Wikipedia:Village pump (technical)/Archive 72

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

Village pump (technical) archive  

This page contains discussions that have been archived from Village pump (technical).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

---

< Older discussions · Archives: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z, AA, AB, AC, AD, AE, AF, AG, AH, AI, AJ, AK, AL, AM, AN, AO, AP, AQ, AR, AS, AT, AU, AV, AW, AX · 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216

'random article' doesn't show in page history?

I don't know if this is a feature or a bug, but when I click the 'random article' link (leads to Special:Random) the article I end up at doesn't show up in the browser page history. well, the first page may show up, but subsequent clicks on the link won't. I don't know if this is Safari specific or true of other browsers, and it's only a problem because I can't navigate back to previous randomly-found pages. --Ludwigs2 17:34, 3 March 2010 (UTC)

Works for me in both firefox and IE. --JokerXtreme (talk) 17:39, 3 March 2010 (UTC)

It does show in Firefox, but I understand why Safari might choose not to. When you visit special:random, MediaWiki doesn't send a page - instead it returns an HTTP response "302 moved temporarily" header, with a "location" set to the actual randomly-selected page. Whether the browser should add special:random to its history is questionable - in this case that would be welcome, but in others (where the website owner uses 302s extensively, for housekeeping purposes) it'll just fill your history up with entries you'd consider to be junk. -- Finlay McWalter • Talk 17:47, 3 March 2010 (UTC)

In Firefox 3.0.18, the page pulled back shows at second position in the history, with Special:Random appearing at the top. In IE7, the page pulled back shows (IE history function is a pain in the ass) but Special:Random doesn't. --Redrose64 (talk) 18:01, 3 March 2010 (UTC)

In webkit this issue is bugzilla ticket 7814. I'm trying to get this bug reopened, because it's clearly not fixed (at least not for Safari). —TheDJ (talk • contribs) 18:39, 3 March 2010 (UTC)

Beginning space in paragraph makes <(pre)> tag?

Why?  —Preceding unsigned comment added by Gloomofdom (talk • contribs) 20:38, 3 March 2010 (UTC) 

It's designed to. See Help:Wiki markup#Limiting formatting / Escaping wiki markup. --Redrose64 (talk) 20:44, 3 March 2010 (UTC)

If you want to indent a line, use the colon per the chart. If you want to preserve formatting without the box, use <poem>...</poem>. ---— Gadget850 (Ed) ^talk 20:52, 3 March 2010 (UTC)

It's note quite a pre tag, mind you. –xeno^talk 20:55, 3 March 2010 (UTC)

WP:SONG

The WP:SONG redirect seems to be redirecting to WikiProject Songs correctly. But if I type WP:SONG into the search field on the left column, then I'm taken to some humor page. I'm not comprehending this. Can someone fix it, please? It's hampering my style. Thanks! – Kerαunoςcopia^◁_galaxies 22:56, 3 March 2010 (UTC)

It's going to WikiProject Songs for me. Did you type WP:WPSONG by accident?♫ Melodia Chaconne ♫ (talk) 23:11, 3 March 2010 (UTC)

(edit conflict) As you type in WP:SONG a scrolly list will appear showing potential matches. When you enter the "G", you'll find that the top two entries are Wikipedia:Song and Wikipedia:SONG. Mouse over the relevant one, and click it. They are two different entries; WP:SONG is a redirect to Wikipedia:WikiProject Songs, and other shortcuts to that are available: try WP:SONGS; this may be entered in the search box without confusion, since the name is unique. --Redrose64 (talk) 23:13, 3 March 2010 (UTC)

Actually now that I look again, I think the culprit is wp:song (or even WP:Song) vs. WP:SONG. Note the caps difference. It's an unfortunate effect from making "WP:" into the same thing as "Wikipedia:", but apparently a redirect that existed beforehand overwrites it. This should probably be fixed. (And incidently, you can turn the auto-suggest off, so assuming that the OP was doing that isn't necessarily fair)23:16, 3 March 2010 (UTC)

Special characters

Maybe this is just an irritated rant, but who in their right minds thought it was a good idea to sort the special characters in the insert list by the letter as opposed to by the accent marks? It's not exactly as though someone who's writing articles about Slavic topics is going to need umlauts or cedillas, while someone editing about German or French topics equally won't need carons or bars. The system was much more efficient when all of the characters were grouped by accent mark type (i.e., all of the vowels with accents aigu together) instead of having to hunt through all of the "e"s to find the right one. That, and for some reason there are two eszetts, thorns, schwas, and eths listed. This really should be fixed. Parsecboy (talk) 23:02, 3 March 2010 (UTC)

MediaWiki talk:Edittools#Mixed case ordering and character identification. Algebraist 23:09, 3 March 2010 (UTC)

Thanks for the link. Parsecboy (talk) 00:12, 4 March 2010 (UTC)

I think the relevant change was in fact discussed in a slightly earlier section on that page: MediaWiki talk:Edittools#Rational and friendly ordering for the Latin characters. To limit confusion, I've put a sub-section heading in front of your comment: MediaWiki talk:Edittools#Accent versus alphabetical ordering. Replies can then follow your comment there without getting mixed up with a largely unrelated aspect of the discussion. — Richardguk (talk) 01:11, 4 March 2010 (UTC)

Malware in image?

While wading through the backlog at Category:Non-free Wikipedia file size reduction request, I came across the image: File:Jcl header.jpg. I downloaded it, then attempted to open it with Photoshop to reduce the resolution. At that point Photoshop froze. Following that I opened it with MS Paint with no problem. I reduced the resolution, then attempted to save the file. When I tried, my anti-virus software popped up and said it contained malware, and deleted it. Now I know there are plenty of false-positives to anti-virus software, but is this even possible? Could someone else take a look? Jujutacular ^T · C 00:58, 4 March 2010 (UTC)

I ran it through http://www.virustotal.com (of which I have no knowledge other than Google found it) which says it runs 42 different malware checkers on uploaded files. It didn't report an issue with that file. -- Finlay McWalter • Talk 01:07, 4 March 2010 (UTC)

What anti-virus are you using? Svick (talk) 01:10, 4 March 2010 (UTC)

McAfee. Jujutacular ^T · C 01:14, 4 March 2010 (UTC)

Well, I solved the problem by doing a print-screen and cropping it. Hopefully if there is malware in it, it will be gone when the higher resolution version is deleted. Thanks for taking a look. Jujutacular ^T · C 01:56, 4 March 2010 (UTC)

Most convenient way to compare sandbox to production?

For templates with a sandbox version, I often want to compare the sandbox version to the production version to see if there are any pre-existing changes changes before I test one. I'm doing it clumsily now: I copy the production source, edit the sandbox, replace all the text by pasting from the clipboard, and then click "Changes". Is there a better way, and in particular, is there a way to get WP to compare the two pages? If so, can one of the template documentation templates be modified to include such a link? — John Cardinal (talk) 15:20, 27 February 2010 (UTC)

When you go to the diff page, in the url (web address) you will see at the end there is a string which looks like "&diff=x&oldid=y". The y value is the page version on the left, and the x the page version on the right. Each edit has it's own "version number". So you can simply copy the most recent version number of the production page, and compare it to the most recent for the sandbox page. Hope that helps. - Kingpin¹³ (talk) 15:26, 27 February 2010 (UTC)

Wow... does that work for any pair of pages? Say there are two articles with a lot of similarity except a few names, could I compare them in this way? --Redrose64 (talk) 15:42, 27 February 2010 (UTC)

Indeed it does ! —TheDJ (talk • contribs) 15:57, 27 February 2010 (UTC)

Thanks. I didn't know that page versions were unique across pages. Makes sense, I guess; I assume the version number is the DB's primary ID for pages.

Is there a way for a template to get the current version number of a page so that it could build the diff link automatically? — John Cardinal (talk) 16:23, 27 February 2010 (UTC)

Left-hand side, toolbox, permanent link (in Monobook, anyway; different skins may place this elsewhere). --Redrose64 (talk) 16:30, 27 February 2010 (UTC)

Wanna script? Here you go:

FlexDiff script

function getID(){
if(document.getElementById('t-permalink')){
return document.getElementById('t-permalink').innerHTML.split("&oldid=")[1].split("\"")[0]
}else{return "No id" 
} 
}
addOnloadHook(
function(){
var diffPortlet='p-cactions'
/*can be:
p-navigation for navigation sidebar
p-tb for toolbox
p-interaction for interaction sidebar
*/
if(window.location.href.indexOf("action=edit")==-1){
addPortletLink(diffPortlet,"javascript:prompt('The Id for this page is:',getID())",'ID', 'ca-id','Page ID', '')
addPortletLink(diffPortlet,"javascript:document.location='http://en.wikipedia.org/w/index.php?title=Main_Page&diff='+getID()+'&oldid='+prompt('Enter ID for other page','')",'FlexDiff', 'ca-diff','Get diffs for this page with any other page', '')
}
} 
)

Add it to your monobook js page (go to vector.js if you use beta) and purge your cache. You'll get two new tabs (or links in your chosen portlet if you change the diffPortlet variable). In beta, the tabs will be hidden under a little down arrow next to the history tab and 'watch this page' star.
To get a diff, first open the page you want on the left of the diff, and click ID. Copy the ID, and open the page you want on the right. Click the FlexDiff button (It should be right next to/under the ID tab), and it will ask for the ID that you copied. Paste it and click OK. The diff will open in the same window (no need for a new window as the script doesn't work on edit pages). Enjoy!
P.S. Isn't it weird that the title parameter in diffs doesn't matter? this has title=Main_Page, while this has title=Wikipedia:Don't_stuff_beans_up_your_nose. Both link to the same diff.ManishEarth^{Talk • Stalk} 03:26, 28 February 2010 (UTC)

Thanks. I'll experiment with that. What I was thinking, though, was if a template could get the ID number of a page, then the sandbox page or the doc page could get the two IDs of interest and include a link to compare them. Knowing you can compare different pages helps a lot, but I was looking for the icing on the cake. — John Cardinal (talk) 04:28, 28 February 2010 (UTC)

Well, there is a magic word for the revision id. It's {{REVISIONID}}. For example, the revid for this current page is -. But I don't think there's any way to get the revision id of another page without adding a magic word which uses AJAX (I doubt this will happen).ManishEarth^{Talk • Stalk} 08:43, 28 February 2010 (UTC)

Why would it have to use AJAX? {{PAGESIZE}} doesn't. T8092. OrangeDog (τ • ε) 21:55, 1 March 2010 (UTC)

PAGESIZE gets the page size of the current page. Every time you save, the displayed value is updated. Same with REVISIONID. But, to fetch the ID of another page, you either have to use AJAX or update the displayed value on all pages with the magic word every time you save. ManishEarth^{Talk • Stalk} 10:28, 2 March 2010 (UTC)

My point being that 326,462 works just fine. All you need to do is purge. OrangeDog (τ • ε) 12:35, 2 March 2010 (UTC)

It does? Maybe we could request the folks at MW......ManishEarth^{Talk • Stalk} 14:18, 2 March 2010 (UTC)

FYI: rev:49575 and rev:51424. --Splarka (rant) 08:06, 3 March 2010 (UTC)

Hmm, looks like it won't happen, as it's been lumped together with some other less-than-sensible changes. OrangeDog (τ • ε) 12:50, 4 March 2010 (UTC)

Practical side of removing autoblock

I maintain a declared alternate account for use on public computers. Since I'm an admin, I don't need to worry about autoblocks, but my alternate was autoblocked until I made a request for autoblock exemption in January; see the log if you're curious. Let's say that some time down the road, I'm using the backup account on a computer that has a keylogger, and the operator decides to start using my account. Will it be as easy for an admin to remove the autoblock as it is to block the account or to grant exemption in the first place? Except for blocking vandals, I almost never do anything with user rights, and I don't want to mess with someone's rights just to find out. Nyttend (talk) 22:35, 2 March 2010 (UTC)

According to Special:ListGroupRights, admins can both add and remove the IP block exemptions group. So it can easily be removed if necessary. Reach Out to the Truth 18:01, 4 March 2010 (UTC)

And they could just directly block the account, which IPBE won't affect. –xeno^talk 18:03, 4 March 2010 (UTC)

Table sorting

When clicking the column headers of the table at List of the largest trading partners of the United States, the data are sorted alphabetically or reverse-alphabetically, rather than being sorted numerically; so a sorted list could be "1", "200", "3.5", and "4,000,000". Is there a way to tell the table to sort numerically? Comet Tuttle (talk) 00:37, 4 March 2010 (UTC)

Yes, don't use non-numbers in table fields. —TheDJ (talk • contribs) 00:45, 4 March 2010 (UTC)

I have fixed it in [1] by blanking cells with no number. See more at Help:Sorting. PrimeHunter (talk) 01:10, 4 March 2010 (UTC)

TheDJ is right, but... is there a way to numerically sort when there are non-numerical characters in the field (such as references)? It'd be quite handy to be able to do that. Fences&Windows 19:17, 4 March 2010 (UTC)

I had that problem a few weeks ago, and I ended up moving the non-numeric data to a separate column. --Redrose64 (talk) 19:34, 4 March 2010 (UTC)

Alternatively, just use {{Sortkey}} to insert whatever sortkey you want. --BrownHairedGirl (talk) • (contribs) 19:39, 4 March 2010 (UTC)

Either {{sortkey}} or {{ntsh}}. So, in this case, {{ntsh|0}}—.—NMajdan•talk 22:47, 4 March 2010 (UTC)

Regex for cite news

I've made some regex for picking up {{cite news}} template in references...but its not very good. Anybody have any suggestions (or know where I can find regex for picking up references)?

