Wikipedia:Village pump (technical)/Archive 37

From Wikipedia, the free encyclopedia

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This page contains discussions that have been archived from Village pump (technical).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

< Older discussions | Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10
11, 12, 13, 14, 15, 16, 17, 18, 19, 20
21, 22, 23, 24, 25, 26, 27, 28, 29, 30
31, 32, 33, 34, 35, 36, 37, 38, 39, 40

Contents

[edit] Bot searches for POV

Is there any bot around or a script that can recurse through a category and its subcats and generate a list of every article that uses certain words? I was looking to search a few cats recursively for "freedom fighter" and "martyr". These are endemic among subcontinental history articles. Blnguyen (bananabucket) 08:05, 30 April 2008 (UTC)

If you haven't already posted at WP:BOTREQ, you might want to do so there. Also, doesn't WP:AWB have functionality that can help? (In some sense, this is no different than searching for a spelling error.) -- John Broughton (♫♫) 14:15, 30 April 2008 (UTC)
if you can give me a list of words to search for I can write a script that does this. βcommand 2 21:21, 30 April 2008 (UTC)
Words:"freedom fighter", "freedom", "martyr", "war crime", "genocide" (and their plurals). Blnguyen (bananabucket) 05:38, 1 May 2008 (UTC)
Categories +recursion: Category:History_of_Bangladesh, Category:History_of_India, History of Pakistan. Blnguyen (bananabucket) 05:38, 1 May 2008 (UTC)

[edit] Lost section at Lisa the Beauty Queen

I have worked on "Lisa the Beauty Queen" today and have encountered a strange error. I have written a section called Reception, which is not display in the article. I tried purging the cache, but it didn't help. It is the same thing in both IE and Firefox. What is wrong?

I fixed it by this edit [1]. There was a / missing in a ref name tag, which made the text following it disappear. It's a very easy slip to make, and hard to spot if you haven't seen it before. DuncanHill (talk) 12:42, 30 April 2008 (UTC)

Thanks a lot. --Maitch (talk) 12:45, 30 April 2008 (UTC)

Talk about "hard to spot"; I spend half an hour combing the history of an article to find what was wrong. I only found it when I clicked "edit" in the problematic section. I don't know how common this problem is, but it sounds like a mistake easy enough to make; perhaps a tip somewhere would be in order? Waltham, The Duke of 14:45, 30 April 2008 (UTC)
That is not an uncommon problem. It is documented at Wikipedia:FAQ/Editing#Why does part of an article not appear, although it's there in the edit screen? and a few other places. I had been wondering if we could hack CSS to highlight the ref tags. Another possibility would be a script to check for ref errors. --— Gadget850 (Ed) talk - 14:54, 30 April 2008 (UTC)
There is some discussion of a bot to find unclosed tags at Wikipedia:Bot_requests#Need_a_bot_to_find_corrupted_REFLIST_outputs. The problem does seem to come up quite frequently, and it is a right bugger trying to find a missing / in a mass of text! DuncanHill (talk) 14:56, 30 April 2008 (UTC)
Oh, dear; I've never really thought about reading the FAQ. How smart of me...
Anyway, the bot is a good idea; this kind of thing is exactly where bots are at their most useful. Waltham, The Duke of 20:00, 30 April 2008 (UTC)

[edit] user talk page redirect

I have a sock puppet (User:Naufana2) for editing on public computers (WP:SOCK#LEGIT) and I've set it so my sock puppet's user talk page redirect's to my talk page. If a bot were to leave my sock puppet a message would it be added to my real talk page (User talk:Naufana) or would be added to the sock puppet's talk page (User talk:Naufana2)? (Thus negating the redirect) Naufana : talk 20:36, 30 April 2008 (UTC)

It would depend on the precise coding of the bot, but in the majority of cases I expect the message would end up at the redirect target (ie, User talk:Naufana). Happy‑melon 21:02, 30 April 2008 (UTC)

Thanks, if I ever notice a bot acting otherwise I'll consider making a suggestion at WP:VP/P. Naufana : talk 21:11, 30 April 2008 (UTC)

You'd be better off mentioning it on the talk page of the operator - as I said, it's entirely a function of how the operator has coded the bot script; but if it doesn't follow redirects it probably should, for account renames, revoked RTV cases, and alternative accounts like this instance. Happy‑melon 21:20, 30 April 2008 (UTC)

[edit] Can't report Spam and/or edit Wikipedia_talk:WikiProject_Spam

I always get this when trying to edit Wikipedia_talk:WikiProject_Spam:

Request: POST http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:WikiProject_Spam&action=submit, from 91.198.174.35 via knsq2.knams.wikimedia.org (squid/2.6.STABLE18) to 208.80.152.40 (208.80.152.40) Error: ERR_READ_TIMEOUT, errno [No Error] at Wed, 30 Apr 2008 22:57:41 GMT

It already messed up that page a bit (double report in the wrong place) and I can't fix it completely; since my request always times out. Any ideas? (I guess it's a server problem)? User Dœ αTΩC 23:11, 30 April 2008 (UTC)

I'm currently attempting to submit an edit to that page, it seems to be on the verge of timing out. Perhaps the page is so completely inundated with external links that it's no longer able to process new edits in a timely manner. Is slightly odd that there'd be a problem since the page is a mere 69 kB long.--VectorPotential Talk 00:31, 1 May 2008 (UTC)
it could also be with the number of templates used. Im going to archive it later. βcommand 2 00:40, 1 May 2008 (UTC)
I didn't even notice all those templates being transcluded there, I just glanced at the page and didn't even look into the source at all. That's defiantly it. --VectorPotential Talk 00:43, 1 May 2008 (UTC)

[edit] Image Upload

HI. I do not know if it's a technical problem, but when I upload a new version of an image, my username replaced the original uploader in the upload history section. My username should be on top of the original uploader. Thanks. --Jackl 01:50, 1 May 2008 (UTC)

The file history on Image:Sprinter 414 UVG HK.jpg looks correct to me, unless there is another image you are referring to. Mr.Z-man 02:01, 1 May 2008 (UTC)

[edit] Interwiki a template

When I interwiki template that use noinclude includeonly onlyinclude, result is error. Why? How to solve it? -- WonYongTalk 02:02, 29 April 2008 (UTC)

