://support.mozilla.com/en-us/kb/error+loading+web+sites#Content_Encoding_Error — Richardguk (talk) 21:00, 7 July 2010 (UTC)

Which license for templates?

Which license do wiki templates fall under? The wiki content license (Creative Commons) or the wiki software license (GPL2)? What about other sites that use MediaWiki? SharkD  Talk  21:29, 7 July 2010 (UTC)

Templates are but data, CC all the way. Gwen Gale (talk) 21:30, 7 July 2010 (UTC)

So, does that mean if a template were copied from Wikipedia to another MediaWiki wiki, some attribution would be required? Also, is this stated somewhere that where people can easily find it? SharkD  Talk  22:29, 7 July 2010 (UTC)

At the Terms of Use linked to at the bottom of every Wikipedia page. --Cybercobra (talk) 22:46, 7 July 2010 (UTC)

Usually when I copy past something, i link to the specific revision id from where I copied, in my edit summary. Then at the very least you have a reconstructible history. —TheDJ (talk • contribs) 00:49, 8 July 2010 (UTC)

Seems weird that templates are CC (well, for some of them it makes sense...), but whatever. SharkD  Talk  05:22, 8 July 2010 (UTC)

External links

Resolved

Usually, external links are written in the way [http://www.google.com.ar/ Google], which produces Google. But what happens when there are brackets inside the URL, and the link gets broken? Is there an alternative way to link? MBelgrano (talk) 02:21, 8 July 2010 (UTC)

Use percent-encoding. For example, replace "[" with "%5B" and "]" with "%5D". Anomie⚔ 02:53, 8 July 2010 (UTC)

(ec) You'll need to replace the offending characters as described here or use the {{urlencode:string}} magic word as described here. —DoRD (talk) 02:57, 8 July 2010 (UTC)

Thank you, the replacement worked MBelgrano (talk) 00:24, 9 July 2010 (UTC)

Image:Transparent.gif

Resolved

 – Resolved by purging

I was looking something up on Hydrogen and noticed a bunch of redlinks in the periodic table in the top right. This turned out to be a failed link to Image:Transparent.gif - which is linked to by a plethora of pages. I then noticed I wasn't logged in, so I logged in. The file appeared. I logged out. The file redlinked.

Something isn't right here - I tried switching browsers to see if it was cookie-related, but the problem persists (newest Opera and IE7). The gif simply isn't there if you're not logged in.

I can imagine that some features and tools should only be presented to admins/bureaucrats/etc, so there's levels of things that are and aren't presented to different types of users. But why this applies to a utility image is beyond me - also why logged-out (ie IP) users not able to access some things that are accessible to logged-in users without any rights, except things like (semi)protection which are handled serverside. --Firien ^{need help?} 09:37, 8 July 2010 (UTC)

A while ago, that image suddenly went missing. These pages when you are not logged in are from the squid caching servers. I have purged the page of the image, and i think that should clear the pages from the squid services. —TheDJ (talk • contribs) 10:44, 8 July 2010 (UTC)

Cheers, that seems to have fixed it --Firien ^{need help?} 12:07, 8 July 2010 (UTC)

String handling

Is there any means of manipulating strings (e.g. chopping the first character)? I need something like this to manipulate a parameter. --Redaktor (talk) 14:24, 8 July 2010 (UTC)

We have an entire category of string manipulation templates. Although I don't have time to check all of them, {{str crop}} seems to be something along the lines of what you're looking for, only backwards. If there's no alternative you could try {{str_right|{{{1}}}|{{#expr:{{str_len|{{{1}}}}}-{{{2}}}}}}} (which is that template with str_right instead of str_left). Bear in mind that all of these templates are expensive and somewhat capricious. Intelligentsock 14:39, 8 July 2010 (UTC)

Many thanks. Yes, that is what I was looking for. In fact, I think {{str right|{{{1}}}|2}} will chop the first two characters, which is what I really need. --Redaktor (talk) 15:03, 8 July 2010 (UTC)

"Do not show page content below diffs" and pending changes

Pending changes seems to ignore this setting, adding &diffonly=0 to the URL. Is there any way to show only the diff to avoid the rather long rendering time when reviewing? (It seems that there should be a setting for this somewhere that furthermore should be independent from the other setting, and I still get the reviewing interface when deleting the &diffonly=0 part of the URL). I am not looking for CSS based fixes since they do not actually prevent rendering and reduce load time. PleaseStand ^(talk) 19:34, 8 July 2010 (UTC)

I'm a little confused by this. The "do not show page content below diffs" preference doesn't add anything to the URL. Additionally, it's certainly working (for me at least) on pending changes too. How is the &diffonly=0 being added? If you want "diffonly", it should be &diffonly=1. Calvin 1998 ^(t·c) 22:21, 8 July 2010 (UTC)

How to reproduce: 1) Set My preferences → Misc → Diffs → Do not show page content below diffs. 2) Go to Special:OldReviewedPages. 3) Click on any diff link (the text of each link is the word "review"). It will ignore your "do not show page content below diffs" setting (each diff URL always has &diffonly=0 at the end, which overrides the preference). There does not seem to be a preferences option to suppress the diffonly=0 or replace it with diffonly=1. PleaseStand ^(talk) 23:05, 8 July 2010 (UTC)

I have opened ticket 24315 for this problem. —TheDJ (talk • contribs) 23:51, 8 July 2010 (UTC)

CSS opacity

Hi there, is CSS opacity still working? See Portal:Cornwall/Random banner/1, it's not working anymore, and there haven't been any changes to that page. If it is still working does anyone know how to fix it on that page? Many thanks, --Joowwww (talk) 22:10, 8 July 2010 (UTC)

The rendered HTML contains style="/* insecure input */". It seems the CSS property filter is considered insecure by MediaWiki and so the whole style is ignored. After I removed the filter, it started working as it should in Firefox, but I think opacity doesn't work in IE. Svick (talk) 23:26, 8 July 2010 (UTC)

Correct, filter is a security problem. See also the bugreport and the commit. —TheDJ (talk • contribs) 23:43, 8 July 2010 (UTC)

I suppose if someone really wanted they could ask for filter:alpha(opacity=...) to be allowed. Changing the regex from filter\s*: to something like filter\s*:(?!\s*alpha\s*\(\s*opacity\s*=\s*\d+\s*\)\s*(?:;|$)) might work. Anomie⚔ 01:18, 9 July 2010 (UTC)

changing a page title

I believe "page view statistics" is incorrect. Shouldn't it be "page-view statistics"? Kdammers (talk) 01:40, 7 July 2010 (UTC)

No. See Page view. Gary King (talk · scripts) 06:11, 8 July 2010 (UTC)

There is nothing in the Page view explanation saying why it should be as it is rather than as I propose. General English grammar brings the two words two together with a hyphen if they are related to each other rather than the second one being related to the last word; if the second word is closer in meaning to the last word (i.e., if the first word modifies the second two words), then no hyphon is used. http://owl.english.purdue.edu/owl/resource/576/01/; "Use a hyphen to join two or more words serving as a single adjective before a noun: a one-way street chocolate-covered peanuts well-known author"

Kdammers (talk) 10:39, 9 July 2010 (UTC)

Slow Loading from Internet to Wikipedia x 2 weeks

Did some sort of over-all structural change happen about 2 weeks ago that caused the Wikipedia to now take eons to open? (Maybe that is when I started to notice it...) When I Google a subject and then from the list provided, click on the Wikipedia link, it takes forever for the page to open - even when I am on Cable or high-speed wifi or on other computers. Sometimes I just open the link in a new window and set it off to the side to eventually load at its pleasure. Wikipedia used to be one of the fastest sites to load, and now it is markedly slower than any other site I access lately. I am just wondering why. Thanks Saudade7 18:46, 8 July 2010 (UTC)

• I should add that once I am *in* the Wiki the links work fine. It's just getting in that's really difficult. And yes, I have cleared my cache, cookies, and history and used various browsers, etc. Saudade7 18:51, 8 July 2010 (UTC)

Having an empty cache would have the effect of making the first load slow, but all subsequent pages fast. OrangeDog (τ • ε) 11:30, 9 July 2010 (UTC)

Recent Changes in Real Time

Firefox 4 beta is out, and it has Websockets support.

For those who have installed it, I wrote a tool that updates the recent changes dynamically.

Here is a demo for en.wikipedia : http://toolserver.org/~thomasv/wprc.html

And here is a script that turns the RC page of any wiki into a self-updating page:

importScriptURI('http://wikisource.org/w/index.php?title=MediaWiki:DynamicRC.js&action=raw&ctype=text/javascript'); 

ThomasV (talk) 07:41, 9 July 2010 (UTC)

It's cool, but it's almost impossible to click on any links, because it scrolls so fast (also Chrome doesn't like the script, it errors out with an origin mismatch) -- Nx / talk 09:33, 9 July 2010 (UTC)

For some reason Chrome requires the page to be on the same server as the websocket. I hope Firefox will not adopt the same policy in the future.

About scrolling : yes, I'm going to add something like that...

ThomasV (talk) 09:49, 9 July 2010 (UTC)

Chrome 5 implements an older version of the specification. It works on Chrome 6. -- Nx / talk 10:28, 9 July 2010 (UTC)

Deep linking to FAMSF website

I need to create links to several of the Thomas Eakins works in the collection of the De Young Museum. To see them, here, search for "Eakins." You will be presented with several paintings, and clicking on them brings you to the description pages I'd like to link to. However, the website does not use canonical URLs. Is it still possible to link to those pages? Raul654 (talk) 05:29, 10 July 2010 (UTC)

It's technically possible. You need to find the URL of the page containing the text for the picture you are interested in. Most browsers you can probably do this by right-clicking on the text and choosing something like "Properties" from the list of options your browser offers. Copy the entire URL and paste it into the wikipage but then delete everything from the first ampersand onwards. For example, if the URL is http://gallery.famsf.org/gallery/artworkDetails.htm?record=61009&idx=3&galleryId=&searchType=advanced&keyword=Search By Keyword&exhibit=&country=&century=&gallery=&building=&barCode=Search By Bar Code&accessionNumber=Search By Accession#, change this to just http://gallery.famsf.org/gallery/artworkDetails.htm?record=61009. The record number is all it needs to display the relevant picture and text, though the resulting page does not include all the menus so you probably ought to include a separate link to the relevant section URL. Not very neat but it works.

As to whether this is permitted by the terms of the website, it seems to be such a bad website that, in my browser, clicking on "terms and conditions" at the bottom simply returns you to the main page. But others here may have some more informed advice about whether deep-linking is in principle acceptable.

