Wikipedia:Village pump (technical)

[[:|WP:VPT
WP:VP/T]]

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, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217

This page is automatically archived by Werdnabot. Any sections older than 7 days are automatically archived to Wikipedia:Village pump (technical)/Archive. Sections without timestamps are not archived.

These discussions will be kept archived for 7 more days. During this period the discussion can be moved to a relevant talk page if appropriate. After 7 days the discussion will be permanently removed.

Getting a partial source, with action=raw?

Is there a way to get the source (ie. raw wikitext) of a section of an article? I tried adding &section=x to the URL (as per how section editing is done) but it still pulls the whole article. If there is no current way to do this via a URL, is there a good way to parse the whole wikitext to split into sections (ie. a regex to split on)? —Daniel Vandersluis^(talk) 21:11, 25 April 2007 (UTC)[reply]

Here's the MediaWiki parser regex for making headings (converted to JavaScript):

       for ( i = 6; i >= 1; --i ) {
           var h = "";
           var r = i;
           
           while (--r >= 0) h += "=";
           
           re = new RegExp("^"+ h +"(.+)"+ h +"(\\s*)$", "gm");
           
           text = text.replace(re, "<h"+ i +">$1</h"+ i +">$2");
       }

You could a split/explode just before the regexp, and you'd have an array of sections. As far as I know, there's no current way to do this via URL. Datrio 07:37, 26 April 2007 (UTC)[reply]

Thanks! —Daniel Vandersluis^(talk) 13:49, 26 April 2007 (UTC)[reply]

dontcountme=s?

Does anybody know why and where the "dontcountme=s" parameter for MediaWiki's index.php originated and what, if any, use it had (or has). This parameter can be found in many installation codes for user scripts, as in:

// install User:Cacycle/wikEd in-browser text editor
document.write('<script type="text/javascript" src="'
+ 'http://en.wikipedia.org/w/index.php?title=User:Cacycle/wikEd.js'
+ '&action=raw&ctype=text/javascript&dontcountme=s"></script>');

I have tried hard to find any reference to it, including the MediaWiki source code. It is also not listed under http://www.mediawiki.org/wiki/Manual:Parameters_to_index.php. Cacycle 03:05, 2 May 2007 (UTC)[reply]

I've also wondered about this... Gracenotes^T § 03:40, 2 May 2007 (UTC)[reply]

I would guess it has something to do with access log processing for statistics and that it has always ignored by the MediaWiki software itself. Mike Dillon 05:53, 2 May 2007 (UTC)[reply]

Well "common selse" would suggest that it relates to MediaWiki's hit counting feature. I can't speak to wheter or not it actualy has any effect in the current MediaWiki version. I know hit counting is disabled on the English Wikipedia and probably most other MediaWiki projects for performance reasons so around here it defenently doesn't do anyting regardles. The way it's used would suggest the intention was to avoid having stylesheet and javascript pages like MediaWiki:Common.css or MediaWiki:Common.js show up as the most viewed pages on the site. People have then just copied the syntax used to load those pages when they created theyr own user css and js pages, but I'm pretty sure it can safely be done away with seeing as the hit counting mechanism is turned off anyway. --Sherool (talk) 06:33, 2 May 2007 (UTC)[reply]

It was a hack for webalizer, which considered all URLs ending in "s" to be JavaScript or CSS and therefore uncountable. Of course we haven't used webalizer for a few of years now. -- Tim Starling 11:08, 2 May 2007 (UTC)[reply]

Just to make sure: we can remove dontcountme=s from our script now, right? (And also from importScript() in MediaWiki:Common.js) — Alex Smotrov 13:27, 2 May 2007 (UTC)[reply]

Yes. -- Tim Starling 12:34, 4 May 2007 (UTC)[reply]

Is there an official guideline/recommendation for the use of class="wikitable"?

Hi, I am wondering whether class="wikitable" is to be prefered over other table layouts. I find that there are way to many lines in wikitable, but maybe there is some rationale behind this? Please see http://en.wikipedia.org/wiki/User:Jasy_jatere/Sinhala_tables for examples of what I mean Jasy jatere 13:33, 2 May 2007 (UTC)[reply]

The second is much better. Use wikitable when it looks good to use wikitable. — Omegatron 13:37, 2 May 2007 (UTC)[reply]

Wikitable is preferable in general, although there are always exceptions to the rule. Feel free to convert tables in articles to wikitable style if they're otherwise plain - they'll look better that way. Nihiltres^(t.c.s) 15:41, 2 May 2007 (UTC)[reply]

superscripts arent "super" in Firefox

They appear below the text. See the screenshot of Toronto Raptors here: http://img95.imageshack.us/my.php?image=ffyg9.png —MC Snowy (Talk / contribs) 21:43, 9 May 2007 (UTC)[reply]

That is odd, but it can't be firefox that's at fault, because the page looks fine in my browser--VectorPotential^Talk 01:27, 10 May 2007 (UTC)[reply]

Then what is it? It looks fine in IE7 and Opera 9.10. I'm using Firefox 2.0.0.3 btw. —MC Snowy (Talk / contribs) 15:33, 10 May 2007 (UTC)[reply]

Can somebody help me? —MC Snowy (Talk / contribs) 01:38, 16 May 2007 (UTC)[reply]

Nobody else has this problem??? —MC Snowy (Talk / contribs) 01:27, 21 May 2007 (UTC)[reply]

Not me. Have you tried clearing your cache? Do you have custom CSS that may interfere? -- ReyBrujo 02:34, 21 May 2007 (UTC)[reply]

Yes I have cleared it. No custom CSS. —MC Snowy (Talk / contribs) 01:09, 23 May 2007 (UTC)[reply]

[1]

This link is very worrisome to me. Perhaps someone could fill out a bug report. I won't, because I'm leaving. However, when I even hover over the contributions history of the user in popups, it really screws up my screen, let alone just going to the page makes it impossible to delete. The Evil Spartan 18:50, 12 May 2007 (UTC)[reply]

Already deleted. Also, get a better browser :-) Just teasing :-) -- ReyBrujo 18:57, 12 May 2007 (UTC)[reply]

random article + refs tag

thought here would be an ok place to ask: is there a way to make a monobook script that adds two tabs to the upper right corner of the UI that are 'yes refs' and 'no refs'? when clicking on the no, it would tag the article with {{references}}, add as such to the edit summary, save, and then go to a random article; the yes tab would simple go to a random article. click-check-yes/no-repeat. a good way to tag articles that need refs eh? a boy do a lot need them. anyone else feel like this would be a good way to assess many articles in short amounts of time in a simple way? JoeSmack ^Talk 13:01, 14 May 2007 (UTC)[reply]

It is a hard thing to see if an article is unsoured in a few seconds. Sometimes (even often) the external links are really meant to be sources. It is hard to call the article unsourced when the sources are just mislabeled. In any event there already is a huge backlog on articles needing sources. I am not sure that we need to find everything that looks unsources and tag it for someone else to deal with. Jon513 15:36, 14 May 2007 (UTC)[reply]

i think it does only take a few seconds to glance for ref tags, non-labelled ELs used as refs in the article, and checking over the EL section (one official website, etc etc) to check. i've got the power of scroll wheel! ;)

there is a huge backlog, but that doesn't mean we should be ignoring labeling work it means we should be doing backlog work more. {{references}} is absolutely for finding everything without a source and tagging it. it isn't a 'let someone else do it' job, it's a 'there, half the work, identifying which articles need attention, is done now' job.

anyone know if this kind of coding is workable? JoeSmack ^Talk 16:28, 14 May 2007 (UTC)[reply]

Here is a quick bit of rough code to do mostly what you want (should work fine in Firefox/Opera, not tested in IE).

if(wgAction == 'edit' || wgAction == 'view') addOnloadHook(refbuttons)
function refbuttons() {
  var yesref='/wiki/Special:Random';
  if(wgAction == 'edit') {
    var noref='javascript:reftag();';
  } else {
    var noref='javascript:edreftag();';
  }
  addPortletLink('p-cactions',yesref,'yes ref','ca-yesref','Go to random article');
  addPortletLink('p-cactions',noref,'no ref','ca-noref','tag with {{references}}');
}

function reftag() {
  if(document.getElementById('wpTextbox1').value.indexOf('{'+'{references}}') != -1) {
    alert('Warning: There is already a references tag present. \nNo action taken.');
    return;
  }
  document.getElementById('wpTextbox1').value = '{'+'{references}}\n' + document.getElementById('wpTextbox1').value;
  var saveme = confirm('A {'+'{references}} tag has been added to the top of this section.\nDo you want to save this page automatically?\n\nClick cancel/escape to abort, to move the tag, or to save manually.');
  if(saveme) document.getElementById('wpPreview').click()
//  if(saveme) document.getElementById('wpSave').click()
}

function edreftag() {
  var url = document.getElementById('ca-edit').firstChild.href;
  document.location.href = url + '&addreftag=true';
}
if(queryString("addreftag") == "true") addOnloadHook(reftag);

function queryString(p) {
 var re = RegExp('[&?]' + p + '=([^&]*)');
  
 var matches;
  
 if (matches = re.exec(document.location)) {
 try { 
 return decodeURI(matches[1]);
 } catch (e) {
 }
 }
  
 return null;
}

Notes:

Taking you to a random page after saving is a bit tricky so is not implemented in this version.
It asks you to confirm that you do want to automatically save.
It works from 'view' (normal top-edit viewing) and 'edit' (editing a page or section). No accesskey assigned, but you can do that (add it to the addportletlink() command as a new parameter).
Even though I wrote it, I do not condone the use of it ^_^.
Automatic save is disabled, it will only preview, unless you change the necessary code (try it out before you do though).