It's a little hard to answer this question without knowing which template you mean. I do know that a lot of Wikipedia templates do not travel gracefully to other MediaWiki wikis because Wikipedia relies on HTML Tidy to clean up some parserfunctions syntax, or some such, and it won't work on other wikis unless they also run HTML Tidy. Other things to check:
  • Did you also copy, to the destination wiki, all the templates that your template calls?
  • Do you have all the necessary extensions running on the destination wiki? Compare Special:Version on Wikipedia to the same page on the destination wiki.
  • You might also need to copy some CSS classes from MediaWiki:Common.css.
Template porting can be a pain. It would be nice if MediaWiki had a feature to export a template plus all of its dependencies, so you could import it all into the destination wiki in one shot. But that would be difficult. --Teratornis (talk) 04:42, 29 April 2008 (UTC)
Ye gods that would be a nightmare! A script that says "oh you want to copy this template? Well you'll need this one, this one and this one as well" would be one thing, but something like {{click}} requires two or three separate extensions to be installed and configured, and {{navbox}}, {{coord}} etc, both use stuff in MediaWiki:Common.css; it would be a pig for a script even to work out when dependencies like that exist, let alone actually chase them across the wiki. Happy‑melon 20:40, 30 April 2008 (UTC)
Yes it would be difficult. That's what programmers do - whatever requires repetitive grunt work, they eventually figure out how to automate, even though automating it might be 10 to 100 times harder than doing it one time by hand. Even though it's a nightmare, it's still a deterministic nightmare. Deterministic nightmares can go into code. I'm not saying this would be a one-afternoon project. But someone might take a shot at it as more and more wikis spin off from Wikipedia, and it becomes correspondingly more important for people to port all the templates, extensions, CSS, and everything else they have gotten used to. It would help, obviously, if MediaWiki kept its entire setup in its database, instead of scattering it between the database and loose files on the filesystem like LocalSettings.php. Another possibility would be to have some turnkey setups for MediaWiki which come with a large starter set of extensions, CSS, templates, and so on all set up and ready to go. I'd expect the driver for all this would be a trend for people to set up their own wiks that, for example, accept original work. I think a lot of people learn to edit on Wikipedia, and then start to wish they could write about stuff that isn't already published, using the efficient wiki tools. There are already several thousand MediaWiki wikis, and probably every one of them required someone to bang through template porting, or somebody wishes someone would. Clearly there's a demand for this, but it's a rather diffuse demand, so it could be a while before someone takes an organized approach to it. I might mention something that would be handy for this, and other things: how about a link you could click on any MediaWiki page, that would open a page that analyzes everything on the page you were just viewing. Every template, every CSS class, every MediaWiki: namespace component, every extension, every image, basically everything the MediaWiki engine had to retrieve to build the page. Then at least you would have a report of all dependencies, if you wanted to be able to duplicate that page on another wiki. --Teratornis (talk) 01:44, 1 May 2008 (UTC)
It should be no problem at all to do this. We just need to issue a single query to templatelinks. By design, that's supposed to contain a list of all templates that any given page includes, for exactly this kind of purpose: tracking dependencies (although the specific application is mostly cache invalidation). Obviously, it would not be possible for such a feature to figure out dependencies on styles, scripts, or extensions, but those can be easily dealt with, once, by anyone who wants to port these things. —Simetrical (talk • contribs) 14:29, 1 May 2008 (UTC)
Actually, isn't this just what the "Include templates" checkbox does? Are there any problems with that? —Simetrical (talk • contribs) 14:36, 1 May 2008 (UTC)

[edit] onclick links in MediaWiki

I run a Wiki at Wikia called Christian Music. We need some links that use the onclick attribute for the <a> tag. I am told there that MediaWiki ignores all <a> tags. Does MediaWiki's link syntax provide an equivalent to onclick? It would help a lot.

Please note that this code will  NOT  be used here at Wikipedia. Will (Talk - contribs) 04:09, 1 May 2008 (UTC)

You could edit your Common.js to add an onload handler which will go over the document and insert the onclick handlers based on something you can set, like the class of a surrounding tag. Something like the following (warning: untested, may have bugs!) ought to do it:
addOnloadHook(function () {
    var myhandler = function () {
        alert("Hello! This is an onclick handler!");
    };
 
    var links = document.getElementsByTagName("a");
    for (var i = 0; i < links.length; i++) {
        if (/(^|\s)xyzzy(\s|$)/.test(links[i].parentNode.className))
            links[i].onlick = myhandler;
    }
});
Then you could use it by writing <span class="xyzzy">[[Some link]]</span> on your pages. —Ilmari Karonen (talk) 05:58, 1 May 2008 (UTC)

The function I need to call has parameters that vary with the link. Each parameter is a parameter to my template. The instructions I was given follow. I only need the last step completed.

Will,

Per our talk to day:


<code><div id="TipLayer" style="visibility:hidden; position:absolute; z-index:1000; top:-100; width: 150px; font-size: 9px;"></div></code>
  • have you refrence template create links with this format:
<a href=[http://net.bible.org/?john+3:16 "http://net.bible.org/?john+3:16"] onclick="getVerse('John 3:16') return false;" target="_blank">John 3:16</a>

Will (Talk - contribs) 06:18, 1 May 2008 (UTC)

I think the code above should have an onclick property rather than an onlick property... x42bn6 Talk Mess 12:30, 1 May 2008 (UTC)
Good idea! That would sure keep my screen from getting all wet. (grin) -- Low Sea (talk) 17:08, 1 May 2008 (UTC)

If you really wanted, you could enable <a> tags on your wiki by converting "&lt;a onclick="whatever()"&gt;" back to <a onclick="whatever()"> but that would require some serious ivy league voodoo. — CharlotteWebb 14:03, 1 May 2008 (UTC)

...and would probably create a security hole big enough to drive an SUV through. —Ilmari Karonen (talk) 23:04, 1 May 2008 (UTC)

[edit] Broken diff

My firefox crashes whenever I try to view a certain diff. On [2], click the top diff and my screen covers with a white fog (like it does on Vista) and my browser stops responding. All others work. Why?...... Dendodge.TalkHelp 17:04, 1 May 2008 (UTC)