— Richardguk (talk) 06:42, 10 July 2010 (UTC)

Great! Thanks a bunch - that was exactly what I needed. Raul654 (talk) 15:10, 10 July 2010 (UTC)

Cite error messages not shown in IE8 – CSS "error" class

I've just learned that cite errors, such as the ones discussed here, are not shown to Internet Explorer 8 users. I see that the text is contained in the CSS class, "Error", and I presume that is the cause but does anyone know if this is an intended effect or a bug? AJCham 13:05, 10 July 2010 (UTC)

I am looking into this. ---— Gadget850 (Ed) ^talk 13:18, 10 July 2010 (UTC)

Is there a version of Wikipedia:Alternative text for images that shows the error messages? The error messages should never have shown in the Wikipedia namespace— they are enabled only for user, main (article), template, category, help and file namespaces by use of {{broken ref}}. We should probably enable error messages for Wikipedia namespace. ---— Gadget850 (Ed) ^talk 13:54, 10 July 2010 (UTC)

Sure, this is the version referred to by the user in the above talk page link. AJCham 14:00, 10 July 2010 (UTC)

I do not see any cite error messages with FF 3.6.4. ---— Gadget850 (Ed) ^talk 14:18, 10 July 2010 (UTC)

I was using FF 3.6.3 and have just now upgraded to 3.6.6. The errors are still visible to me, in both Monobook and Vector. My Firefox version is the Ubuntu build distributed by Canonical – don't know if that would make a difference. AJCham 14:38, 10 July 2010 (UTC)

Try purging the page here or bypass your cache with control shift R. ---— Gadget850 (Ed) ^talk 14:49, 10 July 2010 (UTC)

Do the people seeing the error messages on that history link have their language in their account preferences set to something besides "en"? That could even explain the original problem, if Trafford09 was logged in in IE8 and not in Firefox. For example, compare "en" versus "en-GB". Anomie⚔ 14:52, 10 July 2010 (UTC)

And I even documented that. If they don't see the link at the end to the help page, then that is the likely explanation, as it is custom. Would not explain the IE8 issue though. ---— Gadget850 (Ed) ^talk 14:56, 10 July 2010 (UTC)

Wonder if we could make the en-GB MediWiki pages transclude the en versions? ---— Gadget850 (Ed) ^talk 15:10, 10 July 2010 (UTC)

Good spot Anomie. I do indeed use en-GB. AJCham 15:12, 10 July 2010 (UTC)

I have never figured out what use it is to set the language to en-GB. The only thing that the language setting does is to set the MediaWiki messages. Here is the message you should have seen: [5] The yellow id the default, which shows with en-GB and the green is what you see with en. It isn't just the cite errors, it is a multitude of messages. ---— Gadget850 (Ed) ^talk 15:21, 10 July 2010 (UTC)

There are 23 messages at the moment where the MediaWiki defaults differ between en and en-gb.[6] Things like "color" vs "colour" or "-ize" vs "-ise". Anomie⚔ 16:16, 10 July 2010 (UTC)

Wikipedia robots.txt settings to crawl the live stuff?

I would like to know what are the right robots.txt settings to put in my crawler to be able to download wikipedia from online following wikipedia policy. I will apreciate any information that could help me solve this issue. Here are some screen capture i took when i put the crawler to download wikipedia. http://www.ojoss.com/dumpd/

Thanks odelinespinosa@gmail.com —Preceding unsigned comment added by 66.75.76.204 (talk) 21:43, 10 July 2010 (UTC)

I'm not sure what dumpd is, but this error may be because it doesn't set the User-Agent header. But, unless it's only very few pages, you shouldn't download pages from Wikipedia automatically. You should use the API or database dumps instead. Svick (talk) 22:17, 10 July 2010 (UTC)

Please do not crawl wikipedia in this mannor. Instead consider downloading a database dump (google "wikipedia database dump") or use index.php if you really need the live stuff. Example: To get the live version of the page: Apple point the crawler to http://en.wikipedia.org/w/index.php?title=Apple&action=raw&templates=expand. Tim1357 ^talk 00:03, 11 July 2010 (UTC)

That's assuming he wants the wikitext. Since we're currently not updating static HTML dumps, if he wants the HTML, crawling the site is the best way to get it. Unless the crawler logs in, it will mostly hit the squid cache. Mr.Z-man 00:23, 11 July 2010 (UTC)

stats.grok.se

Stats only 5 days of July. Last stats July 6. Anybody know someone who can get this up? 69.237.157.250 (talk) 01:41, 11 July 2010 (UTC)

Mathbot archiving glitch

Can anyone more clueful that me work out why Wikipedia:Articles_for_deletion/Administrator_abuse_on_Wikipedia isn't being archived by MathBot even though it's closed? I've seen malformed closes on AfDs resulting in MathBot being confused, but I can't see the problem on this one. It's causing the list at Wikipedia:Articles_for_deletion/Old to be stretched longer than it should be. It could be fixed by just commenting out the AfD on the 27 June page, but it'd be nice to fix it properly. Black Kite (t) (c) 21:47, 11 July 2010 (UTC)

Vector search dot (. period) suggestions

If you type Ph.D into the Wikipedia search box in the Vector skin, the first two suggested matches are identical, "Ph.D".

If you try the same in Monobook, the second suggestion is "Ph D" with a space instead of a dot.

In Vector, the second suggestion - although it is "Ph.D" - actually links to Ph D.  Chzz  ►  01:35, 12 July 2010 (UTC)

I think it might be because the software ignores the fullstop character when providing the 'autocomplete' search results. -- œ^™ 10:18, 12 July 2010 (UTC)

It's strange that it would return different results based on your skin, though. Gary King (talk · scripts) 17:23, 12 July 2010 (UTC)

The Vector search uses completely different code than the other skins. This is by design and was never tested before roll-out. OrangeDog (τ • ε) 18:11, 12 July 2010 (UTC)

Hurray for untested code. Gary King (talk · scripts) 18:22, 12 July 2010 (UTC)

Images disabled, the semi-prot notice overlaps the GA note

Forgive me if this is the wrong place; feel free to point me somewhere else.

If I disable images in my browser - Firefox 3.0.19, Windows XP, monobook - and look at a semi-protected page which is a 'Good article', the text This article is semi-protected to prevent violations of Wikipedia's biographies of living persons policy. overlaps both the article title and the text This is a good article. Click here for more information. - I have uploaded a screenshot here. I'm using 1024x768.

Another user, on Firefox 3.6.6 on Ubuntu, using both Vector and Monobook, has the same effect; although at 1600 resolution it does not overlap the title part, but it does if they shrink the window.

A third user checked and saw the problem with Firefox 3.6.6 on Windows 7.

Again, apols if this is a known issue.  Chzz  ►  17:16, 10 July 2010 (UTC)

The topicon templates ({{Good article}}, protection icons, etc.) simply need alt parameter in the image wikicode, like |alt=Good article. Not sure if anybody will be willing to do the change though. — AlexSm 15:38, 13 July 2010 (UTC)

The problem is there precisely because an alt text is used. When images are disabled, the alternate text is displayed. There is not enough space there to visually display those beside one another, but if we removed it, the notices would be useless to blind users, in whose interest we use alt text in the first place. I think this can probably be counted as a known issue. If someone has a good idea of a way around the problem, I'd be interested in hearing it. I would oppose making the messages so short that they are not informative, so solutions that adjusted the icon formatting when images are disabled would be preferable. {{Nihiltres|talk|edits|⚡}} 01:32, 24 September 2010 (UTC)

EditCounterOptIn.js

An external website says that it can only present more organized information about a user's edits if the user opts in by creating a page EditCounterOptIn.js under the user page. When attempting to start that page, the heading says

Code that you insert on this page could contain malicious content capable of compromising your account. If you are unsure whether code you are adding to this page is safe, you can ask at the appropriate village pump. The code will be executed when previewing this page under some skins, including Monobook.

and then

Warning: There is no skin "EditCounterOptIn".

Is this cause for concern? Cephal-odd (talk) 05:59, 13 July 2010 (UTC)

No, not with the edit counter. Graham87 07:35, 13 July 2010 (UTC)

1) For the opt-in, you're not placing any executable code on the page - the tool is just checking for the page's existence, and 2) A .js page is used because .js and .css pages are protected from editing by default from anyone except the owner or admins so you don't have to worry about someone else inserting malicious code. Hope that helps explain it. —DoRD (talk) 13:25, 13 July 2010 (UTC)

Thank you both for the explanation. - Cheers, Cephal-odd (talk) 20:39, 13 July 2010 (UTC)

System errors

Something is amiss with the Wikipedia system.

When I clicked on the link to a diff in the edit history of 2010 G-20 Toronto summit, the page which was displayed was the unrelated Residual-current device? A short while ago, I clicked on another link which produced an anomalous result, e.g.,

• Residual-current device ≠ http://en.wikipedia.org/w/index.php?title=2010_G-20_Toronto_summit&diff=prev&oldid=3732922
• Busento ≠ http://en.wikipedia.org/w/index.php?title=List_of_G-20_summits&curid=221667

In other words, the page displayed on the monitor and the http:// at the top of the screen are mismatched?

If this this not the place to alert someone, what is the better venue?--Tenmei (talk) 18:49, 13 July 2010 (UTC)

When I had finished with the above, I clicked on the links; and a new twist had emerged. Clicking on this hyperlink http://en.wikipedia.org/w/index.php?title=List_of_G-20_summits&curid=221667 produced Bearded lady? --Tenmei (talk) 18:52, 13 July 2010 (UTC)

In all cases, it looks like you're missing a few digits off the end of the diff link. http://en.wikipedia.org/w/index.php?title=2010_G-20_Toronto_summit&diff=prev&oldid=373292266 is a valid diff link for 2010 G-20 Toronto summit, and http://en.wikipedia.org/w/index.php?title=List_of_G-20_summits&curid=22166706 is a valid link to List of G-20 summits (and http://en.wikipedia.org/w/index.php?title=List_of_G-20_summits&curid=2216670 goes to Busento via the redirect at Busentinus River). Anomie⚔ 20:24, 13 July 2010 (UTC)

Yep, that'd be the problem. Diff links must refer to the correct diff ID number to work, and if you have the right ID the article name listed in the URL is rather irrelevant. --erachima talk 20:57, 13 July 2010 (UTC)

Search engine choice

{{ISBN}} uses Special:BookSources to give a listing of various search engines and libraries for books. {{coord}} runs ~geohack on the toolserver which does the same for geomapping. Do we have an equivalent facility for text search? Something that lists general search engines and even better, specialized ones too, with a preloaded query? Franamax (talk) 23:13, 13 July 2010 (UTC)

Template_talk:Infobox_comic_book_title#Non-italics_alternative

Can someone who understands templates, and maybe even nesting of templates, and the best way to do things help us add a field to opt out of italic titles? Thanks. If this request should go somewhere else, please say so. - Peregrine Fisher (talk) 03:57, 14 July 2010 (UTC)

An example can be found in the coding at {{Infobox comics team and title}}. However, given the reason for having the italics template in the the 'box is consistency, if an "opt out" is added, then it would really needs three things:

• A field name other than "opt out", "supress", or "italics off". Of the cuff, "2ndary_box" would be likely the best use.
• An associated tracking category - Category:Comics related articles with secondary infobx
• Clear documentation that the flag is for cases where the 'box is actually a secondary infobox, and not for turning off the comic book series only article because an particular editor dislikes the look. Also that there are composite 'boxes that should be used in place of 2+ 'boxes.