<ref( name[ ]?=[ ]?"+.{1,35}"[ ]?|)>([ ]?){{cite news([ ]?)[|]+.{2,400}?}}[ ]?</ref>

Smallman12q (talk) 02:05, 5 March 2010 (UTC)

I would use something like <[Rr][Ee][Ff][^>]*>\s*{{\s*[Cc]ite\s*news[^}]+}}\s*</[Rr][Ee][Ff]\s*>. It won't work, if there is other template inside the {{cite news}} (more precisely, closing brace), but I think I've never seen that. You might want to change it to allow the <ref> to contain other things than just the template (e.g. I put {{dead link}} there). Svick (talk) 02:41, 5 March 2010 (UTC)

ImageAnnotator gadget on Wikipedia

I would like to propose for ImageAnnotator gadget to be installed on on Wikipedia. It was already discussed once discussed once here. ImageAnnotator gadget is a very popular gadget at Wikimedia commons where it is allowing users to provide more details to specific parts of the image. For example: name objects (and provide links) in panoramas, identify people in group images or highlight hard to see important details of the images. At the moment over 14k images on commons are annotated, for example this one. Please notice annotations identifying people available at Commons but lost on Wikipedia page. I would like to propose to install ImageAnnotator gadget on Wikipedia initially in the File: namespace only to allow more people get familiar with it. That would have an effect of making pages in file namespace on Commons and Wikipedia to carry the same information. --Jarekt (talk) 14:53, 1 March 2010 (UTC)

I have one big problem with this gadget, and that is the enormous amount of JS that it requires. To have such a tool enabled by default and distributed on all pages... I can already hear David complaining with his dialup connection and old computer. —TheDJ (talk • contribs) 16:37, 1 March 2010 (UTC)

I think it is possible to disable the gadget to individual people. --Jarekt (talk) 20:40, 1 March 2010 (UTC)

Oh definitely, but people who actively disable something are not our biggest audience :D It's just something that has me worried. Are there other wiki's besides Commons that have already deployed image-annotator ? —TheDJ (talk • contribs) 21:17, 1 March 2010 (UTC)

Code size should not be such a big issue in practice. Once the code is cached client-side, it wouldn't need to be reloaded on every page view. The code also tries to make as few requests as it can. It's been deployed so far only at the Commons and at the German Wikiversity. AFAIK, the Hungarian Wikipedia has an older version as a gadget. Lupo 11:22, 2 March 2010 (UTC)

Personally, I don't find it especially useful on Commons, and it seems like it would be even less useful here. Mr.Z-man 22:11, 1 March 2010 (UTC)

As I posted in the last thread, my experience is that a substantial portion of "annotations" are useless test edits or "hi mom" stuff that doesn't get cleaned up - and of course if those are loaded here there's no immediate way to clean them up from this wiki. Another point that's been made is that Commons is not an encyclopedia and has only a limited NPOV policy, and some of the annotations are not really appropriate to display here for those reasons. If those hurdles could be overcome, it might be worth it, but I don't see that there's been any attempt to do so. — Gavia immer (talk) 23:30, 1 March 2010 (UTC)

Having said that, we could make the gadget available in the prefs for those who want it of course... —TheDJ (talk • contribs) 00:37, 2 March 2010 (UTC)

I'm not comfortable with automatically importing annotations from commons, because annotation changes make in commons won't show up on wikipedia article editors' watchlists. Identifying people in pictures is a potential BLP issue (e.g. imagine a vandal adding names of living people to pictures like File:Bloody Sunday-Alabama police attack.jpeg). Anyway, the popups could be a magnet for spamming external links onto images.

Technically I quite like parts of it, e.g. using frameless images and animations as popups. But we don't generally use large images or panoramas in wikipedia articles, and the only mainstream use I can see is if we decide to have a new type of encyclopedia page that is principally an image. Because that sort of page would be maintained on wikipedia (rather than on commons) all the usual watchlisting, NPP, neutrality and verifiability processes would limit the potential for problems. That model could be helpful for annotating details in large images such as diagrams, maps, aerial views (example) and maybe group photographs. - Pointillist (talk) 02:57, 2 March 2010 (UTC)

I would love to have it on WP, only that I would like there to be a little 'hide' button in the corner (which itself hides when moused over), which would remove annotations for that image so that it can be seen clearer. (The annotations sometimes bug you especially in an intricate diagram, and removes the beauty from some images)ManishEarth^{Talk • Stalk} 13:06, 2 March 2010 (UTC)

Lack of control over annotations text, different POV and living people policies and similar issues are the same as with any other text in the image description. Wikipedia does not have any control over text in the information template either. To my knowledge it is not a (big) problem now and it should not be any different with other forms of description. What worries me is that, since this tool was introduced a lot of description of image content is done using annotations which are not available to the users on Wikipedia. In the past I and many other people adding images to Commons were much more likely to have in the description field of the information table comments like "9th head in the 5th row (with glasses) was John Doe" (se an example) now we just use annotator with the link to John Doe. It makes live of both parties so much easier, but users on Wikipedia do not even know that more information exist. For example see annotations on this image highlighting hard to see text. We could phase it in first for the users that opt-in and in the future for everyone except for those that opt-out. --Jarekt (talk) 18:51, 2 March 2010 (UTC)

Sorry, Jarekt, but I'm still not convinced that commons annotations should be displayed in the encyclopedia. I can see the value of a zoom tool (see my sandbox experiment), but in your example you are mostly annotating things that aren't important to the encyclopedia, or don't need annotation (see below). IMO there should be a formal proposal, including a list of use cases with examples, to consider whether this gadget would add value to Wikipedia. - Pointillist (talk) 02:10, 3 March 2010 (UTC)

commons:File:Pelion Range from Mt Oakleigh.jpg, commons:File:Lake Pedder From Mt Eliza.jpg, commons:File:Canberra From Black Mountain Tower.jpg are some examples of where this would actually be useful. I don't care for article annotation, just image pages here. Noodle snacks (talk) 08:00, 3 March 2010 (UTC)

I agree. Context is probably the main issue: for example, the annotations in your File:Lake Pedder From Mt Eliza.jpg panorama would be useful in the Southwest National Park article, but unnecessary in Lake Pedder. Adding annotations to a picture on a per-article basis – within wikipedia, not commons – would make more sense than just importing the commons annotations into every article that uses an image. - Pointillist (talk) 12:14, 3 March 2010 (UTC)

They should be by default disabled, and only enabled if you set a param annote=true or somethin like it ([[File:Example.jpg|80px|annote=true]]). There should also be a little cross which removes the annotations from the image temporarily so that it can be seen clearly.ManishEarth^{Talk • Stalk} 14:22, 3 March 2010 (UTC)

I like Manish's parameter idea, especially if annotations could be grouped like footnotes, so you could select a subset of annotations for a specific article. The editor who adds the parameter would then take responsibility for watchlisting the commons page. - Pointillist (talk) 15:06, 3 March 2010 (UTC)

As I mentioned before I was lobbying for annotations limited to File namespace only with possibility of extending it to other namespaces in the future. Those would be visible only after user clicks on the image in the article. I agree that whole-sale inclusion of all the annotations from Commons to articles in Wikipedia might be disruptive and should be avoided. One reason why that could be a problem is that since Commons is a multilingual project the annotations are also in many languages (just like file descriptions) and annotations in different languages would not be helpful on Wikipedia. Since this gadget allows many customization we could separately allow local annotations within wikipedia articles. Those annotations would be fully controlled within Wikipedia and would allow annotating images used in articles with text stored within article wiki code. For those we could have locally agreed on style recommendations and when-to-use guidelines just like we have on Commons. If we agree that such limited release is a good idea than we could verify with User:Lupo that current version of the software supports it. --Jarekt (talk) 18:33, 3 March 2010 (UTC)

Annotations using zoomed sections of image with link(s). The entire panorama is 22MB, this zoomed section is only 620KB.

Thanks for clarifying. As I said earlier, I'm not against all types of commons annotation, e.g. if someone has identified landmarks, vegetation etc in a commons image (e.g. one of Noodle snacks' panoramas) the annotations could be useful in all the localized versions of the encyclopedia, especially if they simply link to relevant articles with no need for language-specific text, like the Scotts Peak screenshot.

The issue is how to bring across annotations from commons on a discretionary basis: deleting some, keeping others and adding new local annotations. Do you have a view on this? - Pointillist (talk) 23:17, 3 March 2010 (UTC)

My position mirrors that of Jarekt (file namespace only, at least at first). I'm not sure how to handle the local annotations etc. We could just use the local image page for annotations here, ignoring the ones at commons. A language (or enwiki hide) parameter for the commons ImageNote template might be an option. Noodle snacks (talk) 10:43, 4 March 2010 (UTC)

It's better to have it centralized at commons and usable at all namespaces. I agree with Pointilist that there should be grouping of tags. Maybe a function to add an annotation only on that instance of the image like an imagemap: [[File:Example.jpg|23px|annote1=title;description;coordinates|annote2=...]] (Yes, I'm going a bit too far, lets get back to the topic). Its best if everything is centralized from the start, as it will be centralized sooner or later, and then there will be problems like the many-one relationship between files on all wikimedias and file names, with two different files with the same filename, on two different wikis. ManishEarth^{Talk • Stalk} 11:08, 4 March 2010 (UTC)

It seems to me we need to check the technical feasibility of parameterized filtering (with a default of "no inclusion") before considering image use outside the xx:file: namespace. If possible the technical discussion should include other points of view e.g. from Mr.Z-man and TheDJ. Once there is a consensus on solutions to the technical issues, the policy questions should be raised elsewhere (presumably VPP). In the meantime, it might make sense to prepare some use cases and really-well-thought-through examples. - Pointillist (talk) 12:04, 4 March 2010 (UTC)

I can try to start the user cases and examples, but since I mostly work on Commons I am unfamiliar with the process. Can someone point me to relevant examples and suggest there to place it? --Jarekt (talk) 16:24, 4 March 2010 (UTC)

This will take some time and resources to achieve. Once you know the technical side is OK, I suggest we start by listing possible cases where annotation might deliver benefits to wikipedia readers. Editors who work on featured articles and pictures should be consulted, but here are some possibilities to start with:

• An aerial view of a city – to identify buildings and parks.
• A panorama annotated with zoomed sections – to identify natural features.
• A photomicrograph of an animal cell – to identify organelles.
• A group photograph – to identify the individuals pictured, maybe with doubt about accuracy (as in the Warsaw Ghetto Uprising image).
• A picture of a painting annotated with zoomed sections – to tie commentary to specific detail in the painting.
• A map of a battle – to provide notes of the dispositions, strengths &c of combatants' units.

