Wikipedia:Village pump (technical)

	Here's an example
This will wrap:	It shows the odd wrapping in Firefox 2.0.0.4, and IE6 and IE7 via Wine
	Anomie 21:26, 28 July 2007 (UTC)[reply]

The technical section of the village pump is used to discuss technical issues about Wikipedia. Bugs and feature requests should be made at the BugZilla since there is no guarantee developers will read this page. Problems with user scripts should not be reported here, but rather to their developers (unless the bug is exigent).

Newcomers to the technical village pump are encouraged to read these guidelines prior to posting here. Questions about MediaWiki in general should be posted at the MediaWiki support desk.

view · edit

Frequently asked questions (see also: Wikipedia:Technical FAQ)

Click "[show]" next to each point to see more details.

If something looks wrong, purge the server's cache, then bypass your browser's cache.
This tends to solve most issues, including improper display of images, user-preferences not loading, and old versions of pages being shown.

No, we will not use JavaScript to set focus on the search box.
This would interfere with usability, accessibility, keyboard navigation and standard forms. See task 3864. There is an `accesskey` property on it (default to `accesskey="f"` in English). Logged-in users can enable the "Focus the cursor in the search bar on loading the Main Page" gadget in their preferences.

No, we will not add a spell-checker, or spell-checking bot.
You can use a web browser such as Firefox, which has a spell checker.

If you have problems making your fancy signature work, check Help:How to fix your signature.

If you changed to another skin and cannot change back, use this link.
Alternatively, you can press Tab until the "Save" button is highlighted, and press Enter. Using Mozilla Firefox also seems to solve the problem.

If an image thumbnail is not showing, try purging its image description page.
If the image is from Wikimedia Commons, you might have to purge there too. If it doesn't work, try again before doing anything else. Some ad blockers, proxies, or firewalls block URLs containing /ad/ or ending in common executable suffixes. This can cause some images or articles to not appear.

For server or network status, please see Wikimedia Status. If you cannot reach Wikipedia services, see Reporting a connectivity issue.

« Archives, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216

Nested tables class="collapsible collapsed"

If someone knows of a workaround for this issue, I'd like to know what it is too. I couldn't find anything for it in Bugzilla either.

I have some nested collapsible tables coded up roughly as follows:

{|class="collapsible collapsed"
!Outer table header row
|-
| interesting information
|-
|
{|class="collapsible collapsed"
!Inner table header row
|-
| more detailed interesting information
|-
|
{|class="collapsible collapsed"
! Inner inner table header row
|-
| Even more detailed information that's not particularly interesting
|}
|}
|}

Here's what it looks like:

Outer table header row

interesting information

Inner table header row

more detailed interesting information

Inner inner table header row
Even more detailed information that's not particularly interesting

All of them should be collapsed by default. But when you expand the outer table, all the inner tables expand as well -- but their "show" links say "show". The Javascript apparently doesn't know they've expanded. When you click on "show" it changes to "hide", but doesn't do anything else. After that, the link works normally.

Any thoughts? TCC (talk) (contribs) 09:58, 1 July 2007 (UTC)[reply]

Looks like you aren't using it in the way it was intended. Looks like it doesn't support nesting. I'll take a look around and get you an answer and the correct code in a bit. —Andrew Hampe ^Talk 23:17, 1 July 2007 (UTC)[reply]

Just so you're aware then, one requirement is to be able to specify the initial state, which is why I'm using the table instead of NavFrame, etc. TCC (talk) (contribs) 01:33, 2 July 2007 (UTC)[reply]

You can hide the nested tables by putting 3 NavFrames inside one collapsible table, e.g. {{Calculus footer}}. Even if you're nesting only 2 or less, you can add an empty NavFrame div to make them hidden. I don't know of a way to make the nested NavFrames expand by default though, unless you're nesting 2 or less. –Pomte 18:55, 24 July 2007 (UTC)[reply]

'[edit]' option for top section of the article

If someone wants to edit top section of the page he/she have to go through 'edit this page' tab which opens whole page for editing and if the page is large then it takes time, similarly preview also takes time. Providing '[edit]' link to the top section is solve this problem. Vjdchauhan 13:42, 18 July 2007 (UTC).

There is a script that does this. It adds a "0" link to the top of every page, since when you edit the lead (top section) of an article, you are editing section 0. The script is located here; just copy and paste the contents of that page into your monobook.js. Gracenotes^T § 14:08, 18 July 2007 (UTC)[reply]

Another script which does the same but creates [edit] link. This was also discussed on bugzilla: bugzilla:156. P.S. I guess it's time to add this question to Wikipedia:Technical FAQ… ∴ Alex Smotrov 14:22, 18 July 2007 (UTC)[reply]

editing large articles causes technical problems

People editing large articles are not always aware of this. For huge articles, say, 400KB+, this can cause breakage on days when the server is very busy. I had intermittent problems editing Wikipedia:List of missing journals until I discovered the section 0 trick and added the edit link. Two suggested workarounds: 1) make the monobook.js with the [0] tab included one of the scripts listed in the preferences, or 2) a robot that automatically adds the edit {{Edit-first-section}} template to articles big enough to cause technical problems. Even if you do the first, the 2nd is very desirable. davidwr/(talk)/(contribs)/(e-mail) 00:18, 25 July 2007 (UTC)[reply]

screening deletion log

Is there any way of screening of filtering the deletion log so it shows only deletions of articles from mainspace? I'm trying to analyze recent speedy deletions WT:CSD/Analysis, and I want to eliminate the 80% or so that are images, talk pages, redirects, user pages, and so forth to speed up work on the rest. DGG (talk) 04:39, 19 July 2007 (UTC)[reply]

From which I gather that the only direct way to do it is currently disabled. Does anyone know a work-around?DGG (talk) 04:23, 24 July 2007 (UTC)[reply]

!vote icons script

I know this page isn't really intended to deal with user scripts, but I figured I would probably get a quick answer here. =)

I'm using the script that adds the !vote icons to vote pages. How could I restrict the pages on which this script initializes, such as only on AFD discussions? GrooveDog (talk) 02:33, 21 July 2007 (UTC)[reply]

See the if(wgNamespaceNumber==0) return; bit? You can just add some more if() or a switch() there to cause the script to abort, if certain conditions are not met. Some useful js variables for this would be wgPageName and wgTitle. --Splarka (rant) 07:31, 21 July 2007 (UTC)[reply]

Since when do we add vote icons to AFD discussions? I thought that idea had been soundly rejected. — Carl (CBM · talk) 12:40, 22 July 2007 (UTC)[reply]

Yes, please do not add icons to your voting script. For some background information, note that {{Support}} and {{Oppose}} are salted, and see these two discussions: at AfD and at RfA. It seems clear they have been rejected. --Iamunknown 01:58, 23 July 2007 (UTC)[reply]

I presume the script being referred to is User:Ais523/votesymbols.js. This only affects the appearance of the page for the person using the script and not other users, which avoids the problems mentioned above. Tra (Talk) 02:19, 23 July 2007 (UTC)[reply]

Ah! Then script away. :P --Iamunknown 02:29, 23 July 2007 (UTC)[reply]

If you are referring to my script (whose whole purpose is for people who like the icons to be able to see them without forcing them on anyone else), contact me on my talk page if you have a specific feature request for it. --ais523 16:29, 24 July 2007 (U T C)

CSS

I have (again) a CSS related question: How can I override the personal userlinks? Specifically, I'd like my userpage link go to my "welcome" page. It must have something to do with the #pt-userpage value, but I have no idea about the proper syntax. —AldeBaer (c) 17:52, 23 July 2007 (UTC)[reply]

I don't think you can do that with CSS, but it shouldn't be too hard with JavaScript. Try copying and pasting the following three lines of code to your monobook.js page:

addOnloadHook(function () {
    document.getElementById("pt-userpage").getElementsByTagName("a")[0].href = "/wiki/User:AldeBaer/welcome";
});

—Ilmari Karonen (talk) 19:04, 23 July 2007 (UTC)[reply]

Works perfectly - you're totally my hero!! If I may ask another question: Where can I best learn stuff like that? —AldeBaer (c) 19:14, 23 July 2007 (UTC)[reply]

A great place is WikiProject User Scripts. Matt/TheFearow (Talk) (Contribs) (Bot) 06:05, 24 July 2007 (UTC)[reply]

Need help!

I need help with regards to templates. For more information, please view this page. I feel that it is paramount to achieve consistency with regards to templates. If anyone knows how to correct this, it would be much appreciated. --S iva1979^{Talk to me} 05:21, 24 July 2007 (UTC)[reply]

The red link is gone now so all articles with cat should end up at Category class articles. --_Hdt83 ^Chat 05:40, 24 July 2007 (UTC)[reply]

Help moving over existing page

I wanted to have the article "Hybrid ARQ" under the expanded title, "Hybrid automatic repeat-request", and redirect to it from the abbreviated title. I created a new page with the expanded title, pasted the content from the old page, and replaced the content in the old page with redirect code. But then I read that this is bad because it does not preserve the content history. So I reverted "Hybrid ARQ" to its old state and deleted the new content of "Hybrid automatic repeat-request". Then I renamed "Hybrid automatic repeat-request" to "Hybrid ARQ (temp)" and planned to rename "Hybrid ARQ" to the expanded title. But I can't because the page "Hybrid automatic repeat-request" still exists (and redirects to "Hybrid ARQ (temp)"). How to solve this mess? -Pgan002 05:14, 24 July 2007 (UTC)[reply]

Wikipedia:Requested moves is the place to be. In the mean time, I've redirected Hybrid automatic repeat-request to Hybrid ARQ. --Edokter (Talk) 08:51, 24 July 2007 (UTC)[reply]

You just need an administrator to delete the leftover redirect. Done. Now you can move the page. — Carl (CBM · talk) 14:33, 24 July 2007 (UTC)[reply]

Interwiki links on the geohack.php

There has recently been discussion on a talk page to change the newly added interwiki links on the geohack. To see an example on how these interwiki links look like go here. The idea is to add a code to the Magnus' toys that finds any interwiki links in the Template and changes them into interwiki links to the geohack.