--Splarka (rant) 07:40, 15 May 2007 (UTC)[reply]

Please split what could be misinterpreted template transclusions in scripts, to stop the scripts showing up in categories with certain methods of installing them. I've done it above. --ais523 07:59, 15 May 2007 (U T C)

It seems like most people will eventually figure out they have to wrap most scripts in <nowiki> without having to baby them. --Splarka (rant) 12:15, 15 May 2007 (UTC)[reply]

(This sort of thing actually screwed up AfD on a couple of occasions before now; people's AfD-filing scripts kept appearing in the 'uncategorised AfDs' section. Even worse, I think it might have been my fault...) --ais523 10:30, 16 May 2007 (U T C)

Awesome! Thanks so much, I've incorporated it into my monobook as placed the script here! It works darn well for me, but I was wondering: is there any way to mark the edits as minor and put {{references}} - this article or section does not cite any references or sources. Please help improve this article by adding citations to reliable sources. in the edit summary? Again, thank you for coding this! :D JoeSmack ^Talk 18:53, 17 May 2007 (UTC)[reply]

Eureka! With some tinkering I got it to do both myself. JoeSmack ^Talk 21:24, 17 May 2007 (UTC)[reply]

Wide Diffs

Hi, don't know if this question has been asked before, but why is it that sometimes a diff will come out many times wider than my browser window, Here's an example of what I mean [2]. I'm using IE 7.0. Is there something I can do to keep these from happening? Thanks! —Elipongo (Talk|contribs) 17:40, 15 May 2007 (UTC)[reply]

The reason for the wide diff is the long URL contained near the end of the diff. There is no whitespace in the URL for the diff text to break on, which causes the display to extend to the width of the URL. I don't know if there's anything that can be done about this. —Daniel Vandersluis^(talk) 18:53, 15 May 2007 (UTC)[reply]

For a while, I used a "diff enhancer" which solved this problem. I stopped using it because it was sometimes too slow. I don't remember where I got it from, so I am unable to give credit where it's due. See [3] for the code. For most people, you need to put this in your monobook.js file, ie in en.wikipedia.org/wiki/User:YOURUSERNAME/monobook.js. I haven't used it for almost a year, so there's a possibility that some change in user interface may have broken it in the meantime.-gadfium 20:46, 15 May 2007 (UTC)[reply]

Found the original: see Wikipedia:WikiProject User scripts/Scripts/Fix diff width. It's probably been updated since the copy I used, so use this link, not the one I gave in the previous paragraph.-gadfium 20:51, 15 May 2007 (UTC)[reply]

Thanks! I'll see if that works for me. —Elipongo (Talk|contribs) 21:43, 15 May 2007 (UTC)[reply]

I also used that for a while, but nowadays I simply use the following on my user CSS: .diff-context, .diff-deletedline, .diff-addedline { overflow: auto; display: block; }. The note on my user CSS's history says it came from Wikipedia talk:WikiProject User scripts/Scripts/Fix diff width#Very simple CSS fix?. Since it uses CSS only, it's much faster (because it uses the browser's highly optimized CSS engine). --cesarb 23:39, 15 May 2007 (UTC)[reply]

I didn't know about that. I've just tried it, and it works well enough that I'll keep it, at least for a while. It's fast, as you say. The only problem for me is that pressing the "End" key to move to the bottom of the article while the diff portion is showing wrecks the screen display. Pressing PgUp and then PgDn again fixes it. There may be further problems if the article and diff was only a couple of screens long. I'm using Firefox 2.0.0.3 on Linux.-gadfium 01:48, 16 May 2007 (UTC)[reply]

This is now fixed. --brion 19:34, 16 May 2007 (UTC)[reply]

Yay! Thank you. --Iamunknown 19:43, 16 May 2007 (UTC)[reply]

I've seen this problem, but not actually thought fixing it, let alone how to. Thanks :) Gracenotes^T § 01:33, 17 May 2007 (UTC)[reply]

So, quick question, how do you disable this? I hate when Sourcesafe does that, and so I hate when Wikipedia does it :-( -- ReyBrujo 01:57, 17 May 2007 (UTC)[reply]

See here. The relevant code is:

table.diff td div {
  overflow:auto;
}

You'd probably want

table.diff td div {
  overflow: visible;
}

Gracenotes^T § 02:15, 17 May 2007 (UTC)[reply]

I've made an additional tweak this morning[4] which makes the extra-wide cells wrap attractively instead of scrolling in Internet Explorer and Safari. Unfortunately Firefox and Opera don't appear to support this style setting, so they'll continue to scroll. --brion 15:11, 17 May 2007 (UTC)[reply]

en.wikipedia article -> about:blank in own app

Hello,

I've recently had a problem with a self-developed app. In this app (developed in Dolphin Smalltalk 6.0), I can click on certain names etc and get the relevant English WP article displayed in an embedded browser window, which (as far as I know) is derived from IE in some way or other that I admit I am not quite familiar with (tbh I just dragged-and-dropped it from a menu in the development environment and, bingo, it worked). Navigation is done quite simply by setting a URL variable to i.e. "http:/en.wikipedia.org/wiki/Frederick_Sanger" and sending a navigate command.

This executable has worked for me for some time. Recently, however, immediately after displaying the article, the thing navigates on to "about:blank" which obviously makes the embedded browser window blank. The strange thing is that this only happens on English WP pages; ie entering eg "http/fr.wikipedia.org/wiki/Frederick_Sanger" (or "/de/", or "/no/", or...) into the address field of the embedded browser window perfectly displays the relevant article from the other-languate WP. All other web addresses I've tried also work fine. In IE itself, everything also works fine.

Now, I admittedly haven't yet had the time to go to the development version (which I haven't looked at for months) to dig out the details of how the browser window is connected to the rest of the app etc. Therefore, I'm not asking for deep technical help or anything -- I'm just wondering if anyone immediately can bring to mind a reason for this new behaviour?

Thanks for any help. OMHalck 20:30, 15 May 2007 (UTC)[reply]

Sounds vaguely like a recurring javascript-ish error. Does your app parse javascript and/or <script src=""> tags? --Splarka (rant) 07:18, 16 May 2007 (UTC)[reply]

My app actually doesn't do anything with the content at all -- just tells the browser window (which I didn't code) to navigate to a certain address. But yeah, some scriptish thing appears to be the trouble -- I get actual javascript error messages on some pages as well.

Anyway, I've just noted that things seem to work using the exact same .exe at my work computer, so I guess I'll just retest when I get back home tonight. If there's still a problem it's then obviously on my end. Thanks for taking the time to answer! OMHalck 08:43, 16 May 2007 (UTC)[reply]

Secure server

The secure server has been desperately slow for the past few hours, to the point of being unusable. It sticks as"waiting for.." It's getting better now - heavy load from the US? 172.206.36.63 07:53, 16 May 2007 (UTC)[reply]

Hidden text

When I type a non-existant page intop the search bar, the resulting text ("Wikipedia does not have an article with this exact name. Please search for..." etc.) is hidden. It's there, but it seems to be pale yellow text on the same colour background. I'm using classic skin in IE6. Could someone investigate this please. – Tivedshambo (talk) 08:51, 16 May 2007 (UTC)[reply]

Seems to be ok again now. – Tivedshambo (talk) 08:55, 16 May 2007 (UTC)[reply]

Capitalization

Capitalization. For some reason I cannot get Wikipedia to capitalize Allan Rechtschaffen's second name in the title of the article. I certainly capitalized the name when I made my entry. This is frustrating.Richard Dates 14:34, 16 May 2007 (UTC)[reply]

I moved the page to Allan Rechtschaffen for you. —Daniel Vandersluis^(talk) 15:00, 16 May 2007 (UTC)[reply]

Watchlist craziness

My Watchlist is suddenly showing a version of the Graham Dilley article instead of, erm, my watchlist. I've tried restarting, to no avail. I did edit said Mr Dilley's (dreadful) article earlier on, but cannot understand what's going on. --Dweller 15:39, 16 May 2007 (UTC)[reply]

Infoboxes and layout

Infoboxes are causing problems w/ layouts. Any article lead w/ an infobox of any nature lost its position as being at the top while the infoboxes got the left top side. Anybody else is experiencing this? -- FayssalF - ^{Wiki me up®} 17:06, 16 May 2007 (UTC)[reply]

No, not really. Try WP:PURGE and WP:CACHE, in that order, and check if that fixes. Titoxd^{(?!? - cool stuff)} 03:27, 18 May 2007 (UTC)[reply]

Suggestion for Search Results.

Perhaps Wikipedia could adopt an approach similar to that of Google. If a search is unsuccessful, due to spelling or wording errors, there could be a suggestion made by Wikipedia (i.e. "Did you mean...?") to guide users.

Sounds like a great idea. But, sadly, while similar ideas have been suggested before, they have generally been turned down, with the explanation that it would cause too heavy a load on the servers. :-( --Tugbug 22:32, 16 May 2007 (UTC)[reply]

Wrapping text around a thumbnail image?

Is there a bit of code to enable a text wrap around a thumbnail image? I have noticed that some pages contain large paragraph breaks due to a thumbnail image being there, and thought maybe I could clean up a bit by using a wrap option. Couldn't find that option on the cheatsheet or in any template that would apply. Thanks. Ebonyskye 19:19, 16 May 2007 (UTC)[reply]

What happened to my diffs?