...and so on. Each of the cases will need an article that would benefit, e.g. perhaps Southwest National Park would benefit from linking to annotations in the File:Lake Pedder From Mt Eliza.jpg panorama. IMO you don't need many examples but they have to be very compelling to readers, so there will have to be some triage at this stage. Next, for each of the shortlisted examples we need to do a great job of preparing the annotations, zoomed sections etc. Once all that is in place you can make a proposal at VPP that the annotator should be enabled for Wikipedia's file: namespace. Does that work for you? If so, it probably needs a wikiproject. I will help, but I am still skeptical so someone else will have to lead! - Pointillist (talk) 17:42, 4 March 2010 (UTC)

To unclog this pump, and also to make a page for consensus, I've made a page at Wikipedia:Village_pump_(technical)/Annotate. ITS INCOMPLETE IT still requires a detail description of the gadget and of the disputed points. Feel free to add some disputed points if I have overlooked any, and some suggestions for the optional extras. ManishEarth^{Talk • Stalk} 01:37, 5 March 2010 (UTC)

Thank you for the stub. I started filling in and researching more current capabilities and customization options of the gadget.--Jarekt (talk) 05:38, 6 March 2010 (UTC)

See the talk page for the current discussion with Lupo. He's decided to implement it on enwiki in a restricted form. ManishEarth^{Talk • Stalk} 06:12, 6 March 2010 (UTC)

Bouncing edit window

This refers to [2]. The computer has Windows 7.

Today, while on the same computer I was on when I posted that, I was adding to something about halfway down the part of the section that was showing in the window. Once I started typing, the line I was adding to jumped to the bottom.Vchimpanzee · talk · contributions · 21:17, 4 March 2010 (UTC)

Later, I was typing and the line jumped down to two lines above the bottom of the edit window. Just now I copied and pasted but couldn't see what I had done since it had jumped below the bottom of the window.Vchimpanzee · talk · contributions · 21:51, 4 March 2010 (UTC)

Ohhh, my edit window did that too today a few times. Fences&Windows 22:39, 4 March 2010 (UTC)

I'd like to mention that Firefox 3.0.18 with Windows XP with Monobook do not give such effects, so which browser, operating system, skin are you using? --Redrose64 (talk) 22:44, 4 March 2010 (UTC)

I had that problem too in FF3.6 in Vista a week ago...I cleared cache and restarted and it went away ^.^ Smallman12q (talk) 02:07, 5 March 2010 (UTC)

Windows 7, Firefox 3.6, Monobook skin. Fences&Windows 14:54, 5 March 2010 (UTC)

On the computer I was on yesterday, it was Explorer, but I refuse to use Firefox except at the one library where that's what they have. There are things I don't like about it. It appears firefox was an option on the computer I was on, but for various reasons I choose Explorer.Vchimpanzee · talk · contributions · 20:36, 5 March 2010 (UTC)

Commons edit keys

Moved from Wikipedia talk:Village pump#Question. Svick (talk) 02:48, 5 March 2010 (UTC)

I was just wondering if there's a way, setting for a user to enable or program the edit keys that commons has at the bottom of thier edit screens ? Mlpearc MESSAGE 02:25, 5 March 2010 (UTC)

We do have that, except that they're links and are categorized (Looks much neater). If you scroll down, you'll see a dropdown box (By default it says 'Insert'), and next to it, this stuff: – — ‘’ “” ° ″ ′ ≈ ≠ ≤ ≥ ± − × ÷ √ ← → · § Sign your posts on talk pages: ~~~~ Cite your sources: <ref></ref>.

If you choose 'Wiki Markup' from the dropdown, you'll get the same thing as in commons, except in link form. ManishEarth^{Talk • Stalk} 12:49, 5 March 2010 (UTC)

• • Aah.... Very good Thanx Mlpearc MESSAGE 21:24, 5 March 2010 (UTC)

Anyone with an eye more Keen than I care to figure out what this is about?

Resolved

 – Cite web syntax fixed. Pointillist (talk) 10:21, 5 March 2010 (UTC)

I'm honestly stumped by this one. My article has 90+ references, but only 57 show up in the reflist at the bottom. What's up? - ʄɭoʏɗiaɲ ^τ _¢ 09:51, 5 March 2010 (UTC)

I don't know why, but it seems that citations 1 to 33 point to footnotes 1 to 33, citations 34 to 78 don't have footnotes, and citations 79 onwards point to footnotes 34 onwards. Hope that helps pinpoint the specific problem. - Pointillist (talk) 10:16, 5 March 2010 (UTC)

Fixed now. The closing }} were missing from the {{cite web}} in ref name="AI75". - Pointillist (talk) 10:21, 5 March 2010 (UTC)

I figured it was something like that. Thank you! - ʄɭoʏɗiaɲ ^τ _¢ 17:10, 5 March 2010 (UTC)

Temporary Autoblocking of Probable Vandalism Only addresses

Discussions are on here ▒ ♪ ♫ Wifione ♫ ♪ ▒ _―Œ ^{♣Łeave Ξ мessage♣} 19:42, 5 March 2010 (UTC)

Template talk:Cite interview#Archivedate, archiveurl

Does anyone know an easy way of including the |archiveurl= and |archivedate= parameters on Template:Cite interview? Further details are at Template talk:Cite interview#Archivedate, archiveurl. Thanks, Cunard (talk) 21:10, 5 March 2010 (UTC)

Alphabetizing in a Category list

Just curious: Why is Paula Froelich alphabetized under "P" instead of "F" in Category:American Jews? (The other entries are sorted by surname.) Casey (talk) 22:37, 5 March 2010 (UTC)

Articles about people aren't automatically sorted by their last name, you have to manually set a sort key. The best way is to use the {{DEFAULTSORT}} magic word, like I did on the article you mentioned. See also Help:Category#Sort order. Svick (talk) 23:15, 5 March 2010 (UTC)

Seeking to create and maintain an up-to-date list of all enwiki page titles

I am seeking to create and maintain an up-to-date list of all enwiki page titles, in furtherance of the RPED extension project. The gist of this part of the project is that once I have a database table of those page titles, I will set up a server that will provide a download of the database to wiki owners, and that will provide those wikis with incremental updates every x seconds to reflect page titles added or removed since the last update.

Anyway, the first step is to get ahold of those page titles. I started out by downloading the 2010-01-16 list of page titles, all-titles-in-ns0.gz and feeding that into my file-reading script. Then I updated my database table to reflect all the page deletions and restores since 2010-01-16, using this script. Then I tried to use the RecentChanges API to get all the new pages since 2010-01-16 by means of this script, but that didn't work because RecentChanges only goes back a few days. (I have since modified that script so that at least it can keep the table up to date with new pages as they are being added in real time.) Special:NewPages only goes back a month, so it can't give me everything back to 2010-01-16.

I considered querying the API for all pages, but that would probably require about 15,000 queries (of 500 page titles apiece). Even though this series of queries would be a one-time thing, if there is an alternative that would put less load on the server, I would prefer to go that route. Please let me know if you have any ideas.

I suppose another option is just to keep getting updates on new, deleted, and restored pages in real time and await the release of the data dump of page titles for mid-February. Then I can splice those two data sources together to get a full, up-to-date page title list. Or is there another, faster way? Thanks, Tisane (talk) 02:00, 6 March 2010 (UTC)

You might also be able to get the initial list of pages by requesting a query from the Toolserver, a live copy of the Wikipedia database. Graham87 07:32, 6 March 2010 (UTC)

If you do want such a list from the toolserver - and it's to be a one-off starting point - you'll find it easier to just request one here. - Jarry1250 ^{[Humorous? Discuss.]} 12:24, 6 March 2010 (UTC)

The best way would probably be to wait for the next update of the titles list, catch up to the current using the API, then do realtime updates using the raw recentchages IRC feed. Mr.Z-man 01:36, 7 March 2010 (UTC)

WP Foundation Forcing Signup using fake "technical problems"?

I used to be editing under IP for reasons I may or may not prefer to detail here.

Anyway, I've recently seen lots of technical failure when doing so, which previously wasn't the case. Specifically, attempts to "Show Preview" or "Show Diff" often resulted in an empty preview (and you bet I would't have liked to "Save" that, either) instead of a proper preview, and attempts to "Save" often resulted in an empty preview as though I hadn't pressed "Save" but as above.

This tech problem has not been around for a long time so I do wonder whether it is fabricated. Seeing as WP is in bed with Google and Google has gotten even more data slurpy and privacy unsafe than it ever used to be (just research "Fuck you Google"), I doubt that this is merely coincidential.

So for now I somewhat signed up by coercion (doubting whether I will use this account for long; no discussion on this please.)

Solid Information anyone?

This issue may get slashdotted, depending on how it clears up or which it proves to be.

-- Telimane (talk) 15:18, 6 March 2010 (UTC)

Update: Just experienced the same tech prob while logged on, with this precise page, so WTF. (Firefox 3.6, unlimited Cookies and JS access granted for all of wiki*.org). Questions posed still apply. -- Telimane (talk) 15:25, 6 March 2010 (UTC)

Basically, there are many possible reason for a problem you are describing. Do you have any extra toolbars or extensions enabled? It is known that the google toolbar for instance, adds data to pagerequests, that can break a submit. Another thing that can be causing this, is if your connection is unstable. Or it might be that your provider has a proxy server that isn't too reliable. —TheDJ (talk • contribs) 15:34, 6 March 2010 (UTC)

If you are worried about data slurping, you're better off registering an account. The account setup requires you to reveal nothing at all about yourself, and using the account means that your IP is not displayed. You suggested that you have other reasons for not registering, and as requested I won't pry ... but if you want to protect your privacy, the IP is a bad choice. --BrownHairedGirl (talk) • (contribs) 15:59, 6 March 2010 (UTC)

This tech problem has not been around for a long time so I do wonder whether it is fabricated. - Um, the problem being new, and unreported (until now) is more evidence that there's an actual problem rather than a conspiracy. If it happened on the same day that Google gave Wikimedia money, I might be able to see it as even fitting the definition of coincidental, but the grant from Google was announced more than 2 weeks ago. If you consider "within a month" to be "coincidentally similar", then you're probably going to be seeing conspiracies everywhere. Perhaps filing a bug report would be more useful than ranting? Mr.Z-man 01:31, 7 March 2010 (UTC)

Editing Question

I am having a problem with the option at the bottom of an edit window, under the "Save Page" / "Show Preview" buttons there is a dropped down window for edit tools with "insert" being the default, my problem is I want to set this option to Wiki markup. I have changed the option many times, the first time it displayed the tool bar, now when I change the option to Wiki markup and nothing happens, the tool bar does'nt show. Also, when I get this working, is this somethig I'll have to set at each edit or once set will become the default ? which is my objective. Thanks Mlpearc MESSAGE 17:34, 6 March 2010 (UTC)

Not sure what the problem is that you are describing.

When you say you "changed the option", do you mean you changed the item in the dropdown list (Insert, Wiki markup, Symbols, etc)? Or do you mean you changed something in your Wikipedia settings (Special:Preferences)?

There was a change to the insert code yesterday, but I think that is not directly related to your problem. But it may have led your browser to download the new code which in turn might have been interrupted and caused the code to stop working properly for you.

If your web browser allows cookies, the dropdown list should default to whichever option you chose last time (Insert, Wiki markup, Symbols, etc).

It might help to clear your browser cache in case the code loaded incompletely. See Wikipedia:Bypass your cache for instructions.

If clearing the cache does not work, it might help to describe the problem more fully and to state:

• what browser and operating system you are using
• whether you are using the "enhanced editing toolbar" (this is chosen in your Preferences under Editing – Experimental features)
• whether you have opted in to the Usability Initiative beta
• any Wikipedia gadgets or custom javascript or CSS you have installed.