It's fine for me (Firefox 2.0.0.14, WinXP), though it takes a ridiculously long time to load. The revision is enormously large (827K) and overspills the expensive parser function count, which could be part of the problem. Algebraist 18:28, 1 May 2008 (UTC)
That could be it, I mght just be impatient. It's just that that blocked user is one of my real-life friends and I wondered what he did to get blocked. OK, thanks...... Dendodge.TalkHelp 18:33, 1 May 2008 (UTC)
Most likely one of the MANY javascripts that you have installed in your monobook.js is hitting a security limit or something within Firefox, when processing such a large diff text. Shouldn't crash of course, but i have seen similar bugs before. --TheDJ (talk • contribs) 18:37, 1 May 2008 (UTC)
Takes an absolute age to load on IE7, doesn't quite crash it but it feels like it wants to. DuncanHill (talk) 20:31, 1 May 2008 (UTC)

[edit] Image issue

Could someone maybe take a peep at this issue? Thanks. —TreasuryTag—t—c 19:58, 1 May 2008 (UTC)

There is code in place to not allow images before the episode has aired. — Edokter • Talk • 20:03, 1 May 2008 (UTC)

[edit] Using nowiki inside a template

I want to create a template that wraps its argument inside code and nowiki tags, like this

<code><nowiki>{{{1}}}</nowiki></code>

where the inner {{{1}}} expands to the template parameter. Invoking the template like this

{{my template|{{another template}}}}

should produce output like this

{{another template}}

just as if the input had been like this:

<code><nowiki>{{another template}}</nowiki></code>

How do I arrange for the {{{1}}} to be substituted, but for the result of the substitution to be protected by a <nowiki>...</nowiki> so that, even if it contains something that looks like a template invocation or wiki markup, it doesn't invoke that template or take the action implied by the markup? —AlanBarrett 20:03, 1 May 2008 (UTC)

<code>{{#tag:nowiki|{{{1}}}}}</code> would do the trick using the #tag PF, otherwise substitution would. The caveat is that any templates or tags equal to {{{1}}} would be expanded, although the nowiki of those expansions will be displayed. I believe there's a template that does this, although I've forgotten the name. GracenotesT § 20:14, 1 May 2008 (UTC)
Sorry, using #tag:nowiki doesn't work any better than using {{nowiki}} (which in turn uses #tag:nowiki). The argument is expanded too early. See User:AlanBarrett/Sandbox for an example, and feel free to edit it. —AlanBarrett 20:39, 1 May 2008 (UTC)
In that case, there's not much you can do. Substitution (which {{code}} already does) will work. GracenotesT § 20:49, 1 May 2008 (UTC)
You can also try to pass the nowiki as a template parameter, eg: <code>{{{1|}}}</code> and {{Template|<nowiki>{{foo}}</nowiki&gt}}. But other than that, it isn't really possible to half-parse something. --Splarka (rant) 20:56, 1 May 2008 (UTC)

OK, so a normal template invocation, or parameter substitution inside a template, will be substituted, re-parsed, substituted again, re-parsed, etc., until either there are no more changes, or a limit is reached. The msgnw parser function will substitute just once, but its argument must be a template name, not a template parameter ({{msgnw:{{{1}}}}} doesn't substitute the value of parameter 1, it substitutes the content of a template whose name is given by parameter 1). {{subst:foo}} substitutes just once, but it happens when the page is saved, not when the page is rendered. Is there no way to substitute a template parameter just once, without re-parsing and re-substituting the result, and do it when the page is rendered rather than when it's parsed? —AlanBarrett 21:34, 1 May 2008 (UTC)

You can probably use {{lessthan}} to kluge this. Usage should be obvious. —Ilmari Karonen (talk) 23:01, 1 May 2008 (UTC)

[edit] Images proposal

I have made a new proposal regarding images here Wikipedia_talk:Manual_of_Style#New_proposal_on_images - I would appreciate any comments, suggestions, advice etc. Thanks Gustav von Humpelschmumpel (talk) 20:48, 1 May 2008 (UTC)

[edit] Deprecation of Template:Qif

Using this template is highly discouraged in the light of #if parser function, yet earlier this month I discovered a couple of templates using it. Currently, it's used only in userspace sandboxes[3] and to avoid further usage of this template, I propose to phase it out completely by disabling its functionality and replacing its contents with an instructions on how to use #if. MaxSem(Han shot first!) 16:20, 28 April 2008 (UTC)

Delete it - keeping it is more trouble than it's worth given that #if: is a universally better solution. Happy‑melon 17:45, 28 April 2008 (UTC)
Yup, it's been deprecated since September 2006. TFD is that-a-way. In the mean time, I think I ought to be able to fix most of the uses with a simple regexp. I already substed all the uses in the Wikipedia: namespace — they were all old substitutions of {{RD}}. (For those interested, the regexp I used was s/\{\{(\s*[Qq]if\s*\|\s*test\s*=)(\{\{\{\w+\|\}\}\})?/{{subst:$1/g.) —Ilmari Karonen (talk) 18:57, 28 April 2008 (UTC)
Also fixed the remaining uses in "Wikipedia talk:" space, plus a couple of uses on archived user talk pages. All remaining transclusions, except for Template talk:Qif itself, are in userspace. In the mean time, though, there's also Template:Switch, which was until today even used in a widely used article space template (Template:Infobox Nuclear power plant). I've fixed that one, but with the job queue at 9 million pages it'll take a while for the change to trickle down. —Ilmari Karonen (talk) 20:05, 28 April 2008 (UTC)
Please don't delete {{qif}}; it's an important part of the history of Wikipedia. Blanking it may be a better idea, or moving it to a Wikipedia:Archive subpage. --ais523 18:54, 1 May 2008 (UTC)
Certainly, I've never proposed to delete it. I've disabled[4] it instead. MaxSem(Han shot first!) 12:00, 2 May 2008 (UTC)

[edit] Some popups functions not working

Hi. I wrote about this at Wikipedia talk:Tools/Navigation popups#Functions stopped working yesterday, but I haven't gotten a response yet (is anyone monitoring that page anymore?}, so I figured I'd cross-post here for a wider audience.