Sometime in the past day or so, the font size used for displaying diffs shrank dramatically. It's now about half the size it used to be, and much harder to read. I'm using the "Classic" skin, if it matters. --Carnildo 20:44, 16 May 2007 (UTC)[reply]

There's now an extra <div> inside each cell in the diff table, but I'm not sure why that would shrink the font size. It sounds as if there was some CSS rule with font-size: 50% that incorrectly matches the new diff format, but if so, I can't find it. —Ilmari Karonen (talk) 20:53, 16 May 2007 (UTC)[reply]

The same default style rules are now used for diffs in all skins. This may be slightly smaller in the classic skin, but certainly it's not 50% smaller than before. --brion 21:08, 16 May 2007 (UTC)[reply]

So, what do I need to add to my standard.css to put things back the way they were? --Carnildo 21:09, 16 May 2007 (UTC)[reply]

Things are back to normal. The addition of diff.css explains why I couldn't find anything in standard.css or common.css. --Carnildo 22:04, 16 May 2007 (UTC)[reply]

(edit conflict x2) It may have been caused by r22204. Try clearing you browser cache and see if it still happens; if it does, we need to find the broken style sheet and fix it. —Ilmari Karonen (talk) 21:10, 16 May 2007 (UTC)[reply]

Scroling Across Diffs

Why all of a sudden, do the diffs have left to right scrollers on each part of the diff instead of just having it longer on the page? --TeckWiz is now R ^ParlateContribs_@ (Let's go Yankees!) 21:07, 16 May 2007 (UTC)[reply]

Because people don't like having 99% of the diff be completely unreadable due to page widening. --brion 21:09, 16 May 2007 (UTC)[reply]

I don't like the scrolling. I like the wide diffs. Is there a css way to override it? --TeckWiz is now R ^ParlateContribs_@ (Let's go Yankees!) 21:11, 16 May 2007 (UTC)[reply]

I'm not sure making 80% of the diff hidden was the answer here; usually Column 1 was narrow and Column 2 was page-width for me, so I just used Column 2 as my guideline for seeing what changed. I hate the scrollbars and they mess with certain wheelmice (which will try to scroll when the cursor is over the bar). -- nae'blis 21:13, 16 May 2007 (UTC)[reply]

If you wish, you may restore the old, broken behavior by customizing your user CSS. See the new diff.css for the current styles; the bits you'd want to "reverse" are the overflow: auto (change it to visible, I think) and the table-layout: fixed (change it to auto). --brion 21:22, 16 May 2007 (UTC)[reply]

Personally, I think it's a huge improvement. Hats off to the developers. Thanks for your efforts on our behalf. --Dweller 22:05, 16 May 2007 (UTC)[reply]

I think it is as well: I found it annoying when checking differences that I had to scroll the broswer. This new feature makes it a lot easier to check differences. Acalamari 23:51, 16 May 2007 (UTC)[reply]

Only the first line of each paragraph is being shown in Mozilla, with no scroll bars. (I can still sort of scroll by highlighting the text and moving the mouse up and down.) It's kind of extremely annoying. -Nogood 06:08, 17 May 2007 (UTC)[reply]

Which browser version? This particular failure mode is a new one for me. —Ilmari Karonen (talk) 06:51, 17 May 2007 (UTC)[reply]

This failure (Just the first n bytes of the diff being displayed) is currently occurring with my browser: Netscape 7.0 on Solaris 8. It makes the diff essentially useless for any paragraph with more than n bytes. It's clearly a CSS problem as the entire paragraph is there in the HTML source of the page.

Atlant 17:00, 21 May 2007 (UTC)[reply]

Time-based indexing of articles

I've begun work on forming a possible proposal for a temporal/time-based indexing effort for Wikipedia articles so that the the date when something happened could help categorize other articles of possible interest. It isn't immediately clear whether an implementation could be done with existing MediaWiki categorization or if it requires new features, so technical input would be welcome on the subject.

I've created a workspace at Wikipedia:Temporal Indexing for review. Best regards, CHAIRBOY (☎) 22:21, 16 May 2007 (UTC)[reply]

IP Spoofing vandal

A user has been vandalizing a number of articles related to Jerry Falwell today from a large number of IPs. In this edit [5], he/she claims to be able to spoof his IP address. Can a dev/checkuser look into this and see if this individual is really spoofing IPs or if they are just exploiting open proxies? --BigDT 00:06, 17 May 2007 (UTC)[reply]

This was once possible, see User:Brion VIBBER/Cool Cat incident report, however, it was fixed. It is still possible someone could have gotten around this, so it can't hurt for a dev / checkuser to look (assuming we have the full header feature on our checkuser). Prodego ^talk 00:17, 17 May 2007 (UTC)[reply]

CheckUser does now have full support for xff headers; however, these are also not entirely impossible to spoof. My guess would be that a checkuser would not reveal much, but it may not hurt. More likely is that they are open proxies -- I saw the request on WP:OP, and I'll take a look at them in a bit. AmiDaniel (talk) 01:14, 17 May 2007 (UTC)[reply]

Woops, looks like Prodego beat me to it, and found all of them open except for one. AmiDaniel (talk) 01:34, 17 May 2007 (UTC)[reply]

I'm gratified to see that. I'm amazed that the same vandal knew about all those open proxies. The one address that wasn't open was an address the vandal was boasting he spoofed. If he does that again, how does one locate the open proxy through which he's spoofing? =Axlq 14:25, 17 May 2007 (UTC)[reply]

Upload file links?

Why are there two upload file links? One is in the toolbox for "interaction" and another in "toolbox". Are we planning to change how the upload process works? If so the one in the "interaction" box looks more friendly to new users. The older one should then be deleted. -- _Hdt83 ^Chat 01:44, 17 May 2007 (UTC)[reply]

The discussion of this has been announced repeatedly at the Village Pump. Please contribute to that discussion. Uncle G 08:51, 17 May 2007 (UTC)[reply]

Subpages link in the toolbox

This was recently added on commons and has proven very useful. It is a javascript that when added to MediaWiki:Common.js adds a "Subpages" button under the toolbox which links them to the page that shows all subpages of that page. You can see an example of this here on commons (look at the sidebar on the left, at the last link in the toolbox). Yonatan ^talk 02:29, 17 May 2007 (UTC)[reply]

Lupin's anti-vandal tool broken?

Recently, this anti-vandal tool is not showing any edits of vandalism (even on level 3 Wikidefcon). Is it a bug? Can you please fix this anti-vandal tool? Thank you very much in advance!--PrestonH^{(Review Me!) • (Sign Here!)} 03:20, 17 May 2007 (UTC)[reply]

P.S. I would normally contact Lupin for this situation, but he is inactive for nearly a month.

I e-mailed Lupin. Does anyone know about changes in the recent changes RSS? -- lucasbfr ^talk 17:07, 17 May 2007 (UTC)[reply]

This is such a helpful tool. Having it down is putting a damper on my vandal bashing! Strangely, it still filters my watchlist for me. There are a few others commenting about this at User talk:Lupin/Anti-vandal tool.—Gaff ^ταλκ 22:49, 17 May 2007 (UTC)[reply]

Map images not updating despite purges and BYC

Four most recent versions of maps on Commons are not updating on Wikipedia (or Commons) as the images or thumbnails or in templates. User:VerruckteDan and I have tried purging the maps and bypassing our caches. For example this map Image:Pennsylvania Locator Map.PNG is used in state park infoboxes (like Little Pine State Park).

The DE, IL and LA maps display the correct version on their own, but have similar issues of displaying an older version when used in Geoboxes as locator maps:

DE: Image:Delaware Locator Map.PNG used in Dover, Delaware and many others
IL: Image:Illinois Locator Map.PNG used in Springfield, Illinois and many others
LA: Image:Louisiana Locator Map.PNG used in Alexandria, Louisiana and many others

Update: In each of the three cases (DE, IL, LA) the map (first link) displays the most recent version, but the Geobox City (included in the second link) displays an older version (different border colors, edge of water not a darker blue). These are calibrated maps (so entering the latitude and longitude in the Geobox automatically calculates the position of the locator dot) so could that somehow be freezing in an older version?

Thanks in advance for any help / advice with this problem, Ruhrfisch _><>°^° 03:59, 17 May 2007 (UTC) I updated the description above to offer more information. Ruhrfisch _><>°^° 14:25, 17 May 2007 (UTC)[reply]

I added the Delaware map to White Clay Creek State Park today. The current version shows when I look at the map itself, but an older version showed up in the infobox. This is not a Geobox template. Ruhrfisch _><>°^° 03:22, 24 May 2007 (UTC)[reply]

Protection status in wikicode

Is it possible to access the protection status of a page in wikicode? It would be nice if the templates in Category:Protection templates could notice when they are transcluded on unprotected pages, but I couldn't find any way to do that. --Derlay 09:25, 17 May 2007 (UTC)[reply]

You can see it in a pages logs (link on the history page) or when you try to edit. -- John Reaves (talk) 10:38, 17 May 2007 (UTC)[reply]

But the question was if there is some sort of magic word, variable, or parser function that could be used automatically in a template. Far as I see there isn't. Femto 12:28, 17 May 2007 (UTC)[reply]

This might make an interesting feature request (make requests at mediazilla:). It could make the small versions of the pp-whatever templates obsolete if used in MediaWiki space, among other things. --ais523 12:53, 17 May 2007 (U T C)

Filed as bugzilla:9947. --Derlay 23:24, 17 May 2007 (UTC)[reply]

Scroll bars on diffs?

Is there some script I can put in my monobook to turn this update off?--VectorPotential^Talk 10:30, 17 May 2007 (UTC)[reply]

The answer is above. --cesarb 10:47, 17 May 2007 (UTC)[reply]

Pretty-printing of Special:Whatlinkshere

I'm a bit frustrated with randomly ordered output of Special:Whatlinkshere. I'd like to see the following features, so that it's more useful:

Default alphabetical sorting
More important, linking through template transclusion should be indicated. When the page contains e.g. navboxes transcluded through numerous pages, it's almost trivial to fix the link (in order to fix double redirects or link over redirect) by editing the template, but one can't see it by Whatlinkshere. Take this for an example: the sheer majority of links comes through Template:Municipalities of Serbia. How difficult is for developers is to fix the page so that it looks like:

The following pages link to Irig

Template:Municipalities of Serbia
- Aleksandrovac (transclusion)
- Alibunar (transclusion)

...

etc. Duja► 11:04, 17 May 2007 (UTC)[reply]

I don't think the report is random, I think it's chronological. Corvus cornix 20:52, 17 May 2007 (UTC)[reply]

Whatever it is now, alphabetical would be much easier to search/scan. -- nae'blis 22:03, 17 May 2007 (UTC)[reply]

If the output was in a table with two columns, and it used the class="sortable wikitable" tag, then we could easily sort the content to our liking. --Arcadian 22:12, 17 May 2007 (UTC)[reply]

Compare selected versions button

I have a script which used to fix the compare button so that it became a link capable of being middle clicked in Firefox, so that it opens in a new tab. Recently it has stopped working. Is there another way of doing this, maybe a css hack? Steve block Talk 11:07, 17 May 2007 (UTC)[reply]

Works for me (on test.wikipedia and a local copy of en.wp). What browser/version are you using? Do you get any JS errors? --Splarka (rant) 07:47, 18 May 2007 (UTC)[reply]

Firefox 2.0.0.3, and no, no js errors. Steve block Talk 21:21, 22 May 2007 (UTC)[reply]

Research Design vs. Study Design

When one enters the term "Research Design", one is directed to the term "Study Design" The article discusses the design of epidemiological studies in medicine. The term "Research Design" has a much broader meaning which includes research in all the sciences, in the social sciences, and in other fields as well. It is a general term which encompasses a lot of types of research. I would like to start an article on "Research Design" but I don's want to combine it with "Study Design" because that doesn't make sense. How do I start a new article on "Research Design"?Richard Dates 18:44, 17 May 2007 (UTC)[reply]

When you enter "Research Design" and then get redirected to "Study design", there is a little line saying (Redirected from Research design) at the top, where "Research design" is a link. Click on the link, it will take you to the redirect page, which you can then edit. --Ideogram 23:32, 17 May 2007 (UTC)[reply]

Setting focus to the input box after loading a page

The focus should be set to the search input box after loading a page. Most importantly the search page.