— Richardguk (talk) 21:18, 6 March 2010 (UTC)

• • OK, for future reference: first, I was refering to the Dropdown window not Preferences. Clearing my cache did not work (at that moment), so I started to look for the answers to your questions, I went to Special:Preferences, looked at my edit settings the first thing to catch my eye was the "show edit toolbar"(req.java) I knew this was for the bar at the top of the edit window(wich I don't want to loose) anyway I unchecked it, came back to the dropdown option and selected Wiki markup and the toolbar appeared, but now I don't have the bar at the top, so I went back and checked the box for "show edit toolbar"(req.java). Anyway what fied the problem I don't know but all's well. Thank you for your time, sorry for the rambling and Happy Editing Mlpearc MESSAGE 22:09, 6 March 2010 (UTC)
    • most likely you had a corrupt cookie in your browser - unsetting and resetting the preference would have rewritten the cookie. glad it worked out. --Ludwigs2 22:32, 6 March 2010 (UTC)

Spamlist auto-alarm

Not sure who to talk to about this one. Since Wikipedia:Bug reports and feature requests says to bring it up here when one is not sure, here we are. Recently there was edit warring going on over in the France article. An editor was able to repeatedly add a website to the Reflist that was on the spamlist. When I tried to use that same spamlink on the Talk:France#FRENCH . . . page, the spam alarm went off and did not allow me to add it. (That's how I found out the link was on the spamlist.) So apparently there is a bug that may allow some editors to add spamlist websites to article namespace without receiving the spamlist auto-alarm and without their edit being disallowed by that alarm. If I'm not in the right place, then please point me in the correct direction.

• Ref. Wikipedia's SPAM list! – the "\binternationalliving\.com\b" entry
• Ref. This edit as one of several examples when the editor was able to edit the International Living spamlink into the France article
  

 —  Paine (Ellsworth's Climax)  22:32, 6 March 2010 (UTC)

It's because the spam blacklist works only for URLs (usually starting with http://), so the edit wasn't blocked, because it didn't actually create clickable link. Svick (talk) 22:53, 6 March 2010 (UTC)

You're right. I went back and checked each case, and each time the editor began the non-link with "www" instead of "http://". Thank you very much, Svick!

 —  Paine (Ellsworth's Climax)  01:27, 7 March 2010 (UTC)

{{quote}}

Here external links are not working. Fix please.100110100 (talk) 23:58, 6 March 2010 (UTC)

Looks like editor Father Goose fixed this for you. The use of some symbols, such as the equal sign (=), will sometimes do weird things to the Quote template. In this case, the {{=}} did the trick. Always use it within any template when you need to type in an equal sign.

 —  Paine (Ellsworth's Climax)  01:43, 7 March 2010 (UTC)

Requires Answer

Resolved

Hopefully someone can fix this.174.3.110.108 (talk) 05:31, 7 March 2010 (UTC)

HTML

The syntax in all these templates should appear the same:

• {{Commons}}
• {{Commons category}}
• {{Wikispecies}}

There shouldn't be any reason they are different, right?174.3.110.108 (talk) 05:36, 7 March 2010 (UTC)

The graphics and wording are different, and this can have unpredictable effects depending on the browser, default machine font, wikipedia skin, and other issues. They should all look pretty much the same, yes, but there may be variations. --Ludwigs2 06:02, 7 March 2010 (UTC)

Extra Space

There is extra space at Langia. Can someone fix this?174.3.110.108 (talk) 05:42, 7 March 2010 (UTC)

The one between the wikispecies tag and the stub tag? Fixed it. You could have done it yourself, even non-logged in users can edit most pages. 59.183.141.117 (talk) 06:20, 7 March 2010 (UTC)

Photo on talk page covers content

I found a solution to this once before but a search of the Village Pump archives didn't show what I had done, at least not on the early pages of the search.

User talk:Scarce has a bear in the last topic. I created a temporary solution, which was the second-to-last topic, but I don't know how to fix it permanently. Also, the topic I created shows that when I added to the topic, the computer I was on didn't show the bear.Vchimpanzee · talk · contributions · 19:25, 24 February 2010 (UTC)

Sorry, what? OrangeDog (τ • ε) 13:56, 25 February 2010 (UTC)

I can't see the bear today. I'm on a computer with Windows 7. The details of my home computer, where I can see the bear, are here.Vchimpanzee · talk · contributions · 21:58, 25 February 2010 (UTC)

What bear? That discussion you linked to is about previous problems with your computer and Yahoo, not with Wikipedia, and nothing about a bear. OrangeDog (τ • ε) 23:05, 25 February 2010 (UTC)

Perhaps it's a bearelephant in the room? --Redrose64 (talk) 23:13, 25 February 2010 (UTC)

Orangedog, you have to scroll down within the discussion to see what my computer is.Vchimpanzee · talk · contributions · 21:48, 26 February 2010 (UTC)

I just checked the page and there's no bear, though there's nothing in the history to indicate it was deleted.Vchimpanzee · talk · contributions · 21:51, 26 February 2010 (UTC)

WHAT BEAR!? OrangeDog (τ • ε) 00:11, 27 February 2010 (UTC)

Was it perhaps this bear or maybe ? Peachey88 ^{(Talk Page · Contribs)} 00:20, 27 February 2010 (UTC)

I think it was the second one.Vchimpanzee · talk · contributions · 19:13, 27 February 2010 (UTC)

Facepalm. OrangeDog (τ • ε) 22:04, 28 February 2010 (UTC)

And this means what?

Scarce has archived his/her talk page. I didn't see the bear.Vchimpanzee · talk · contributions · 21:03, 3 March 2010 (UTC)

Okay. The bear is here and always will be.Vchimpanzee · talk · contributions · 21:08, 3 March 2010 (UTC)

The bear is also here and very easy to see.Vchimpanzee · talk · contributions · 21:10, 3 March 2010 (UTC)

Okay, cancel my first statement. The bear is in the current revision when I go through the history. I was looking for the bear's removal but didn't see it.Vchimpanzee · talk · contributions · 21:14, 3 March 2010 (UTC)

Still can't see a bear on any of those revisions on the page or in the source, including in this diff. The evidence suggests that you were either imagining it or it has nothing to do with Wikipedia. OrangeDog (τ • ε) 12:58, 4 March 2010 (UTC)

Neither can I, and I've been watching this thread in puzzlement for over a week. Have you perhaps got a virus? Somebody once sent me a trojan in an email, and after opening it, I found I had a little cartoon sheep wandering all over my screen, and occasionally bleating. --Redrose64 (talk) 13:48, 4 March 2010 (UTC)

Are you guys for real? It's occluding almost the entire box. Just scroll down, you can't miss it. — Bility (talk) 16:36, 4 March 2010 (UTC)

Right. You've got me doing some serious timewasting now, and if I was in employment it would be a carpetable offence. Scarce has a rather complicated user page, ditto talk page, which are both built up from several subpage transclusions. The talk page pulls in User talk:Scarce/talk; this in turn pulls in User:Scarce/box which brings in {{User:Scarce/Userboxes/Bear}}, which does contain a bear, but a very small one - File:Bear Square.JPG at 50px. However, on Scarce's main and talk pages, the only way that I can see this pic is by going to the "Userboxen" scrolly list, and scrolling down inside that. It seems quite normal for a 50px image, and certainly doesn't show unless I scroll inside the "Userboxen". --Redrose64 (talk) 17:03, 4 March 2010 (UTC)

← Could it be something to do with the anchor link and image element for the bear userbox pic being enclosed in a <div> with class="floatleft" (presumably part of the userbox template), unlike the other userbox images there, which perhaps renders differently between browsers? –Whitehorse1 20:38, 7 March 2010 (UTC)

Strange behavior of en.m.wikipedia.org on Amazon Kindle

When the default bookmark on the Amazon Kindle (http://en.m.wikipedia.com) is used to access Wikipedia, the search field is pre-filled with words that seem somewhat random in nature. This appears to be based on some sort of server-side processing based on the received User-Agent HTTP header, as it can be replicated in any web browser by overriding its User-Agent string to that used by the Kindle, which is "Mozilla/4.0 (compatible; Linux 2.6.10) NetFront/3.3 Kindle/1.0 (screen 600x800)". Please see http://www.kindleboards.com/index.php/topic,20966.0.html for more info on the odd behavior being seen when the site is accessed this way, along with some screen shots. —Preceding unsigned comment added by NogDog (talk • contribs) 18:00, 6 March 2010 (UTC)

I will look into this tomorrow. Seems like a type of caching that goes wrong somewhere. —TheDJ (talk • contribs) 01:29, 7 March 2010 (UTC)

bugzilla:22757. —TheDJ (talk • contribs) 16:45, 7 March 2010 (UTC)

Where can I make proposals for features I would like to be added in MediaWiki

For instance, it would be handy to know if some template is being transcluded or not, or how many times it was trancluded. This would eliminate a lot of problems concerning deletion templates. I was trying to find where exactly in MediaWiki I could do that, but I didn't find anything. --JokerXtreme (talk) 12:26, 7 March 2010 (UTC)

bugzilla: —TheDJ (talk • contribs) 12:29, 7 March 2010 (UTC)

What links here contains list of transclusions of a template. Svick (talk) 12:33, 7 March 2010 (UTC)

It's not a bug and "What links here" cannot help. I'm talking about something that can be used within the template script. See here:[3]. I want the ombox to be trancluded just once. There seem to be some workarounds for these, which include complicate substs, noincludes and nowikis, but this is too messy. There should be some more robust and easy way to do that. --JokerXtreme (talk) 12:44, 7 March 2010 (UTC)

Bugzilla is also for suggesting new features, not just for reporting bugs. Svick (talk) 12:48, 7 March 2010 (UTC)

Aha, ok thanx then :) --JokerXtreme (talk) 12:54, 7 March 2010 (UTC) Oh, that's also mentioned in the top of the page, pardon me XD --JokerXtreme (talk) 12:56, 7 March 2010 (UTC)

<rant location="wrong, but I'm not personally motivated to do anything else about it; maybe someone reading this is">In that case maybe someone should suggest a new feature for Bugzilla itself—its main page talks only of bugs, bugs, bugs. Based on that, people turning up there with enhancement suggestions are effectively turned away. Yes, they could create an account just to see if enhancements are really accepted too, but many won't get that far.</rant> PL290 (talk) 13:14, 7 March 2010 (UTC)

Because it's designed as a bug system that is what it commonly says in most cases. And the only really custom text we have on there is the for lines next to the big "bug" which offers the link to this handy guide which shows how to use it and such and does mention feature requests as well. Peachey88 ^{(Talk Page · Contribs)} 13:26, 7 March 2010 (UTC)

As for the original question, it's already half (the counting uses) on Bugzilla as bugzilla:4394, It just really hasn't be implemented due to what appears to be database load issues. Peachey88 ^{(Talk Page · Contribs)} 13:29, 7 March 2010 (UTC)

Mmm, no, that's not what I'm asking. Here's the problem. The Tfd template is transcluded in template pages that are nominated. Those templates are transcluded in articles, so if they are nominated, the Tfd template gets retranscluded (for the second time) into the article. This is how notifications about templates that are to be deleted, are inserted in articles. Which is a good thing.

Now, the bad thing is, that they also get transcluded in documents, as they are used as examples. See here:Template:Rating-big. So I'd like to be able to have sections in the Tfd, such as the red ombox that won't get transcluded more than once. Or to be able to transclude noincludes somehow. I believe there is a way, that is used in the prod template, but it's too messy and I don't fully understand it. --JokerXtreme (talk) 13:53, 7 March 2010 (UTC)

(Bugzilla seems crammed up with reports that go back to 2005. I'm not sure if I should even bother. ) --JokerXtreme (talk) 13:55, 7 March 2010 (UTC)

The problem was not visible for sometime, because of some changes. You can see it now here:

Template:Rating-big and this is how it looks like inside the article: When_It's_All_Over_We_Still_Have_to_Clear_Up. This is because as you can see here: [4] this is included:

{{#ifeq:Wikipedia|Template

which means if the namespace is template and the basepage is not Tfd, print the ombox. Otherwise print the appropriate notification. The second part (basepage is not Tfd)is used, so the same problem won't happen here Template:Tfd.

I guess an easy solution would be to just remove the ombox and not display it at all, although I'd prefer a better solution. --JokerXtreme (talk) 17:08, 7 March 2010 (UTC)

I filed a proposal in bugzilla: [5] [6]. If you think it is a good idea, please vote for it. Thanks --JokerXtreme (talk) 19:45, 7 March 2010 (UTC)

Why is this malformed maintenance category being populated?