For the past few days, the "diff my edit" (cumulative changes since my last edit) and the "latest edits" (cumulative changes by last editor) have both stopped working. They used to create a new tab with the relevant diff, but now nothing happens. I've tried several things to see if I could fix this, without success. Is it just me and something in my configurations, or is anyone else having this problem? Thanks! Ashanda (talk) 17:43, 28 April 2008 (UTC)

Heh, I didn't even know it would do that! Anyway, works fine for me under IE7 (after I enabled popup windows from en.wikipedia.org). Sorry for the null result, but thanks for the tip! :) Franamax (talk) 21:05, 28 April 2008 (UTC)
But now looking at your monobook.js, your FriendlyConfig section looks to have invalid syntax, comma-followed-by-right-brace will almost certainly fail. Take that whole section out. Franamax (talk) 21:11, 28 April 2008 (UTC)
And strike my first comment, after purging my browser cache I DO have the same problem. Thanks a lot! diffmyedit and latestedits do nothing at all now. (I still don't like your latest FriendlyConfig section though). Franamax (talk) 21:39, 28 April 2008 (UTC)
Thanks for the advice. When I looked closer at the Friendly configuration code, I noticed that they do leave off that final comma. They list it along with multiple previous configuration settings that do use commas, so I missed it in my ignorance of javascript. So I erased the comma, not that it seems to have made any particular difference- as you pointed out. I've tried completely taking out everything but the popups configs from my monobook.js; working in IE7 & Firefox; changing the selected popupstructure; logging off then back on to Wikipedia; and restarting the computer. Always purging my own and the server's cache each time. With your report that you're having the problem too, I'm at the point of wondering whether something's changed here on the wiki. I checked and the popups coding hasn't been changed lately and Vista hasn't downloaded any updates in more than two weeks. So does anyone know if there's something new on Wikipedia that's breaking these functions? Thanks! Ashanda (talk) 05:52, 29 April 2008 (UTC)
My guess would be changes in the api.php function, but the popups code is a monster to look through. I think User:Lupin is the god of navpopups, but it looks like Lupin hasn't been online for several months. I guess we'll have to try to track down a lesser being who has some understanding of that code. Franamax (talk) 18:58, 29 April 2008 (UTC)
I investigated the issue and I cannot find anything wrong. I used the functions on a link of my talk page and it worked exactly like it should (including after cache cleaning). Are you sure you did not install or activate a popup blocker that is preventing the new window/tab to be opened? Can you give me a specific link that I can test for the popup? Perhaps it has something to do when revisions or something are specified.... --TheDJ (talk • contribs) 21:37, 29 April 2008 (UTC)
No new popup blockers, and Wikipedia's on my whitelists for both browsers. I use the Google toolbar and there haven't been any updates on those for months and anything I may have downloaded inadvertently would have only affected one browser or the other, wouldn't it have? I have this problem on any link anywhere- talk pages, mainspace, revision histories, watchlist, recent changes, etc. If the problem isn't on here, then the only other thing I can think of is that it's a Vista or a McAfee issue. It does seem odd that sub-popups work (i.e. hovering over "History" within a popup brings up a new, "sub", popup) but that the two functions that require a new page/tab to open won't. Ashanda (talk) 03:50, 30 April 2008 (UTC)
I've run into this same issue. My popups work except for the diff tab. If anyone has any advice, please let me know. I've purged my cache, allowed popups from en.wikipedia.org, turned it on and off. Please drop me a line on my talk page. Thanks in advance! Tnxman307 (talk) 17:47, 1 May 2008 (UTC)
@Ashanda. Do you perhaps have the popups enabled, both as a gadget (selected in your preferences ) and trough your monobook.js ? Perhaps they are conflicting. If not, can you try removing the importScript from your monobook.js and enabling the popups Gadget instead? Perhaps the problem is limited to the version of Lupin and the problem is not present in the version that is installed as a gadget ? --TheDJ (talk • contribs) 18:48, 1 May 2008 (UTC)
I have only the gadget enabled, IE7 on XP Pro, Trend Micro AV, popups allowed and have the problem. It was working fine until I purged my cache on 28 Apr investigating this thread, then stopped working. Franamax (talk) 19:02, 1 May 2008 (UTC)

<unindent>Nope, didn't have it selected in both my Gadgets and Monobook.js, so no conflicts. I tried using it just from the Gadgets, though I had to leave the structure configuration in my monobook.js to even have the diff options available- no dice. I tried logging in using Safari today, still no diffs. I checked in at the discussion I started at Wikipedia talk:Tools/Navigation popups#Functions stopped working, and some editors say that things started working for them again yesterday. I don't know what's up. Ashanda (talk) 18:31, 2 May 2008 (UTC)

[edit] Diff coloring again

We should consider adopting the Afrikaans Wikipedia diff coloring style. It is unobtrusive yet still allows you to spot small changes like addition or removal of spaces. For example: [5]. The code they are using is:

/* Mise en évidence des diffs par du gras et de la couleur. */
table.diff {
 padding:.5em;
}
table.diff td { 
 vertical-align:top;
}
td.diff-addedline { 
 background:#D8E4F6; 
}
td.diff-addedline .diffchange {
 background:#B0C0F0;
 color:#001040;
 font-weight:bold;
}
td.diff-deletedline {
 background:#E4F6D8;
}
td.diff-deletedline .diffchange {
 background:#B0E897;
 color:#104000;
 font-weight:bold;
}
td.diff-context {
 background:#FEFEFE;
}

—Remember the dot (talk) 19:34, 28 April 2008 (UTC)

Those almost-black-but-still-not-completely-black letters in changed text are ugly. Simple black would do better. As a variant, here is the coloring I use (and adopted it for AWB):
td.diff-deletedline span.diffchange {
    background-color: #FFD754; color:black;
 }
 
 td.diff-addedline span.diffchange {
    background-color: #73E5A1; color:black;
 }
 
 .diffchange-inline {
   border: none;
 }
this colouring was invented by Mithgol. MaxSem(Han shot first!) 19:53, 28 April 2008 (UTC)
Red background
Red background
Those are quite... different colors. I only have one line in my monobook for the same effect (but only using red, see image to the right):
/* Red background for changed text. */
 