Search sites like Yahoo, Google or MSN do it. — Preceding unsigned comment added by 217.228.171.112 (talk • contribs) 17:48, May 17, 2007 (UTC)

Yes, and it should autofocus to summaries when deleting or moving. ^High_InBC^{(Need help? Ask me)} 17:49, 17 May 2007 (UTC)[reply]

There's a good reason why it doesn't do this on the Main Page (it's so people can scroll with the arrow keys), but Special:Search is less of a problem. This might make a good feature request (see mediazilla: for feature requests). --ais523 17:51, 17 May 2007 (U T C)

I've made a feature request with bugzilla. Most search related sites set focus even for scrollable pages, if somebody really wants to scroll with up/down pressing tab to move the focus away from the input box should not be a issue.

Your claim is inconsistent with reality, as a simple test of major search sites will show. --brion 20:42, 17 May 2007 (UTC)[reply]

All three major search sites (google, msn, yahoo) set the focus, msn and yahoo are scrollable. Other info pages who set the focus and are scrollable: answers.com, dictionary.com. Which major search or info pages did your reality check yield? --Frank

Yahoo, at least, doesn't set the focus (there's no insertion point in the search box, and scrolling is possible). Somehow, though, the focus becomes set when I start typing, so presumably they're using some JavaScript/browser-dependent trick to cause the focus to act as if it were set. --ais523 14:39, 18 May 2007 (U T C)

www.yahoo.com sets focus here, at userscripts.org there is script for the greasemonkey firefox extension to set the focus to the input box for any page -frank

Search input box should be at located top/center

Almost anybody except wikipedia got this right: google, yahoo, amazon, msn, even IMDb fixed it some time ago. —The preceding unsigned comment was added by 217.228.171.112 (talk)

If you get an account, you can go to Special:Preferences, and under the "skin" options, select the skin called "Standard". That has the search input box located at the top. Or, if that's not enough for you, you can write your own skin at a myskin user subpage, once again requiring an account. (Here is the default skin for positioning, called "monobook".) Gracenotes^T § 19:00, 17 May 2007 (UTC)[reply]

Thanks for this info, it should be default however since most people don't know wikipedia is skinnable.

Yep, certainly is. People can even install their own scripts (see WP:JS) in order to enhance viewing and editing articles. It is useful to know, though. I actually prefer the search bar at the side, but that may be because I'm used to it.

I just logged out to look again. (I'm hardly ever logged out anymore, IMHO wikipedia "shure looks purdier", but is becoming practically unusable for anons :-/ ). Ayup, the default skin sucks as bad as ever ^^;;. It uses sans-serif fonts for body-text, and indeed doesn't have search boxes at top and bottom of page. Urk. Think of the anons, folks! --Kim Bruning 19:35, 17 May 2007 (UTC) I wonder if it was designed by a committee O:-)[reply]

Ignoring Sections in TOC

Is there any way to ignore a series of sections in the table of contents - I'm integrated my archive with my talk page - and so far it's not working.daniel folsom 04:55, 17 May 2007 (UTC)[reply]

I've added class toclimit-2, which only shows level 1 headers. However, that also hides current discussions. (Revert it when you take a look.) One solution might be to put all archived discussions under level 3 headers once they become archived, and then use toclimit-3 for the TOC. Gracenotes^T § 20:32, 17 May 2007 (UTC)[reply]

Bug in article

the "Western Massachusetts" article appears to have a bug that freezes the page (at least with IE6). Looks like it's halting at <table id="toc

The "edit this page" tab never appears, so you kind of get stuck there. LADave 20:18, 17 May 2007 (UTC)[reply]

Looks fine here (tested IE6, IE7, Firefox). --brion 20:35, 17 May 2007 (UTC)[reply]

Captcha on login

I gotta say.... I'm all for strengthening the login process but multi-step logins suck. If there is no one-step alternative I'll only be logging in to create articles and I'll be doing all my editing as an IP. It will suck not using a true signature - but I'm sure I can come up with some off-line signature that I can drop in to the end of a message easily enough, like this: I'm not signed in right now but this is me Ga rr ie

Ga rr ie 21:47, 17 May 2007 (UTC)[reply]

Mutlistep logins? I've noticed no change in the login process recently. --Deskana (AFK 47) 21:49, 17 May 2007 (UTC)[reply]

There must be something weird with cookies. Garrie, you may want to try clearing your cookies. I'm not sure the exact steps to reproduce it, but I have noticed that when I have changed my password and then go to another machine where I was logged in, I am first prompted for my password as normal, but then, after I enter it, I get prompted to enter it again with a captcha. --BigDT 22:00, 17 May 2007 (UTC)[reply]

I guess I was spooked by all the increased talk of strong passwords... at pretty much the same time as my post I had recovered from a phishing attack which partially dumped a rootkit on my PC. I think BigDT was spot on with the cookies because I lost a couple of cookies for a little while. Everything is "normal" now.Ga rr ie 22:57, 22 May 2007 (UTC)[reply]

MAJOR IP issues

Apologies for the Caps, but there's some really odd stuff going on. There's been a flurry of edits from IP's like 66.230.200.145 (this is currently the most consolidated discussion of this event), which is registered to Wikimedia. Autoblocks have been affected a wide range of users, coming from as far apart as Illinois and Montana. Someone connected with the Wikimedia servers needs to investigate. --Ybbor^Talk 22:13, 17 May 2007 (UTC)[reply]

Wikipedia:Administrators'_noticeboard#Anomalous_auto-block_message seems to indicate the problem has cleared up. --Ybbor^Talk 22:20, 17 May 2007 (UTC)[reply]

Beat me to it :) Martinp23 22:21, 17 May 2007 (UTC)[reply]

This discussion is really scattered. Also some at Wikipedia:Administrators'_noticeboard/Incidents#Category:Requests_for_unblock. --Ybbor^Talk 22:24, 17 May 2007 (UTC)[reply]

I am using Wikimedia as an open proxy? Yay! Either that or I just got really lucky and found a non-hardblocked TOR exit node. Apologies for being happy, I know this is hard on others. Thanks, Armed Blowfish (mail) 01:35, 18 May 2007 (UTC)[reply]

This may or may not be related to Wikipedia_talk:Administrator_intervention_against_vandalism#Bug_ID_9213. The bug report there also mentions a problem with the Squid caches that results in anons not receiving messages on IP talk pages. -- _Hdt83 ^Chat 01:41, 18 May 2007 (UTC)[reply]

Maybe a WP:BEANS but this sounds like a Wikimedia squid that was temporarily unlisted on the trusted XFF servers (all the squids are technically open proxies, but are always trusted not to forge XFF headers, so are always transparent).

No WP:BEANS there. That is essentially what happened. Nothing was unlisted, just new machines were not listed. Titoxd^{(?!? - cool stuff)} 19:41, 18 May 2007 (UTC)[reply]

Wikipedia Logo

segment copied from end of Wikipedia:Village pump (proposals)#Wikipedia Logo.

There are at least 2 major character errors in the current Wikipedia puzzle-globe logo, and other minor problems.
User talk:Ambuj.Saxena/Wikipedia-logo is the most centralized discussion/link compilation that I know of. (There's even a petition at that link's projectpage)
Nohat has explained the problems with correcting the errors. But noone seems to have a solution.
Somebody with patience and brains (and either delegating or computer-graphics skills), needs to adopt this problem as a personal mission, lest it remain unsolved for another year. --Quiddity 02:53, 18 May 2007 (UTC)[reply]

This is a trademarked image. You have to consider those implications as well. You may want to ask whether there would be any issues replacing the old logo with a corrected one. Titoxd^{(?!? - cool stuff)} 03:16, 18 May 2007 (UTC)[reply]

Infobox not displaying

Someone with time to check Seemless? The infobox is messed up, and can't see what is exactly not working (plus I am a little bit busy right now). -- ReyBrujo 03:46, 18 May 2007 (UTC)[reply]

Done. Cheers. --MZMcBride 03:54, 18 May 2007 (UTC)[reply]

Internal Server Error

I get the following message when I try to move more than two pages in a row: "Internal Server Error: Sorry, the server has encountered an internal error. Please wait a moment and hit "refresh" to submit the request again." I have to wait few minutes in order to move the next page. It's been like this for few days. Does anybody know what's wrong? Jogers (talk) 12:41, 18 May 2007 (UTC)[reply]

A move rate limit of 2 moves per minute for non-sysop or non-bot users was introduced by Tim Starling on May 15th "in response to vandalism on Wikiquote". 86.140.128.210 22:32, 18 May 2007 (UTC)[reply]

It makes my work much harder. Is it going to be permanent? Jogers (talk) 05:45, 19 May 2007 (UTC)[reply]

http://en.wikipedia.org/w/COPYING

http://en.wikipedia.org/w/COPYING (linked from Special:Version) has media type application/octet-stream instead of text/plain, and this makes it a bit hard to read with many browsers. --Derlay 22:24, 18 May 2007 (UTC)[reply]

URGENT: Twinkle bug

There's a bug crept up in twinkle. After each revert, the first thing that appears except the restored version is "undefined". No clue why that is. Look here, here, and as part of the nonsense text, and here after I separated undefined from the rest of nonsense.

This is pretty urgent, because you have to remove this by hand. I can't write a little bot that automatically removes "undefined" from every twinkle revert, nor fix the bug. This may turn into more than a little mess. Evilclown93 00:59, 19 May 2007 (UTC)[reply]

An article in the namespace with the same bug.