Category:Articles with unsourced statements from November 2,009 is clearly a non-standard maintenance category, so I was going to depopulate it and delete it ... but I can't figure out why either Nellore or North Malabar are being included in this category.

I have tracked the problem down to the tag {{Citation needed|date=November 2009}}, which I removed in this edit: no "articles with unsourced statements from November 2,009". However, when I reverted the test, the category reappears.

Can anyone figure out why {{Citation needed}} (or one of its helper templates) is having this bizarre effect on these two articles, but no others? --BrownHairedGirl (talk) • (contribs) 20:04, 7 March 2010 (UTC)

The comma(s) are being added because the text is being treated as a number, and it is automatically putting them in (i.e. it changed January 20000009 into January 20,000,009). This is happening because they are being added by the formatnum magicword, which takes place in Template:Infobox Indian jurisdiction, where it says, {{#if: {{{population_total|}}} |<small>{{formatnum: {{{population_total}}}}}</small>|}}. I'm not sure whether this is standard behaviour; I don't often deal with infoboxes. Ale_Jrb^talk 20:28, 7 March 2010 (UTC)

Adding that this effect occurs because, while the categories always appear at the bottom of the page when rendered, they can of course appear anywhere on the page. As they are placed within the population_total option, the category text (complete with date) that is generated by Template:citation needed from Template:Fix, is seen by the parser as part of the population_total input that is sent to formatnum in the infobox code. Ale_Jrb^talk 20:33, 7 March 2010 (UTC)

Thanks, Ale_jrb; that explains it nicely. I guess it's one of the limits of parser functions, that they aren't bright enough to spot glitches such as that part of the input data is a template call and should not be formatted

I don't like removing citation tags, and the only fix I can see is to remove formatnum from that field in {{Infobox Indian jurisdiction}}. It's not used in the global {{Infobox settlement}}, so I guess it's liveable-without. --BrownHairedGirl (talk) • (contribs) 21:39, 7 March 2010 (UTC)

Previous discussions: Wikipedia:Village_pump_(technical)/Archive_47#what_causes_this_abnormal_categorization.3F and Template talk:Infobox settlement/Archive 11#Infobox code ruins output of Template:Fact. I'm sure there was another discussion on this where there was a suggestion to add a separate reference parameter for each of those fields. ---— Gadget850 (Ed) ^talk 21:54, 7 March 2010 (UTC)

In the meantime, I have just removed the formatnum, which has fixed things. Fortmatnum is a nice idea in theory, but IMRHO it's not acceptable for the template to be incapable of accommodating {{citation needed}} in a straightforward manner. I have left a note about this at Template talk:Infobox Indian jurisdiction#Formatnum_removed. --BrownHairedGirl (talk) • (contribs) 22:16, 7 March 2010 (UTC)

Also T23054. Anomie⚔ 00:15, 8 March 2010 (UTC)

syntaxhighlight

The tag for this extension appears to ignore attributes other than lang:

examples
input output html <syntaxhighlight lang="cpp"> int main(void); </syntaxhighlight> int main(void); <div dir="ltr" class="mw-geshi" style="text-align: left;"> <div class="cpp source-cpp" style="font-family:monospace;"> <pre class="de1"> <span class="kw4">int</span> main<span class="br0">(</span><span class="kw4">void</span><span class="br0">)</span><span class="sy4">;</span> < /pre> </div> </div> <syntaxhighlight lang="cpp" id="func_prototype_example" style="padding:0px; white-space:normal; border:none; display:inline;"> int main(void); </syntaxhighlight> int main(void); <div dir="ltr" class="mw-geshi" style="text-align: left;"> <div class="cpp source-cpp" style="font-family:monospace;"> <pre class="de1"> <span class="kw4">int</span> main<span class="br0">(</span><span class="kw4">void</span><span class="br0">)</span><span class="sy4">;</span> < /pre> </div> </div>

I can only hope this behavior is not intentional. I would like some way to invoke the mw:Extension:SyntaxHighlight GeSHi to colorize inline code excerpts (that is, to highlight arbitrary code such as int main(void); mid-paragraph using the same set of rules).

Would I have to write a separate extension? ―AoV² 00:48, 8 March 2010 (UTC)

You can use the parameter enclose="none": int main(void);. Svick (talk) 01:22, 8 March 2010 (UTC)

Nice, very nice, yet nothing I would have guessed. Thanks. ―AoV² 02:05, 8 March 2010 (UTC)

WikiCite Builder

I've created a script at http://smallin.freeiz.com/beta/cite1.php that can take New York Time article urls such as http://www.nytimes.com/2010/02/25/nyregion/25paterson.html and produce a {{cite news}} template like this

{{Cite news|title=Question of Influencein Abuse Case of Paterson Aide |author=William K. Rashbaum |author2=Danny Hakim |author3=David Kocieniewski |author4=Serge F. Kovaleski |author5=Kitty Bennett |author6=Alain Delaquérière |author7=Barbara Gray |url=http://www.nytimes.com/2010/02/25/nyregion/25paterson.html |publisher=The New York Times |date=25 February 2010 |accessdate=6 March 2010 }}

Anybody think this will be useful?Smallman12q (talk) 01:56, 7 March 2010 (UTC)

http://toolserver.org/~dispenser/view/Reflinks Sole Soul (talk) 03:02, 7 March 2010 (UTC)

That tool uses page metadata, whereas mine use the NYT article search API. It also appears that reflinks doesn't capture multiple authors...for me it returns

<ref>Rashbaum, William K.. (2010-02-24) [http://www.nytimes.com/2010/02/25/nyregion/25paterson.html Question of Influence in Abuse Case of Paterson Aide]. NYTimes.com. Retrieved on 2010-03-07.</ref>.

I'll write a more advanced interface/bot later...Smallman12q (talk) 13:20, 7 March 2010 (UTC)

refToolbar Plus now supports this. ---— Gadget850 (Ed) ^talk 00:50, 9 March 2010 (UTC)

Account hacked?

I have had strange things in my observation list. Some of them imply that my account has somehow been hacked. When I tried to delete them, some of them still remained in the list. So I really would appreciate your help --David Liuzzo 18:20, 7 March 2010 (UTC) —Preceding unsigned comment added by David Liuzzo (talk • contribs)

There are a number of ways pages can sneak on to your watchlist. If a page is moved, then moved back, for example, you will be watching both pages. In the case of page move vandalism, that could lead to pages with offensive titles being on your watchlist. Could that be the case? Prodego ^talk 22:30, 7 March 2010 (UTC)

Maybe, but why am I unable to delete the rest of those entries on my list? --David Liuzzo 11:18, 8 March 2010 (UTC) —Preceding unsigned comment added by David Liuzzo (talk • contribs)

Tag

There's extra space below the cleanup tag here. I don't know how to fix it. Can you fix it, thanks.174.3.110.108 (talk) 22:08, 7 March 2010 (UTC)

 Done Gary King (talk) 04:42, 9 March 2010 (UTC)

"Split" edit view

I have created a user script to split the edit textbox into a separate box for each section, somewhat similar to how wikiHow has multiple edit boxes in its "Guided Editor". However, the script is very incomplete and quite hackish. For example, in the script I change the IDs of textboxes to get the toolbar (for bold, italic, etc.) to work. wikEd compatibility is even more difficult, so the script uses a 100 ms timer to synchronize textboxes (quite inefficient). Unsolved is that the script should work with other editing tools and not lose any changes (who knows what will happen due to the changing IDs). Please note that the script is still very much a work in progress and should not be used yet except for testing purposes. You will need to copy the text javascript:SpliteditJsSplit() into the address bar to activate the script, as right now, it adds no tab or button to the page. So how can I make this script less hackish and more compatible, even if it means major changes? PleaseStand (talk) 04:02, 8 March 2010 (UTC)

Here:

addOnloadHook(function{

addPortletLink('p-cactions', "javascript:SpliteditJsSplit()",
'split', 'ca-split', 'Split edit window', '0')
})

Add this to the script in the top, it'll add a tab in the cactions bar (the one where edit, talk, history tabs are) ManishEarth^{Talk • Stalk} 10:32, 8 March 2010 (UTC)

I just did this. Works perfectly under Vector skin, adding the tab and keyboard shortcut under the drop-down arrow. The script is still hackish though. I am considering having it put a select section drop-down below the main textbox, and only use the main textbox, to avoid the hackishness of using multiple textboxes with changing ID's (see the original post). That would also eliminate scroll wheel (mouse) focus issues. PleaseStand (talk) 12:07, 8 March 2010 (UTC)

RfC on Temporary Autoblocking of vandalism only IP addresses

The RFC is progressing here ▒ ♪ ♫ Wifione ♫ ♪ ▒ _―Œ ^{♣Łeave Ξ мessage♣} 19:38, 8 March 2010 (UTC)

on white-space

The lower-most part of Wikipedia talk:Accessibility renders with <p><br /></p> tags of unknown origin, creating an unpleasant gap of two blank lines. How does one make this go away, and/or prevent the same thing from happening at the ends of articles? ―AoV² 23:01, 8 March 2010 (UTC)

It's because there are <ref> tags on that talk page, but no <references />. This causes an error message on the bottom of the page, but that message is empty on talk pages. (That error message is MediaWiki:Cite error refs without references). Svick (talk) 01:40, 9 March 2010 (UTC)

Interesting. Why is the message empty on talk-pages? I understand well how to achieve this effect but can′t guess confidently as to the reason. Surely suppressing only the category—or using perhaps an alternative category for non-article cite errors, without the message—would be more appropriate. Either option would provide a hint regarding the corrupted page lay-out at least. ―AoV² 03:56, 9 March 2010 (UTC)

I think that cite errors on talk pages don't matter at all (the talk page you mentioned is a good example of that), so there is no need for them to be in the error category, nor for the error message. And while I would try to remove the <p> too, if it was relatively simple, I think it's not worth it in this case (it would require change in Cite.php). Svick (talk) 11:42, 9 March 2010 (UTC)

User contributions really slow

I can look at an article's history and it displays quickly, but if I try to see a User's contributions, it takes upwards of two minutes to display. Woogee (talk) 00:10, 9 March 2010 (UTC)

I would imagine that this is due to the structure of the WP database system, whatever that is, so that gathering the history of an article is fast because the history is associated with the article, whereas looking up the contributions of a User requires scanning all the articles for a match (there are a lot of articles, to put it mildly). I would doubt that this could be fixed without adding a new table (collection of fields) to the database, which would require some programming (I think there are only two official programmers). David Spector 17:22, 9 March 2010 (UTC)

There are more than two paid programmers, and many more volunteers. The query in question should be about as efficient as a history view: that's why we have the user_timestamp and usertext_timestamp indices on the revision table. —Aryeh Gregor (talk • contribs) 19:06, 9 March 2010 (UTC)

Which user is taking a long time? Your contributions load almost immediately for me. —Aryeh Gregor (talk • contribs) 19:06, 9 March 2010 (UTC)

Works fine for me. When I look at my own contributions, it takes a second or 3; I've got over 24,000 contributions, so there doesn't seem to be a backend problem. If you've got a skin that has to do a lot of javascript work for each entry, that could be the problem. Studerby (talk) 22:39, 9 March 2010 (UTC)

Help:Table#Vertically_oriented_column_headers

Is there a way to turn the phrase around, and is there a way to make the words read up and down, like in a crossword? —Preceding unsigned comment added by 174.3.110.108 (talk) 07:52, 9 March 2010 (UTC)

You can use the same procedure as in the help page you mentioned: create images of the text you want to display in the orientation you need and using them like ordinary images. Svick (talk) 11:57, 9 March 2010 (UTC)

Ogg doesn't work in Firefox 3.0.17

Ogg doesn't work in Firefox 3.0.17. Clicking an Ogg-playing button just results in the error message Sorry, your system does not appear to have any supported player software. Please download a player. Firefox is supposed to support "ogg theora" (sic) without a download.