- J Greb (talk) 04:09, 14 July 2010 (UTC)

Wikipedia Topicon Problem

{{Autopatrolled}} and {{Reviewer_topicon}}, these two templates are transcluded in my userpage. When I open my userpage, sometimes the topicons seem like this: File:Mybug 02.png (the left light blue bar is the vertical scroll bar; I am using Mozilla Firefox) and sometimes like this one: File:Mybug 01.png. I have also applied it to other pages (ie my userspace and User:Alexanderps), but i have got the same result. Is there anyone having the same problem? How can I fix this problem? -- Amit6 (talk) 11:08, 11 July 2010 (UTC)

I'm guessing you're referring to this version in monobook skin. The answer is in MediaWiki:Monobook.js (current version), which was created after this discussion: MediaWiki talk:Common.js/Archive 16#topicon_part_deux. It applies the fix to vertical position of these absolutely positioned icons. The 2nd screenshot is when the fix is not active: either you have JS off in your browser or there was no CentralNotice at the moment. Which doesn't make any sense to me, since that notice is not the only thing that can screw with the icons (watch/unwatch actions make them look as ugly). Imho you need to ask sysops at MediaWiki talk:Common.js to improve the code. — AlexSm 15:28, 13 July 2010 (UTC)

The problem is that people keep changing how centralnotice works. So the 'fix' is basically useless in most cases anyway. —TheDJ (talk • contribs) 09:08, 14 July 2010 (UTC)

So, apply the fix on a permanent basis, dropping CentralNotice/wgAction checks and also removing addOnloadHook which makes icons "jump" after the page is loaded. — AlexSm 14:08, 14 July 2010 (UTC)

How I stay signed in

Referring to this question, I discovered that with Internet Explorer 8 I can stay signed in if I delete my history without "Cookies" checked.Vchimpanzee · talk · contributions · 21:45, 12 July 2010 (UTC)

Facepalm OrangeDog (τ • ε) 12:48, 13 July 2010 (UTC)

Well, excuse me. I was advised to delete cookies, but obviously I have to do the opposite for this to work.Vchimpanzee · talk · contributions · 14:31, 14 July 2010 (UTC)

Have you considered a superior browser? –xeno^talk 14:32, 14 July 2010 (UTC)

There's nothing wrong with Internet Explorer 8. Changing browsers causes more problems than it solves.Vchimpanzee · talk · contributions · 15:14, 14 July 2010 (UTC)

I don't recall saying there was anything wrong with IE8 (per se), I just suggested a superior alternative for Wikipedia editing. –xeno^talk 15:15, 14 July 2010 (UTC)

Watchlist bot exclusion

The user preferences panel gives an option to exclude bot edits from your watchlist, which is all well and good. However, is it possible to somehow exclude specific users from your watchlist instead? (I would ignore the archival bots and signpost distributor, specifically.) Ignoring all bots is not acceptable to me because I often have images watchlisted which I did not upload, and therefore would not receive the deletion notices for, but which I would not see bot deletion tagging for if I excluded all bots from my watchlist. Thanks. --erachima talk 20:57, 13 July 2010 (UTC)

You could periodically check for changes in file namespace only, with bots edit shown, and hide bots edit in your usual watchlist. — AlexSm 21:28, 13 July 2010 (UTC)

Javascript could be written to do this, possibly modifying User:Gary King/hide pages in watchlist.js. –xeno^talk 21:32, 13 July 2010 (UTC)

Maybe. That script is old and not really well written so a new one would probably have to be written. Probably not by me, though. Gary King (talk · scripts) 19:05, 14 July 2010 (UTC)

Compare selected revision problems

I don't know if it's just me, but I can't compare revisions from a page history at the moment. I get the error messge You have either not specified a target revision(s) to perform this function, the specified revision does not exist, or you are attempting to hide the current revision.. It may just be coincidence, but I've just had the admin bit restored to my account [7] - maybe this has screwed something up. Any thoughts? (For info, I'm in classic skin running on IE6) —  Tivedshambo  (t/c) 12:51, 14 July 2010 (UTC)

Are you clicking the button, or hitting "enter"? If the latter, it is related to Wikipedia:Village pump (technical)/Archive 76#After Revision delete, default action changed for pressing "enter" on history talk page and you'll need to press the button for now. If the former, maybe a better browser might help? –xeno^talk 12:54, 14 July 2010 (UTC)

I'm pressing the button. I'll try it on Firefox tonight, but at present I'm stuck in the office where there's no choice of browser :-( —  Tivedshambo  (t/c) 13:28, 14 July 2010 (UTC)

Ouch. Maybe this script might help as a workaround? –xeno^talk 13:30, 14 July 2010 (UTC)

importScript('User:Superm401/Compare_link.js'); //   turn "compare selected revisions" into a copyable link

• Bugzilla should probably be filed on this iff bugzilla:23747 going live doesn't fix it. –xeno^talk 13:38, 14 July 2010 (UTC)

• On another wiki I made the RevDel interface hidden by default because it's a bit distracting and not needed 99% of the time. Just a simple script in Sysop.js for all sysops. — AlexSm 16:42, 14 July 2010 (UTC)

I don't have any problems under Firefox and IE8, so it's probably a browser issue. —  Tivedshambo  (t/c) 18:17, 14 July 2010 (UTC)

---

Since auto update to Firefox 7.0.1 (been a while), whenever I try to 'Compare Selected Revisions', my system doesn't know what to do with the return index.php, and asks whether to save it or select the program to open it. I suspected was Firefox, but checking on IE 8.0.6001.18702 returning same error. I'm running on XP - Service Pack 3.

Not a show stopper, but quite frustrating. Any suggestions for settings / installations gratefully received.

Best wishes Haruth (talk) 20:50, 14 October 2011 (UTC)

IE update and gadget issues

Resolved

My employer's IT recently updated our browsers from IE6 to IE7 (yeah, a 'lil behind the times), but in doing so, most of my gadgets no longer work, especially popups, HotCat, and the RefToolbar. Since I've never had an issue with this on my home computer on IE7 or IE8, I'm guessing it must be that some of the settings may have changed, but I'm not sure which options to begin fiddling with. Both computers use Windows XP. Any idea where to start, like security and javascript? bahamut0013^words_deeds 17:19, 9 July 2010 (UTC)

I don't know about those three gadgets specifically, but a lot of scripts don't work in Internet Explorer because the browser acts differently from all other browsers. Specifically, it does not count empty spaces as content (other browsers read whitespace as text nodes). Gary King (talk · scripts) 18:16, 9 July 2010 (UTC)

Well, I can rule out it being the browser itself, since it does work on the same browser on my home computer, as I said. But I'm sure that the options and settings are wildly different, being a military system. I'm also pretty sure it's not my network, because they all worked fine on the Navy/Marine Corps Intranet before the update. bahamut0013^words_deeds 18:28, 9 July 2010 (UTC)

Maybe a different JavaScript engine, or security settings that affect page display. And even if both your computers say they are running "IE7", it's possible for them to have different patches to IE7 or to Windows itself. — Richardguk (talk) 06:51, 10 July 2010 (UTC)

Is there anything I can look at to verify this, or any specific settings you think I could try tweaking? bahamut0013^words_deeds 17:51, 13 July 2010 (UTC)

Your best bet is to use a JavaScript debugger, which will tell you if the gadgets are returning any errors when they aren't working. I've never used IE7 before; it appears that it doesn't have a built-in JavaScript debugger, so you'll have to use some of the options available. Here's a decent page that should help you on your way. Gary King (talk · scripts) 16:24, 15 July 2010 (UTC)

NMCI pushed another update, and it seems to have resolved the issue (and created others, but that's a different matter). Thanks anyway! bahamut0013^words_deeds 19:40, 15 July 2010 (UTC)

Content Based Searching

http://docs.google.com/document/edit?id=1luPp50mVg6d1CXAs_Qrdn7f6eJZmAO5tBT2roVyL1o0&hl=en
http://docs.google.com/document/edit?id=1vce2qdjCVVaEEN7ODssSBRjLU1DR0Cew_DkrdJaUESQ&hl=en

Can we use a content based search engine for document search in Wikipedia. The above link will guide you through a mathematical search engine, which uses a Principle Component Analysis to provide most appropriate results. —Preceding unsigned comment added by Prashant.prashn (talk • contribs) 03:56, 14 July 2010 (UTC)

I suggest you make an attempt to make it scalable, and actually test it on real data (like a Wikipedia dump maybe). Theoretical papers on search are all well and good, but no-one cares if you don't have practical working code. OrangeDog (τ • ε) 15:24, 15 July 2010 (UTC)

What's flashing by at the top of my watchlist?

For the past few days, some message has been appearing for a tenth of a second at the top of my watchlist, maybe twice a day. It's not in the page source. I'm still using monobook, so my wild guess is that they're trying to sell me on vector ... anyone know? Anyone know a url where I can see the message? - Dank (push to talk) 15:43, 15 July 2010 (UTC)

I suggest logging out, deleting all your cookies for all Wikipedia and Wikimedia sites (or just all of your cookies), and logging back in. It's likely a message that you've clicked "hide" on and forgotten about. Incidentally, I'm still using Monobook too and haven't noticed anything. --Deskana (talk) 15:47, 15 July 2010 (UTC)

Thanks. In the past, the messages I've hidden in my watchlist have been visible in the page source; this one isn't. - Dank (push to talk) 16:07, 15 July 2010 (UTC)

I notice this all the time. What I do to hide it is, when I'm loading a page, I stop the page from loading before the message disappears, then I click the "Hide" link to hide the box. It usually hides the box for a few weeks before it reappears again. Yes, it is annoying. Gary King (talk · scripts) 19:36, 15 July 2010 (UTC)

Thanks Gary, I'll try that. - Dank (push to talk) 20:58, 15 July 2010 (UTC)

That would be Central and Local notices, Due to how they are loaded, they are done then hidden as appropriate (for example if they are no longer meant to be displayed or are hidden by the user), What you are seeing are the past "OMG! We are on vector now" messages. Peachey88 ^{(Talk Page · Contribs)} 01:31, 16 July 2010 (UTC)

User Subpage

Where exactly do I go to create a user subpage?--Woogie10w (talk) 20:45, 15 July 2010 (UTC)

Somewhere like User:Woogie10w/Sandbox. OrangeDog (τ • ε) 20:57, 15 July 2010 (UTC)

Thanks--Woogie10w (talk) 21:09, 15 July 2010 (UTC)

Changing class information in a template section

If mediawiki:common.css has coding such as this for a class:

table.bigbox{
   border: 1px solid #aaaaaa;
   background-color: #f9f9f9;
   color: black;
}

...and I create this template:

{{bigbox
| image       = {{{image|}}}
| headerstyle = background-color:#FFFFFF
| data1 = {{{helpmeplease|}}
}}

And with the "data1/helpmeplease" parameter I don't want:

border: 1px solid #aaaaaa;

What type of coding can I use to make sure there is no border?

I tried style formatting and nothing worked.

I hope I wrote this clearly, thank you so much in advance for your help! Adamtheclown (talk) 20:12, 15 July 2010 (UTC)

Example
This is an infobox without border.