Please contact the creator of the script. The script is not a part of the MediaWiki coding and as such, you are unlikely to receieve a response to your problem from this page. --Deskana (AFK 47) 01:04, 19 May 2007 (UTC)[reply]

I know that. The creator is offline, and I doubt he'll be online until later (around 12-14 UTC, maybe), and I'm just drawing attention to the fact the Twinkle is a bit bugged. I'm almost done with the email and going to post a message to Azatoth's talk, no worries. Evilclown93 01:08, 19 May 2007 (UTC)[reply]

We can shut down the script if needed, by blanking it. Do you think that is necessary? Prodego ^talk 01:23, 19 May 2007 (UTC)[reply]

Hmm, maybe I'll review the code. Voice-of-All 01:34, 19 May 2007 (UTC)[reply]

I have an article called Dragons escape the game and everybody keeps posting it for that stupid 'speedy deletion' and that is innapropriete to delete aticles on wikipedia because they will be blocked and eventualy be a vandal to delete and mark articles for speedy deletion.any posts of comments must be sent to --Tub city adventures 03:16, 19 May 2007 (UTC)'s talk page.[reply]

I believe that the problem is that

 var form = this.responseXML.getElementById( 'editform' );
 var text;

should be

 var form = this.responseXML.getElementById( 'editform' );
 var text = '';

"text" was concatenated with the article text, although it was not initialized to anything, so it was represented as a string with the value of "undefined" (and not "null", since it's a primitive data type). Gracenotes^T § 04:34, 19 May 2007 (UTC)[reply]

Workaround for watchlist limit?

I'm having trouble with my watchlist. Even when I set it to show 7 days of edits, I'm getting considerably less than a day. I'm monitoring and active in editing a couple of hundred pages, but the artificially set 250-edit cap is making it so that I can't even see edits from before I went to bed last, some 12 hours ago. Is there a workaround for the 250-edit cap, other than just trimming down my watchlist? I can't see a single page that lit up today that I feel comfortable removing. MrZaius^talk 05:50, 19 May 2007 (UTC)[reply]

Go to my preferences (upper right) → watchlist and change the value Number of edits to show in expanded watchlist: to some number greater than 250. -- MarcoTolo 02:32, 20 May 2007 (UTC)[reply]

undo/revert

Just wondering how you can revert/undo edits; such as if someone vandelises a page, to undo the vandelism edit to take it back to how it was. Either leave a comment here or on my talk page please. Thanks :) J S Firefox 11:13, 19 May 2007 (UTC)[reply]

I have answered on their talk page- feel free to expand on it. GDonato (talk) 14:05, 19 May 2007 (UTC)[reply]

New patent template - suggestions wanted

I just set up a very simple template "patent" that takes a patent number and publication code (US/EP/WO) to create a reference string. An example of its use is at SEAgel. I was wondering what people think of this idea - am I reinventing the wheel, or can something be changed or improved? For instance the interaction with ref or cite might be improved. Also, I currently have the nationality of patent after the number so that if it's omitted it defaults to "US", but maybe that's silly - it might be more standard to have the prefix before and not bother with having a default. Mike Serfas 15:10, 19 May 2007 (UTC)[reply]

Operation_Sea_Scan

Why doesn't this redirect work? Nardman1 17:12, 19 May 2007 (UTC)[reply]

It should be working now as I took out the percent-encoding, which can break wikilinks sometimes. - MTC 17:21, 19 May 2007 (UTC)[reply]

problem with IE

The login pages for orkut and gmail are not getting opened.Please help me out to solve the problem.Is there really any problem with IE or that to OS or to the internet connection?—The preceding unsigned comment was added by Jayamailbox (talk • contribs) 03:45, May 20, 2007 (UTC)

Sorry, but this board is for issues related to Wikipedia itself. You need to look elsewhere if you're having problems with other websites.
That said, I'd recommend you switch to Firefox. So much better than IE in every conceivable way. EVula // talk // ☯ // 05:06, 20 May 2007 (UTC)[reply]

Issue with User:Lupin/Filter recent changes

User:Lupin/Filter recent changes does not appear to be working. At least one user has confirmed this over there. Any ideas? Navou 00:27, 20 May 2007 (UTC)[reply]

I have fixed the recent2.js script and filter recent changes works again. Cacycle 06:21, 20 May 2007 (UTC)[reply]

Displaying only part of an image

I remember seeing a page (can't find it now), where an image was displayed "cropped", i.e. the total height of the actual image was about 1200px, but of those, only a segment of 100-200px was displayed. What markup achieves that, b/c I couldn't find anything on Wikipedia:Picture tutorial. Thanks in advance, —Alde Baer 13:28, 20 May 2007 (UTC)[reply]

You pretty much have to "cheat". There are two basic options AFAIK, you can do a CSS hack, putting the image into a fixed sized div element with the "overflow:hidden" style set (not ideal the full image is still loaded, and won't work in all browsers). Alternatively you can just upload a cropped version of the image and use it as the thumbnail for the full version (so that clicking gives the full version) using the thumb=image syntax: [[Image:Example.jpg|thumb=Example_cropped.jpg|some camption]]. --Sherool (talk) 13:59, 20 May 2007 (UTC)[reply]

I did not know about the thumb= command. Aldebear, I used to have a template that did that, but all it did was display the entire image but only allow part of it to be visible. I deleted the template when I realized that trick is really hard on peoples browsers. ^(H) 14:03, 20 May 2007 (UTC)[reply]

Thanks a lot to both of you. Going to play around for a bit, see what works best. —Alde Baer 15:09, 20 May 2007 (UTC)[reply]

templates with span style="display:none"

I'm getting some curious behaviour with {{Mathematician data}}. Its used to record data about mathematicians in {{Talk:Blaise Pascal/Data}} which are themselves templates, and allow extraction of the data. For instance

{{Talk:Blaise Pascal/Data|key=dates}} gives 1623 – 1662
{{Talk:Blaise Pascal/Data|key=sortname}} gives Pascal, Blaise

And with no parameter a longer description is produced

{{Talk:Blaise Pascal/Data}} gives This subpage contains data about Village pump (technical) which is used by the Wikipedia:WikiProject Mathematics to produce tables of mathematicians. Edit this page to add or update the sortable name (sortname), date of birth or death (dates) and field (contribs) of Village pump (technical).

Name: Pascal, Blaise
Dates: 1623 – 1662
Contribs:

{{Talk:Blaise Pascal/Data}} looks like

{{Mathematician data|{{{key}}}
|dates= 1623 – 1662
|sortname= Pascal, Blaise
|contribs= 
}}

Everything works fine, until we have tried to work with BCE dates. To get these to sort properly in Wikipedia:WikiProject Mathematics/Wikipedia 1.0/Mathematicians, we have include a working sort key inside span tags. For instance

{{Mathematician data|{{{key}}}
|dates= <span style="display:none">*615</span>384 – 322 BCE
|sortname= Aristotle
|contribs= 
}}

Extracting the date field works fine

{{Talk:Aristotle/Data|key=dates}} gives 384 – 322 BCE

but no output is produced in the no-parameter case.

{{Talk:Aristotle/Data}} gives This subpage contains data about Village pump (technical) which is used by the Wikipedia:WikiProject Mathematics to produce tables of mathematicians. Edit this page to add or update the sortable name (sortname), date of birth or death (dates) and field (contribs) of Village pump (technical).

Name: Aristotle
Dates: 384 – 322 BCE
Contribs:

I've been scratching my head trying to workout why the span tag causes problems. Any ideas? --Salix alba (talk) 13:55, 20 May 2007 (UTC)[reply]

Well, anything with its display set to "none" will never show up in any circumstance; saying as the span tag is properly closed, though, that shouldn't be an issue (I tried adding the closing semicolon at the end to Artistotle, but that didn't do anything). Are the quotes (in the span tag) throwing it off, maybe? EVula // talk // ☯ // 15:19, 20 May 2007 (UTC)[reply]

Not showing the *615 is the intended action. What we would like is for blah...384 – 322 BCE...blah to appear, but for the full text to be there for sorting purposes. Whats curious is that no text appears even in the html source. --Salix alba (talk) 15:38, 20 May 2007 (UTC)[reply]

(edit conflict) What I've noticed is that the text won't show up at all, even when looking at the page source. I think it's the equals sign that's causing the problem, since removing it causes it to display correctly and adding it in other locations to other parameters causes a blank output. This would probably therefore indicate a problem with the template parameters or perhaps the #switch: ParserFunction, rather than the formatting. Tra (Talk) 15:40, 20 May 2007 (UTC)[reply]

I did experiment to see if it was the equals sign. Changing the span to <span foo="bar"> while still having a = sign actually displayed correctly. --Salix alba (talk) 19:45, 20 May 2007 (UTC)[reply]

I'm guessing that might be because MediaWiki doesn't recognise foo as a valid HTML tag parameter and consequently strips it out before evaluating the wiki-code. If you change it to <span style="display:inline"> which is valid, then it doesn't work. Tra (Talk) 20:34, 20 May 2007 (UTC)[reply]

Urgh, not HiddenStructure again. display: none does show up in several circumstances; text browsers, plaintext rendering, search engines, and anything else which ignores either CSS or its display property. It would be better, if possible, to use a separate field as the sort key, defaulting to the date if it's not available. --cesarb 18:32, 20 May 2007 (UTC)[reply]

The problem, for sortable tables, is that there is not currently other way to customize the sorting of certain cells. For example it is common to see things like:

{| class="wikitable sortable"
|-
| <span style="display:none;">Bush, George W.</span> [[George W. Bush]] || 2001-present

If we can get a built-in way of doing this, it would be easy enough to convert all the display:none's at the same time. Until then it's better than nothing. — CharlotteWebb 19:32, 20 May 2007 (UTC)[reply]

Yes I'm coming to that conclusion about having a sortdate key. Hidden structures seems to be the only way to get the table to sort right and display sensible dates, indeed they are discussed extensivly in Help:Sorting. None of the files are in main space, so the associated problems are minimised. --Salix alba (talk) 19:45, 20 May 2007 (UTC)[reply]