If this is a problem with my computer, do you have any ideas what it might be? David Spector 17:19, 9 March 2010 (UTC)

Sorry to be dense, but why not go to FF 3.5+? Wknight94 ^talk 17:21, 9 March 2010 (UTC)

A quick Google search turned up this. According to that blog, Ogg Theora support was added in Firefox 3.1. Are you unable to upgrade your browser? Reach Out to the Truth 17:45, 9 March 2010 (UTC)

I'd recommend going straight to FF 3.5 ---- works great for me. Wknight94 ^talk 17:56, 9 March 2010 (UTC)

Native playback in Firefox will only work in 3.5 or later. (3.1 doesn't exist, it was renamed to 3.5.) I've updated the page to reflect this. If you have Java installed, however, it should still work using Cortado. If you don't have Firefox or Chrome or Java installed, and can't install them, you'll have to download the file and view it in a media player that supports Ogg Theora/Vorbis. —Simetrical (talk • contribs) 19:02, 9 March 2010 (UTC)

Yeah, that's what I get for relying on outdated blogs. :D Firefox 3.5's video support is pretty awesome, so it's definitely worth upgrading if possible. Reach Out to the Truth 19:16, 9 March 2010 (UTC)

3.6 is latest stable ATM. :) ¦ Reisio (talk) 04:53, 10 March 2010 (UTC)

Editing problem

I have recently switched to IE8, and have noticed a problem that whilst editing, the editing window sometimes automatically scroll up when I type, so that my typing ends up at the bottom of the box, or sometimes even off-screen (so I have to scroll down). Is there any way of stopping this, as it makes editing very difficult! Cheers, пﮟოьεԻ 57 19:33, 9 March 2010 (UTC)

This is bug #19334, which I have fixed in r62191, but said fix hasn't been deployed yet. Also, as noted in the code review, there are some minor issues (at least with the Cologne Blue skin) with my fix. --Jack Phoenix _(Contact) 22:37, 9 March 2010 (UTC)

OK, cheers. I'll make sure compatibility mode is on for now. пﮟოьεԻ 57 22:43, 9 March 2010 (UTC)

Is Template:Imagemap now entirely redundant or does it offer some increased functionality over and above simply calling an image ?

Can someone help answer my question at Wikipedia:Templates for discussion/Log/2010 March 2#Template:Imagemap? Thanks, –xeno^talk 22:25, 9 March 2010 (UTC)

• I've added a comment on the TfD page. –droll [chat] 23:13, 9 March 2010 (UTC)

• I think there is a more important issue with this template and others. Its really an issue with [[Image:]] itself. Any image can be set to link to any page on the net. Government websites usually notify a user when they are following a link outside the .gov domain. In Wikipedia the user is generally informed when a link is external by the use of the external link icon. [[Image:]] can be used to link to any site with no notification. The user might expect to go to an image description page and end up who knows where. IMHO such links should be clearly identified or be limited to wiki media sites. I am aware that spam and attack sites are usually blocked but how about sites that are not yet identified.

  If this is and old topic then "never mind" as Roseanne Roseannadanna used to say. Try this  –droll [chat] 05:07, 10 March 2010 (UTC)

Watching page sections?

So is there a way to watch just a section of a discussion page. It seems that this ability would be welcomed by many and the advantage seems obvious. Could such a thing be implemented? –droll [chat] 22:10, 9 March 2010 (UTC)

It's not possible currently, but the planned LiquidThreads will be able to do this. Svick (talk) 22:22, 9 March 2010 (UTC)

There was a guy holding his breath for LiquidThreads to come - unfortunately he died. –xeno^talk 22:25, 9 March 2010 (UTC)

To be fair, I once had the exact same though about SUL being rolled out... EVula // talk // ☯ // 05:12, 10 March 2010 (UTC)

LiquidThreads really must come one day... it is sorely needed. :-( I wish they'd accept my help with it but they don't seem interested. Dcoetzee 06:20, 10 March 2010 (UTC)

Who's "they", and what help of yours are they not interested in accepting? —Aryeh Gregor (talk • contribs) 23:24, 10 March 2010 (UTC)

LT is the best. They better hurry up! ManishEarth^{Talk • Stalk} 14:08, 10 March 2010 (UTC)

Finding Unsourced BLPs

Is there any way to track unsourced BLPs (i.e. those tagged with Template:BLP unsourced) via category or WikiProject? I'm asking because I'd like to give some focused help on athletics-based articles because I know where sourcing for this sub-section of articles can likely be found. Any ideas? Sillyfolkboy (talk) (edits)^{Join WikiProject Athletics!} 03:45, 11 March 2010 (UTC)

Yes, Category:All unreferenced BLPs. – ukexpat (talk) 04:04, 11 March 2010 (UTC)

For a project specfic output, you will need to do a category intersection using a tool (such as Cat Scan (Help Guide)) and have it compare two categories such as Category:All unreferenced BLPs and Category:WikiProject Athletics articles. Peachey88 ^{(Talk Page · Contribs)} 14:08, 11 March 2010 (UTC)

For anyone following this up, User talk:DASHBot/Wikiprojects seems to be providing this kind of information. Sillyfolkboy (talk) (edits)^{Join WikiProject Athletics!} 10:46, 17 March 2010 (UTC)

Red category being populated by a template

There's a "fooian fooers" template being used in a lot of categories which puts them into a red category Category:Australian. Anyone know how to fix the template? DuncanHill (talk) 09:01, 11 March 2010 (UTC)

These changes, perhaps? I've pinged BrownHairedGirl. Chris Cunningham (not at work) - talk 10:04, 11 March 2010 (UTC)

Sorry, my bad. I'll sort it out. --BrownHairedGirl (talk) • (contribs) 10:07, 11 March 2010 (UTC)

Thanks. DuncanHill (talk) 10:12, 11 March 2010 (UTC)

Just testing a fix (now nearly complete), but just wanted to say huge apologies for screwing up such a widely-used template (I didn't test one scenario). --BrownHairedGirl (talk) • (contribs) 10:34, 11 March 2010 (UTC)

Now fixed, and documentation updated. Unfortunately, it's going to take 24 hours or so for the template to be purged, and in the meantime hundreds of categories will display inappropriate parent categories. :( --BrownHairedGirl (talk) • (contribs) 11:13, 11 March 2010 (UTC)

Many thanks for the prompt attention. DuncanHill (talk) 11:28, 11 March 2010 (UTC)

No prob. It's just what's expected of any of us: make a mistake, fix it ASAP.

Anyway, it seems that I am in luck, and the template changes have been purged promptly. Category:Australian is now empty, as is Category:Czech, and the other categories listed at User:RussBot/category redirect log. Phew!

Now, gotta go and find an industrial-strength scrape-egg-off-face device :( --BrownHairedGirl (talk) • (contribs) 11:32, 11 March 2010 (UTC)

File:Socks and PJ.JPG thumbnail

Resolved

This file has refused to generate a thumbnail for over a week now on sock. Any ideas why? Chris Cunningham (not at work) - talk 10:01, 11 March 2010 (UTC)

Something must have happened to http://upload.wikimedia.org/wikipedia/commons/thumb/b/b1/Socks_and_PJ.JPG/220px-Socks_and_PJ.JPG somehow, but the system still thinks it's good so won't regenerate it, I'm guessing. I bet if someone deleted that file it would be fixed next time someone called it. ¦ Reisio (talk) 11:26, 11 March 2010 (UTC)

I reuploaded it and purged the cache. For some reason, a thumbnail on commons links here, while the wikipedia thumbnail links here (Same link as above). ManishEarth^{Talk • Stalk} 12:10, 11 March 2010 (UTC)

Because we have different default thumbnail sizes presumably. OrangeDog (τ • ε) 13:20, 11 March 2010 (UTC)

That's what I thought, too, but shouldn't all WMF wikis have the same thing? I mean why change the default sizes? ManishEarth^{Talk • Stalk} 15:07, 11 March 2010 (UTC)

Or you could have overridden the default in your user prefs. - Jarry1250 ^{[Humorous? Discuss.]} 18:02, 11 March 2010 (UTC)

Software change affects older "Logo fur" template?

The "Purpose of use" portion of an older "Logo fur" template is now causing an error to display and is attracting attention of editors who are flagging fair use images for deletion. See herefor an example of the old template and the error. The error drew a deletion notice. Replacing with a more current tag, as here, fixes the problem, but many older fair use images posted by numerous editors are now being flagged (and deleted). For example, see this discussion. Since this just started occurring, I assume a software change has caused this, but I'm not sure. Geoff ^{Who, me?} 15:26, 11 March 2010 (UTC)

This is by design. File:4th World Scout Jamboree.png had a FUR added on July 10, 2009 that left |Use= blank, resulting in the 'Purpose of use field' showing the error "No purpose specified. Please edit this image description and provide a purpose" and putting it into Category:Wikipedia files that transclude the Non-free media rationale template with no Purpose specified. Same problem with File:Arnott's Shapes-pizza flavour.jpg. A bot trawls the category and eventually tags the file for deletion and we fix the empty parameter. What is probably confusing is that |Purpose= simply overrides the default text in the 'Purpose of use field' but does not fix the blank |Use= error message. ---— Gadget850 (Ed) ^talk 15:43, 11 March 2010 (UTC)

BTW: FURME still has a bug where it sometimes adds a weird website by default; I fixed it on your last edit.

That is just bad template design then. If you want to fix the usage of a parameter, add error checking. This is just plain incorrect. —TheDJ (talk • contribs) 16:12, 11 March 2010 (UTC)

It now renders an error that describes what the problem is, but does not put it into a category that is food for incorrect deletions. —TheDJ (talk • contribs) 16:40, 11 March 2010 (UTC)

Looks like the error checking for that is in {{Non-free use rationale}}. ---— Gadget850 (Ed) ^talk 17:05, 11 March 2010 (UTC)

That check is only for not specifying Purpose. It should not be the same check as "specifying Purpose, but not Use". At least in my humble opinion. —TheDJ (talk • contribs) 19:31, 11 March 2010 (UTC)

.NET installation for AWB

I use windows xp home. In preperation of future use of AWB I was downloading .Net Framework 2.0 and got this error message, "The .NET Framework 2.0 Service Pack 1 is a later version of the .NET Framework 2.0. The earlier version cannot be installed over the later version." Will the 2.0 Service Pack 1 version be sufficient. I do assume it is but no harm in asking ? Mlpearc MESSAGE 18:48, 11 March 2010 (UTC)

Ye SP1 should be fine.Smallman12q (talk) 01:26, 12 March 2010 (UTC)

Thank you Mlpearc MESSAGE 01:31, 12 March 2010 (UTC)

Server Lag again

Resolved

 – Issued is fixed and replag is back to 0, For the nerdy reasoning behind it see the sys admin log. Peachey88 ^{(Talk Page · Contribs)} 09:29, 12 March 2010 (UTC)

In case of panic, please press.