There is no Template:Bigbox, so I suppose you want general solution for all box meta-templates (like {{infobox}} or {{navbox}})? I'm afraid there is no single solution, but most of those meta-templates should provide a parameter to override the style. For example, in the case of {{infobox}}, you can use this code (the result is in on the right):

{{Infobox
| bodystyle = border: none;
| above = Example
| data1 = This is an infobox without border.
}}

Svick (talk) 11:18, 16 July 2010 (UTC)

Diagram

I would like to make a diagram listing size ranges for Miniopterus bats from the data at List of bats of Madagascar#Family Miniopteridae. This would look similar to File:Subfossil lemur C14 ranges.svg. Is there a way to do this on-Wiki, using some intricate kind of formatting or the EasyTimeline extension, or do I need to create an image? Ucucha 16:19, 16 July 2010 (UTC)

I'm not in favor of (ab)using <timeline> for other purposes, so I would suggest creating an image. SVG is probably the best format, because it's scalable and can be relatively easily edited. Svick (talk) 16:31, 16 July 2010 (UTC)

Wouldn't it be possible to do this via some sort of table syntax, though? I would prefer that, because it would be easier to edit (I'm almost certain new species of Malagasy Miniopterus will continue to be discovered) and to adapt to other similar purposes. Ucucha 16:37, 16 July 2010 (UTC)

How to publish completed article

HOW ON EARTH do I publish my edited page?!? Or make the page "go live"? I cannot find this topic in ANY help sections, ANYWHERE. I've been looking for hours. It must just be ridiculously simple, obvious, and right in front of my face. I just want to move it from my userspace to become an actual Wikipedia article. This is very frustrating. Please help! —Preceding unsigned comment added by SportsScienes (talk • contribs) 16:53, 16 July 2010 (UTC)

International Sports Sciences Association has already been created. Ruslik_Zero 16:57, 16 July 2010 (UTC)

Also see WP:Your first article. However, since the article already exists, you can edit the one above to include all the additional material in your draft. —DoRD (talk) 17:01, 16 July 2010 (UTC)

Skin stats

Are there publicly available stats on the popularity of the various skins among Wikipedians? If so, where are they? I imagine those figures could be quite interesting in light of the recent Vector implementation. Thanks for the help. —Anonymous Dissident^Talk 03:35, 16 July 2010 (UTC)

Not publicly available, I do know what it looked like before vector, March 21st 2009 for en.wp. The figures from back then are here, because I once requested them. I'll try to ask for an update of those numbers. —TheDJ (talk • contribs) 12:20, 16 July 2010 (UTC)

With the help of User:DaB., I now have some new numbers. I first have to explain something. The database is currently in a transition from one format of preferences storage, to another format. Because of this transition, these numbers are on the 269787 users who in the past months have made 'a' change to their preferences. The numbers probably also include all users who in the past had explicitly chosen the monobook skin in their preferences, these users have all been changed to vector. Users who had default skin and didn't change their preferences are probably not included in this data. As such it is very difficult to make any definitive explanations on this data, but it seems that about 34.000 users have chosen to go back to monobook. (16.000 users use modern skin). There are about 600.000 en.wp Wikipedians (people with more than 5 edits) in total. [8] The list with current known skin selections. —TheDJ (talk • contribs) 13:00, 16 July 2010 (UTC)

Is 600000 the right number? It looks like that came from column A in the first table on the page you linked, but that's accounts with at least 10 edits; column C might be more useful, which indicates that we've been trending about 40000 accounts with at least 5 edits per month recently.

It would be interesting to know skin statistics for "active" users, for some value of active (e.g. user_touched in the past few months, if the comment in tables.sql is accurate). For example, vandals and socks that managed to rack up more than 5 edits before being blocked years ago would be counted in those numbers, as will formerly-active editors who have left long before Vector. Anomie⚔ 13:43, 16 July 2010 (UTC)

I copied the stats to here: Wikipedia:Skin#Raw_skin_usage_data. Figure they'd be useful. Gary King (talk · scripts) 03:51, 17 July 2010 (UTC)

Marvelous. Please continue to update and expand the information at that page. -- Quiddity (talk) 17:22, 17 July 2010 (UTC)

Also, there is a Toolserver issue request is at https://jira.toolserver.org/browse/TS-592 in case that helps anyone coordinate on this. -- Quiddity (talk) 17:22, 17 July 2010 (UTC)

AutoWikiCite

I've written a tool in vb.net 2.0 that allows automatic/assisted {{cite news}} citations with the following websites:

• foxnews.com
• foxbusiness.com
• reuters.com
• nytimes.com (this one is broken)
• bbc.co.uk
• washingtonpost.com
• cnn.com
• latimes.com
• wsj.com
• informationweek.com
• time.com
• nypost.com
• businessweek.com
• usatoday.com/news/

The tool and its source are available at sourceforge. If you find it useful/worthwhile, please leave a note and I'll consider further development...Enjoy.Smallman12q (talk) 23:39, 16 July 2010 (UTC)

You should add it to Wikipedia:Citation tools

Looks intriguing, although Windows-only. If I wasn't confined to linux, I'd give it a try.

I've been using the bookmarklet at User:Bazzargh/citemark for a few months. It's very simple, but does what it says on the box. -- Quiddity (talk) 17:33, 17 July 2010 (UTC)

I've developed something similar for AWB: User:Rjwilmsi/CiteCompletion. Rjwilmsi 18:38, 17 July 2010 (UTC)

It's written in .net so I will make a mono port for the 1.0 release.=DSmallman12q (talk) 21:02, 17 July 2010 (UTC)

Citation needed and paragraphs

The {{citation needed}} template is causing the paragraph following the one to which it is attached to run together. The bug is only evident when the second paragraph begins with a wikilink. For instance,

Poodles make great bacon sandwiches.

Animal cruelty should not be allowed on Wikipedia.

works fine. But,

Poodles make great bacon sandwiches.

Animal cruelty should not be allowed on Wikipedia.

results in the two lines being run together. This is not happening here on this page: it appears to only be a problem in the main namespace, which is very odd, but it is easily verified by pasting the above text in to an article and viewing it in preview. I have tried this in IE6, Firefox 3.5 and 3.6 - all with the same result.

I have no idea how to fix this. SpinningSpark 14:03, 17 July 2010 (UTC)

{{Citation needed}} uses {{fix}} which uses namespace detection, so whatever is causing the problem only shows in mainspace. I reported this two years ago; see Template talk:Fix/Archive 1#Fix at end of paragraph. I think a space after {{citation needed}} is a workaround. ---— Gadget850 (Ed) ^talk 14:32, 17 July 2010 (UTC)

I've no idea why, but the general problem seems to be that whitespace is ignored between a category declaration and a wikilink (and the "fix" template ends with category declarations). A workaround would probably be to add a no-break space after the category declarations in the fix template.--Kotniski (talk) 14:49, 17 July 2010 (UTC)

PS - {{fix}} suppresses the categories outside mainspace, which is why the problem only shows up in mainspace.--Kotniski (talk) 14:54, 17 July 2010 (UTC)

I've made an editprotected request at Template talk:Fix/Archive 1#Fix category breakage; someone recently moved the categories output by {{fix}} from before to after the text output. Moving them back before causes the categories to no longer be at the end of the paragraph, so it all works fine. Anomie⚔ 17:38, 17 July 2010 (UTC)

The fix has been applied by Ucucha in this edit. If you still see the problem, try purging the page. Anomie⚔ 17:52, 17 July 2010 (UTC)

Thanks, that got it. I thought it best not to apply your fix myself as I did not have the faintest idea what I was doing. SpinningSpark 19:12, 17 July 2010 (UTC)

Preference to mark all edits minor by default asked to be removed in bugzilla:24313

Per the above-linked bugzilla, and this discussion, this preference will likely be removed shortly (as users often set it and forget it, causing non-minor edits to be marked minor).

The preference will be reset to the default; those who used previously this preference could probably use a script to restore the functionality, if desired. –xeno^talk 21:30, 8 July 2010 (UTC)

This change will be occurring very shortly; in order to assist users who had been appropriately using this, is anyone available to write a simple script to restore the functionality? –xeno^talk 22:38, 14 July 2010 (UTC)

So was there a point to removing it if you are also after a script to add it back? -- WOSlinker (talk) 22:59, 14 July 2010 (UTC)

if(wgAction == 'edit') {
    addOnloadHook(function minorEdit() {
       document.getElementById('wpMinoredit').checked = true;
    });
}

Several points. Firstly, the existence of a preference in the interface caused some problems by implying that using the preference was always allowed and could not be restricted. Secondly, editors who misuse a script can be restricted by having an administrator remove the script and protect their skin.js. Because there's no way to stop an editor from changing their preferences, the only way to stop abuse of the preference is by blocking. Thirdly, installing a script is a bit more difficult than checking off a preference, and requires the editor to view a message reminding them that they are responsible for the use of the script, so it there should be fewer editors doing so without thinking of the consequences. — Gavia immer (talk) 23:16, 14 July 2010 (UTC)

Thank you for writing the script, WOSlinker. Gavia sums it up well above. I'm sure there were a few folks out there who were happily gnoming along without incident that may be put off by this bugzilla, so that's why I've requested the script. Much obliged! –xeno^talk 00:26, 15 July 2010 (UTC)

I haven't tried the above script yet, but, as is, won't it forget your current unchecked value on a refresh due to preview or showing changes? Mark Hurd (talk) 05:58, 15 July 2010 (UTC)

I don't think so. The minor-edit checkbox retains its condition while showing preview and changes. Someguy1221 (talk) 06:08, 15 July 2010 (UTC)

I think Mark was asking whether the script would re-check the checkbox after pressing "Preview". I haven't tested it, but I don't think it will as it seems wgAction is "submit" in those cases, rather than "edit". Anomie⚔ 11:46, 15 July 2010 (UTC)

If it's checked when you click the "Preview" button, it will be checked on the preview page, because the form contents are always reproduced when you preview. If you had the script auto-check on the preview page too, then if you unchecked it and previewed, it would become re-checked, which is not the desired behavior. —Aryeh Gregor (talk • contribs) 20:08, 16 July 2010 (UTC)

Anomie has it right and I have now confirmed Preview and Changes both work as expected. Mark Hurd (talk) 04:38, 19 July 2010 (UTC)

Here's something better. It remembers your last minor edit setting in a cookie. -- WOSlinker (talk) 08:20, 15 July 2010 (UTC)

if(wgAction == 'edit') {
    addOnloadHook(function minorEdit() {
       addHandler(document.getElementById('wpMinoredit'), 'change', minorEdit_change);
       if (document.cookie.indexOf("minorEdit=true")!=-1) {document.getElementById('wpMinoredit').checked = true };
    } )
}

function minorEdit_change() {
    var e = new Date();
    e.setTime( e.getTime() + (24*60*60*1000) ); // one day
    document.cookie = 'minorEdit='+document.getElementById('wpMinoredit').checked+';expires=' + e.toGMTString();
}

The above script sets future edits to the same setting as your most recent edit. It does this for one day before reverting to marking all edits to major edits. I don't see the purpose of this script and it seems extremely error prone to me, as there is no true default in this case and the "default" that is set expires without warning. I recommend against its use. Bigmantonyd (talk) 23:35, 21 March 2011 (UTC)

Could you please instruct the non-technical folks (including me) exactly what steps to take to use the script? atakdoug (talk) 06:20, 16 March 2011 (UTC)

As per Atakdoug - where in wikipedia does one apply this script - I am a gnome so most of what I do is minor. I have some understanding of Java but where to write it to? Mark^Dask 12:27, 17 March 2011 (UTC)

(edit conflict)Here's my experience: First, I tried this in my Common.js. When it didn't work, I then put it in my Vector.js, but it still didn't work. I gave up and let it be for a few minutes. Then I thought I'd try the first example. The weird part is that just as I began to edit Common.js (I'd already removed the code from Vector.js), I noticed that it had started to work! The minor box was checked! So I left things as they were and continued general editing. At this point, the second example above was in my Common.js, my Vector.js was empty, and all was just fine. Then a few minutes later, I noticed that it had stopped working! So, it appeared as if the code had to be in the Vector.js to work. And that is correct. I put it back into Vector.js, and it started working a few minutes later. So the easy steps are:

• Cut and paste the second example above to your Vector.js file...
  • go to "My preferences"
  • click the "Appearance" tab
  • click and open the Vector skin's "Custom JavaScript" link
  • click "Edit"
  • paste the second example above into the Edit screen
  • save the edit
• And you're finished! Remember that...