I've now implemented a more elegant workaround. An extra sortdates key is used for pre 1000AD dates. The table code in wraps this up in the hidden span, sort sorting works and dates display correctly. Thanks for everyones help. --Salix alba (talk) 20:32, 20 May 2007 (UTC)[reply]

Oldest in logs

Can the devs make "oldest" and "newest" buttons in user logs that go to the oldest logs of a user and the newest ones, the same way it does for contributions? --R ^ParlateContribs_@ (Let's go Yankees!) 18:03, 20 May 2007 (UTC)[reply]

Maybe if we move logs over to use the Pager class, but there is no unique key to sort on. Voice-of-All 00:46, 21 May 2007 (UTC)[reply]

Prose size script

I installed Dr pda's prose size script a while back and it worked fine for a while. However, recently it stopped working on certain pages such as Lawrence Taylor and History of the New York Giants among others. I asked Dr pda and he said it works fine for him on those pages. The bottom left of my computer screen says done but with errors on page, with a yellow triangle with a black exclamation point in the middle. The page size button doesn't appear at all. Can someone help me with this? Quadzilla99 18:07, 20 May 2007 (UTC)[reply]

In case anyone can help, the problem has been narrowed down (Dr pda's script is now broken when section headings begin with numbers, due to some change at the meta level). See discussion here. SandyGeorgia (Talk) 22:48, 20 May 2007 (UTC)[reply]

One issue I can see is that the output of the search page now says "KiB" instead of "kB". This was introduced by Cyde in two changes on May 13 and 14. I don't have Internet Explorer to reproduce the actual issue, but this change will break the getWikiText() function as it currently stands. Mike Dillon 01:13, 21 May 2007 (UTC)[reply]

I think I may have committed a fix; reload your JavaScript and see if it's working now? --Cyde Weys 12:26, 21 May 2007 (UTC)[reply]

Cyde, sorry, but you're talking to a techno-dummie here. I did ctrl-F5 on my monobook — is that what you mean? Or does Dr pda have to do something to his script? No, not fixed yet. SandyGeorgia (Talk) 12:29, 21 May 2007 (UTC)[reply]

Ah well, had my changes in the MediaWiki namespace been what caused this bug, that would have fixed it. --Cyde Weys 16:19, 21 May 2007 (UTC)[reply]

I didn't mean to imply that fixing the "KiB" thing would fix the problem, I just noticed that it was broken.

I just attempted to recreate the problem in Internet Explorer and I got a JavaScript error about "link.href" not being defined, but the document statistics were still shown. I was using Internet Explorer 6 and I was not logged in, so I put:

javascript:importScript("User:Dr pda/prosesize.js");getDocumentSize()

into the URL bar to test it. The error about link.href was supposedly on line 99, column 5 of something, but I couldn't find it. There is a reference to link.href in wikibits.js, but It wasn't on line 99... Can someone exactly describe what happens when this "breaks" and which specific version of Internet Explorer is affected? Mike Dillon 00:11, 22 May 2007 (UTC)[reply]

I have the problem on both IE6 and IE7, and the "break" is simply that the Prose size link is gone from my toolbox for any article that has a section heading beginning with a number (but not for article titles beginning with numbers). Dr pda's ArticleHistory script is also gone — they are both there for all other articles. He says he made no changes, and they were working before, and they continue to work on other articles. SandyGeorgia (Talk) 00:22, 22 May 2007 (UTC)[reply]

Would you mind getting the JavaScript error message and posting it here? You can find it by double-clicking the error icon in the status bar and clicking "Show details" (not sure about the exact wording since I don't have IE handy at the moment). Mike Dillon 03:47, 22 May 2007 (UTC)[reply]

I suspect that this issue is related to the recent addition of MediaWiki:Wikimediaplayer.js to MediaWiki:Common.js (added on May 13 by Cyde). When I load History of the New York Giants in Internet Explorer, I get the aforementioned error about "link.href" on line 99. If you look at line 99 of MediaWiki:Wikimediaplayer.js, the code is attempting to do something with "link.href.substr". I'm not exactly sure why this would only happen in articles with numeric headings, but that does indeed seem to be the case.

...Time passes

I'm pretty sure I just figured out the problem with MediaWiki:Wikimediaplayer.js. The for loop that iterates over the HTMLCollection of "A" tags ends up having fairly weird values for the keys. Unlike the behavior in Mozilla, where the keys are sequential numbers, Internet Explorer has "length" as the first key. In the case of "A" tags, the keys used in a for each loop start with the sequential numbers and then are followed by text values related to the link href or anchor value. I think that code can be fixed by changing the loop from this:

for( var key in links ) {
    link = links[key];

To this:

for (var key = 0; key < links.length; key++) {
    link = links[key];

The reason that it breaks when there are numeric headings is that Explorer is attempting to use any string that starts with a number as a numeric index. For example, "1986: Super Bowl Champions" is treated as the number 1986. If there is no link at that array index, the "link" variable is undefined and thus calls to "link.href" generate an error. I don't know if this is causing the problem with Dr pda's scripts, but I could see an error in this script preventing subsequent JavaScript from executing. Mike Dillon 15:43, 22 May 2007 (UTC)[reply]

Mike, I'm not sure if you still need input from me, or if I should show this to Dr pda to see if what you describe fixes it? I'm sorry; I don't really speak this language. Pr pda is busy in real life for about two weeks, but if you think that's it, should I ask him to peek in here? I'm not getting an error anywhere: I just get ... nothing. SandyGeorgia (Talk) 16:37, 22 May 2007 (UTC)[reply]

Sorry for getting overly technical. The fix needs to be made in MediaWiki:Wikimediaplayer.js by an admin. I don't know if it will fix the issue with Dr pda's scripts, but it will fix the "link.href" error I was seeing in IE6 on the History of the New York Giants page. One step at a time.

P.S. Would you mind getting the error message that I asked for earlier? Mike Dillon 17:00, 22 May 2007 (UTC)[reply]

I don't have an error message; I don't know where/what to get. I just have a missing link in my toolbox. But Quadzilla99 (talk · contribs) does get an error message, I think; I'll ask him to look in here. Thanks so much for all your help. SandyGeorgia (Talk) 17:29, 22 May 2007 (UTC)[reply]

If you're getting the yellow icon with the exclamation point in the status bar, double click it to see the error message. When the dialog comes up, click "Show Details". If there is more than one error message, hit the "Next" button to see additional messages. Mike Dillon 17:47, 22 May 2007 (UTC)[reply]

(undent)Hi Mike, the message says (when I double click on the icon and hit show details):

Line: 99
Char: 5
Error: 'link.href' is null or not an object
Code: 0
URL: http://en.wikipedia.org/wiki/History_of_the_New_York_Giants

Hope this helps. Quadzilla99 18:04, 22 May 2007 (UTC)[reply]

If that's the only message, that's the same one I got. I believe my fix above will correct that one. Hopefully that's actually the source of the problems with the prose size script. Mike Dillon 18:30, 22 May 2007 (UTC)[reply]

So I should cut and paste this into my monobook? Or does the change need to be made by Dr pda? Quadzilla99 18:56, 22 May 2007 (UTC)[reply]

It needs to be fixed in MediaWiki:Wikimediaplayer.js by an admin. Mike Dillon 20:19, 22 May 2007 (UTC)[reply]

I looked into this briefly, and noticed that in the source html, most anchor tags have ids, except those with names starting with numbers. Not sure what the id is used for or if this is related, but it seemed a curious coincidence. Gimmetrow 19:34, 22 May 2007 (UTC)[reply]

The ids are added by HTML Tidy; they are not generated by the MediaWiki source. The reason that the ones starting with numbers don't have ids is that XML ids can only start with a letter or underscore:

Definition: A Name is a token beginning with a letter or one of a few punctuation characters, and continuing with letters, digits, hyphens, underscores, colons, or full stops, together known as name characters.] Names beginning with the string "xml", or with any string which would match (('X'|'x') ('M'|'m') ('L'|'l')), are reserved for standardization in this or future versions of this specification.

See http://www.w3.org/TR/REC-xml/#sec-common-syn. Mike Dillon 22:40, 22 May 2007 (UTC)[reply]

So what's the status on this?Quadzilla99 19:11, 23 May 2007 (UTC)[reply]

It should be fixed. Bypass your cache to see if things are cool now. Mike Dillon 20:53, 23 May 2007 (UTC)[reply]

Onlinesource template

Can anyone add the small=yes parameter to {{Onlinesource}}? There has been a request at the talk page for a long time, but no one has done it. SandyGeorgia (Talk) 22:47, 20 May 2007 (UTC)[reply]

It looks like it does. However until just now {{Onlinesource2005}} and the other year specific templates did not. --Salix alba (talk) 23:56, 20 May 2007 (UTC)[reply]

Thanks! SandyGeorgia (Talk) 12:14, 21 May 2007 (UTC)[reply]

Cascading Protection

Can a page be semi-protected with cascading protection? Also, how many levels does it go cascade? --R ^ParlateContribs_@ (Let's go Yankees!) 00:03, 21 May 2007 (UTC)[reply]

Semi-protection does not work with cascading. After all, if it did, an autoconfirmed non-admin could semi-protect any article he/she wanted by transcluding it on a semi-cascade-protected page. Cascading goes as many levels as there is transclusion, although I'd assume there is a numerical limit; I'd imagine pre-expand size kicks in at some point as well. Gracenotes^T § 00:15, 21 May 2007 (UTC)[reply]

I thought they changed it so that cascading sprotected pages were sprotected? --Deskana (AFK 47) 14:44, 22 May 2007 (UTC)[reply]

Obviously I was mistaken. God knows where I came up with that idea. --Deskana (AFK 47) 14:46, 22 May 2007 (UTC)[reply]

Well, that is sort of true. I disabled cascading protection for non-full, but changed the way it works so that *if* it was enabled, it would actually cascade semi-protection. Voice-of-All 15:43, 22 May 2007 (UTC)[reply]

Weird bits of text in various places

In the last couple days, I've been getting weird chunks of text in random parts of pages. I'd say it's happened maybe 10 times. Looks as though some database field is getting randomly included in an article, or parts of HTML tags are getting stripped out. Refreshing the page usually makes it go away. For example, here are a few lines from my watchlist; look at the second line:

    20:07 Hanford Site (diff; hist) . .  (+500) . . 70.20.190.160 (Talk) (fire balloon attack)
    20:05 torial_elections%2C_2010" title="United States gubernatorial elections, 2010">United States gubernatorial elections, 2010 (diff; hist) . .  (+7) . . 67.162.154.78 (Talk)
m   19:55 Pierce v. Society of Sisters (diff; hist) . .  (-4) . . Davidwr (Talk | contribs) (→Background - wikilink fixup of Supreme Court of the United States)

-Pete 06:30, 21 May 2007 (UTC)[reply]

That looks like part of the start of an HTML tag is missing; you're seeing the second half of an <A> tag there, for instance (it's exactly where the A tag should be, it's just that the first half of the tag is missing). I'm not sure what's causing it. --ais523 09:53, 21 May 2007 (U T C)

Yes, I think your analysis is correct, but when I refresh the page, the problem goes away. It's only been happening in the last couple days.