.diffchange {
  background: #fdd;
}
— Edokter • Talk • 19:56, 28 April 2008 (UTC)
Yoink, I'll be having that! I like the Afrikaans one the best, I think. How do I get rid of the ugly white table background - is it possible, or does it need a change to the source? Happy‑melon 19:58, 28 April 2008 (UTC)
The Afrikaans guys just stole it from French Wikipedia, btw, so we shouldn't credit them. MaxSem(Han shot first!) 21:44, 28 April 2008 (UTC)

Very nice! I've been using these CSS rules for a couple of days, and I have to say they look excellent. They do take a little bit of getting used to, particularly the white context background instead of the default gray, but that passes quickly. I'd certainly support adding these to Common.css. —Ilmari Karonen (talk) 16:11, 30 April 2008 (UTC)

[edit] Implemented

Per discussion here, I've added the Afrikaans/French variant to MediaWiki:common.css, with modification proposed by me - strictly black font color. Please purge your cache and see if you like it. MaxSem(Han shot first!) 12:06, 2 May 2008 (UTC)

It looks lovely (I use the greenscreen gadget). Thank you. DuncanHill (talk) 12:19, 2 May 2008 (UTC)

OK, there is one problem with this... You can no longer see empty lines, which is an essential part of the diffs; non-changed lines should have a background. I think this was implemented a bit too soon. — Edokter • Talk • 12:42, 2 May 2008 (UTC)

I agree. There wasn't much discussion. Prior to reading your comment, I switched to a more familiar color scheme (by reassigning the green hues to the new page text and changing the old page text to a shade somewhat similar to the conventional yellow), but it might be a good idea to simply revert to the standard setup if a solution isn't devised soon. —David Levy 13:10, 2 May 2008 (UTC)
I've restored the grey background for td.diff-context. — Edokter • Talk • 13:22, 2 May 2008 (UTC)

I can't see the changes in diffs anymore with these horrible new colors. I have a visual impairment. The old diff colors were very clear, high contrast, and easy to see. I don't think I am only one who will prefer the original longstanding diff colors. Please revert the change until pending further discussion, or at least provide a user preference to disable the horrible new diff colors. I see there really hasn't been much discussion yet. - Neparis (talk) 13:56, 2 May 2008 (UTC)

I'm glad it was reverted; I was surprised to see suck lack luster colors in a diff. EVula // talk // ☯ // 15:01, 2 May 2008 (UTC)
I agree that there was insufficient discussion and support the reversion. A change that reduces accessibility is a bad one.
Regarding the goal of making small differences easier to spot, I suggest using the following code:
.diffchange {padding: 0px 2px 0px 2px; border: 1px dashed red; margin: 0px 1px 0px 0px}
This simply adds a dashed line (without otherwise altering the appearance in any way). —David Levy 16:16, 2 May 2008 (UTC)
I think the dashed border was shot down pretty hard not a month ago when a dev implemented that change. — Edokter • Talk • 16:29, 2 May 2008 (UTC)
Oh, I wasn't aware of that. What was/were the complaint(s)? I've been using the above code for more than a year, and it's worked very well for me. —David Levy 16:34, 2 May 2008 (UTC)
Old discussion is here: Wikipedia:Village pump (technical)/Archive 36#Dotted borders. — Edokter • Talk • 16:43, 2 May 2008 (UTC)

[edit] Account clearance

The following discussion is archived. Please do not modify it. Subsequent comments should be made in a new section.

Would it be possible to do a one-time clearance of all accounts older than say...6 months with no edits? I just noticed there are over 7 million, so most good usernames will have been taken--Phoenix-wiki 19:53, 30 April 2008 (UTC)

If anyone wants to use one of those usernames, they may visit WP:USURP. Nakon 19:55, 30 April 2008 (UTC)
Yes but many people don't know about it and nearly allnew users get denied for having too few edits...--Phoenix-wiki 19:57, 30 April 2008 (UTC)
The problem is that accounts can't easily be deleted, certainly not through the mediawiki software (AFAIK). This would require a dev to manually delete these accounts from the user table (or more likely, use a script to do the same). However, this is not an inherently bad thing: in fact, I would strongly support a proposal to delete accounts which are more than, say, 6 months old, with no edits whatsoever. It would stop our "number of users" statistic being so misleadingly high, and as you say, reduce the chances of people wanting to create an account getting a name conflict with a useless account. Happy‑melon 20:58, 30 April 2008 (UTC)
  • strong support. a simple SQL query is all that is needed. βcommand 2 21:18, 30 April 2008 (UTC)
  • Support good housekeeping, good for new editors, more realistic user numbers are good for the Wikipedia. DuncanHill (talk) 21:50, 30 April 2008 (UTC)
  • Support - more accurate statistics, more available usernames (which will also reduce load on bureaucrats doing usurping), will clear all the "JOE IS A <expletive>" usernames blocked before making any edits out of Special:Listusers, and may also delete some sleeper accounts. Mr.Z-man 22:01, 30 April 2008 (UTC)
  • Support Very excellent idea. MBisanz talk 22:23, 30 April 2008 (UTC)
  • Support – There might be many people who registered accounts with the intention to use them in the future... After six months, it is more likely than not that they have forgotten all about them. Waltham, The Duke of 22:41, 30 April 2008 (UTC)
  • OPPOSE - willing to change pending answer to this legal question... Wouldn't a blanket deletion like this run afoul of the GNU and WIKI attribution licenses? USURP at least leaves a trail when the username changed hands but if new users were able to re-create accounts with usernames that were previously in use it could cause all sorts of attribution confusion. -- Low Sea (talk) 22:48, 30 April 2008 (UTC)
I would of course support cleaning away any accounts that have ZERO edits. -- Low Sea (talk) 22:52, 30 April 2008 (UTC)
This is strictly about accounts with no edits whatsoever. No legal problems here. Waltham, The Duke of 22:53, 30 April 2008 (UTC)
  • Question - how many such users are there (with zero edits)? And BTW, even if they don't edit they might be very active readers who want to set their own preferences or watchlist - which require an account. Sbowers3 (talk) 22:58, 30 April 2008 (UTC)
  • Comment - I would think that there are a large number of people who create accounts just for the sake of browsing, after all you can't have a watchlist unless you have an account. I suspect that if we were to go ahead with this we would suddenly find a parade of confused, good faith users, long time readers, lurkers, and other relatively harmless people trying to contact us to find out what happened to their accounts.--VectorPotential Talk 23:01, 30 April 2008 (UTC)
    • It would be trivial to put something in the sitenotice or the watchlist notice a couple weeks beforehand asking people to make an edit if they want to keep their account. Mr.Z-man 23:07, 30 April 2008 (UTC)
    • Alternately, if the sysadmins use a script to do it, they could also check if the user has anything on their watchlist. Mr.Z-man 23:12, 30 April 2008 (UTC)
  • Support: But ask users first. 1 week, no response, delete...... Dendodge.TalkHelp 18:35, 1 May 2008 (UTC)
Please note that we do not intend to do this at this time or any planned time in the future. Note however that unused accounts can be easily usurped as necessary, *if* necessary. --brion (talk) 23:31, 30 April 2008 (UTC)
Some people may just make accounts for preferences, so it's best not to have them not seem to have randomly disappeared by next login attempt ;). And the extra rows cost almost no space/overhead. Perhaps what we could do is make decent automated "active user" lists... Voice-of-All 04:18, 1 May 2008 (UTC)
  • I don't see a need for this. Why delete all the inactive accounts, just to have more people create the accounts, and then be inactive with them again? We have usurpation policies in place. If an editor is serious about staying on the project, they can usurp the name. If they aren't, do you really want them to be the one to claim the in-demand name, which is what this would allow them to do? I don't. seresin ( ¡? ) 04:22, 1 May 2008 (UTC)
    • The thing is, most users do not use these accounts at all, therefore they will not be re-created. If they did use them, there would be little point in deleting them, would there not? Besides, most new editors, no matter how serious they are about joining Wikipedia, have probably never heard of the capability to usurp accounts. Waltham, The Duke of 08:33, 2 May 2008 (UTC)