A few flaws are in the current interwiki links.

It's unusually placed at the top of the page instead of the normal bottom right bar.
It doesn't render the entire link, such as type, scale and region.
It has problems understanding some of the country codes
Adding new interwiki links the right way is difficult, you will have to go trough 10 different templates and add the link.

I believe it would be best if we would avoid the discussion to spread trough many places and continued the discussion at the original talk page, instead of here. Thank you. --Steinn inn 13:07, 24 July 2007 (UTC)[reply]

monobook.js

My monobook.js stopped responding. It doesn't work at all. How do I fix this? -Flubeca^Talk 15:15, 24 July 2007 (UTC)[reply]

Strings in JavaScript can't contain literal newlines. I think I've fixed your monobook.js for you; try bypassing your cache and seeing if it works then. --ais523 16:22, 24 July 2007 (U T C)

WooHoo! It works! Thanks! -Flubeca^Talk 20:20, 24 July 2007 (UTC)[reply]

Clearing my cache is something I always forget to do too! Mind-you, I've tried installing two anti-vandal tools via my monobook.js, and even with the cache cleared, I still can't get them to work... :-( Lra drama 13:22, 27 July 2007 (UTC)[reply]

Sign-in problems ...& MSIE "vs." Mozilla

Hello! I have a technical problem. I am a newcomer on Wiki and the exact opposite of a nerd, so this may be the wrong place to ask or I may be making some other mistake, and I know you all have other things to do. Please forgive me in that case. I have asked real-life friends for help, but they doubt the problem is with my computer.

To the point:
1. I created my Wiki account about a week ago. I got an e-mail, which I was asked to browse to confirm my account. I did that (any number of times), but only got to the "confirmation invalid" page. Now, I happened to be logged in over the deadline and a nice older colleague told me my account was verified by that. But I had also asked for the e-mail connection, and that part gave the same result when I asked for it anew today. I had protracted problems with my e-mail server last week, but when it did function, only this seemed not to.
2.a) Browsing with Microsoft Explorer: Every time I turn pages (whether by search&go, topflap, link without or -in a box), I get error messages (which I just okay away). Most often: "Line: 672 / Character: 35 / Error: '}' expected / Code: 0". But some hour ago, it changed to: "Line: 673 / Character: 8 / Error: '}' expected / Code: 0". With the former error, I seemed to have no further problem within a page. At the same time as the latter turned up (connection or not?), it became impossible to make a redirect. I checked my typing many times over, but all I got was a normal article text reading: "1. REDIRECT", and then the target word in normal size.
2.b) Using Mozilla Firefox, the page turning works allright. Instead, as far as I have discovered, some section "edit" marks are displaced up or down, and on the "edit" subpage, a hotchpotch of words (from those handled by the last few editors, it seems) appear underlined in red. The editing works somewhat erratically.
Both these browser problems appear, as far as I have discovered, only in the English Wikipedia. Not on other Wikipedias I have peeped into, nor on my e-mail or Google.
Can somebody help me, or repost this letter to the right authority? I would be infinitely grateful. Yours sincerely, Keinstein 20:29, 24 July 2007 (UTC)[reply]

I cannot answer your question on IE errors, as I don't trust MSIE enough to run it on anything but Microsoft's update site (and, most of the time, the computer I'm using doesn't even have MSIE), but as to your other questions:

The email verification is only for sending/receiving email messages via the "email this user" function, and for recovering a lost password. You could try going to your preferences and requesting the email confirmation again there.
Redirects do show as "1. REDIRECT [[Example]]" when edited directly. It should still work, unless you added extra whitespace before the redirect.
The misplaced section edit links are a known issue; see Wikipedia:How to fix bunched-up section edit links for more information.

--cesarb 00:37, 25 July 2007 (UTC)[reply]

Many thanks for taking time with me!
And yes, I have to remember to check the whole process before I assume something is wrong.

...but...please get me right, I'm not blaming you - I still can't make the e-mail service work. Perhaps I have to let my real-life friends supervise if I'm consistently missing some obscure button after all...

...and as to MSIE, I must confess I am a little puzzled. I've understood it's not very popular, but to me, this far, these error messages apart, it seems to work better than Mozilla. (??) Keinstein 11:44, 25 July 2007 (UTC)[reply]

See Criticism of Internet Explorer for more detail. --cesarb 23:39, 26 July 2007 (UTC)[reply]

Thanks, I'll do so! Keinstein 16:01, 27 July 2007 (UTC)[reply]

Main Page modifications

Per Wikipedia:Help_desk#Technical_Questions and MediaWiki_talk:Common.css#Main_Page_title_coding. I was looking at other outlets to get this issue answered. Thank you. --User:Charitwo/Sig 01:06, 25 July 2007 (UTC)

Image with an & in the file name shown only as text link

On the page Johnny Fingers the image Image:JX, JP & KG.jpg appears as a link rather than displaying the actual image. This is the markup used:

[[Image:JX, JP & KG.jpg|thumb|right|Johnnie Fingers with colleagues and friends JPR and Miumiu, August 2006]]

The actual image should appear. I assume this is related to the & in the filename. I asked this at MW:Project:Support desk#Image with an & in the file name shown only as text link and the only response I got was that it may be a caching problem. Is this a known issue? Thanks Patleahy 01:25, 25 July 2007 (UTC)[reply]

Same problem as above. Anomie 02:02, 25 July 2007 (UTC)[reply]

Thanks, Patleahy 02:33, 25 July 2007 (UTC)[reply]

Change Common.js to fix NavFrame bug

Moved from MediaWiki talk:Common.js#Please set NavigationBarShowDefault .3D 0 (an editprotected request):

In the unmaintained but very widely used Dynamic Navigation Bars (NavFrame) section, the variable NavigationBarShowDefault controls how many NavFrames need to appear before they are all supposed to be initially collapsed. Currently it is set to the value of autoCollapse (== 2), which controls the analogous behavior for collapsible tables above.

Sadly, the templates which use NavFrames, e.g. {{hidden}} and {{ifd}} contain <div class="NavContent" style="... display:none; ..."> which, when there are 2 or fewer NavFrames on a page, results in a "[hide]" toggle which doesn't do anything but change to "[show]" the first time it is clicked, which then acts normally. This bug has been reported multiple times, e.g., on Template talk:Hidden and Template talk:Ifd.

This problem can be corrected by changing var NavigationBarShowDefault = autoCollapse; to var NavigationBarShowDefault = 0; which will have no adverse effects (in fact 0 is suggested as a value in comments just above the assignment statement.)

A superior but more complcated fix would have createNavigationBarToggleButton() examine the .style.display of its NavContent and NavPic children, and have it set the NavToggle to be a NavigationBarShow instead of a NavigationBarHide if any child's .style.display == 'none'. Code showing how to query those children's values is already in the previous function toggleNavigationBar() where they are toggled. Then you could remove the NavigationBarShowDefault variable and the loop it controls at the end of createNavigationBarToggleButton(). This would be far superior because it will allow effective control of the toggle's initial state, but you would want to debug it in your personal Common.js file first, of course. I have tested the fixed code below in my monobook.js Thank you. BenB4 01:45, 23 July 2007 (UTC)[reply]

Totally agree, current implementation needs to be fixed ∴ Alex Smotrov 15:34, 24 July 2007 (UTC)[reply]I'd go with the complex fix, as the quick fix will unnecessarily collapse navigation boxes on articles which have only one. —Ruud 22:24, 24 July 2007 (UTC)[reply]Although I prefer the complex fix, I haven't been able to find any use of NavFrames which don't have style display:none. Do you know of any? ←Ben^B4 22:42, 24 July 2007 (UTC)[reply] Nevermind, {{hidden begin}} does. Complex solution attempt follows. ←Ben^B4 01:05, 25 July 2007 (UTC)[reply]