Not sure if this is related, but I've also been having random chunks of text appear in my edits, also in the last few days. It's the kind of thing I might assume results from a stray "paste" keyboard shortcut on my part, except that it's happened several times, in areas of text I haven't even been editing. For instance, the following weird text showed up after this edit; I hadn't even cliicked in that paragraph. (I bolded the extra text below to draw attention to it.)

The most famous and prominent example of TABOR legislation is in the state of Colorado.[1] In 1992, the voters of the state amended Article X of the Colorado Constitution to the effect that any tax increase resulting in the increase of governmental revenues at a rate faster than the combined rate of population increase and inflation as measured by the either the cost of living index at the state level, or growth in property values at the local level, would be subjected to a popular vote in a referendum, a process referred to as &quo<link rel="stylesheet" type="text/css" href="http://en.wikipedia.org/w/index.php?title=User:Lupin/navpop.css&action=raw&ctype=text/css&dontcountme=s">t;de-Brucing" after Douglas Bruce, the author of the amendment. This applies to any cities and counties in Colorado as well as the state itself. Additionally, any "natural growth" in revenues that exceeded this rate was to be either earmarked for educational improvements or rebated to the taxpayers once an adequate reserve ("rainy day") fund was established. This has led to a decrease in actual tax revenue (relative to population and inflation) for two reasons. Because the law does not adjust for rising productivity, additional income from year to year among the same population can not be effectively taxed. Secondly, the law only looks at the previous year, leading to a "ratchet-effect", wherein if tax revenue temporarily lowers in a recession, revenue can not rise back to pre-recession levels without a referendum. In Colorado, these factors have led to a decreasing overall tax revenue in the state.

-Pete 19:44, 21 May 2007 (UTC)[reply]

It looks very related; that's a chunk of text from the page you received being included in the information you sent (that's the styling information for navigation popups, as it happens, but given what you've said upthread it's probably just a bit of text plucked out from the page's HTML at random). One possibility might be network problems somewhere between you and Wikipedia, but I'm not sure what would case the symptoms you've seen exactly. --ais523 12:19, 22 May 2007 (U T C)

Thanks again for your thoughts. Network weirdness sounds plausible, except that there are so many error checks built into TCP transmission that it seems unlikely. Also, I haven't noticed this posting to sites other than Wikipedia. In case it's relevant, I have POPUPS installed...I've been thinking of ditching it though, and trying a different tool. Twinkle seems popular, I may give that a try. -Pete 20:03, 22 May 2007 (UTC)[reply]

One more weirdness - this time an explicit error, upon trying to save an edit to the John E. Frohnmayer article:

A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was:
(SQL query hidden)
from within function "SiteStatsUpdate::doUpdate". MySQL returned error "1205: Lock wait timeout exceeded; Try restarting transaction (10.0.0.237)".

-Pete 22:19, 22 May 2007 (UTC)[reply]

afd and mfd closings

What is the category or page listing the templates to close an afd, an mfd, etc? I want to try my hand at some non-admin closures of these (which are allowed in unambiguous keep cases). Thanks. -N 11:00, 21 May 2007 (UTC)[reply]

Wikipedia:Deletion process gives instructions for closing XfDs. --ais523 11:01, 21 May 2007 (U T C)

Hebrew Link bug?

Screenshot: [6]

Any ideas what might be causing this? Cheers, Tompsci 13:06, 21 May 2007 (UTC)[reply]

Notes in Firefox overlap

More and more pages have notes in two columns, and often the notes of the columns overlap to the extent that they become unreadable and even unclickable. In Opera there still seems to be the one column version of notes. I checked logged off and with a new Firefox profile, but the problem is still there.[7]

What can I do to make this go away? Do a need a special stylesheet or something? Or is there actually a problem with that two colomn layout that seems to be more and more common.--Kornelis 14:01, 21 May 2007 (UTC)[reply]

I changed that page to single-column footnotes, since some of those URLs are clearly too long for the two-column form. I'm not sure what should be done to fix the general problem, though. --Derlay 14:21, 21 May 2007 (UTC)[reply]

Oof, those URLs do look ugly. The best solution, in my opinion, is to change the encyclopedia, rather than change your browser. {{cite web}} is useful. Gracenotes^T § 18:52, 21 May 2007 (UTC)[reply]

Exactly. Replace the long plain url's with the proper external link formatting, which will result in entries that can linewrap properly. Either {{cite web}} or just external links format. --Quiddity 19:10, 21 May 2007 (UTC)[reply]

Thanks. Since Opera displayed differently I suspected a browser related issue (being on Linux I could not check msie, IE NetRenderer[8] works only at the top of a page). Also I wasn't aware that the ref mechanism had an optional two column option (I now see it is CSS3). I find it difficult to keep track of all the options, so when other people use them I won't touch them even when the output is wrong (reading up on cite I see that that's an acknowledged criticism). To understand the modification made by Derlay I not only had to update my knowledge of cite.php, I also had to find a page on mozilla's work on css3 two columns and read the relevant page at www.w3.org. Not that I mind that (interesting read) but I didn't expect it would lead me there. :)

Thanks again. --Kornelis 08:51, 23 May 2007 (UTC)[reply]

An idea to fix the backlog at Category:Disputed fair use images

This category is tough to clean out because any admin who goes through it has to check manually on each image to make sure it's been there for 7 days. It also contained, at my last pass through the beginning of the alphabet, disputes which were over 6 months old. I think it'd be a good idea to date the associated template, {{non-free use disputed}}, and have a bot create dated categories for it. Problem is I don't have the technical expertise as far as template coding or bot-making to actually implement this myself. Is anyone out there who does, and wants to help? (ESkog)^(Talk) 18:46, 21 May 2007 (UTC)[reply]

MediaWiki:Lag-warn-normal

Why is MediaWiki:Lag-warn-normal showing up on the watchlist and contrib list? Is it really necessary to know that the list is 1 or 2 seconds slow? --- RockMFR 20:37, 21 May 2007 (UTC)[reply]

This was a result of mediazilla:9628. For lag of less than 30 seconds, MediaWiki:Lag-warn-normal is displayed. MediaWiki:Lag-warn-high is displayed for significant lag. I'd imagine some users might want the following in their user css:

div.mw-lag-warn-normal { display: none; }

Gracenotes^T § 21:52, 21 May 2007 (UTC)[reply]

In response to your question, though, this feature is a bit too useful to disable. (Check the archives of VPT for many, many threads on watchlist lag.) If you don't like it, don't display it! Gracenotes^T § 21:56, 21 May 2007 (UTC)[reply]

informing the user of a one-second lag is clearly a bit ridiculous. it'll only show up for >15 seconds now. (that number can be changed if needed). kate.

The dark background color is really irritating. What happened to the light color? Night Gyr (talk/Oy) 02:55, 22 May 2007 (UTC)[reply]

I think even the light color was emphasising it too much. Can't it be toned down? Petros471 19:46, 22 May 2007 (UTC)[reply]

This should be at the bottom of the page or somewhere not in the way... lag notices are a very minor thing and it's not something that needs to be in your face. --W.marsh 19:49, 22 May 2007 (UTC)[reply]

What would be nice is if we could change the formatting in the MediaWiki page. Right now, all it has is the text, and the box with its colors looks like it might be hardcoded somewhere? —Centrx→talk • 21:50, 22 May 2007 (UTC)[reply]

Yep, in main.css... could be overwritten, of course. Gracenotes^T § 23:47, 22 May 2007 (UTC)[reply]

Request for Information.

Hello,

Is there a way to scan an article for keywords? I was looking into the idea of possibly auto-formatting an article after scanning for desired keywords. --Aarktica 22:11, 21 May 2007 (UTC)[reply]

It's possible in Mozilla (scans the keywords in the article's body within the edit mode), but not in IE. --Brand спойт 22:52, 21 May 2007 (UTC)[reply]

How about within WP? For example, is there a way to scan multiple articles for the word "tundra" and format it with italics? --Aarktica 23:10, 21 May 2007 (UTC)[reply]

You could potentially find a bot that performs a similar function, or maybe any bot and ask the operator for advice, or get them to incorporate the function into the bot. -- Tompsci 23:43, 21 May 2007 (UTC)[reply]

I think theres a feature in Google to search for keywords within an article. -- _Hdt83 ^Chat 03:25, 22 May 2007 (UTC)[reply]

I think wikipedia:AutoWikiBrowser has this built in. Night Gyr (talk/Oy) 03:28, 22 May 2007 (UTC)[reply]

Thanks for the feedback; hopefully one of these options will allow for mass-formatting of multiple articles. --Aarktica 23:39, 22 May 2007 (UTC)[reply]

Animated GIF not displaying

Hi,

I uploaded an animated GIF to Uncyclopedia here, but thumbnails of the image do not display. In the place where a thumbnail would be, it says, 'Error creating thumbnail: /usr/wikia/source/wiki16svn/bin/ulimit-tvf.sh: line 4: 23995 Segmentation fault "$@".' You can see the message here. Anyone know what could be causing the problem?