1. If you use one of the other skins to edit, you'll paste the code into the "Custom JavaScript" for that skin,
2. I could not get this to work in the "All skins" Custom JavaScript,
3. AND MOST IMPORTANTLY wait a few minutes while you continue to edit, because the change might not be immediate.  — Paine Ellsworth ( CLIMAX )  12:47, 17 March 2011 (UTC)

Excellent - thank you Paine Mark^Dask 04:51, 18 March 2011 (UTC)

Well, don't thank me yet, because it looks as though this doesn't work all the time. Sometimes I open an edit page and the Minor box is checked, and other times it's not checked. I don't know if the devs are still working on this or what. – Paine Ellsworth ( CLIMAX )  07:08, 19 March 2011 (UTC)

Okay, I just converted back to the first choice, the:

if(wgAction == 'edit') {
    addOnloadHook(function minorEdit() {
       document.getElementById('wpMinoredit').checked = true;
    }
)}

That works (so far). – Paine Ellsworth ( CLIMAX )  07:45, 19 March 2011 (UTC)

I don't like this. The choice was made, by me, to set my edits as default. This choice was made by me. This choice is something I stand by. This choice should be respected. I do not like the notion that my edits will have to be manually set to minor while this is my default behavior. I am appalled by this move. Maki (talk) 20:33, 19 March 2011 (UTC)

I also think this is absurd. Maybe someone could convert one of these experimental scripts into a simple checkbox on my Editor Preferences that would let me toggle the minordefault value. Perhaps just unhide the one that is still there. A different repair would be to just set the value in my user record back to true and forget about either the script or checkbox. Anyway -- I've used the 1st (3rd) script in Vector skin's "Custom JavaScript" . -- LantzR (talk) 08:48, 10 June 2011 (UTC)

Switched script to [|Mark_minor_edit] (same code with comments) LantzR (talk) 10:57, 10 June 2011 (UTC)

I am currently assisting a non-technical user who is stuck with having every edit marked as minor with no way to turn it off. That hidden box needs to be turned off for all users by a 'bot. Giving the users a way to turn it back on would seem to be a Good Thing for users such as Makitk above who appear to know what they are doing, but the hidden box needs to be unchecked for users who don't do something special to keep it turned on. I am sure that there are many other users who, like mine, have been spending years inadvertently marking all edits - including major changes - as minor without really knowing what a minor edit is or why their is an m next the every edit. There is a larger issue as well: when a user interface option is hidden, careful thought needs to be given to the users who now have an unchangeable user preference. Guy Macon (talk) 15:04, 10 June 2011 (UTC)

GeSHi clarification before I go to bugzilla

I tried to use the syntax highlighting code "<syntaxhighight lang="rsplus"></syntaxhighight> (for the R programming language, but it appears that rsplus is not supported in wikipedia. Can anyone confirm that "rsplus" doesn't work or let me know if there is a bug listed for GeSHi to update the languages on wikipedia? Thanks. Protonk (talk) 01:50, 15 July 2010 (UTC)

I get the same behavior you do - the docs say rsplus is supported, but trying to use it gives an error message that doesn't list it as one of the supported options. I can't find a bug report for this, so I'd go ahead and make a new one if I were you. — Gavia immer (talk) 02:23, 15 July 2010 (UTC)

• Listed at buzilla #24383. Here's hoping I didn't miss something obvious. Protonk (talk) 03:11, 15 July 2010 (UTC)
  • I've tagged it as "1.16wmf4", which means it should get higher priority for being pushed to the main site. (X! · talk)  · @671  ·  15:06, 18 July 2010 (UTC)

Template:Location map

I create new Template:Location map2 with ideas from de, ru, fr and uk wiki. This template works also with location map templates with y, x parameters, for example: Template:Location map China1, Template:Location map Canada1, Template:Location map Russia1, Template:Location map Africa1. May be it need to include code of Template:Location map2 to Template:Location map?--Амба (talk) 21:17, 16 July 2010 (UTC)

This is a fork of {{Location map}} and is being discussed at Templates for discussion.  –droll [chat] 06:53, 18 July 2010 (UTC)

Two search boxes - any way to make them gracefully switch from side by side to one below the other?

Hey all, recently we redesigned the header for Wikipedia:Reference desk pages and added another search box for the whole of wikipedia in case people miss the one at the top. These are put side by side, see for example Wikipedia:Reference desk/Computing. Problem with this is it increases the minimum horizontal resolution for the header to above 800 pixels.

While I left a suggestion on the talk page, no one has yet came up with what would seem to me to be the obvious solution, make them gracefully switch from being side to side to one below the other as necessary. Does anyone know if this is possible and if so, how to do it?

Nil Einne (talk) 06:33, 18 July 2010 (UTC)

Put them into two divs and float them left: User:Nx/Sandbox -- Nx / talk 06:42, 18 July 2010 (UTC)

The search boxes are contained within a table in Wikipedia:Reference desk/header/howtoask. It looks straightforward to reformat the table if there's consensus to do so. —  Tivedshambo  (t/c) 06:48, 18 July 2010 (UTC)

Btw, the inputboxes are wider than 300px (at least on my computer), so the table cells were wider than 300px despite the header style suggesting that they are 300px wide. -- Nx / talk 06:52, 18 July 2010 (UTC)

Image links

Normally, if I click on a Wikipedia image, it takes me to the description page for that image. Recently, I haven't been able to click on any image at all. I'm still using the MonoBook skin, in IE8. --S-man (talk) 19:06, 18 July 2010 (UTC)

Nevermind. I reset the browser; it's all good now. --S-man (talk) 20:26, 18 July 2010 (UTC)

Cannot use parameters to create a reference inside a template

Resolved

I have been experimenting with the use of a sub-template to create references to an external site which uses structured URLs, created by an existing template {{London Gazette}}.

My intention was that the parameters in my new template which creates the row in the data table would include three for the reference: {{{issue}}}, {{{date}}} and {{{page}}}. Those three parameters would be used by the row-building template to call a sub-template which would create a named reference for one of the columns, as follows: <ref name="{{{issue}}-{{{date}}}-{{{page}}}">{{London Gazette |issue = {{{issue}}} | startpage = {{{page}}} | date={{{date}}} }}</ref>

That would give me properly formatted references, with names which allow cite.php to merge refs to the same page

In my tests I have incorporated error-checking to ensure that all parameters are present, and have successfully passed all the parameters. {{London Gazette}} is called ... but all the parameters in <ref name="{{{issue}}-{{{date}}}-{{{page}}}">{{London Gazette |issue = {{{issue}}} | startpage = {{{page}}} | date={{{date}}} }}</ref> are empty.

To demonstrate this, I have set up a simple test at User:BrownHairedGirl/sandbox, in which the template {{User:BrownHairedGirl/myref]]}} is passed a one-word parameter. As you can see the parameter has a null value both in creating the reference name and in creating the text of the ref.

Is this a known bug, and is there any workaround?

It may sound like an esoteric issue, but in this case it would make it much easier to create and maintain a long series of references, particularly since each reference will be shared by an average of about 30 entries. --BrownHairedGirl (talk) • (contribs) 14:41, 18 July 2010 (UTC)

Using {{#tag:ref}} in place of <ref name="etc.">content</ref> seems to fix it. Intelligentsock 15:13, 18 July 2010 (UTC)

See WP:REFNEST for full syntax and bugs. ---— Gadget850 (Ed) ^talk 15:26, 18 July 2010 (UTC)

Big thanks to both of you. I didn't know about {{#tag:ref}}, but that fixes it all. --BrownHairedGirl (talk) • (contribs) 11:37, 19 July 2010 (UTC)

Prompt me when entering a blank edit summary

Resolved

If the "Prompt me when entering a blank edit summary" feature is enabled in preferences and an edit summary is not entered, it prevents non-autoconfirmed users from saving edits which trigger the captcha. The edit window gets locked into a cycle of entering the captcha, being informed you didn't enter an edit summary and clicking continue to save without one, and being asked for a captcha again. Obviously this is easy to overcome by disabling the feature in preferences or entering an edit summary, but new users may be unaware of what they are doing wrong, and may assume the captcha is the problem 1230049-0012394-C (talk) 16:29, 19 July 2010 (UTC)

See bugzilla:10729 -- Nx / talk 16:33, 19 July 2010 (UTC)

Good to know it's being worked on, thanks! 1230049-0012394-C (talk) 16:35, 19 July 2010 (UTC)

Strange glitch

This appeared while viewing User talk:White Shadows. I'm running The latest Firefox on Mac OS X 10.6.3. ~NerdyScienceDude (✉ • ✐ • ✍) 18:15, 19 July 2010 (UTC)

I'm running exactly the same browser on the same operating system and I'm getting what it's supposed to look like. I'm using Vector. Mr. R00t ^Talk 18:36, 19 July 2010 (UTC)

Yes, it's a recognised bug when transcluding special pages - bugzilla:23293. Nothing to worry about really, just, as you say, a glitch. - Jarry1250 ^{[Humorous? Discuss.]} 18:40, 19 July 2010 (UTC)

Moving to a lowercase article title

Hello. However it is that we are able to display articles such as iPhone with a leading lowercase letter, could someone please do that to Inext? - Richard Cavell (talk) 14:41, 16 July 2010 (UTC)

 Done - thanks to EmilJ. - Richard Cavell (talk) 14:45, 16 July 2010 (UTC)

I have write IPhone. I do it in that way: I clicked in editor the "Insert Link" button, typed on the keyboard "iPo" (in that cases!!!) and clicked in combobox "IPhone" item. So i can't understand: Where is the iPhone article you talking about??? And why the script change my "iPo" to "IPo"+"ne"??? Please fix that bug!

[[9]]

--W.M.drossel (talk) 11:17, 20 July 2010 (UTC)

The script is an intended feature to allow you to find pages quickly. You can always not use the "Insert Link" button and just type [[iPhone]] instead.