This is the code for the complex fix; I tested it in my monobook.js /** Dynamic Navigation Bars (experimental) ************************************* * * Description: See [[Wikipedia:NavFrame]]. * Maintainers: UNMAINTAINED */ // set up the words in your language var NavigationBarHide = '[' + collapseCaption + ']'; var NavigationBarShow = '[' + expandCaption + ']'; // shows and hides content and picture (if available) of navigation bars // Parameters: // indexNavigationBar: the index of navigation bar to be toggled function toggleNavigationBar(indexNavigationBar) { var NavToggle = document.getElementById("NavToggle" + indexNavigationBar); var NavFrame = document.getElementById("NavFrame" + indexNavigationBar); if (!NavFrame || !NavToggle) { return false; } // if shown now if (NavToggle.firstChild.data == NavigationBarHide) { for ( var NavChild = NavFrame.firstChild; NavChild != null; NavChild = NavChild.nextSibling ) { if ( hasClass( NavChild, 'NavPic' ) ) { NavChild.style.display = 'none'; } if ( hasClass( NavChild, 'NavContent') ) { NavChild.style.display = 'none'; } } NavToggle.firstChild.data = NavigationBarShow; // if hidden now } else if (NavToggle.firstChild.data == NavigationBarShow) { for ( var NavChild = NavFrame.firstChild; NavChild != null; NavChild = NavChild.nextSibling ) { if (hasClass(NavChild, 'NavPic')) { NavChild.style.display = 'block'; } if (hasClass(NavChild, 'NavContent')) { NavChild.style.display = 'block'; } } NavToggle.firstChild.data = NavigationBarHide; } } // adds show/hide-button to navigation bars function createNavigationBarToggleButton() { var indexNavigationBar = 0; // iterate over all < div >-elements var divs = document.getElementsByTagName("div"); for( var i=0; NavFrame = divs[i]; i++ ) { // if found a navigation bar if (hasClass(NavFrame, "NavFrame")) { indexNavigationBar++; var NavToggle = document.createElement("a"); NavToggle.className = 'NavToggle'; NavToggle.setAttribute('id', 'NavToggle' + indexNavigationBar); NavToggle.setAttribute('href', 'javascript:toggleNavigationBar(' + indexNavigationBar + ');'); var NavToggleText = document.createTextNode(NavigationBarHide); for ( var NavChild = NavFrame.firstChild; NavChild != null; NavChild = NavChild.nextSibling ) { if ( hasClass( NavChild, 'NavPic' ) || hasClass( NavChild, 'NavContent' ) ) { if (NavChild.style.display == 'none') { NavToggleText = document.createTextNode(NavigationBarShow); break; } } } NavToggle.appendChild(NavToggleText); // Find the NavHead and attach the toggle link (Must be this complicated because Moz's firstChild handling is borked) for( var j=0; j < NavFrame.childNodes.length; j++ ) { if (hasClass(NavFrame.childNodes[j], "NavHead")) { NavFrame.childNodes[j].appendChild(NavToggle); } } NavFrame.setAttribute('id', 'NavFrame' + indexNavigationBar); } } } addOnloadHook( createNavigationBarToggleButton );The first and last lines match the already-existing section beginning and end lines. Note that this solution will not fix the bug if there are nested NavFrames, as many of our templates, e.g., {{hidden}} have for some reason. I will fix those separately. ←Ben^B4 01:33, 25 July 2007 (UTC)[reply]The reason these templates have nested NavFrames is to artificially increase the count of NavFrames in the page, thus pushing it over the threshold and making all the NavFrames on the page hidden by default. Yes, it's an ugly hack. --cesarb 03:23, 25 July 2007 (UTC)[reply]Ah! So that's what they are for. It's funny, {{ifd}} only has two, which doesn't work because images usually don't have any other NavFrames, and two or less don't get collapsed. Anyway, this change would fix that problem, as long as the affected templates are unkluged. ←Ben^B4 03:25, 25 July 2007 (UTC)[reply]Two or less? When adding the double nested NavFrames was being done everywhere, the code was configured so that one or less didn't get collapsed; the default must have been changed later (probably after other, less kludgy, ways of having a default-collapsed box were added). --cesarb 13:24, 25 July 2007 (UTC)[reply]Currently, NavigationBarShowDefault = autoCollapse; and autoCollapse, from the completely independent collapsible table implementation just prior in Common.js, autoCollapse = 2. ←Ben^B4 16:32, 25 July 2007 (UTC)[reply]There are two supporters above, are there any objections or more support? ←Ben^B4 07:04, 30 July 2007 (UTC)[reply]Idea for a possibly useful templateThe context would be article content something like:Dawkins says that Caddis larvae are among most remarkable creatures on earth.<ref>{{harv|Dawkins|2006|p=238}}</ref> [...] ==References== <references/> [...] ==Bibliography== *{{Citation | last =Dawkins | first =Richard | author-link =Richard Dawkins | year =2006 | edition =30th Anniversary Edition | publication-place =United States | place =New York | publisher =Oxford Press, Inc. | id =0-19-929115-2 }}.All of that currently exists, of course. The first bit with the {{Harv}} puts an entry in the References list containing an item which wikilinks to the matching citation in the Bibliography section. There may be several of these linking to the same book citation with different page numbers. If one follows the wikilinks, it is necessary to use the browser's page history (via the Back button, etc.) to back up from the Bibliography list entry to the Reference section item from whence one came. There is a clickable backlink at the References section item which links back to the ref in the article text. Most navigation to between refs and notes/cites is done with clickable links and backlinks. This need for the Back button to be used is an exception to that user interface convention. My idea is that it might be useful to have a template ( {{Back}}? ) which could be placed immediately prior to {{Citation}} in the above context example and which would insert an icon ( ^ would do fine, like other backlink icons ) which, if clicked, went one page back in the article history (back to the page showing the References list item from whece the user came). My javascript skills are very rusty, and I'm not sure whether javascript can be used inside of a template. Comments? Have I OD'd on wikiedits today? -- Boracay Bill 08:05, 25 July 2007 (UTC)[reply]I've browsed around a bit and AFAICT, scripting which would accomplish this is not supported in Wikipedia, including in templates. This could be accomplished via an external link to an outside helper page but, though I haven't seen that disallowed by policy or by guidelines, I expect that this would meet with disapproval (besides being flakey). -- Boracay Bill 02:01, 27 July 2007 (UTC)[reply]Image problemsI was wondering if someone could take a look at the problem I posted about Here. After several people trying to help, the issue is still unresolved. I'd like to add that I've disabled Adblocking in Firefox 2.0.0.5, and have no add-ons running in IE7 (7.0.5730.11), and have disabled all of the native blocking services that came with it (I use IE7 specifically for various testing purposes, so nothing is blocked with it). I'm very computer savvy, software savvy, and I have done everything on my end I know of that would cause this issue. (Also I'd like to reiterate that I do see all other images on Wikipedia. I've never been to a page and not seen the images that were supposed to be there.) Any help you could provide would be appreciated. Thanks! ^Ariel♥Gold 10:31, 25 July 2007 (UTC)[reply]Very interesting problem. Seeing that you're an experienced user and considering your detailed description, I think it's safe to assume that you've considered the usual suspects. FWIW, the image does render fine for me in the latest Firefox stable, as well as developer nightly. The same is true for the six versions of Internet Explorer I have handy on this machine and for an older version of Opera. That's what leads me to believe that it's not the Gecko engine. Hmm, is it possible that there's a transparent ad-filtering proxy on your network or (this is less likely) on your upstream provider's network? Another possibility would be deliberate nullrouting (for instance via /etc/hosts or its Win32 equivalent)? I'm afraid that's pretty much all I can come up with at this point but I'll try to check if there's anything weird about the image (ISO 10918-1/2 non-compliance, etc.) when I get a minute. Cheers S ^up? 11:03, 25 July 2007 (UTC)[reply]First, thank you very much for the prompt reply! I too am using Nightly tools, so I'm unsure what the issue could be. Here's the thing, if this were some blocking done from my end, then none of Wikipedia's images should be showing up for me, right? And it is definitely not a proxy issue on my end or my provider, lol, they don't block a single thing. I'm not sure what nullrouting is, but I strongly doubt that's it, (I'm running Win XPSP2 with all updates, and I have not edited my hosts file at all.)I suspect that it is related to the /ad/ tag that is in the image's URL, but after someone else tried to re-upload it without success in removing that tag, I don't know what else to do. (FWIW, the image is completely mine, taken by me, with my camera, and has never been distributed other than my own private rarely visited web site.)^Ariel♥Gold 11:17, 25 July 2007 (UTC)[reply]It's not a tag; IIRC, it's the first two characters of the hexadecimal representation of a cryptographic hash on either the image's content or the image's name (I don't remember which). It's done that way to avoid the performance problems caused by having too many files in a single directory (it's a very common trick on Unix-based servers). If it's a hash of the image name, try uploading with a different name; if it's a hash of the image file, only changing the image itself would change the value. And to check if there is a proxy in the way, try running something like tcptraceroute. --cesarb 13:35, 25 July 2007 (UTC)[reply]

What do you see when you go directly to the image? If there is a transparent proxy in there somewhere, it might be replacing the image with an HTML explanation page which you might be able to see that way. If that doesn't work, you could also try installing the Firefox Live HTTP Headers extension to let us know what you see in the server response when using Ctrl-F5 on the above image link. For example, I see:

HTTP/1.x 200 OK Content-Type: image/jpeg Etag: "7261208650148119589" Accept-Ranges: bytes Last-Modified: Wed, 25 Jul 2007 09:42:36 GMT Content-Length: 8620 Date: Wed, 25 Jul 2007 09:43:23 GMT Server: lighttpd/1.4.13 Age: 2720 X-Cache: HIT from sq43.wikimedia.org, MISS from sq41.wikimedia.org X-Cache-Lookup: HIT from sq43.wikimedia.org:3128, MISS from sq41.wikimedia.org:80 Via: 1.0 sq43.wikimedia.org:3128 (squid/2.6.STABLE12), 1.0 sq41.wikimedia.org:80 (squid/2.6.STABLE12) Connection: close

Anomie 13:45, 25 July 2007 (UTC)[reply]