Thanks,

Squaw 93 00:45, 22 May 2007 (UTC)[reply]

I'm only ¾ sure, but I think I heard that MediaWiki can't scale animated GIFs. Nihiltres^(t.c.s) 03:31, 22 May 2007 (UTC)[reply]

It can, it just fails on ones that have been optimized by dropping unchanged portions of the frame, because it doesn't scale the droppings correctly. Uncyclopedia is entirely separate from wikipedia, anyway, so this isn't the place to ask. Night Gyr (talk/Oy) 03:33, 22 May 2007 (UTC)[reply]

Downloading a font to view one of its characters

Even though this is more of a matter of policy (etended from Wikipedia:Village pump (policy)#Is it appropriate to make people download a font to see an "unfree" unicode codepoint?), it is appropriate to require an additional font in order to view just one character within the font? I am asking this from the technical point of view, because I cannot recall a need to download nonstandard fonts in order to view Wikipedia articles correctly. The situation is about the International Symbol of Access and involves Template:Access icon and Help:Displaying the international wheelchair symbol. Tinlinkin 04:52, 22 May 2007 (UTC)[reply]

I think that it is appropriate. We do it in other instances. --Iamunknown 05:04, 22 May 2007 (UTC)[reply]

Those examples are for language support, which is an expected download if you want to view particular languages and alphabets/characters. (I have those fonts, by the way, and I did not have to do anything special to get them–but that may be because I specified a full installation...) The fonts are included with and easy to install with Firefox and IE (I am assuming). In this case, I think the expectation to download one font to support one character (this is my main contention) as opposed to an entire character set is close to nonexistent, and impractical as you have to go out of your way to install it. Tinlinkin 05:24, 22 May 2007 (UTC)[reply]

The East Asian character sets are not provided in IE, one must install them with a Windows installation CD. I guess I don't see that your main contention as something to be contended. If editors insists upon using the International Symbol of Access, if said symbol is not freely licensed, if using the Unicode font is what we need to do, then so be it. (That is my opinion.) --Iamunknown 06:02, 22 May 2007 (UTC)[reply]

If I want one egg, but eggs only come packaged by the dozen, I have no choice but to buy the dozen. I don't see why that sense has to apply here. That's how I see it, and even I don't think that is the best rationale against the font. If the final resolution is to use the font, I'll bow to it, but not at this time. Tinlinkin 11:04, 22 May 2007 (UTC)[reply]

Usage calculation tool

Is there any tool to calculate the usage of an article? --♪♫ ĽąĦĩŘǔ ♫♪ ^{walkie-talkie} 08:41, 22 May 2007 (UTC)[reply]

See Wikipedia:Technical FAQ#Can I add a page hit counter to a Wikipedia page? Tinlinkin 11:07, 22 May 2007 (UTC)[reply]

Thanks buddy :) --♪♫ ĽąĦĩŘǔ ♫♪ ^{walkie-talkie} 13:25, 22 May 2007 (UTC)[reply]

There is this. Quadzilla99 18:23, 22 May 2007 (UTC)[reply]

Deleted edits of renamed users stay associated with the old name

I was browsing deleted edits (unfortunately I don't remember the article) when I noticed that deleted edits of renamed users stay associated with the old name. This can be explained as follows: 1-User:X edits an article. 2-The article is deleted (after an AfD for example). 3-User:X is renamed to User:Y. 4-The article is restored (after a deletion review). Now what happens when the deleted edits are restored? Are they automatically assigned to the new name or what? I saw such deleted edits stay with the old name but I don't know what will happen when they are restored. Is this normal or is this a known bug? --Meno25 12:25, 22 May 2007 (UTC)[reply]

It's known. I think the code to reattribute deleted edits during renaming exists but is disabled for performance reasons. --ais523 12:40, 22 May 2007 (U T C)

Right, there is no INDEX on user_text for the archive table, the same reason why we don't let people browse user's deleted contributions. Voice-of-All 14:41, 22 May 2007 (UTC)[reply]

And once the edits are restored, they remain attributed to the old username. See special:contributions/Pianoman87 which is my old username - in the case of both Template talk:If and Air Force Amy, my edits were deleted when I was user:Pianoman87 and then restored after my username change. Graham87 12:56, 22 May 2007 (UTC)[reply]

I don't know that much PHP, but I believe that the code ais523 is referring to is:

global $wgRenameUserQuick;
if( !$wgRenameUserQuick )
	$this->tables['archive'] = 'ar_user_text';

...I think. Gracenotes^T § 13:54, 22 May 2007 (UTC)[reply]

Watchlist

What's with the watchlist? I mean the lag and the message? Can the message be turned off and will this lag be around from now on? Aaron Bowen 17:56, 22 May 2007 (UTC)[reply]

Set items of the class "mw-lag-warn-high" as invisible or hidden or such using your user .css page. Voice-of-All 19:30, 22 May 2007 (UTC)[reply]

Hi Aaron. The message is a notice which explains a preexisting, intermittent phenomenon which sometimes causes the watchlist and other lists to not display recent changes while various servers catch up with large amounts of changes. It only shows while there is a delay, and does not show at other times.

If you wanted, you could hide it anyway, though; scroll up for some hints on that. --brion 19:31, 22 May 2007 (UTC)[reply]

Vanishing watchlist

Hi!

I was referred to this page at the helpdesk with my problem. And my problem is that recently some articles have become "unwatched" i.e. vanished from my watchlist. One fellow told me that nobody can remove items from my watchlist (except for those who have access to the page's database I guess...) and restore them either. Do you think that there's something which can be done about it? Coolkoon 18:55, 22 May 2007 (UTC)[reply]

I believe there's a not-quite-yet-resolved issue where some operations such as protection and unprotection don't show up properly in the watchlist. Try checking the "Expand watchlist to show all applicable changes" option in your preferences; this may help work around that by displaying the preceding edits for those pages. --brion 19:36, 22 May 2007 (UTC)[reply]

Thanks for the hint, now the watchlist appears to be complete, but do you think that it affects the number of the items I have in the watchlist? And also, how can I prevent the vanishing of the articles from the watchlist?

Hiding the contents box

Hello. I was wondering how I could hide the contents box in one of my user subpages. Please could someone enlighten me?--Diniz 19:38, 22 May 2007 (UTC)[reply]

Put __NOTOC__ at the top of the page. x42bn6 Talk Mess 20:14, 22 May 2007 (UTC)[reply]

Thanks!--Diniz 21:19, 22 May 2007 (UTC)[reply]

Point of interest: it can be placed anywhere on the page, not just the top, and it'll still work. Minor detail, but I just thought I'd mention it. EVula // talk // ☯ // 22:23, 22 May 2007 (UTC)[reply]

New Messages

The orange bar won't go away, even after I cleared my browser cache. 75.20.220.134 01:26, 23 May 2007 (UTC)[reply]

This is why. Many people are frustrated at this. --R ^ParlateContribs_@ (Red Sux!) 02:04, 23 May 2007 (UTC)[reply]

Lag issue?

Generally I have Wikipedia loading fast, but it takes a time for preview and saving (a long waiting for wikipedia.org). Is it due to server lag or simply because of my LAN? --Brand спойт 08:24, 23 May 2007 (UTC)[reply]

AFAIK, it is due to the server lag. -- FayssalF - ^{Wiki me up®} 14:57, 23 May 2007 (UTC)[reply]

Proposed change to "dot" template

A recent change to {{·}} was proposed, regarding the arrangement of spaces. As this is a widely used template, I thought I should link the discussion from here. Current discussion is at Template talk:·#trailing_space. – Luna Santin (talk) 00:58, 24 May 2007 (UTC)[reply]

Rollback on watchlist

Admins do not get a rollback button on their watchlist, and for good reason, you cannot tell what the edit actually is unless you look.

But that has changed, now there are automatic edit summaries and we see things like "←Replaced page with 'donkey balls!!!!'" showing up on our watchlist. It would be really handy to be able to one-click repair this type of vandalism instead of loading the diff first. What do people think? ^(H) 01:03, 24 May 2007 (UTC)[reply]

I don't know if this would be possible to sort out, but I would say only have it if the automatic edit summary arrow is displayed. --R ^ParlateContribs_@ (Red Sux!) 01:09, 24 May 2007 (UTC)[reply]

A distinct possibility. ^(H) 01:10, 24 May 2007 (UTC)[reply]

I find that I misclick random links on my watchlist far too frequently to be comfortable with this. (Though I've been meaning to make a javascript "are you sure?" popup for rollback anyway.) —Cryptic 01:11, 24 May 2007 (UTC)[reply]

They could be placed in a CSS class for easy removal, size changing, or script access. ^(H) 01:17, 24 May 2007 (UTC)[reply]

How about if the rollback button were on the left instead of the right? I probably accidentally hit it 10 times in my first month as an admin. I was used to being able to hit the wheel button to go into scroll mode. Well, in Firefox and IE7, wheel opens a link in a new tab. So if you hit the wheel wanting to go into scroll mode and just so happen to be over top of this brand new rollback button, you fire off the rollback. It would be great, though, if there were a rollback on the left side by the diff/history links, both on contributions AND on the watchlist. --BigDT 01:18, 24 May 2007 (UTC)[reply]

Hmm... I use mouse gestures, with a top-down gesture to open in a new tab. So, I put the mouse cursor at the top and gesture down through all the (diff) links, opening them in as many tabs. A rollback link there may be dangerous for me. Talk about massively rollbacking by mistake... -- ReyBrujo 01:22, 24 May 2007 (UTC)[reply]

Tableless Web Design?

Does anyone know if there's movement anywhere in the Wikimedia world to move to Tableless web design? Vagary 02:22, 24 May 2007 (UTC)[reply]

The default Monobook skin is already a tableless design. In general, I don't see many places on Wikipedia or other Wikimedia sites that use tables for things other than tabular data. Mike Dillon 02:48, 24 May 2007 (UTC)[reply]

Autoblocks

If a user is unblocked, and the autoblock hasn't expired yet, is the autoblock automatically undone also. And since autoblocks only last 24 hours, a user can just create new accounts and edit again from their IP in 24 hours? --R ^ParlateContribs_@ (Red Sux!) 02:23, 24 May 2007 (UTC)[reply]

No and yes. Prodego ^talk 02:31, 24 May 2007 (UTC)[reply]

Thanks. --R ^ParlateContribs_@ (Let's Go Yankees!) 02:33, 24 May 2007 (UTC)[reply]