Once again, there's a high database server lag. On my watchlist, it says "Due to high database server lag, changes newer than 1,303 seconds may not appear in this list." Hopefully someone can fix this. Thanks. --Hadger 01:23, 12 March 2010 (UTC)

Same here. It apparently lags back to about 1:00 UTC. Brambleclawx 01:24, 12 March 2010 (UTC)

Something must have broken...I'm getting a 25 minute delay now?Smallman12q (talk) 01:27, 12 March 2010 (UTC)

I believe this is the second lag this year (or the first one I experienced was in late 2009; I'm not sure when I experienced my first high database lag, all I know is that it was either this year or late 2009). --Hadger 01:29, 12 March 2010 (UTC)

Okay! It's fixed! --Hadger 01:38, 12 March 2010 (UTC)

Same here. Usually, I only experience lags of about 30 seconds. I actually have a screenshot of the message saying that there was a lag of 2,021 seconds. And why is it that every time I post in here about the special page problem, nobody bothers to reply? Brambleclawx 01:41, 12 March 2010 (UTC)

I don't know why. (Hey, I just replied to you! XD!) --Hadger 02:07, 12 March 2010 (UTC)

• This happens sometimes, occasionally much worse, but it always get sorted out eventually. The reason nobody is answering is because only those running the servers will have any clue why there is a lag, and they're not all regular editors--Jac16888^Talk 02:12, 12 March 2010 (UTC)
  • The information is in the sysadmin log like usual. The problem was with db12, due to "locking of pagelinks table" by a snapshot process (makes something like a backup of data). —TheDJ (talk • contribs) 02:32, 12 March 2010 (UTC)

Hey it got fixed! Whoever got it fixed, thank you! (Maybe the server lag was caused by them fixing this problem?) Brambleclawx 02:20, 12 March 2010 (UTC)

Except that its somewhat still messed up, as several diffs of the same article are listed. Brambleclawx 02:24, 12 March 2010 (UTC)

See also your original section above :D —TheDJ (talk • contribs) 02:28, 12 March 2010 (UTC)

Server lagging again, so I assume someone is trying to fix the bug then? Brambleclawx 03:29, 12 March 2010 (UTC)

I wonder why the server is lagging. --Hadger 03:47, 12 March 2010 (UTC)

As explained above, the db is getting dumped for a backup and its causing the pagelinks table which controls the watchlist to get locked up. Peachey88 ^{(Talk Page · Contribs)} 04:24, 12 March 2010 (UTC)

"Due to high database server lag, changes newer than 3,234 seconds may not appear in this list." That's quite some lag, and still going as of this posting. - DustFormsWords (talk) 04:25, 12 March 2010 (UTC)

Looks like it might be time to panic. Again. Equazcion ^(talk) 04:26, 12 Mar 2010 (UTC)

Now at 4,239 seconds...also being discussed at ANI, for the curious. -- Collectonian (talk · contribs) 04:41, 12 March 2010 (UTC)

I just got 4,917 5,444 seconds (÷ 60 = 81.95 90.73 minutes) on my watchlist at approximately 11:53PM 12:00AM EST. — SpikeToronto 11:55 05:03, 12 March 2010 (UTC)

Support total panic. 6,077, as of this posting. - DustFormsWords (talk) 05:11, 12 March 2010 (UTC)

• Is anyone working on this or is everyone asleep? - NeutralHomer • Talk • 05:13, 12 March 2010 (UTC)

Maybe there's a 1-800 number we can call for tech support? Wine Guy~Talk 05:17, 12 March 2010 (UTC) —The preceding comment was intended to be humorous.

Well, someone in the relevant part of the world should probably carefully read WP:SOFIXIT, and then head on over to Wiki HQ and boldly give the servers a good kick. :-) - DustFormsWords (talk) 05:22, 12 March 2010 (UTC)

Is anyone else constantly getting the new messages bar? I figure they're related, because I was the last person to edit my talk page, and every few minutes the orange bar reappears. Bradjamesbrown (talk) 05:26, 12 March 2010 (UTC)

I am experiencing lags of 7,223 seconds (2 hours 0 minutes 23 seconds) but I am not getting the new message bar. The contribution lists are doing the same thing, making vandalism reversion a bit harder. NotAnonymous0 (talk) 05:31, 12 March 2010 (UTC)

• As of this post we are at 8,465 seconds....AND FALLING! Whoohooo! At least it didn't go over 9000. - NeutralHomer • Talk • 06:08, 12 March 2010 (UTC)

• Oh crap :( Back up to 466 seconds. OK, three in one night, someone needs to fix the thing once and for all before the mad morning rush starts in a couple hours from the East Coast. - NeutralHomer • Talk • 07:34, 12 March 2010 (UTC)

Possible to change "show/hide" color

Is it possible the change the default link color of the "show/hide" text in the month NavFrames located here? I'd like to make them more visible if possible. Thanks --Torsodog^Talk 04:04, 12 March 2010 (UTC)

Think I'd just change the heading background to something less dark. 2¢ ¦ Reisio (talk) 17:12, 12 March 2010 (UTC)

That's definitely an option, but I'd like to change the font color if possible as all of the navboxes concerning this team use that color blue. --Torsodog^Talk 22:56, 12 March 2010 (UTC)

It would take a modification of MediaWiki:Common.js to make it fully configurable. Something you can do without an admin's help is:

<div class="NavHead" style="background:#006699; color:gold;">April<div style="position: absolute; right: 0; top: 0; width: 6ex; height: 100%; border: 0; padding: 0;" class="toccolours"></div></div>

Kludgy, but then so is the whole system. ¦ Reisio (talk) 05:13, 13 March 2010 (UTC)

Secure Links

The secure links script doesn't seem to work at Wikipedia:Abuse_response#Open_cases. For example it links Wikipedia:Abuse reports/167.21.254.12 to https://secure.wikimedia.org/wiki/Wikipedia:Abuse_reports/167.21.254.12 rather than https://secure.wikimedia.org/wikipedia/en/wiki/Wikipedia:Abuse_reports/167.21.254.12 . This may have to do with the fact that its a category tree...though I'm not sure...Smallman12q (talk) 23:32, 12 March 2010 (UTC)

Yes, those links are loaded on demand, so the script should not be able to adapt them initially. It is known (Bug) that this extension at times has also done the reverse (generate https links when not using the secure server), so this is likely a caching issue in the categorytree extension, rather than a bug in the link fixing script. —TheDJ (talk • contribs) 23:41, 12 March 2010 (UTC)

I have written a patch to fix this problem and hopefully it will be fixed soon. —TheDJ (talk • contribs) 00:55, 13 March 2010 (UTC)

Proposal of gadget for highlighting your comments

Please see Wikipedia:Gadgets/proposals#User:PleaseStand/highlight-comments.js. PleaseStand ^(talk) 23:54, 12 March 2010 (UTC)

Right-click section editing

Is it possible to set up my account so that I can edit sections by double right-clicking? Right now I have the right-click section edit option turned on and I end up on the edit page by mistake when trying to go to the subject of an AfD debate, for example. I know that I can open the link in a new tab by middle-clicking, but right-clicking is what I am used to. It would also be great if I could double right-click the article title as well and be taken to editing the lead section only. PleaseStand ^(talk) 01:12, 13 March 2010 (UTC)

I'm sure there's a script or gadget for this...I currently have this but I've had this feature so long I forgot how I added it 0.oSmallman12q (talk) 02:16, 13 March 2010 (UTC)

Nvm, I found it...go to my preferences->Editing->Advanced Options, check the box for "Edit pages on double click (requires JavaScript)" and hit "Save" at the bottom of the screen. Hope this helps!Smallman12q (talk) 02:21, 13 March 2010 (UTC)

That's helpful but not exactly what I was asking for, the option to edit a section by right-clicking, but by double right-clicking in order to be able to right-click links in section headings. Now that I look at it...it seems that rightclickedit.js was written with this concern in mind: "We don't want to deprive the noble reader of a context menu for the section edit link, do we? (Might want to extend this to all <a>'s?)" (emphasis added). So, if the right-clicked element is not a link (not an "a" element) or its parent does not have a class attribute of "editsection", it proceeds to the edit screen. Unfortunately, for a heading on an AfD log page (not especially problematic in this case given that there is another link right below), the class is "mw-headline". OK, I just found the problem, so how do I get it fixed, eliminating the problems with opening the link in a new tab while still allowing easy editing of the transcluded AfD debate (for example, as I suggested with the double right-click)? PleaseStand ^(talk) 04:01, 13 March 2010 (UTC)

Talk subpages and TOCs

With a talk page that includes an assessment comments subpage, is there a fix for the subpage sections being listed in the main talk pages' contents list?

For example, if Talk:Foo/Comments exists with 3 sections, then Talk:Foo will include those 3 sections in its TOC; clicking on the links has no effect. The 3 sections can appear multiple times, presumably once for each projbanner that makes use of {{WPBannerMeta}}'s Comments parameter. Thanks. –Whitehorse1 20:13, 7 March 2010 (UTC)

This is a frequently asked question, but doesn't seem to be properly documented anywhere. To avoid these problems it is best not to have any headings on comments subpages. This is one of the reasons that the community has decided to stop using these subpages in the future. — Martin (MSGJ · talk) 20:41, 7 March 2010 (UTC)

Thanks, M. I vaguely recall the village pump discussion on phasing them out. Would changing the existing headings to, say, bold larger type plus <hr> tags work as an interim step do you think?
I ask because the suggestion at the page you linked, of subst'ing the comments subpage to the talk page and then deleting it, might cause attribution problems with the page being larger than one or two comments. For that matter it looks like 2 sections were copypasted into it, without histmerging, from the wikiproject's talkpage. Or would license compliance not be a problem since regardless of where they were originally submitted, the comments themselves are signed? –Whitehorse1 22:06, 7 March 2010 (UTC)

I think the last discussion was Wikipedia:Village pump (technical)/Archive 57#Talk pages: the repeating table of contents (TOC) problem. ---— Gadget850 (Ed) ^talk 21:58, 7 March 2010 (UTC)

I've just changed them to use ; and that sorted out the repeated TOC on that page. It might be better to move those comments to the main talk page and then blank or redirect the subpage. — Martin (MSGJ · talk) 22:02, 7 March 2010 (UTC)

Thanks, Martin and Gadget. I was just looking at the last discussion link suggesting '''<big>''' with ; after submitting that addition to my comment, when you added a note you'd kindly fixed the markup. Thanks! Ideally, the deleting the subpage route sounds preferable to save having to put one of those {{copied}} templates on the Talk pages. Will the attribution thing above be an issue with that solution? –Whitehorse1 22:12, 7 March 2010 (UTC)

I don't think there is an issue with attribution on talk pages, because editors' signatures would seem to do it perfectly well. — Martin (MSGJ · talk) 22:18, 7 March 2010 (UTC)

It would seem so to me as well. (Unsigned comments until they're fixed aside obviously.) Anyone know for sure, or are you confident on that point? –Whitehorse1 22:22, 7 March 2010 (UTC)

There's no problem with attribution, as long as the comments are signed. Of course a short note would be nice, saying that it was copied from the "/Comments" subpage, but most people leave notes when they're copying discussions from one page to another. However, there's no reason to delete the page history of those "/Comments" subpages, and they should probably be redirected rather than blanked, since blank pages are liable to be deleted in the future. Graham87 00:50, 8 March 2010 (UTC)

Actually redirecting is problematic because then the wikiproject banners will try to self-transclude. — Martin (MSGJ · talk) 07:03, 8 March 2010 (UTC)

I was thinking that there should be a way to noinclude a redirect so that it would redirect if you visited the page, but it wouldn't transclude the target of the redirect. Alas the following didn't work on Talk:Halloween/Comments:

<noinclude>
#REDIRECT[[{{FULLROOTPAGENAME}}
</noinclude>

Any suggestions? — Martin (MSGJ · talk) 13:50, 9 March 2010 (UTC)

Aren't these being deprecated anyway? –xeno^talk 13:54, 9 March 2010 (UTC)

Yes they are, but the question of what to do with the pages has not been fully decided upon.

1. Delete (loses history)
2. Redirect (causes problems with self-transclusion)
3. Blank (may be confusing)
4. Replace with message (seems a bit overkill)
   — Martin (MSGJ · talk) 14:48, 9 March 2010 (UTC)

Thanks all. AIUI the general view seems to be it's best to subst the Comments page contents to the main Talk page, which satisfies history & attribution through the signed-dated comments themselves, followed by deleting the—now-redundant Comments subpage using noinclude tags, as outlined at the discussion Martin linked. –Whitehorse1 19:59, 13 March 2010 (UTC)

Special Pages: Problem?!?!