I see nothing. No box, and no image placemark (I have my browser set to display image placeholders). There's just... nothing, lol. I'll install that addon and update once I get some sleep. Thank you everyone for the assistance, I'll update later tonight. ^Ariel♥Gold 14:26, 25 July 2007 (UTC)[reply]Alright, I installed the Firefox Live HTTP Headers extension, but it is not showing up in my add-ons list, nor in the "view" menu, so I'm unsure what to do with it. Pressing ctrl-f5 when viewing the page, does nothing, the image is still not there for me. *shrug* :) ^Ariel♥Gold 10:53, 26 July 2007 (UTC)[reply]It should give you an option under Tools to bring up the window. Anomie 12:51, 26 July 2007 (UTC)[reply] Okay, for some reason it did not install the first time. I reinstalled it, but when I click on it, there's absolutely nothing in the HTTP Headers box. (As compared to when I look at this page or any other page there is plenty of info there.) I'm stumped, LOL! ^Ariel♥Gold 08:12, 27 July 2007 (UTC)[reply] Nothing at all for the absentee image, even if you Ctrl-F5 on the image page with the HTTP Headers box open? Anomie 12:07, 27 July 2007 (UTC)[reply] Okay, I guess I did it wrong or something, I've never used that add on. I opened the page, Image:Amy Grant in 1998.jpg and then I opened the HTTP Headers and did not see anything. This time, I opened the HTTP headers first, and then loaded the page and hit ctrl-f5. I got a whole bunch of stuff, too much to paste here. What should I be looking for? ^Ariel♥Gold 12:20, 27 July 2007 (UTC)[reply] There should be a section beginning GET /wikipedia/en/a/ad/Amy_Grant_in_1998.jpg HTTP/1.1, or you could just load the image directly after clearing the box and it should be the only section. Anomie 13:22, 27 July 2007 (UTC)[reply]I'm sorry, I don't mean to be dense, but I do not understand what this add on does, or is supposed to do. I'm still not seeing the image, and there is a whole ton of info in the box when I hit ctrl-F5. Clearing the box does just that, clears it. Pressing ctrl-F5 again fills it up with the same amount of stuff. I am still unsure what you mean by "you could just after clearing the box" I'm not sure what I'm supposed to do :) Again, I'm sorry I'm having a hard time figuring this out.^Ariel♥Gold 13:37, 27 July 2007 (UTC)[reply]Hrm, I really think you're having trouble with some sort of ad-removing proxy. The text I posted says "you could just load the image directly after clearing the box", with "load the image directly" being an external link to the image. If you're not seeing that, the proxy must have edited the page to remove the "ad". As for what the add-on does: When you want to see a webpage, your browser sends a request to the server. The server responds with some information (like what I posted earlier) followed by the HTML or image data that you actually see. See HTTP for more information. This extension shows you the actual request and the response information that you otherwise cannot see. I'm hoping that the response information you're seeing will have an explanation of why the image is being blocked for you. Anomie 14:03, 27 July 2007 (UTC)[reply]That is very strange that you have text I'm not seeing. I don't see it in IE7 either. I have never had any problems with text, or with links being removed from any websites, so I'm at a complete loss as to why yours was. Now that I'm in the edit box, I see the orginal text you put in, the link directly to the upload:wikimedia site. So, using that link from this box, here is the result:http://upload.wikimedia.org/wikipedia/en/a/ad/Amy_Grant_in_1998.jpg GET /wikipedia/en/a/ad/Amy_Grant_in_1998.jpg HTTP/1.1 Host: upload.wikimedia.org User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.5) Gecko/20070713 Firefox/2.0.0.5 Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5 Accept-Language: en-us,en;q=0.5 Accept-Encoding: gzip,deflate Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 Keep-Alive: 300 Connection: keep-alive If-Modified-Since: Fri, 25 Aug 2000 01:00:00 GMT Cache-Control: max-age=0 HTTP/1.x 200 OK Server: Netscape-Enterprise/2.0a Pragma: No-cache Date: Fri, 25 Aug 2000 23:00:00 GMT Last-Modified: Fri, 25 Aug 2000 01:00:00 GMT Accept-Ranges: bytes Content-Length: 881 Content-Type: image/gif ---------------------------------------------------------- http://sb.google.com/safebrowsing/update?client=navclient-auto-ffox&appver= 2.0.0.5&version=goog-white-domain:1:23,goog-white-url:1:371,goog-black-url:1:12900,goog-black-enchash:1:30188 GET /safebrowsing/update?client=navclient-auto-ffox&appver=2.0.0.5&version= goog-white-domain:1:23,goog-white-url:1:371,goog-black-url:1:12900,goog-black-enchash:1:30188 HTTP/1.1 Host: sb.google.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.5) Gecko/20070713 Firefox/2.0.0.5 Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5 Accept-Language: en-us,en;q=0.5 Accept-Encoding: gzip,deflate Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7 Keep-Alive: 300 Connection: keep-alive Cookie: PREF=ID=7e348dffb4cac7e3:TM=1185535849:LM=1185535849:S=o7OEtEZQjZIKTZYR HTTP/1.x 200 OK Content-Type: text/html; charset=UTF-8 Server: TrustRank Frontend Transfer-Encoding: chunked ----------------: ---- Date: Fri, 27 Jul 2007 17:19:44 GMT Cache-Control: private, x-gzip-ok="" ----------------------------------------------------------So, the big thing that jumps out at me is this sb.google.com thing, which bothers me for a number of reasons. First, I do not have google toolbar installed, (*Holds up garlic, the sign of the cross and any other items to ward off evil* lol) I have never installed any google software, nor do I have any other google add-ons installed. I do a comprehensive virus scan each night with two different programs, as well as spyware/adware scans with two separate programs each morning. In over 10 years of being online, I've never once had a virus, or any type of adware/spyware. I just ran both scans, and again, 0 threats found, even on deep scans. I found this site but it still would not explain why the image will not load in IE7, to which I have added no add-ons, and I have disabled the built-in pop-up blocker, and the built-in phishing filter. I use it for testing purposes, and therefore nothing is blocked in IE7. The add-ons I have installed on Firefox relating to advertisements (adblock, noscript) I've disabled, restarted Firefox, and tried again and still see no image. At this point, I am ready to just give up and say "oh well, I can't upload images if I want to see them" lol.^Ariel♥Gold 17:44, 27 July 2007 (UTC)[reply]It looks like your request for the image is giving you back an 881 byte GIF file (image/gif) instead of the requested JPEG. The headers indicate that it was served by "Netscape-Enterprise/2.0a" and that the image was created in August 2000. This definitely sounds like an ad-blocking proxy that is replacing the requested image with a transparent one-pixel GIF. What does the title bar in Firefox say when you load this image? Does it say "Amy_Grant_in_1998.jpg (GIF Image, 1x1 pixel) - Mozilla Firefox"? Mike Dillon 17:59, 27 July 2007 (UTC)[reply]The Google thing is an anti-phishing thing built into Firefox, you should be able to configure or disable it by opening Edit→Preferences and looking in the Security section for "Tell me if the site I'm visiting is a suspected forgery". This is the response from when you request the image, and it doesn't look at all like what I see:HTTP/1.x 200 OK Server: Netscape-Enterprise/2.0a Pragma: No-cache Date: Fri, 25 Aug 2000 23:00:00 GMT Last-Modified: Fri, 25 Aug 2000 01:00:00 GMT Accept-Ranges: bytes Content-Length: 881 Content-Type: image/gifDo you have a proxy configured in your browser settings? In Firefox, that would be in Edit→Preferences, then Security, then Network, then "Configure how Firefox connects to the Internet". Anomie 18:12, 27 July 2007 (UTC)[reply]Tools-Options-Advanced-Network = No, no proxy. Never have. I disabled the Anti Phishing thing. ^Ariel♥GoldMike: It says Amy_Grant_in_1998.jpg (GIF Image, 32x32 pixels) - Mozilla Firefox (Build 2007071317) I did notice this, and thought it odd, but I'm not sure why it is doing that. As I said, I've never once had any issues with Wikipedia images showing up (or any other images on any other site) and it is doing the same thing in IE7, so... lolAs a note, there is no "source" to view, I tried that. I tried saving the image, and I can save the image, but when I open it, it is a 32x32 white box. I'm not altogether moronic regarding things, I've created a number of websites, so while I don't know the intricacies of how the scripting stuff works, I have a basic knowledge of pages, images, html, etc. And I'm baffled, lol.I do wish to say thank you to everyone who has been working with me on this issue over the past few days, I appreciate all of your time and effort, and you have my most sincere thanks, even if I don't solve the problem. :) ^Ariel♥Gold 18:26, 27 July 2007 (UTC)[reply]Sorry for just responding now - things are a bit busy here at the moment and I'd simply forgotten to reply. If this is really a transparent proxy (or, well, technically it's an intercepting proxy, I guess), you may have more luck with actually using a third party proxy. There are several commercial services (mostly privacy-related, some of which use SSL) and there are obviously tor nodes and open proxies (which may or may not be blocked). And this doesn't necessarily have to be ad-related (even though I assume it is). I know of a number of bigish sites that do still employ some form of relatively simple content filtering solution at their gateways. Since that is mostly signature or heuristics-based there's always the possibility of a false positive. Another possibility would be an IDS. Unfortunately, that's all I can come up with and I'm afraid it's not much of a solution. Sorry. S ^up? 23:55, 27 July 2007 (UTC)[reply]Did you try the (slower) secure server, which you can find a link to in the FAQ at the top of this page? It should bypass any proxies you have, and probably will workaround the problem. If you can, try doing a tcptraceroute to the en.wikipedia.org server on port 80; it should stop at the transparent proxy instead of getting all the way to the Wikimedia clusters, and show you where the problematic proxy is. --cesarb 00:53, 28 July 2007 (UTC)[reply]Alright, thank you to everyone for the suggestions and assistance! It is most greatly appreciated.