Signing up to a site, going away for some indefinite period, and then having your login not work when you return is extremely annoying. The members figure can be augmented with an active members figure if desired. As for usernames being taken, well, that's what happens on big websites, be more creative. —Simetrical (talk • contribs) 14:33, 1 May 2008 (UTC)

Yes, but six months? This is rather implausible. And even if it is not, how would you feel about one year's inactivity? That would remove all doubt, I think; after an entire year, there are three possibilities: one has forgotten about the account or stopped using it, or one has been using it to read (in which case one will in all probability use a watchlist and see the notice), or one has not resisted the lure not to make even a single edit. In every case, there will hardly be any disturbance, and we get to clean up our user logs. Waltham, The Duke of 08:33, 2 May 2008 (UTC)

This proposal has been and gone, years and years ago — I think I remember it from 2005. As Brion Vibber, our CTO and the final authority on all things technical has said, this will not happen, regardless of any on-wiki consensus that may be formed. — Werdna talk 14:06, 2 May 2008 (UTC)

Closed as Brion will not implement this system regardless of on-wiki consensus. Nakon 14:08, 2 May 2008 (UTC)
The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made in a new section.

[edit] Popups gadget problem

I use the navigation popups gadget, on Safari on WinXP. It appears to be malfunctioning - it works when I point at the title of a page, but when I point to "diff" or "hist" or "last" nothing now happens. Any help much appreciated. DuncanHill (talk) 21:18, 30 April 2008 (UTC)

I have tried switching them off and on again, clearing my cache, don't know what else to try. DuncanHill (talk) 08:37, 1 May 2008 (UTC)
They do appear to be working normally in IE7. DuncanHill (talk) 08:53, 1 May 2008 (UTC)
I just posted in a section above this one. I've run into the same problem and it just started today. I can't think of any changes I've made. I use Firefox2 on WinXP. Any suggestions? Tnxman307 (talk) 17:51, 1 May 2008 (UTC)
I don't know what happened, but it seems to be working again. How odd. Tnxman307 (talk) 18:36, 1 May 2008 (UTC)
Still not working for me on Safari. DuncanHill (talk) 11:17, 2 May 2008 (UTC)
And now (after turning off, clearing cache, turning on and clearing cache yet again) they are working! DuncanHill (talk) 12:40, 2 May 2008 (UTC)

[edit] Hidden section headers

Is there any way to have the section headers not show up by still achieve section editing (i.e., have &section=section_number work? Wrapping the section header (==section header==) in a display:none div does not work as it also disables anchoring to the section as well as section editing. Any ideas? --soum talk 18:20, 1 May 2008 (UTC)

Why? -- John Broughton (♫♫) 21:20, 1 May 2008 (UTC)
Take a look at 2008 Indian Premier League. The list of fixtures becomes a bit too cumbersome for editing. Sections would be of help. But section headers would mess up the display. --soum talk 13:38, 2 May 2008 (UTC)

[edit] Line of text shooting off the right-hand edge of the screen

In the article Onan there is a line of text which extends far beyond the right-hand edge of the screen. I think it might have something to do with the {{dn}} templates in it, but can't really work it out. Any help please? DuncanHill (talk) 12:50, 2 May 2008 (UTC)

There is an unclosed non-wrapping span somewhere, probably in {{fix}}, or {{dn}}... Having trouble figuring it out. — Edokter • Talk • 13:03, 2 May 2008 (UTC)
The italic markup in {{fix}} is terminating the italic markup in the article prematurely; either the article should use inline CSS or the template should. I've left a couple examples at User:TangentCube/sandbox5. —TangentCube, Dialogues 13:16, 2 May 2008 (UTC)
Thanks for fixing it - I must confess I don't really understand how, but it is still appreciated! DuncanHill (talk) 13:18, 2 May 2008 (UTC)
I had just figured that out and switched the article to use blockquote. --— Gadget850 (Ed) talk - 13:19, 2 May 2008 (UTC)
...or that works too. —TangentCube, Dialogues 13:21, 2 May 2008 (UTC)
It must be in {{fix}}, because {{cn}} causes it as well. --— Gadget850 (Ed) talk - 13:23, 2 May 2008 (UTC)
Reported the issue at Template talk:Fix --— Gadget850 (Ed) talk - 13:28, 2 May 2008 (UTC)