For technical reasons, all articles start with an upper-case letter, but you can add {{Lower case title}} to make it look like it starts with a lower-case letter. OrangeDog (τ • ε) 12:04, 20 July 2010 (UTC)

"Changes since I last edited" scipt

Is this script still working for other people? All I get is the "You have not edited this page! (recently)" message (even when I have). I'm still using the monobook skin.--Dodo bird (talk) 23:32, 18 July 2010 (UTC)

It has not been working for me recently in Vector either. —Ost (talk) 12:54, 19 July 2010 (UTC)

Here are the fixes necessary to make it work again. I'll see if I can make it slightly more robust against future changes. -- Nx / talk 19:44, 19 July 2010 (UTC)

Wikipedia talk:WikiProject User scripts/Scripts/Changes since I last edited#Fixed version -- Nx / talk 20:33, 19 July 2010 (UTC)

Thanks for the fix. It's working now. --Dodo bird (talk) 18:33, 20 July 2010 (UTC)

Skin checking

Is it possible to write a script (or is there one already?) that will check which skin and browser you're using and show them a different version of the page based on that? I'm currently talking about the problems at HJ Mitchell's userpage with the English flag at top. It won't look right in Vector when it look right in MonoBook and looks bad in MonoBook when it looks good in Vector. I know that there are scripts that you can add to your monobook.js or vector.js. Mr. R00t ^Talk 16:49, 19 July 2010 (UTC)

You mean something like like the GENDER magicword? No, not that I know of. –xeno^talk 16:55, 19 July 2010 (UTC)

How about we just outlaw elements that cover parts of the userinterface ? The topicons are bad enough in my opinion. —TheDJ (talk • contribs) 17:16, 19 July 2010 (UTC)

Not a bad idea. I've moved the "toolbox" to the top location and some users have elements that then obscure my ability to click "user contributions", "logs", etc. (though this is not the case here). –xeno^talk 17:18, 19 July 2010 (UTC)

Something like this was what I was thinking of. Would something like that even work? Mr. R00t ^Talk 18:05, 19 July 2010 (UTC)

For javascript, the skin global variable contains the skin name (right click, view source, and you'll see a bunch of globals in the js code block at the very top, including the skin), for css, the body element has a skin-specific class, e.g. "skin-vector". Of course this is not very useful, because you can only add the script or css for yourself, not anyone else visiting your userpage, so it will be broken for everyone else. -- Nx / talk 18:13, 19 July 2010 (UTC)

I know that. I need to be able to be able to access whatever program it uses and put it on a page. I'm hoping that it will work similar to {{REVISIONUSER}}. Mr. R00t ^Talk 18:19, 19 July 2010 (UTC)

File a bugzilla:. –xeno^talk 18:23, 19 July 2010 (UTC)

Alright. Mr. R00t ^Talk 18:24, 19 July 2010 (UTC)

There's a mediawiki extension that introduces a magic word to output the skin name, but the problem with that is that it breaks the parser cache - whenever someone with a different skin visits the page, the page would have to be reparsed (alternatively, multiple copies of the page would be stored in the cache, one for each skin). It's not installed and probably will never be installed at WP for this reason. -- Nx / talk 18:24, 19 July 2010 (UTC)

bug 24444. I was thinking of placing different versions in the cache to make it work. Mr. R00t ^Talk 18:34, 19 July 2010 (UTC)

That would be a lot of cache duplication for a pretty much useless (IMO) feature . It would be more efficient to just do what was done for topicon, and add some css to the global skin-specific css files to position the image correctly. But I don't see how that would be useful except for userpages. -- Nx / talk 18:52, 19 July 2010 (UTC)

There is no dev that will do this I suspect. It is a stupid little feature for myspace-like userpages with a big performance hit. —TheDJ (talk • contribs) 12:17, 20 July 2010 (UTC)

Desription of wikifunctions.dll

Is there any desription to AWB library - wikifunctions.dll? Besuglov.S ^cont / _talk 16:34, 20 July 2010 (UTC)

Exporting from OpenOffice

Are you familiar with the function of exporting from OpenOffice to MediaWiki? See Wikipedia talk:Tools#Exporting from OpenOffice.

Thank you. --Amir E. Aharoni (talk) 21:27, 20 July 2010 (UTC)

Note: I've responded at the linked talk page. AJCham 21:39, 20 July 2010 (UTC)

Telling articles from redirects

I've been around a few years and have a bloated watchlist (about 12,361 entries). I'd guess that many hundreds of them are redirects that were created automatically when pages were moved. When I sit down to delete some entries from the list I get discouraged at the thought of checking all of those to find the redirects. Does anyone know of a way to determine which articles are redirects automatically? Perhaps starting with pasting the watchlist into a page? Or using some tool? (This was inspired by trying out Huggle and finding it timing out when checking my watchlist, though I found a work-around for that).   Will Beback  talk  07:43, 18 July 2010 (UTC)

Add the line .watchlistredir { font-style:italic; } to your personal css file. This will display redirects in italics. —  Tivedshambo  (t/c) 07:58, 18 July 2010 (UTC)

Brilliant. I deleted 87 redirects from the first two letters of the alphabet alone, not counting some that I kept. BTW, do you know if there is a listing of those Wikipedia-specific CSS elements? There are some other things I'd like to change.   Will Beback  talk  08:13, 18 July 2010 (UTC)

Yes, Wikipedia:Catalogue of CSS classes. Graham87 08:40, 18 July 2010 (UTC)

Thanks!   Will Beback  talk  20:15, 18 July 2010 (UTC)

Wow, this made me actually go and finally use the CSS thing for the first time, despite seeing a lot of various other useful hints and stuff that I could have over the years. Thanks. ♫ Melodia Chaconne ♫ (talk) 14:05, 18 July 2010 (UTC)

I personally also like to add (in addition to the above):

.allpagesredirect:after { color: #808080; content: " (redirect)"}

It adds a little something more. Color can, of course, be tweaked to suit your preference. --Izno (talk) 13:22, 19 July 2010 (UTC)

Or, use Anomie's link classifier script:

importScript('User:Anomie/linkclassifier.js'); // Linkback: [[User:Anomie/linkclassifier.js]]
importStylesheet('User:Anomie/linkclassifier.css'); // Linkback: [[User:Anomie/linkclassifier.css]]

The follow the instructions at the top of that page to bypass the cache.

See User:Anomie/linkclassifier.css for a list of the other link colors. ---— Gadget850 (Ed) ^talk 17:51, 19 July 2010 (UTC)

I am a firm believer in Link classifer. I said it before and I'll say it again "I think this script should be stanard issue to all editors". To your question this script will turn all redirected links Green, and all Disambig's to Yellow and AFD's to pink. doesn't get any eiser than that. Mlpearc powwow 18:03, 19 July 2010 (UTC)

I don't think it should be standard issue. Changing so many elements on a page, can be rather problematic for older browsers and computers. If it was JUST css, it be a different thing, but it is using Javascript, and a script that on a page like Barack Obama can take a rather long while to execute. If you want to use it, that's your prerogative, but there are many more people who do NOT want to use it, and to who it would actually be a problematic script. So add it as a gadget perhaps, but never enabled by default. —TheDJ (talk • contribs) 12:14, 20 July 2010 (UTC)

It would probably confuse the frack out of most casual editors. ---— Gadget850 (Ed) ^talk 12:20, 20 July 2010 (UTC)

Linkclassifier was recently tweaked to enable run-on-demand (instead of every pageload): User:Anomie/linkclassifier#Usage. I use "p-tb" instead of "p-cactions", to place the link in the "toolbox", instead of in the tabs. None of the slow-down, all of the power :) -- Quiddity (talk) 19:41, 20 July 2010 (UTC)

I certainly don't like the idea of highlighting redirects differently from direct links, by default, in articles (as opposed to watchlists, where it might make some sense). If that were to happen, my prediction would be that we'd see a run on people "fixing" redirects by changing them to pipes. --Trovatore (talk) 21:31, 20 July 2010 (UTC)

Actually, now an idea has occurred to me. It just might make sense to color piped links differently. One of the most objectionable things about piped links is that the reader has no warning that the link may not go to an article that would be expected to have the title indicated. With a different color, that would be different. Of course it would still be bad practice to use pipes in such a way that information that ought to be given to the reader is encoded in the pipe.

So does anyone have a gadget that colors piped links differently? I might use that one. --Trovatore (talk) 21:36, 20 July 2010 (UTC)

I think you are referring to WP:EASTEREGG. ---— Gadget850 (Ed) ^talk 03:13, 21 July 2010 (UTC)

Also, redirected pages don't generally have talkpages, so where the talkpage is redlinked on your edit watchlist page, you can spot them straight away. Lugnuts (talk) 12:01, 21 July 2010 (UTC)

Contribs page

I'm not sure if this belongs here, and I ask to be excused if it does not. I noticed something odd that is there when I open anyone's contributions page.

(latest | earliest) View (newer 250  | older 250) (20 | 50 | 100 | 250 | 500)

There are two spaces (an extra space) between "newer 250" and the vertical bar. This applies always, no matter whose contributions I'm seeing, no matter how many edits I've chosen to view, and no matter whether I'm logged in or not. Why is it so? --Theurgist (talk) 07:05, 21 July 2010 (UTC)

I think I have found the error, see Wikipedia:MediaWiki messages#MediaWiki:Viewprevnext. Svick (talk) 09:45, 21 July 2010 (UTC)

It should be fixed now. Svick (talk) 16:29, 21 July 2010 (UTC)

Double space after a period

The MOS:FULLSTOP guideline says: "The number of spaces following the terminal punctuation of a sentence in the wiki markup makes no difference on Wikipedia because the MediaWiki software condenses any number of spaces to just one when rendering the page." Is that true regardless of what browser the reader is using? Long previous discussion here. Art LaPella (talk) 05:52, 21 July 2010 (UTC)

Yes, because this is an underlying feature of HTML in general; in most cases any amount of whitespace is treated identically to one unit of whitespace and rendered as one space. Note that doublespacing does make a difference in the formatting of the edit window text, so it's not entirely without effect. — Gavia immer (talk) 06:00, 21 July 2010 (UTC)

Yes, that seems to be the standard answer. However, it might be useful to refer to this page where the original discussion started. The technical discussion began near the end of the thread. WP:MoS#Specific statement with respect to double sentence spacing or double spacing after terminal punctuation is needed. Basically, the discussion centers over whether is it possible for someone to view the spacing in a Wikipedia article with a browser that shows HTML as it was typed, as in "wysiwyg". --Airborne84 (talk) 12:33, 21 July 2010 (UTC)

In short: No, it's not possible.

While it is certainly possible for someone to create a browser that treats white space in a non-standard way, all of the major browsers (for English, anyway) treat runs of multiple whitespace characters as a single space unless specifically instructed otherwise by tags (e.g. <pre>) or CSS. And even were someone to use such a non-standard browser or edit their user css to apply the appropriate CSS directives to the entire page, MediaWiki as installed here (or, more likely, tidy) will itself collapse runs of multiple whitespace characters before serving the page to the browser. So even though there "should" be 5 spaces in [ ] according to the wikitext, if you look at the rendered page output there is only one. Anomie⚔ 15:51, 21 July 2010 (UTC)

MediaWiki would have to be changed for this to happen, but it's not impossible. —Aryeh Gregor (talk • contribs) 16:02, 21 July 2010 (UTC)

True, but the question was whether it could be done by using an unusual browser. Perhaps it was an unwarranted assumption on my part that they wouldn't care about a possibility outside their power to directly implement. Anomie⚔ 16:41, 21 July 2010 (UTC)

Yes, it would be perfectly possible in principle. The obvious way to do it would be to have MediaWiki replace a double space with a space followed by a non-breaking space. There are some HTML-based applications that do this. However, it's unlikely anyone would be interested in doing the coding for that, since on the web, the universally accepted convention is single spaces between sentences. Also, it would make a lot of pages look odd where people wrote several spaces in a row and expected it to show up as a single space in the actual article. —Aryeh Gregor (talk • contribs) 16:02, 21 July 2010 (UTC)

For that matter, the request would be possible if MediaWiki would just stop collapsing multiple spaces. Normal browsers would continue to collapse the whitespace, but the hypothetical non-standard browser mentioned would be able to choose to display them. I personally agree that changing MediaWiki in this way is not worth doing. Anomie⚔ 16:41, 21 July 2010 (UTC)

Sentence spacing was a recently featured article on the Main page. T'would be remiss not to mention :) -- Quiddity (talk) 20:18, 22 July 2010 (UTC)

image annotations

Why do images on wikimedia have option to annotate images but on wikipedia they do not? Kallimachus (talk) 01:14, 22 July 2010 (UTC)

Because the ImageAnnotator gadget is only enabled at Wikimedia Commons. Graham87 02:25, 22 July 2010 (UTC)

It seems very useful, especially when looking at complex images. We need to enable that for wikipedia. Kallimachus (talk) 05:26, 22 July 2010 (UTC)

This has been discussed here at length between December 2009 and March 2010:

• Wikipedia:Village pump (technical)/Archive 68#Image annotation in WP
• Wikipedia:Village pump (technical)/Archive 71#Image Annotation
• Wikipedia:Village pump (technical)/Archive 72#ImageAnnotator gadget on Wikipedia
• Wikipedia:Village pump (technical)/Annotate
• Wikipedia talk:Village pump (technical)/Annotate

Perhaps someone who is interested could read the discussions and summarise the current situation here. I think there was concern that vandals could add inappropriate annotations without easily being detected.