^Ariel♥Gold 17:53, 28 July 2007 (UTC)[reply]Polar co-ordinates and boundaries (template:co-ord)Can anyone with an insight into technical issues involving maps have a look at Template_talk:Coord#Co-ordinates_for_polar_locations and the discussion below that one as well, please? Thanks. Carcharoth 13:15, 25 July 2007 (UTC)[reply]External link iconA small CSS bug (I think) in IE6: when an external link wrap to the next line, the link icon disappears. Actually, it moves to the very end of the next line, and you can still see the top of the icon. I can't actually find where this is defined; I found a whole slew of a.external classes in main.css, except for regular http: links. But I think it is related to the following:background: url(external.png) center right no-repeat;Firefox displayes it correctly. --Edokter (Talk) 13:21, 25 July 2007 (UTC)[reply]Simple solution: don't use IE6. :P All seriousness aside, IE6 has numerous CSS-related issues. The CSS code you provided really is proper for the effect that the software is going for; the fact that it breaks in IE6 is unfortunate (plus, external link text probably shouldn't be two lines anyway; almost always, you can provide shorter link text and then a description next to it). EVula // talk // ☯ // 20:54, 25 July 2007 (UTC)[reply]alt attributesIs it possible, and if so how, to specify the alt attribute of an image a) when uploading the image and B) for an individual use of that image in an article? Andy Mabbett | Talk to Andy Mabbett 13:21, 25 July 2007 (UTC)[reply]Just set a caption when adding the image to the article. --cesarb 13:37, 25 July 2007 (UTC)[reply]Thanks, that's better than nothing, but the caption often isn't suitable for use as alt text (which should not simply duplicate text on the page); and some templates don't have a linked caption facility (example: Delirious?), or editors don't want to use captions for aesthetic reasons. Andy Mabbett | Talk to Andy Mabbett 16:09, 25 July 2007 (UTC)[reply]Use it without 'thumb' to hide the caption. --Edokter (Talk) 16:19, 25 July 2007 (UTC)[reply]Thank you, but I don't understand what you mean. Andy Mabbett | Talk to Andy Mabbett 18:34, 25 July 2007 (UTC)[reply]From Wikipedia:Alternative text for images#Conflict between caption and alt text: "the image syntax has no way of specifying alt text that is different from the caption". -- Patleahy 20:47, 25 July 2007 (UTC)[reply]Which is a pity, really, because users of text readers often benefit from a more specific, literal description than you would normally expect to see as a caption. Adrian M. H. 20:51, 25 July 2007 (UTC)[reply]I agree. This was suggested as a new MediaWiki feature a long time ago, see [1]. -- Patleahy 21:24, 25 July 2007 (UTC)[reply]"a more specific, literal description" sounds more like a "longdesc" attribute than "alt". Andy Mabbett | Talk to Andy Mabbett 23:33, 25 July 2007 (UTC)[reply] It does not have to be longer than a caption, it just helps at times if it is different. The ability to enter some longdesc text would be useful, but may not be worth implementing in practice. Adrian M. H. 17:11, 26 July 2007 (UTC)[reply]Editprotected request at Cite webSee Template talk:Cite web#Parameter language. The span wrapping $language was specifying an apparently superfluous position, which I'm told was causing display errors in one or more versions of IE. I've removed that, for the time being, if anybody wants to have a look. The rest of that request has more to do with whether $language should be bolded or not -- for that, I figure we need some more discussion. Input welcomed and requested. – Luna Santin (talk) 20:13, 25 July 2007 (UTC)[reply]WatchlistI'm not sure if this has ever been discussed, but I was wondering if its possible to include the time/date of the edits made on a watchlist. Pats Sox Princess 20:35, 25 July 2007 (UTC)[reply]It already does, or am I being dim and missing your point? (Which happens!) Adrian M. H. 20:46, 25 July 2007 (UTC)[reply] To clarify, when I view my watchlist, I see a list of edits under a date heading, but the time the edit was made isn't shown. Pats Sox Princess 20:49, 25 July 2007 (UTC)[reply]Oh, then I guess it's your watchlist that is missing something rather than me! It should show the time of each edit, regardless of whether you use enhanced changes, all recent edits, or whatever setting you choose. Adrian M. H. 20:52, 25 July 2007 (UTC)[reply]Do you know how to correct that problem then? Pats Sox Princess 21:11, 25 July 2007 (UTC)[reply]No, otherwise I would have provided the answer. I have never seen this before. Adrian M. H. 21:13, 25 July 2007 (UTC)[reply]Are you sure? Here is an entry from my watchlist (pulled off the top):(diff) (hist) . . User talk:Moe Epsilon‎; 22:41 . . (+1,168) . . Rbellin (Talk | contribs | block) (→Lists - an agony in eight fits - reply)Note the time: 22:41. Prodego ^talk 22:43, 25 July 2007 (UTC)[reply]OMG. I can't believe I never noticed the time. I'm used to looking at Special Contribs, and just learning to use the watchlist. I guess the fact that the time is in a different place in the watchlist than Special Contribs through me off. Pats Sox Princess 23:34, 25 July 2007 (UTC)[reply]Yeah, we really need to standardize all this crap. If contribs lists entries one way, wl and logs should probably do the same. AmiDaniel (talk) 17:32, 26 July 2007 (UTC)[reply]That's bugged me for a long time too. The fixed length fields should be first in the line so more items line up. Note also that a history item contains the date. Or maybe that's specific to my settings, as mine shows seconds too: Recent changes:(diff) (hist) . . Oneworld destinations‎; 11:17:07 . . (+4) . . 213.158.196.65 (Talk) (IB fly's to warsaw now:))Article History:(cur) (last) 2007-07-26T11:17:31 Malkageffen (Talk | contribs) (6,740 bytes) (undo)The difference between (cur) and (last) has always puzzled me, as does the meaning of the labels. (I muddle through by ignoring (cur)). —EncMstr 18:31, 26 July 2007 (UTC)[reply] (cur) is the difference to the current revision; (last) is the difference to the revision before the one you clicked last on. --ais523 08:24, 27 July 2007 (UTC)Well, if its possible, everything should be standardized. Make time go to the right of the pagename, and get rid of (cur) and (last)--just use (diff) Pats Sox Princess 19:38, 26 July 2007 (UTC)[reply]Oooooooo nooooooo! Please put the times to the left of the page name, so they line up. And delta sizes to the left of the name too. —EncMstr 20:02, 26 July 2007 (UTC)[reply]Yeah I meant left. Right wouldn't line them up correctly. Pats Sox Princess 21:13, 26 July 2007 (UTC)[reply]Captchas when revertingI've found that when I try to revert simple page blankings I'm required to fill in a captcha because of the external links that are being added back. When did this happen? This is becomming a large annoyance when fighting vandals and my patience for this recent change is wearing thin. --Android Mouse 23:28, 25 July 2007 (UTC)[reply]As far as I was aware it's only supposed to affect anons and newly registered users, it's one of the reasons I hardly ever edit as an anon anymore. --VectorPotential^Talk 23:37, 25 July 2007 (UTC)[reply] That makes sense, since I tried reverting on this account and didn't get it. How long until I can use the account I created today for vandal fighting without having to deal with the captcha? --Android Mouse 23:47, 25 July 2007 (UTC)[reply] Should be 4 days from whenever you registered your new account until you get an autoconfirmed flag, after that you should be able to revert without getting the CAPTCHA. --VectorPotential^Talk 23:49, 25 July 2007 (UTC)[reply] Ok, thanks. --Android Mouse 23:50, 25 July 2007 (UTC)[reply]New Favicon?I am seeing a weird favicon now in my browser when i come to wp, its like a 3d box with red blue and light blue sides? Is that supposed to be there? When I request the url directly its the correct one. Is this a bug in IE or something? pw 13:14, 26 July 2007 (UTC)[reply]I'm sure not seeing it in Safari, so probably, yes. Nihiltres^(t.l) 13:47, 26 July 2007 (UTC)[reply] To be fair, Safari chaches favicons something awful. The image itself could change (or even the hard-coded reference to it) and you'd keep using the cached version. EVula // talk // ☯ // 21:23, 26 July 2007 (UTC)[reply]User creation log doesn't go back past 25th JulyI was trying to look up user accounts created between 15:00 and 15:20 on 20th July, but the log doesn't go back past 25th. Despite that, usernames created earlier do show up when searched for. Is there any way around that problem? I neeed this information because it may be relevant to a vandal who utilized sockpuppets created several days before the actual attack in order to circumvent partial page protection. There is a gap between the creation times of two of his accounts - which may mean he has created one or more in between - so I want to check the list of accounts created in the corresponding period to see if any other were used in a manner that may suggest a connection with the vandal.