[edit] Add only article space pages I edit to watchlist

Is there a way of only adding articles and article talk pages which I edit to my watchlist, not every page I edit? Hiding T 13:30, 2 May 2008 (UTC)

Easy to do with a userscript. Remove this option from your preferences and then add the code below to your monobook.js (or other subpage if you use other skin). —AlexSm 13:56, 2 May 2008 (UTC)
//mark "Watch this page" checkbox when editing article or article talk
if (wgAction=='edit' && (wgNamespaceNumber==0 || wgNamespaceNumber==1)) 
  addOnloadHook(markWatchThis)
function markWatchThis(){
  var watchThis = document.getElementById('wpWatchthis')
  if (watchThis) watchThis.checked = true
}

Thankyou. Hiding T 14:12, 2 May 2008 (UTC)

Might not work on preview – perhaps wgAction=='edit' → (wgAction=='edit' || wgAction=='submit')... GracenotesT § 14:24, 2 May 2008 (UTC)
The idea is to set the initial state of the checkbox only when you start editing. If you uncheck it manually after that, I don't think the script should overwrite your decision. —AlexSm 14:31, 2 May 2008 (UTC)
You should consider getting this added to Help:Watching pages. --— Gadget850 (Ed) talk - 16:49, 2 May 2008 (UTC)

[edit] removing numbers from Table of contents

Is it possible to remove the numbers from the Table of Contents, and use my own numbering system for the sections? How? --Filll (talk) 18:41, 2 May 2008 (UTC)

<div class="nonumtoc">__TOC__</div>
--Splarka (rant) 20:46, 2 May 2008 (UTC)

[edit] Titles

In Spanish, the titles of works (musical books, films, pictures, sculptures, pieces, programs of radio or television, etc.) they are governed by the same norms that any text. This means that only they must accentuate the first word, and the names.

Examples: Últimas tardes con Teresa (Teresa it's a name); La vida es sueño; La lección de anatomía; El galo moribundo; Las cuatro estaciones; Las mañanas de la radio; Informe semanal. See Dictionary of doubts (mirror), Real Academia Española and Association of Spanish Language Academies.

Also see: Wikipedia:Naming conventions (capitalization); and other examples of Spanish capitalization are The Circular Ruins/Las ruinas ciruclares and Funes the Memorious/Funes el memorioso.

Then, an article titled in Spanish would have to fulfill the Spanish norms, and not to take capital letters (unless it is a name). Or I am mistaken?

Eloy (talk) 21:40, 2 May 2008 (UTC)
How is this relevant to VPT? Happy‑melon 21:57, 2 May 2008 (UTC)
Well... this is the English Wikipedia. As long as the title of the article is English, I don't see what it matters. We generally provide the "localized" names in the articles themselves. EVula // talk // ☯ // 22:06, 2 May 2008 (UTC)
This page is not really the place to discuss it, but Wikipedia:Naming conventions (capitalization)#Capitalization of expressions borrowed from other languages says: If the article is about a work in a foreign language (such as a book or other written work, movie, album, or song), using the capitalization found in most English language reliable sources is recommended. So I think you are mistaken. It does not have to fulfill the Spanish norms, but it may do so depending on what English sources call it. PrimeHunter (talk) 00:06, 3 May 2008 (UTC)

[edit] Can a template do this?

The {{quote request}} aka {{request quote}} template has a problem and I think I have a solution IF templates can do the following:

1. template used inline generates inline comment (that is what QR does now)

and

2. beggining of paragraph would also add ONE block saying something like:

The following text includes one or more requests for explicit quotations from the cited source to address questions of interpretation. Please place the quotation on the TALK page for this article, not in the article text itself.

I guess what I am asking is

(A) Can a template create both an inline comment and a block comment above itself?

and

(B) Can a template not produce duplicate block comments if one already exists.

I think what I envision is having 2 tags: {{QR}} (alias for the current "quote request") and {{QR-EXPLAINED}}. When QR is added then as the template software expands the template it would look above (backwards search) for either the QR-EXPLAINED tag or the end of the previous paragraph (whichever comes first). If it finds no QR-EXPLAINED tag then it adds one.

Can this work ? -- Low Sea (talk) 17:20, 30 April 2008 (UTC)

No. (A) is just about possible if you train users to put the template around the paragraph, but that's counterintuitive; (B) is impossible without adding sitewide scripts. If you don't mind it being just you who can add templates like that, you could request a user script for it (see WikiProject User scripts' requests page); but that fix would only work for you, and for other users the templates would work the same as currently. --ais523 18:38, 1 May 2008 (UTC)
I agree with ais523; I don't think you can do that. You want your template to use information from the page the template is transcluded into. As far as I know, a template can only use information that's passed into it as arguments. Sorry. Why not just have people add the QR-EXPLAINED template at the beginning of the paragraph themselves? ☺ Coppertwig (talk) 00:57, 2 May 2008 (UTC)
That is plan B ... just was wondering if it could be done. Thanks folks. -- Low Sea (talk) 04:52, 4 May 2008 (UTC)

[edit] Software I have developed

Resolved. Blocked sockpuppet spouting nonsense

Pedro :  Chat  15:21, 2 May 2008 (UTC)

Hi there, I have developed a piece of software that allows me to see any Ip address from the entire wikipedia without being a checkuser. I have written the program in Visual Basic and I have tested it and works fine. I was wondering if anybody would know who to contact about the software possibly being given to the checkusers in order to reduce the work load of continually checking each individual IP address when all you have to do is you put the IP address in that you want to search for and it will take you to the exact computer that it is assigned to rather than a given geographical location. I beleive that this software will greatly save upon the amount of time that a checkuser is having to take up searching through all of the contributions that the editor has made. Anyway a little about myself I am a Computer Forensic student i passed my A Level computing 100% and i am predicted an A at university. Any comments are welcome on my talkpage and i will get back to you ASAP Computermadgeek (talk) 11:24, 2 May 2008 (UTC)