— Richardguk (talk) 23:11, 22 July 2010 (UTC)

Transwiki Watchlists

There's way too much separation between Wiki projects. I want to have a transwiki watchlist where I can keep an eye on Wikipedia Pages in various languages as well as files I keep an eye on in the commons and entries I've contributed to in the dictionary. It's way too much of a pain to keep an eye on everything, so certain things go unwatched or unresponded to.

Am I the first one to bring this up? Don't many of you have the same complaint/problem? Chrisrus (talk) 05:40, 22 July 2010 (UTC)

There is an external tool called Watchlistr that does exactly this. And there is also an open bug requesting this. Svick (talk) 09:48, 22 July 2010 (UTC)

Edit summary moved

I've been away for a while. Today I return to make an edit and found the edit summary hidden away(on a mobile device) to the right of the page. if I wasn't really aware from before that edit summaries are desirable I would not of even realised entering an edit summary was possible now. Where did the discussion happen on moving it's position? What where people thinking? Is entering an edit summary something no longer desirable? Don't we want newbies especially to enter edit summaries as much as they can? Regards, SunCreator ^(talk) 10:09, 22 July 2010 (UTC)

Edit summaries are certainly very desirable, not only from newbies. I don't see any such change in Firefox (even with small window width). What kind of device and browser are you using. Could you provide screenshot/photo of the problem? Svick (talk) 10:32, 22 July 2010 (UTC)

Parser ifeq used with Magic Words

Resolved

For some reason this coding does not work:

{{#ifeq:{{NAMESPACE}}|Template||'''[[New York City]]'''}}

I want to create a condition, that:

If a the namespace IS a template, '''[[New York City]]''' will NOT be shown.

If the namespace is NOT a template, '''[[New York City]]''' WILL be shown.

Any suggestions? Can I use magicwords this way? If not, is there another way to do this? Thank you for your time. Adamtheclown (talk) 14:27, 22 July 2010 (UTC)

Your code works fine, at least when I tested it on Template:Template sandbox (displayed nothing) versus Wikipedia:Sandbox displayed New York City. –xeno^talk 14:32, 22 July 2010 (UTC)

Your right xeno, I just double checked myself, and came here to report this. Thank you for taking the time to do this. 14:35, 22 July 2010 (UTC)

Citing an annotation

Not sure if this is the right place...but how do you cite/reference an annotation?Smallman12q (talk) 16:03, 22 July 2010 (UTC)

We need more context to help you. An annotation where? – ukexpat (talk) 16:07, 22 July 2010 (UTC)

An annotation in an article...for example in Edmund_Evans#Process_and_techniques the last line...I'd like to do

<ref group="a">On page viii of ''The Caldecott Aesop: Twenty Fables : A Facsimile of the 1883 Edition'', a production note states that the "majority of color plates were made from the first woodblock renderings of Caldecott's work." The statement is not sufficiently clear to indicate to what extant the original blocks were used, however the cracks in the color plates suggest that the original blocks may in fact have been used.<ref>Richardson, p. 33</ref></ref>

but it gives me an error. I'd like to cite a page for the annotation.Smallman12q (talk) 16:42, 22 July 2010 (UTC)

WP:REFNEST is probably what you're after. - Jarry1250 ^{[Humorous? Discuss.]} 16:55, 22 July 2010 (UTC)

Ye nested ref is what I needed. Thanks.

{{#tag:ref|On page viii of ''The Caldecott Aesop: Twenty Fables : A Facsimile of the 1883 Edition'', a production note states that the "majority of color plates were made from the first woodblock renderings of Caldecott's work." The statement is not sufficiently clear to indicate to what extant the original blocks were used, however the cracks in the color plates suggest that the original blocks may in fact have been used.{{#tag:ref|Richardson, p. 33}}|group="a"}} This worked =).Smallman12q (talk) 17:46, 22 July 2010 (UTC)

DBPedia links

Shouldn't each article link to its DBPedia equivalent? Either at the end of the interwiki links, and/ or through a meta header: <link rel="alternate" href="http://dbpedia.org/page/Birmingham"> for Birmingham, for example? Andy Mabbett (User:Pigsonthewing); Andy's talk; Andy's edits 16:30, 22 July 2010 (UTC)

Well it's not a sister project (so far), and link attributes would require some rather big software changes. Having said that, I'd prefer linking to an open database that contains verified sets of data. So instead of collecting all the data in wikipedia in a separate DB, collecting open data in a DB, is probably more useful in the long term and a better match to our goals. Matching from wikipedia to dbpedia is probably already rather simple to do with DBpedia itself, so I don't see a good reason to do that from within Wikipedia 'manually'. —TheDJ (talk • contribs) 20:13, 22 July 2010 (UTC)

Uploading to Commons

User:guillom of the Wikimedia Multimedia Usability team has published some results on his blog of the first testing done of the usability of Wikimedia Commons upload process, and the results of the first prototype for the new upload wizard that is currently under development. See also the following movies:

• File:Multimedia_usability_project_2010_-_Current_interface_testing.ogv
• File:Multimedia_usability_project_2010_-_Prototype_testing.ogv
• File:Multimedia_usability_project_2010_-_Room_for_improvement.ogv

Please contribute your ideas, and you can test the prototype if you want. I note that some of the results might apply to other processes as well. The "i'll just skip all this, because it is too much text" being the most notable usability issue that will likely apply to many other pages. —TheDJ (talk • contribs) 20:08, 22 July 2010 (UTC)

Thank you. As a follow-up, I'd like to add that I'm currently working on a Questions & Answers page to address the most frequently asked questions. Feedback on the prototype is of course welcome, but you will understand that we may not have the resources to answer every comment individually, especially if many are similar. The Q&A page isn't ready yet, so it'll take some time before I can publish it. So, please consider this a "soft launch": we don't want to make a lot of publicity about our prototype yet. If you happen to know about it and you want to share your opinion, that's fine. But we'll officially invite the community later to try it out, when the Q&A page lets us focus on the most useful comments. Thanks, Guillaume Paumier 20:32, 22 July 2010 (UTC)

Strange "redirect"

I was looking at a very long article, but if I went back to it the computer would have to scroll down to get back to what I was reading. However, it didn't do that. It jumped forward to an article whose link I clicked on earlier while looking at the very long article, and the back button was no longer blue, meaning it couldn't be used.

I've asked similar questions before but no one seems to have an answer for this. Sometimes with very long articles or emails the back button won't go back to them, or the forward button won't go back to them.Vchimpanzee · talk · contributions · 20:11, 16 July 2010 (UTC)

And it just happened again. I used the back button several times and it wouldn't go forward again because THIS page is so long. I should also add I have IE8 and Vista.Vchimpanzee · talk · contributions · 20:13, 16 July 2010 (UTC)

Have you ever tried Firefox? I used to think Internet Explorer was the bees knees until I actually gave Firefox a chance. Never looked back. This is surely a browser issue, not Wikipedia's. –xeno^talk 20:15, 16 July 2010 (UTC)

If this is what I think it is, then it isn't specific to Wikipedia. What might be happening is that when you hit "Back", the previous page is a redirect that sends you to another URL. This then wipes out your "Forward" pages since you are creating a new series of pages, starting with the redirected page, which is essentially the same as you clicking on a link, at least for the purpose of this example. Gary King (talk · scripts) 20:19, 16 July 2010 (UTC)

I don't like Firefox. I've had to use it at libraries. Also, making a major change at home will just make things worse for me.

I know there wasn't an actual redirect, but the behavior of the computer was like when there is one.Vchimpanzee · talk · contributions · 20:30, 16 July 2010 (UTC)

Some wikilinks redirect to a section in an article, which is the same as a redirect and would wipe your Forward history, too. Gary King (talk · scripts) 20:35, 16 July 2010 (UTC)

This is, again, a browser issue but it's like when I click on "Back" and get something I've already backed away from. I wonder if there's a glitch in the function of IE8? Only this time, instead of scrolling down to the specific point in the Wikipedia article (which is sort of like a redirect), it sent me back to an article I had clicked on "back" from.Vchimpanzee · talk · contributions · 20:55, 16 July 2010 (UTC)

I'm seeing the forward button turn gray instead of black when the computer scrolls down in a long article or other type of page. It doesn't usually happen in a short article.Vchimpanzee · talk · contributions · 20:58, 18 July 2010 (UTC)

There is probably nothing we can do here - this seems like an IE issue, not a Wikipedia issue. –xeno^talk 16:44, 19 July 2010 (UTC)

I posted this here but didn't get a lot of help. Someone, somewhere, must know the answer to this.Vchimpanzee · talk · contributions · 18:44, 19 July 2010 (UTC)

At the risk of being repetitive and annoying, the answer is that you are using a sub-standard browser. I'm sure that you can customize Firefox to function how you like. If it's just tabbed browsing that you don't want, install the Tab Killer add-on. –xeno^talk 18:50, 19 July 2010 (UTC)

I don't like change. Why does no one understand that? IE8 is perfectly good and this is just an annoying glitch that makes me curious. A wholesale change would be much worse.Vchimpanzee · talk · contributions · 20:05, 19 July 2010 (UTC)

Ok. You should probably try to find some kind of dedicated IE support forum; most of our technically-minded users do not use IE and probably won't be inclined to help find a solution to this browser problem. –xeno^talk 20:20, 19 July 2010 (UTC)

Or perhaps give a specific example of a page producing the error; I for one am not going to try to find a page that is "very long" with a redirect to a section near the end of the page. I can't even understand the statement of the problem, so I'm glad there are other editors that seem to be able to interpret the issue. —Ost (talk) 21:05, 19 July 2010 (UTC)

I'm sure I mentioned Deepwater Horizon oil spill. I planned to reproduce the sequence of events on the computing reference desk (people were responding more here), but I never got around to it. I can't today because there's not enough time.Vchimpanzee · talk · contributions · 13:14, 20 July 2010 (UTC)

And have you tried Chrome, a rising browser that is getting popular these days ? Or Opera ? There are 5 leading browsers at the moment, and IE is, as a matter of fact, the worst of them. It's only popular and used because Microsoft is using his monopoly in a very unfair fashion.