--The Fifth Horseman 15:44, 26 July 2007 (UTC)[reply]There appears to be a limitation on viewing logs past a certain point in the past, most likely for performance reasons. Aren't there database dumps you could download? -Nard 15:51, 26 July 2007 (UTC)[reply]Didn't know about that. After I checked the descriptions for the dump contents, however, none of them indicate - to my understanding - that they contain the user creation logs. Or am I just plain missing something I should have been aware of? --The Fifth Horseman 09:52, 27 July 2007 (UTC)[reply] You want the "logging" dump. That contains all log entries, including user creation. Matt/TheFearow (Talk) (Contribs) (Bot) 10:43, 27 July 2007 (UTC)[reply]semi-protection indicatorThe article Sea turtle is current semi-protected, yet I don't see the usual indicator on the top of the rendered page (while I'm logged on) that indicates that it is indeed semi-protected. Why is Sea turtle different from other semi-protected pages like George Bush ? - Bevo 16:30, 26 July 2007 (UTC)[reply]Because {{pp-semi-protected}} was not put on Sea turtle page. Personally, I find it very inefficient that protection requires 2 edits in history (protect + put template). Developers might fix it in future, see bugzilla:10347. Also automatic protection notice now can be done with Javascript. I do not propose this at Wikipedia:Village pump (proposals) simply because nobody else will care, as usual ∴ Alex Smotrov 16:55, 26 July 2007 (UTC)[reply] I've added the template. Tra (Talk) 16:59, 26 July 2007 (UTC)[reply] Yes, you can see it now. I think that problem has occured on various other pages around Wikipedia, I can't name any, but I'm sure I've come accross some on RC Patrol. Lradrama 13:27, 27 July 2007 (UTC)[reply]Auto watch list questionI am interested in a system that might allow a group of editors to add a large number of selected articles relating to a specific Project to their watch list. Is there a simple way to find out if any existing bots do this? Is there an alternative method anyone knows of? Cheers Ben MacDui (Talk) 19:17, 26 July 2007 (UTC)[reply]http://en.wikipedia.org/w/index.php?title=Special:Watchlist&action=raw ←Ben^B4 19:20, 26 July 2007 (UTC)[reply] Gotcha - thanks. Ben MacDui (Talk) 19:33, 26 July 2007 (UTC)[reply]Open in New TabHaving now become addicted to tabbed browsing, I find it frustrating that the Wikipedia search entry does not support this. I doubt that I am the first to raise this point. Is this the correct forum to mention this? --CloudSurfer 21:14, 26 July 2007 (UTC)[reply]Depends on your browser. I use Safari/Mac, and I can get it to show up in a tab just fine. EVula // talk // ☯ // 21:22, 26 July 2007 (UTC)[reply]Try this in your javascript. --Splarka (rant) 07:43, 27 July 2007 (UTC)[reply]addOnloadHook(function(){ if(!document.getElementById('searchform')) return var search = document.getElementById('searchform'); var check = document.createElement('input'); check.setAttribute('type','checkbox'); check.setAttribute('name','newwindow'); check.setAttribute('id','newwindow'); check.setAttribute('value','true'); check.setAttribute('onclick','searchTargetSwap(this.checked);'); search.appendChild(check); search.appendChild(document.createTextNode('new window')); }); function searchTargetSwap(chkd) { var search = document.getElementById('searchform'); if(chkd) { search.setAttribute('target','_blank'); } else { search.setAttribute('target',''); } }Alternative approach (without extra checkbox). Tested in IE 6 (hold Shift) and Firefox 1.5 (hold either Shift or Ctrl). You do not need the script in Opera 9.2, it has this functionality built-in (hold Shift) ∴ Alex Smotrov 15:17, 27 July 2007 (UTC)[reply]//hold shift to open Go/Search in new windows addOnloadHook(function(){ var but if (but=document.getElementById('searchGoButton')) but.onclick = newWinSearch if (but=document.getElementById('mw-searchButton')) but.onclick = newWinSearch }) function newWinSearch(e){ if (!e) var e = window.event this.form.target = (e.shiftKey || (is_gecko && e.ctrlKey)) ? '_blank' : '' }I've been running something like Sparkla's checkbox version for a few months: User:Mike Dillon/Scripts/searchNewWindow.js. Mike Dillon 18:03, 27 July 2007 (UTC)[reply]P.S. Whether or not this opens in a tab is controlled by browser settings for whether or not "_blank" windows open in a tab or window by default. There is no way in JavaScript to create a tab explicitly, at least no cross-browser way that I've seen. Mike Dillon 18:06, 27 July 2007 (UTC)[reply]Bugzilla?Hi, all! I reported a request (not actually a bug) to Bugzilla.wikimedia.org several days ago. Will I hear something back from them? What's the process? Bugzilla doesn't seem well documented except for How to report a bug - everything else is a bit unclear. -- SatyrTN (talk | contribs) 21:20, 26 July 2007 (UTC)[reply]Assuming you used a real email address, you'll get an email whenever there is a comment or status change posted. Given the work load of the devs, low priority requests can sometimes be ignored for very long periods of time, though. Dragons flight 21:31, 26 July 2007 (UTC)[reply]Autoblock durationI've always assumed the autoblocker sets a duration of 24 hours, when we block somebody with autoblocks enabled. I've seen a few cases where someone's autoblock seemed to outlast their original block, though. In cases where fewer than 24 hours remain on the original block, is the autoblock duration still 24 hours? – Luna Santin (talk) 22:04, 26 July 2007 (UTC)[reply]Yes. --cesarb 23:36, 26 July 2007 (UTC)[reply]I thought that was fixed a long while back? Dragons flight 23:39, 26 July 2007 (UTC)[reply]New usermessageI originally had this in WP:VPA but decided to move it here. I was wondering if there is a way to make a usermessage similar to the orange "You have new messages" one. The new one would appear when a user page was edited by anyone other than the owner of the page. The message would probably be red and say something along the lines of "You have new vandalism." I realize that there probably would be a large amount of people against this as an official feature, but is there a way for it to be implemented as a user script? — Bob • (talk) • 22:40, July 26, 2007 (UTC)The problem with such a feature is that there is no "owner" of a page. Wikipedia:Ownership of articles might be of interest to discussion participants. — Madman bum and angel (talk – desk) 00:15, 27 July 2007 (UTC)[reply] Yes, it's more-or-less possible to do it with JavaScript, but I hesitate to write such a script for the reason Madman mentions. Gracenotes^T § 00:43, 27 July 2007 (UTC)[reply] Ah, I thought it'd be okay since it would be only to a userpage, but I see that they too belong to the community. Thanks anyways, — Bob • (talk) • 02:06, July 27, 2007 (UTC) I actually quite like your idea. My page has been vandalised quite a lot, and sometimes, it hasn't been noticed by others on RC Patrol, meaning someone could look at my page and see some vulgar language (or something else that usually gets put on by vandals) and I won't know unless I look at my watchlist. I think it's a good idea. Lradrama 13:30, 27 July 2007 (UTC)[reply]New Shortcut Prefix ProposalI would think it would be excellent if we could create a shortcut like WP: for user pages. U: would be my suggestion for user pages and UT: for user talk pages.. Thoughts? Jmfangio| ►Chat 09:22, 27 July 2007 (UTC)[reply]Bad idea. WP: shortcuts do not work magically, a redirect has to be created for each: Special:Prefixindex/WP:. Allowing potentially millions of users to have a mainspace redirect for every page and subpage of their user and talk namespace would be... silly. --Splarka (rant) 10:05, 27 July 2007 (UTC)[reply]I like the idea, but it would only be used by reasonably active users. Probably a rule that you cannot create one to your own - leave it to if someone else finds it necessary. Matt/TheFearow (Talk) (Contribs) (Bot) 10:41, 27 July 2007 (UTC)[reply]That's a good point MtF. My thinking is that there are a million instances when I'm jumping back and forth between different user talk pages and it would be easier to just type U:username into the search box. Jmfangio| ►Chat 10:58, 27 July 2007 (UTC)[reply] It's a nice thought, but in practice it's a bad idea. Littering mainspace with tons of userspace redirects isn't a good idea. When users become inactive the redirects will stil be there, and we'd need to patrol them. Overall it wouldn't work. --Deskana (talk) 11:21, 27 July 2007 (UTC)[reply] Yeah, I see your point. Do you think setting up M:Help:Magic words would be possible for it? I understand prefixes, but I have no idea how magic words work so maybe that's a practical question. Jmfangio| ►Chat 11:26, 27 July 2007 (UTC)[reply] Magic words don't do what you seem to think they do. If I put __NOTOC__ in here, there'd be no table of contents, for example. They alter things when you put them on pages, they don't have anything to do with shortcuts or redirects. --Deskana (talk) 11:31, 27 July 2007 (UTC)[reply] I was thinking there would be someway to create a template that relies on magic words so that if someone entered the template - like TL:U:username, it would work. That's where i was thinking magic words could come in, but again - i don't know a darn thing about implementing them. Thanks for your comments, alas i will have to stick to the current system :-) Jmfangio| ►Chat 11:39, 27 July 2007 (UTC)[reply] It seems shortcuts like "u:" are possible with the current software by defining them as interwiki links. That could be convenient in the address box of the browser. If used in a page the appearance according to CSS class "extiw" could be confusing.--Patrick 12:39, 27 July 2007 (UTC)[reply]But wouldn't that still require setting it up for each user? Juan Miguel Fangio| ►Chat 12:52, 27 July 2007 (UTC)[reply]It's a good idea, and I would certainly use it. It saves using the search function or typing User:(username) in all the time. But I have to admit, User:Deskana ;-) has a very good point. Lradrama 13:31, 27 July 2007 (UTC)[reply]Oh duh - I wasn't reading. I somehow skipped over the interwiki links! :-) Juan Miguel Fangio| ►Chat 13:34, 27 July 2007 (UTC)[reply] I think it's a great idea, although one problem is that if it's set up as an interwiki link and searched for in the search box, it would be necessary to click on the 'you searched for' message before going to the page. Tra (Talk) 15:12, 27 July 2007 (UTC)[reply] That is why putting it in the address bar (replacing the current page name) might be preferable.