A checkuser has the capability to match up IPs to a certain username; I doubt your software can do that. Anyone can see the IP adresses of editors without a username and do a lookup, and there are other tools that can assist in that, like WikiScanner. — Edokter • Talk • 13:30, 2 May 2008 (UTC)
Yes my software is able to do that. But I would also point out that there is a major security leak in one of your servers as my program was able to tap on the door and say hey you know me dont you and the server let it in. Computermadgeek (talk) 13:42, 2 May 2008 (UTC)
User blocked indef as a sock of User:Chris19910. He also has software that can do the work of 10 stewards. Nakon 13:56, 2 May 2008 (UTC)
Wow, that was quick. For future reference, my edit-conflict:
If you have genuinely developed a method of identifying the IP address used to make an edit to wikipedia as a logged in user without checkuser access, you are in significant breach of Foundation policy (section 1, "Only persons whose identity is known to the Wikimedia Foundation shall be permitted to have access to any nonpublic data or other nonpublic information produced, collected, or otherwise held by the Wikimedia Foundation..."), which effectively terminates your rights to use Wikipedia. If you think that your program violates this criterion then you must contact the Wikimedia Foundation immediately (info@wikimedia.org) or the Arbitration Committee (arbcom-llists.wikimedia.org). If, as I suspect, this program is just an automated amalgamation of various WHOIS or RDNS queries, then there is no privacy issue; the checkuser community would still be interested to hear about the tool - you should e-mail the ArbCom at the e-mail address above (as the majority of the ArbCom are also checkusers). Happy‑melon 14:07, 2 May 2008 (UTC)

For future reference, Happy-melon, that should have gone to security@wikimedia.org instead. — Werdna talk 00:38, 3 May 2008 (UTC)

My bad (I just went to the foundation and looked for "contact us". I guess if it had been sent to info@ it would have ended up in the right hands eventually. AS it turns out, no need to worry. Happy‑melon 18:09, 3 May 2008 (UTC)

[edit] Request tool to discover articles without any images for Wikipedia Takes The City photo contests

I would like to request a tool to discover all articles without any images (except on templates) within a given category. It would also be useful to exclude specific subcategories. This would be useful for the Wikipedia Takes The City photo contests. For example, it would be great to have a list of all such articles in Category:Manhattan, including its subcategories, but excluding the specific subcategory Category:People from Manhattan. It would be fantastic if someone could help me with this. Thanks.--Pharos (talk) 04:41, 4 May 2008 (UTC)

Not sure if this is useful as a tool. But I can run it for you as a one time query. Aaron Schulz 04:47, 4 May 2008 (UTC)
There are a number of different groups considering Wikipedia Takes The City events. We already held one in New York (without the benefit of such a tool), there will be one in Nashville this month, and a second in New York this fall. There has also been discussion about holding such events in the Netherlands, Hungary, South Korea and France. I hope that if was organized as a tool, it could be used by all these groups, and also that we would be able to try multiple queries to make sure we're formulating it right. Thanks.--Pharos (talk) 05:01, 4 May 2008 (UTC)
Take a look at Category:Wikipedia requested photographs in the United States and its subcats. DuncanHill (talk) 07:01, 4 May 2008 (UTC)
Yeah, I'm familiar with that process, and used it to some extent in Wikipedia Takes Manhattan. But because it depends on people actually going through articles and manually adding the tag to the talk page, this only catches a fraction of the total. An "Imagless Tool", as described, would not only help the Wikipedia Takes The City events, but be of general use in cataloging articles without images (within a category or perhaps within a WikiProject), and we could also use a bot to mark their imageless status on the talk pages automagically (thus benefiting the encyclopedia-creation process generally).--Pharos (talk) 01:28, 5 May 2008 (UTC)
Make a bugzilla request, and I'll look into efficiency issues. Any special page probably won't have the exclude bit though. If it's OK, I'll put it on my todo list. Aaron Schulz 04:51, 5 May 2008 (UTC)
This is efficient enough for an isExpensive() special page, but probably not for an inexpensive one. It would be page LEFT JOIN imagelinks ... WHERE il_from IS NULL or similar, same as uncategorized pages or whatnot. Here are the first 1000 alphabetically:
MySQL query
(If anyone knows of a magical template to put this in a box or something so it doesn't take up 1000+ lines of the page, feel free to use it.) —Simetrical (talk • contribs) 14:30, 5 May 2008 (UTC)

[edit] Belize

Resolved. Gary King (talk) 02:56, 6 May 2008 (UTC)

Does anyone else see this:

km² (List of countries and outlying territories by total area)
Expression error: Unexpected div operator sq mi

at the top? Seems to have been there for sometime - has a template screwed up? asenine t/c\r (fc: f2abr04) 14:04, 4 May 2008 (UTC)

The problem was with {{infobox country}}, and now seems to be fixed. Algebraist 14:38, 4 May 2008 (UTC)

[edit] Why is this article in CAT:CSD?

The Celestial Plot contains nothing to cause it to be in CAT:CSD, but it appears there as of my timestamp. Can anyone identify why? Splash - tk 15:48, 4 May 2008 (UTC)

Caching issue - null edit cleared it. No idea how it got into the category in the first place. Happy‑melon 15:54, 4 May 2008 (UTC)
From WP:HD#Category Bug?: probably transclusion of HasSpoilers (now deleted). Algebraist 16:23, 4 May 2008 (UTC)

[edit] Possible FF3b5 render bug

Try this page, [6]. The line on the left-hand side is rendering black instead of gray with FF3b5 (Linux). It renders gray on Opera 9.27 (Linux). Could a few other people tell me their result. If it's a true FF3 render bug, I'll file a report at bugzilla. Jason Quinn (talk) 19:05, 4 May 2008 (UTC)

Where am I looking? The left-hand border of the table? Happy‑melon 20:54, 4 May 2008 (UTC)
No. The left hand border of the article itself (or the right hand border of the navigation window if you like). Jason Quinn (talk) 01:24, 5 May 2008 (UTC)
Seems no different to me than the border on any other mainspace page (IE7 on Windows). I can't work out whether it's grey or black. Happy‑melon 10:53, 5 May 2008 (UTC)
I can confirm this, Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9b5) Gecko/2008041514 Firefox/3.0b5. If you file a bug, post the bug number so I can vote for it. A smaller page as a test case would be good, that one is pretty huge. Firebug seems to choke on it. —Simetrical (talk • contribs) 14:48, 5 May 2008 (UTC)
Nothing appears out of sorts here (FF 3.0b5 / Mac OS X). --brion (talk) 17:57, 5 May 2008 (UTC)