Basically, you can ask a question to the Microsoft Community. I absolutely won't guarantee that your issue will be fixed (IE isn't popular for solving issues at all, quite the contrary), but you're likely to get an answer. Dodoïste (talk) 08:40, 21 July 2010 (UTC)

My problems will be worse if I add these strange browsers. That I guarantee.Vchimpanzee · talk · contributions · 14:17, 21 July 2010 (UTC)

Also, Firefox has behaved very poorly at the library where I have to use it if ads cause problems. When the computer freezes up there is an ad on the screen with moving information but not what I came to the site to see, or everything is ther but I can't do anything. Or there is some information at the bottom of the screen about some web site relating to an ad. If I try to click or go anywhere else, I get the message "stopped". As if that communicated any information.Vchimpanzee · talk · contributions · 15:16, 21 July 2010 (UTC)

I know I promised I'd stop pestering you about Firefox, but just to response to your issue there, Adblock is an excellent Firefox add-on that can rid you of those meddlesome ads. –xeno^talk 15:18, 21 July 2010 (UTC)

I use Firefox at a library. So it's not up to me. As for that being a solution at home, it's yet another way the solution would be worse than the problem.Vchimpanzee · talk · contributions · 16:55, 21 July 2010 (UTC)

More here.Vchimpanzee · talk · contributions · 20:29, 23 July 2010 (UTC)

robots.txt, __ NOINDEX__ and {{userspace draft}}

The __NOINDEX__ and {{noindex}} directives aren't that well-known or well documented.

I was creating subpages under User:Geo Swan for years before I became aware of them.

__NOINDEX__ directive and robots.txt seem to work differently. A robots.txt, with the appropriate contents, tells well-behaved web-crawling robots, like those that search engines use to find contents, will honor the robots from indexing the files in that directory, and all subdirectories -- while the __NOINDEX__ directive only applies to the article on which it has been placed.

I wonder, since, to a web-browser's robots, the subpages of User pages are just files and directories, couldn't a properly drafted robots.txt eliminate the need to put the __NOINDEX__ directive on all our subpages?

If a file has already been hidden from web-browsers, then a {{userspace draft}} would be redundant, and unnecessary, wouldn't it?

Thanks! Geo Swan (talk) 23:21, 21 July 2010 (UTC)

An admin would need to put every user who used noindex on the robots.txt, so it would be rather inconvenient (users couldn't do it themselves) and robots.txt would become rather huge {{userspace draft}} is used on more than 10,000 userpages; even if everyone who used the template had 10 subpages, that would still be 1,000 more things to list on robots.txt. Mr.Z-man 01:05, 22 July 2010 (UTC)

I thought that robots that honored robots.txt directives, look for a robots.txt in every directory they traverse. So, if I create a subpage User:Geo Swan/Guantanamo/robots.txt shouldn't it be able to block compliant robots from everything in that directory?

I would still appreciate feedback on whether the {{Userspace draft}} tag is redundant when web browsers have already been blocked, via robots or NOINDEX. Geo Swan (talk) 02:29, 23 July 2010 (UTC)

From what I've seen, robots.txt only applies in the root directory. Additionally, robots.txt files need to follow a very specific format, just listing entries on a wiki page won't work because of all the other HTML. {{userspace draft}} is the only one of the 3 that provides a visual note in addition to noindexing; if you don't care about the box, then yes, its redundant. Mr.Z-man 03:01, 23 July 2010 (UTC)

Only one robots.txt works per domain. en.wikipedia.org gets only one for the whole site. —Aryeh Gregor (talk • contribs) 21:09, 23 July 2010 (UTC)

Wikitext weirdness

I noticed a minor fault in how (presumably) MediaWiki renders a page, and will describe it in case anyone wants to follow up. The events took place at WT:Requests for comment/Jagged 85#Cleaning up the problem:

1. At 06:06, 20 June 2010, a new section was added, and the wikitext included a <p> tag (with no closing tag) (diff). Looking at the html source for the page (permalink) shows that something (HTML Tidy?) added a closing </p>, and everything was good.
2. Several comments were later added with varying indents (colons at left margin). All good.
3. At 09:08, 28 June 2010, I appended a new comment: blank line and no colon (diff). All good.
4. At 08:25, 22 July 2010, John Vandenberg appended a new comment: blank line and no colon (diff). Bug! His comment is rendered with no line break before mine (as if there were no blank line between the two comments) (permalink). I tried adding ?action=purge to the page URL, and I examined the wikitext for any strange characters, but can't see any. Bug still present.
5. I mentioned this at User talk:John Vandenberg#Wikitext weirdness.
6. At 11:00, 22 July 2010, Jack Merridew fixed the problem by replacing the <p> with a blank line (diff).

Conclusion: the <p> caused MediaWiki to misrender two paragraphs below. Johnuniq (talk) 00:08, 23 July 2010 (UTC)

Not unexpected: the unclosed <p> element contains the indented text, which is parsed according to its own wikitext rules, then the paragraph tag is implicitly closed by HTML Tidy at the next section heading. HTML tag soup does not mix well with wikitext porridge. In fact wikitext porridge does not always mix well with wikitext porridge. It's a quirky syntax at the best of times. — Richardguk (talk) 00:36, 23 July 2010 (UTC)

MediaWiki did rework the initial post by ImperfectlyInformed; it properly wrapped the text as two paragraphs.^{view source} As the thread continued with indented replies, it kept working fine; the indented stuff was not contained by the paragraphs (aside: the whole mechanism of using definition lists for indented replies is semantic mush; fix *that*). It worked when Johnuniq stared a new paragraph, too. But somehow the embedded <p> element addled things such that when JohnV added another implied paragraph (i.e. blank line), MedaiWiki got confused; it failed to parse Johnuniq's && JohnV's posts such that it generated two paragraph; it goofed, and wrapped them in just one; JohnV's blank line being reduced to a mere space. MediaWiki got it wrong, here, and the improperly formed embedded <p> was what triggered the mis-generation of served-code.

Embedded markup can be useful, but most people can't produce well-formed markup; that's why we have wiki-syntax: it's very simple. There's something amiss in sanitizer.php, or nearby, that needs fixing-up. Cheers, Jack Merridew 22:46, 23 July 2010 (UTC)

Need CSS/javascript help with Template:ref_label

Template:ref label wraps its content in a span, rather than a div. This causes problems when the background color is set, because if the line spacing is wide enough then the underlying background color seeps through the cracks. There is a thread about this on the template talk page.

The CSS that is set in Common.css [10] refers directly to the span tag, not just the .citation class name. I have not looked up the javascript that takes care of changing the background color.

I don't know enough about our CSS/javascript setup to try to fix this problem. If someone with the knowledge has time to fix it, it would be appreciated. — Carl (CBM · talk) 00:56, 23 July 2010 (UTC)

I believe this was fixed for {{ref}}; see Template talk:Ref#Text color. ---— Gadget850 (Ed) ^talk 02:05, 23 July 2010 (UTC)

I don't think that's the same problem. Here's an exaggerated example of what's wrong with the style in question. Make sure it wraps onto multiple lines and look at the gaps between them. — Carl (CBM · talk) 02:13, 23 July 2010 (UTC)

xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx xxx

Looking at the other discussion, it looks like you are talking about the spacing between the lines. ---— Gadget850 (Ed) ^talk 03:12, 23 July 2010 (UTC)

It is not just that template. Click on (b 2010):

b, a (2010).

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

The above uses {{harvrefcol}} which is classed with citation, so I suspect it is something to do with that class. Compare to this link.

References

The above is classed with reference. The reference list is an ordered list inside a div, whereas the template uses a span. I think you are going to have to convert it to use a div. If you get it working, the same fix should be applied to other reference templates. ---— Gadget850 (Ed) ^talk 03:38, 23 July 2010 (UTC)

Yes, that's what I said originally: it needs to be updated to use a div. That means Common.css has to be edited to match, and also possibly the javascript that does the highlighting needs to be edited. — Carl (CBM · talk) 03:57, 23 July 2010 (UTC)

You can do a sandbox version and add the CSS to your personal CSS for testing, then request that it be added to the site CSS. I don't believe there is any JS involved in the highlighting. ---— Gadget850 (Ed) ^talk 04:15, 23 July 2010 (UTC)

I don't think that's a good idea. This is a typical example of what should be an inline element, changing it to a block element for the sake of a highlight color, does not make sense. Is not having a continious background over multiple lines really such a big problem ? —TheDJ (talk • contribs) 07:26, 23 July 2010 (UTC)

Cite places the list inside a div, so I don't understand the issue. I have a test at User:Gadget850/t8. ---— Gadget850 (Ed) ^talk 11:50, 23 July 2010 (UTC)

Citation Checking in Show Preview

When I make changes to an article, I click on Show preview to see my changes. If I put in a Wiki link, for example, I can click on the link in the preview and make sure it works. However, if I make a citation change and click on the superscript citation, I can't see how the citation resolves. Thus, I have to Save page, check the resolved citation, and if I screwed up, go back in and fix it. Is there a way to do what I want without saving the page (short of using a sandbox)?--Bbb23 (talk) 01:14, 23 July 2010 (UTC)

Add this to Special:MyPage/skin.js and purge the page per the instructions at the top of the page:

importScript('User:Anomie/ajaxpreview.js'); // Linkback: [[User:Anomie/ajaxpreview.js]]

At the bottom of the edit window, you will find a button marked Ajax Preview w/Refs. ---— Gadget850 (Ed) ^talk 02:02, 23 July 2010 (UTC)

Or simply type a temporary references tag at the end of your wikitext:

<references />

then preview – but don't do this unless you can remember to remove the temporary line before saving! — Richardguk (talk) 05:46, 23 July 2010 (UTC)

You can also simply avoid section edit on those occasions. Editing the whole page can be slower, but preview then shows the whole page including the way the cites have resolved. (Anomie's script sounds useful though; must have a look at that myself.) PL290 (talk) 12:11, 23 July 2010 (UTC)

Thanks very much for the suggestions.--Bbb23 (talk) 22:55, 23 July 2010 (UTC)

Main page problem

A recurring problem that makes WP look amateurish is the main page spacing. On the right is a story about Kosovo but there is a picture of a man that looks (pardon my ignorance) like he might be from south India. If the pictures can be lined up, this would be better. Suomi Finland 2009 (talk) 22:11, 23 July 2010 (UTC)

The image belongs to the entry with "(pictured)":

• "In the final Test match of his career, Sri Lankan international cricketer Muttiah Muralitharan (pictured) becomes the first bowler to take 800 Test wickets."

See Wikipedia:FAQ/Main Page#Why are the images on "In the news" and "On this day" not aligned next to each relevant entry? PrimeHunter (talk) 23:04, 23 July 2010 (UTC)

Possible issue with alt-left-arrow in IE8

For a number of months now, a nontechnical Wikipedia editor named Vchimpanzee has occasionally reported problems to the Computing Reference Desk involving a disappearing forward-history in IE8. (Here is his most recent thread, a lengthy attempt to be meticulous in his bug report.) I just reproduced the problem on my machine under IE8 but am having trouble reproducing the problem reliably. My one repro case was that I edited WP:SANDBOX, clicked Link A, then clicked Link B, then pressed Alt-left-arrow on the keyboard, which took me to Link A as expected, but the browser's "Forward" button turned from blue to gray, indicating the forward-history had been lost. Then, the Back command brought me, surprisingly, to Link B.

Is it possible that some feature of Wikipedia's use of AJAX is interfering with the alt-left-arrow or alt-right-arrow keyboard shortcuts of IE8? Without a good repro case that is my first guess at the source of the issue. Comet Tuttle (talk) 21:15, 23 July 2010 (UTC)

<insert obligatory get a better browser remark here> If you suspect it is ajax, disable javascript and try reproducing it again. But this looks more like a browser bug. -- Nx / talk 22:05, 23 July 2010 (UTC)

Something else that might 'invalidate' the forward history is a change in the windows hash (using the TOC to navigate to a section for instance). —TheDJ (talk • contribs) 15:58, 24 July 2010 (UTC)