--Patrick 22:54, 27 July 2007 (UTC)[reply]There are millions of Wikipedia users, it would be very likely that more then 1 user would want the same shortcut, thousands of users would fight over U:Model Wikipedian. Tcrow777 talk 23:49, 29 July 2007 (UTC)[reply]I don't think userpage redirects would be a good idea, but would it be possible to make shortcuts to special pages? I know you can't redirect to special pages (is that a bug or intentional?), but even soft redirects could be useful. — Bob • (talk) • 04:13, July 30, 2007 (UTC)MediaWiki microformat mark-up issuesFor issues encountered when adding microformats to Wikipedia and other sites, using Media Wiki mark-up, please see [2]. Andy Mabbett | Talk to Andy Mabbett 16:50, 27 July 2007 (UTC)[reply]Wikilink problemI tried to nominate my recently created article Bernard d'Abrera at Did You Know but I was unable to make the wikilink work properly: [3] --Filll 17:55, 27 July 2007 (UTC)[reply]Fixed it. What happened was that the link name had this ` and the article name had this'. It was recognized as two different names. I redirected the article so it always goes to the correct name. --_Hdt83 ^Chat 18:59, 27 July 2007 (UTC)[reply]Thank you very much.--Filll 20:15, 27 July 2007 (UTC)[reply]Master-Slave language in db lock messageI was attempting to edit a page when I received a "Database locked" error message ("The database has been automatically locked while the slave database servers catch up to the master" [controlled by MediaWiki:Readonly lag]). I know there has been a bit of controversy surrounding those terms, and I think it would be best for us to at least interwiki link the words "master" and "slave", not only as a political correct approach, but also to be less cryptic and have the error message more understandable/useful. Thoughts? -Mysekurity 21:00, 27 July 2007 (UTC)Due to the frequency that I have recently been seeing those messages, I can see why it would be a good idea. I understand them but I am a programmer. I know many people who would have no idea what that really meant and would think something was broken. Perhaps a clarified version might be a good idea if possible. -- Chrislk02 (Chris Kreider) 21:03, 27 July 2007 (UTC)[reply] Well, we could wikilink the phrases to Master-slave (computers). EVula // talk // ☯ // 21:22, 27 July 2007 (UTC)[reply] I like EVula's idea: let's wikilink it. It avoids changing the wording while adding meaning. Nihiltres^(t.l) 21:25, 27 July 2007 (UTC)[reply] I think EVula has the right idea. Until(1 == 2) 21:37, 27 July 2007 (UTC)[reply] I was about to edit the page myself, but then realized it'd probably be better to check here first. I was going to wikilink both slave and master to master-slave (computers), but wondered if there was something more appropriate. If someone gives me the go-ahead, I'll edit it now. -Mysekurity 22:12, 27 July 2007 (UTC) [ec]Since everyone seems to agree that I'm awesome for making the suggestion (mmm, ego), I made the change to MediaWiki:Readonly lag. I apparently created the page, though, so I'm not sure if it will override stuff or not. I'm actually kind of confused by it... EVula // talk // ☯ // 22:13, 27 July 2007 (UTC)[reply] That should be fine, the MediaWiki namespace is for deviations from the default system messages.--Patrick 23:03, 27 July 2007 (UTC)[reply] /me sings EVula theme song Gracenotes^T § 22:36, 27 July 2007 (UTC)[reply] Now all we have to do is lock up the database to test it. Anyone feel like deleting the sandbox? -Mysekurity 22:48, 27 July 2007 (UTC) EVula, from what I understand of the MediaWiki namespace, the creation oddity is because a nonexistent MediaWiki page has content: the MediaWiki default message (which is null by default, for some). Creating the page overrides the default with the content of the page. As for deleting the sandbox... ouch that would lag. We'll have to wait and see, I don't think anyone really wants to try to lock up the servers. Nihiltres^(t.l) 00:57, 28 July 2007 (UTC)[reply] Just wait until a few users are renamed. I'm pretty sure someone has tried to delete the sandbox before, causing about 15 minutes of lag, so please don't do that. You will get several developers vying for your head. Titoxd^{(?!? - cool stuff)} 01:05, 28 July 2007 (UTC)[reply] I guess facetiousness doesn't come across over the interweb. I've never understood why they don't just lock up deleted pages that default to something else (Commons images on a project, e.g.). Is there a technical reason they can't stop someone from uploading naughty pictures over one from the Commons, or is it just policy? -Mysekurity 07:49, 28 July 2007 (UTC) I don't think such a feature (autoprotection of images that already exist on Commons) exists, but it shouldn't really be too hard to implement. File a bug? As for MediaWiki: pages, those are always fully protected (and cannot be unprotected) anyway. —Ilmari Karonen (talk) 10:05, 28 July 2007 (UTC)[reply] Per m:Help:User rights it seems "upload", "reupload" (overwrite), "reupload-shared" (overwrite on Commons?) are separate rights, so it should be technically easy, if not politically. --Splarka (rant) 07:14, 29 July 2007 (UTC)[reply](arbitrary undent) I just got a database locked notice while deleting stuff at CAT:CSD, and the words didn't seem to be linked, despite the message having been changed. Any idea what happened? Nihiltres^(t.l) 18:59, 28 July 2007 (UTC)[reply]Don't quote me on this, but I believe some MediaWiki pages don't accept Wikisyntax, and instead have to use standard HTML (a href tags, e.g.) to achieve wikilinking. Maybe I'll try to change the wikilinks to plainlink externals and see what happens. -Mysekurity 08:22, 29 July 2007 (UTC) I've seen the same thing on some page somewhere, but I was expecting broken [[formatting]] in that case, I suppose. Well, trying out HTML links can't hurt all that much, it's probably worth it. Nihiltres^(t.l) 14:22, 29 July 2007 (UTC)[reply]It really is not that important. The terms "slave" and "master" are only insensitive if we apply them to people, it is a perfectly healthy relationship that 2 pieces of hardware can share. Hardware thrives on this sort of relationship, and does not do well in the relationship of equals humans have come to expect. Has anyone actually complained about this wording? Until(1 == 2) 14:29, 29 July 2007 (UTC)[reply] Right. I don't think you, I, nor anyone else reading this page (or most of the tech aspects of Wikipedia) would object to the language, and I think it's written in such a way that it isn't likely to make someone think we were talking about slavery, but it's enough to make a new user question. I was more concerned about the jargon being unfamiliar to the layperson and therefore upsetting than about the political correctness. That said, there really isn't any reason that "primary" and "secondary" wouldn't work just as well. -Mysekurity 20:32, 29 July 2007 (UTC)That makes sense. Until(1 == 2) 20:52, 29 July 2007 (UTC)[reply] Except for the reason that "primary and secondary" are not used much in that regard, "master and slave is". I do support a link to Master-slave (computers) if really necessary. Garion96 (talk) 21:02, 29 July 2007 (UTC)[reply] Yes, but the purpose of the message is not to provide the devs with anything. They already know what's happening. It's the layperson who we're more concerned about, and whom this page will actually be useful. Cryptic error messages and unfamiliar jargon detract from that usefulness and should be avoided or explained if Wikipedia is to become usable by the tech-illiterate. -Mysekurity 23:29, 29 July 2007 (UTC) Correct, so a link to Master and slave should educate the tech-illiterate, much better than using a term which is not the actual term for it. But if that is too much work for the dev's (or perhaps they think it's pointless work on which I tent to agree) it should stay the way it is. Garion96 (talk) 19:07, 30 July 2007 (UTC)[reply] I just got the locked database message and the words weren't linked. — Bob • (talk) • 01:31, July 31, 2007 (UTC) You could try raw HTML, which might work. I don't think it will though. Prodego ^talk 03:00, 31 July 2007 (UTC)[reply] I've added some title attributes to the HTML markup that Prodego added to the page. That should help finish the "faux wikilink" effect. EVula // talk // ☯ // 03:51, 31 July 2007 (UTC)[reply]BAG membership nominationAs is the custom, to ensure openness and encourage community participation, I'm informing the technical community of a pending BAG membership nomination. All comments are welcomed, and all are treated equally. Madman bum and angel is the user in question, and you can view the debate here. As always, any comments you may wish to make on open BRFAs will be very welcome on their respective pages :). Thanks, Martinp23 18:55, 28 July 2007 (UTC)[reply]Ditto, for Daniel as well. Daniel 03:05, 29 July 2007 (UTC)[reply]Monobook CSS infobox class: text-align:leftHere's an example This will wrap: It shows the odd wrapping in Firefox 2.0.0.4, and IE6 and IE7 via Wine Anomie 21:26, 28 July 2007 (UTC)[reply]Currently infoboxes inherit text-align:justify from the bodyContent class (in MediaWiki:Monobook.css). This can look ugly, because it creates huge gaps; see for example Universal Serial Bus, where in "Number of devices" there is such a big space between "Number" and "of" that "of" is closer to the text in the right hand column than to "Number". I think it makes sense to add text-align:left to infoboxes to fix this. Headings should be unaffected since they have text-align:center set manually on them. Hairy Dude 20:46, 28 July 2007 (UTC)[reply]Hrm, it looks just fine in my browser (Safari 2.0/Mac). What are you using? Can you provide a screenshot? EVula // talk // ☯ // 20:48, 28 July 2007 (UTC)[reply] This may be a result of having Misc > Justify paragraphs enabled in Special:Preferences. Tra (Talk) 20:56, 28 July 2007 (UTC)[reply] I think you're right. For the record, I'm using Firefox 2.0.0.5 on Ubuntu Feisty. I still think the change is a good idea - I'm not going to give up nicely justified article text for slightly better-looking infoboxes. Hairy Dude 21:00, 28 July 2007 (UTC)[reply] You could put .infobox {text-align:left} into your own monobook.css so that it appears just for you. Tra (Talk) 21:25, 28 July 2007 (UTC)[reply]Deleting orphaned categories.I have recently moved all the articles from Category:Max Ernst paintings to the new Category:Paintings by Max Ernst to be more in keeping with the relevant naming convention. I assumed that categories disappear after you remove the parent categories, but the old category still exists with no links or contents. How do I get rid of it? Justin Foote 23:12, 28 July 2007 (UTC)[reply]See WP:CSD#Categories. I think #1 should be applicable, which needs {{Db-catempty}}. Adrian M. H. 00:01, 29 July 2007 (UTC)[reply]Deleting user scriptsIt seems that {{db-userreq}} no longer works to delete user scripts, as the template is not expanded. Similarly, manually adding the script page to the deletion request categories fails. Any suggestions? Anomie 23:19, 28 July 2007 (UTC)[reply]As far as I know, on .js/.css pages templates and links are not rendered but they are still "expanded" in some sense. Your userscript page does appear in Category:Candidates_for_speedy_deletion_by_user ∴ Alex Smotrov 23:40, 28 July 2007 (UTC)[reply] How confusing... Thanks! Anomie 01:02, 29 July 2007 (UTC)[reply]State Farm Insurance looks blankThe page State Farm Insurance looks blank, even though it seems to have valid content. --Amir E. Aharoni 14:57, 29 July 2007 (UTC)[reply]Move permissionsThe option to "unlock move permission" in the protection window for currently unprotected pages does not seem to working properly for me. Specifically, the move window stays disabled even after clicking the checkbox. Dragons flight 22:16, 29 July 2007 (UTC)[reply]Confirmed. It looks like Rob Church broke it somewhere in here. Prodego ^talk 22:17, 29 July 2007 (UTC)[reply] It was accidentially broken earlier while fixing mediazilla:10642; actually Rob Church is the one who fixed it, see mediazilla:10732. Now you just have to keep checking Special:Version and wait for that revision. ∴ Alex Smotrov 22:36, 29 July 2007 (UTC)[reply] Ah, that would explain why I couldn't find exactly what broke it. :) Prodego ^talk 22:42, 29 July 2007 (UTC)[reply]My edits taken awayI was just editing Dir en grey discography, wikifying it. I was going by section, so as not to wikify it all, and then have an edit conflict. Anyways, I did the Albums section, the EPs section, the Remix albums section, and was working on the Singles section. I was done with the Singles section, clicked "Save page" and it said "Wikipedia is currently down, so you will not be able to save your edits." or something like that. I assumed it just meant the Singles section, since all the other sections had already been saved. I edited the page, and the other sections were wikified, and I started to work on the Singles section again. I was done, clicked "Save page" and then the only section that was wikified was the Singles section. All the other sections had been reverted to a list format. Help? *silver* 23:47, 29 July 2007 (UTC)[reply]This is your problem, I think. --Deskana (banana) 23:49, 29 July 2007 (UTC)[reply] Any idea why Cyrus XIII reverted all my hard work? I've wikified other discography pages and they were not tampered with. I'm really sad that all my hard work just went down the drain. *silver* 23:51, 29 July 2007 (UTC)[reply] "restoring full release dates, no need for tables in a name/date-only-scenario" was the given reason. If that doesn't make sense to you, you should ask Cyrus. --Deskana (banana) 23:53, 29 July 2007 (UTC)[reply] Well, it makes sense, but the full release date is on the page for the album/single/EP/remix album/compilation album! Besides, a lot of discography pages do not have the full release date. That was a lot of hard work.alpha channel for faviconWhy not map the white colour of the favicon to an alpha channel so it appears with transparency? I have uploaded a version of the Wikipedia favicon with transparency to this location... http://wikicode.frihost.net/bt/favicon.ico I have tested this in Firefox and the transparency does function correctly so there is not that silly white box surrounding the W in tabs and in the bookmarks menu. --Anthony5429 03:05, 30 July 2007 (UTC)[reply]It used to be that way, looks like someone broke it. Prodego ^talk 03:47, 30 July 2007 (UTC)[reply]Boise is devoid of intelligence?I couldn't help but laugh at this. Guess Google caught a vandalized version again. See the entry for Wikipedia, and you'll know what I mean... --_Hdt83 ^Chat 05:54, 30 July 2007 (UTC) http://www.google.com/search?hl=en&lr=&as_qdr=all&q=boise&btnG=Search.[reply]Quick! To WP:BJAODN! — Bob • (talk) • 08:09, July 30, 2007 (UTC)My signatureSorry to post on the village pump again, but I have another problem. Silly me. Anyways, I want *silver*► to be my signature for whenever I type ~~~~, with the date, of course. Anyhow, I entered that into the signature slot on "My Preferences," and checked "Raw signature." It was too big, but I'm positive I've seen people with signatures bigger than that. Help? *silver* 16:19, 30 July 2007 (UTC)[reply]Just use <font color="#003333">blah</font> instead of <span style="background-color:White;color:#003333">. Until(1 == 2) 16:25, 30 July 2007 (UTC)[reply] <font> instead of CSS? Gracenotes goes into combat mode ;) It's not a length problem, actually: it's an issue with invalid XHTML. The last </small> tag isn't matched with a starting tag, and the first <span> isn't matched with a closing tag. Regardless, the following works, and it's a bit shorter than your previous one: *silver*► Hope this helps, Gracenotes^T § 21:30, 30 July 2007 (UTC)[reply] You could further shrink the size by making "background-color" just "background". EVula // talk // ☯ // 22:07, 30 July 2007 (UTC)[reply]Check out this recently archived thread too: Wikipedia:Village pump (proposals)/Archive#Reduce allowed length for signatures and usernames. --Quiddity 16:55, 30 July 2007 (UTC)[reply]Links in - slightly confusedSorry to be vague, which I suspect is the case, but I want to link to some wiki pages and need to use a complete URL. I have looked at all the policy and help I can find on links but haven't found what I need. In order for my app, a PDF printer driver, to recognise a link it must contain “.htm”, “.html” … etc which Wiki url,s don’t. I tried adding various extensions directly and also /index.htm but can't get anything to work. Please help, I only need a document name, default or otherwise and I am feeling a little silly now, after all links are hardly complicated.To automatically generate links to wikipedia, http://en.wikipedia.org/w/index.php?title=name%20of%20page (where you replace name%20of%20page with the percent-encoded name of the page) is the easiest method. The links don't end in .html; one workaround you could use to make a page name that did end in .html is this: http://en.wikipedia.org/w/index.php?title=Main%20page&fakeparameter=.html (which is a somewhat cheeky way to do it, but a similar method has been used to fool other programs in the past). --ais523 18:04, 30 July 2007 (UTC)Help with uploading SVGsI uploaded Image:Questionmark.svg, and is it just me, or is there a large white rectangle in the bottom-right corner? Just in case it is just me, I'm running Firefox 2.x and made the picture in Inkscape. Does anyone know how to fix it? I tried uploading new versions a couple of times with slightly different saved files, and this always happens. Here's the XML if that helps (I don't know if it does). I uploaded Image:Questionmarkblue.png as a bitmap alternative to it (that's what it should look like). — Bob • (talk) • 21:09, July 30, 2007 (UTC)Hi there :) It's possible that the server doesn't have the font you used for the "?". You might want to convert the "?" to a path. -- JaeSharp 07:17, 31 July 2007 (UTC)[reply]sortable wikitables and SafariAs some might know, there have been some issues with sortable wikitables left on Safari. I have found the cause of some of my problems. Some tables use a sortkey trick. Take the following tablecell: <td><span style="display:none">02005-06-02</span> <a href="/wiki/June_2" title="June 2">June 2</a>, <a href="/wiki/2005" title="2005">2005</a></td> The idea that the span is hidden from the reader, but is used in sorting. This was not working in Safari (2 and 3). The cause lies in wikibits.js in function ts_getInnerText(). This function retrieves the textual contents stripped from the html tags.function ts_getInnerText(el) { if (typeof el == "string") return el; if (typeof el == "undefined") { return el }; if (el.innerText) return el.innerText; // Not needed but it is faster var str = ""; var cs = el.childNodes; var l = cs.length; for (var i = 0; i < l; i++) { switch (cs[i].nodeType) { case 1: //ELEMENT_NODE str += ts_getInnerText(cs[i]); break; case 3: //TEXT_NODE str += cs[i].nodeValue; break; } } return str; }The problems is in the line "if (el.innerText) return el.innerText;" innerText is an IE extension. on FF it is not used because it is undefined. On Safari the innerText call IS present but unfortunately, it is not the same as the IE version. On Safari it does not return "display:none" text. Therefore the whole sortkey is unusable on Safari. There are 2 ways to deal with this issue:remove the entire line. I'm not sure if it's actually that much faster on any platform, and the FF way seems to work just fine. add "if (el.textContent) return el.textContent;" right before it. This basically is the FF variant of innerText and Safari supports it as well. Some very old browsers might not have it, but they can use our current FF for-loop. I tested textContent on Safari 2/3, FF 2.0 and Opera 9.I'd file a bugticket, but i'm too lazy to register in yet another bugzilla system. --TheDJ (talk • contribs) 21:46, 30 July 2007 (UTC)[reply]Well, I guess we're too lazy to fix your bug. :) --brion 01:27, 31 July 2007 (UTC)[reply]Dead link checkingI've completed much of the work on my dead link checking script. With this script, you may check a single page or entire category of pages for bad and dead links. Each link detected will be given a score based on its error code, content-type, and redirection. This way it can elevate 'soft' 404 error or servers sending out a bad MIME type. The results are colored according to their rank progressing from red to green and group by page. If you’re using Firefox, click the "Use javascript" button to enable enhancements that let you preview the page, built-in search types for locating the link again, and submitting the changes to Wikipedia using your account. I'm currently looking for some feed back over the interface and methods for searching replacement links. —Dispenser 23:20, 30 July 2007 (UTC)[reply]