All special pages under the maintenance section, appear to have a problem. They are not listed under inactive, so I assume they have not been disabled for technical reasons, as well as not having any indication of why it does not work. The pages are still regularly cached, but there are no results. Can someone explain, or find out what happened? Brambleclawx 23:51, 9 March 2010 (UTC)

Thx for the reminder below :D. I actually asked about this on IRC 2 days ago, which led to a very interesting conversation that concluded with "nobody knows". After that MZMcBride tracked the bugticket bugzilla:21837 as being related, and after more and more talk, the cause was found (apparently some process didn't fully finish during an upgrade of the mysql databases). This problem has been corrected, and after the scripts have run (which can take 2 weeks for some pages), the pages should start to come back online again. —TheDJ (talk • contribs) 02:27, 12 March 2010 (UTC)

They are back online, however, they list several diffs of the same page. Will this resolve itself because the pages are just back up, or is there another bug requiring fixing? And thanks for looking into it! Brambleclawx 02:32, 12 March 2010 (UTC)

Asking. —TheDJ (talk • contribs) 02:49, 12 March 2010 (UTC)

It has been resolved. Brambleclawx 18:45, 13 March 2010 (UTC)

Good to hear !! —TheDJ (talk • contribs) 19:59, 13 March 2010 (UTC)

Redirect links

We have a very useful disambigious finding tool which when given a wikipedia page will tell you which links are disambigious on it. Do we have an similiar tool for finding links that are redirects? Reason being that recentrelated changes don't show changes for pages that are a page from redirects. Regards, SunCreator (talk) 13:52, 13 March 2010 (UTC)

I think you mean "related changes" right ? (i'm personally not aware of such a tool btw but my CSS always shows redirects as green links in a page). —TheDJ (talk • contribs) 15:46, 13 March 2010 (UTC)

Yes, related. Where do I get a CSS/how do I use it show I can see redirects in green? Regards, SunCreator (talk) 15:50, 13 March 2010 (UTC)

Remember that linking to redirects is perfectly OK; there is no general reason to change links to bypass redirects. — Carl (CBM · talk) 15:54, 13 March 2010 (UTC)

Yes. The purpose here is for knowing 'related changes' in a similar way to watchlists. Regards, SunCreator (talk) 15:58, 13 March 2010 (UTC)

Just answered this at Wikipedia:Help desk#Gadget to display disambiguation links?. ---— Gadget850 (Ed) ^talk 16:08, 13 March 2010 (UTC)

Redirects often exist because of typos— fixing those is perfectly acceptable. ---— Gadget850 (Ed) ^talk 16:09, 13 March 2010 (UTC)

a.mw-redirect { color:green} a.mw-redirect:visited { color:darkgreen}

(Edit conflict) added to your skin css (usually monobook.css) will color redirects green. There's also User:Dschwen/highlightredirects.js that adds a switch, but it's disabled on special pages. Remember that redirects are cheap. — Dispenser 16:21, 13 March 2010 (UTC)

Thanks all! Redirects are cheap and good things. Regards, SunCreator (talk) 16:46, 13 March 2010 (UTC)

User:Anomie/linkclassifier.js is wonderful; identifies redirects, disambigs, and also fair-use images and pages marked for deletion. Happy‑melon 21:27, 13 March 2010 (UTC)

Search problem

From here:

• http://en.wikipedia.org/w/index.php?title=Special:Search&search=census_nigerian_copyright_scholars_photo_variety_context_limbaugh_samarkand_klutz_prefix%3AWikipedia_talk%3A&fulltext=Search&ns0=1&redirs=1
• http://en.wikipedia.org/w/index.php?title=Special:Search&search=census_nigerian_copyright_scholars_photo_variety_context_limbaugh_samarkand_klutz_prefix%3AWikipedia_talk%3AI&fulltext=Search&ns0=1&redirs=1

As soon as I add even the "I" from "Identifiying reliable resources/Archive 21", the search breaks. Could someone with a wrench have a look at this? Paradoctor (talk) 21:33, 13 March 2010 (UTC)

That's because it is not under Identifiying reliable resources/Archive 21, it is at "Wikipedia talk:Reliable resources/Archive 21". At least it was the last time the search index was updated. And that is not as often as the pages are edited/moved/deleted. —TheDJ (talk • contribs) 22:18, 13 March 2010 (UTC)

So my first guess was on the money, thanks. BTB, how often is the search index updated, typically? Paradoctor (talk) 23:18, 13 March 2010 (UTC)

Found it, thanks. My brain is in slomo today. Paradoctor (talk) 00:03, 14 March 2010 (UTC)

testing templates (continued)

for anyone interested in commenting bugzilla:22135. I would have posted this the previous discussion here, but the discussion has already been archived. —Preceding unsigned comment added by Stmrlbs (talk • contribs) 21:52, 17 January 2010 (UTC)

Job queue problem?

I modified a project's template to add additional classes last Thursday. Afterward, it removed nearly every article from the project. I can still go to an article's talk page and the correct category is there, but the article is not listed in the category. If I make a null edit (or any edit), then it populates it in the category but it is impractical to have to do that to every article in the project. The project shows very low, if any, rated articles and nearly 200 unrated. It shows 0 stub-rated articles and I'm pretty sure we have our share of stubs. I have brought this up on the assessment bot talk page and I was told it was more than likely a job queue problem and to give it a few days to a week. It will be one week tomorrow. I've never seen anything job queue related take 6 days. Should I be more patient and give it more time or is there another problem?—NMajdan•talk 17:30, 10 March 2010 (UTC)

You are not the first to make this observation. The jobqueue seems almost empty, but some of these category changes take ages none the less. Up to 2 weeks these days (1 week was definitely the max in the past) before some categories are updated. I'll ask around if anything changed. —TheDJ (talk • contribs) 19:51, 11 March 2010 (UTC)

I left a message several days ago at Help talk:Job queue#Stuck queue which is being ignored. Where is the proper place for requesting job queue assistance? --Redrose64 (talk) 12:41, 14 March 2010 (UTC)

Probably bugzilla:. or mail one of the system administrators directly. —TheDJ (talk • contribs) 13:18, 14 March 2010 (UTC)

Who would be a suitable sysadmin? --Redrose64 (talk) 17:19, 14 March 2010 (UTC)

bunching problem

I noticed that carrot did not have section editing enabled, and usually this is a bunching problem caused by lots of images, which this article does indeed have. I tried to fix it twice, first with Template:stack and then with Template:FixBunching, and neither one did the trick. Anyone got any other ideas? Beeblebrox (talk) 00:07, 13 March 2010 (UTC)

Cenarium fixed it by removeing __NOEDITSECTION__. ---— Gadget850 (Ed) ^talk 00:14, 13 March 2010 (UTC)

(e/c) Yes, the magic word __NOEDITSECTION__ was on it. Though I see that it's also on a hundred of articles which would need fixing too. Cenarium (talk) 00:19, 13 March 2010 (UTC)

What is the purpose of that anyway? I don't recall ever seeing it before. Beeblebrox (talk) 00:59, 13 March 2010 (UTC)

It removes all of the edit section links; it's used most notably for archives (archive templates like {{talkarchive}} include it), on the main page and other pages where showing edit section links wouldn't make sense. Some lists like List of woredas in the Amhara Region or List of current NFL team rosters use it too, though it's pretty rare. Otherwise it should definitely not be used in articles. Cenarium (talk) 02:48, 13 March 2010 (UTC)

I removed most of them, except where it seemed that the __NOEDITSECTION__ is reasonable, like “List of …” articles, or Results of the Canadian federal election, 2000. Svick (talk) 17:18, 13 March 2010 (UTC)

Ah, that makes sense. I still wonder why someone chose to put in the carrot article, but that's another matter. Thanks for clarifying, no matter how long you edit here, there's always something else to learn about how it works... Beeblebrox (talk) 19:17, 14 March 2010 (UTC)

Searching for unrendered wikitext

Can anyone suggest a way to search for articles containing specific unrendered wikitext? I'm interested in identifying articles which contain "<cite id=" or "<Cite id=". Wtmitchell (talk) (earlier Boracay Bill) 02:05, 13 March 2010 (UTC)

Unrendered wiki-text? Could you be more specific? Do you mean hidden comments, templates...? (The best approach would be to use regex).Smallman12q (talk) 02:10, 13 March 2010 (UTC)

I believe that he is probably referring to incorrectly entered wikicode which appears raw as the software doesn't recognize it. Please correct me if I'm wrong. Thanks. -- IRP ☎ 03:28, 13 March 2010 (UTC)

The database scanner in AWB will search within unrendered wikitext. Graham87 04:01, 13 March 2010 (UTC)

I clarify: Up until recently, HTML constructs like <cite id="whatever">hand-crafted cite</cite> were used to supply hand-crafted cites with linkable anchors. Presuming browser support, such cites would highlight in blue when accessed via the specified anchor (e.g., like [[#whatever|wikilink to cite]]). This was recently deprecated due to conflicts with HTML 5. (See here, here, here, here, here, and here for some background. )

The (one? a?) current HTML idiom for accomplishing this seems to be e.g., <span class="citation wikicite" id="Reference-refDoe2010">Doe, John (2010) ''A fictional account''</span>. However, articles exist with the now-broken and now-deprecated <cite id=... tags in them. I'm asking whether anyone knows how I could search for such articles.

Graham87, thanks for the pointer to AWB's database scanner. I currently favor an Ubuntu Linux platform and haven't used AWB lately. I think I've got too much on my plate just now to consider coming back up to speed on AWB and dealing with multi-gigabyte downloads of database dumps. Wtmitchell (talk) (earlier Boracay Bill) 02:16, 14 March 2010 (UTC)

If I can get a download, I will take a stab at it. This was on my todo list anyway. We will probably find more wonderful reference templates not on my list. ---— Gadget850 (Ed) ^talk 02:03, 14 March 2010 (UTC)

Thanks. I'd appreciate a note on my talk page about whatever you turn up. Wtmitchell (talk) (earlier Boracay Bill) 02:11, 14 March 2010 (UTC)

Downloaded in about 2.5 hours; extracting now. I will play with it later tomorrow. ---— Gadget850 (Ed) ^talk 05:56, 14 March 2010 (UTC)

See User:Gadget850/dbsearch/cite 2010-Feb-03. About 75 templates and 2500 articles. ---— Gadget850 (Ed) ^talk 01:42, 15 March 2010 (UTC)

Entire ISPs unable to read Wikipedia

OTRS has received communications from employees of two different ISPs (in two different countries) that indicate that their users are unable to even read Wikipedia from any IP starting with 175. Attempting to do so returns an error "the server where this page is located is not responding". This problem has been ongoing for nearly a month. Does anyone have any idea what could be causing this, or what more information is required to tease that out? Thank you. Someguy1221 (talk) 08:25, 13 March 2010 (UTC)

Best if they join #wikimedia-tech on the IRC freenode network and ask mark or domas. Otherwise they might be helped with the wikitech-l @lists.wikimedia.org but that will take longer, this route also works but is still longer :D

Helpful information for solving this problem, is the best estimation of the affected range, their ASN and traceroutes. Are there issues with just wikipedia, or also commons.wikimedia.org ? —TheDJ (talk • contribs) 11:44, 13 March 2010 (UTC)

I've passed along the information you gave me and have asked the questions you suggested. We just today received a complaint from a third ISP in yet a third country. Someguy1221 (talk) 06:01, 14 March 2010 (UTC)

Do we not tell our OTRS responders who to talk to about these issues? Nobody on the English Wikipedia village pump is able to resolve this, and I doubt anybody would know where to start if they could. E-mail a system administrator, or track somebody down in #wikimedia-tech. — Werdna • talk 06:33, 14 March 2010 (UTC)

Actually, no, we are not told what to do with these issues. I always thought it was conspicuously missing from the OTRS instructions. Is there a particular SysAdmin that would be best to receive these types of complaints? I do have all of their email addresses. Someguy1221 (talk) 06:48, 14 March 2010 (UTC)

Just track someone down on #wikimedia-tech like Werdna suggested. Or if it's not urgent, try e-mailing wikitech-l; makes sure all the right people see it, even if replies aren't instant. ^demon^{[omg plz]} 17:07, 14 March 2010 (UTC)

Including alt text in the File: namespace

Please assist First off, I am never quite sure what belongs here, mediawiki.org, or at Meta. Forgive me if this proposal is outside the scope of Wikipedia or Wikimedia. Presently, the situation for the application of the HTML attribute alt is as follows:

• Mediawiki software generates alt text for every image. To use the example of the current main page and to further chose one image from that page at random, we have File:Icos Laboratories.JPG. On the main page, the file has no caption, but the image link does have the title "Laboratories at Icos" and the image itself has identical alt text. If you go to the page in question, this image is used with the caption "Laboratories at Icos", but the alt text reads "A woman in a lab coat takes glass vials out of a machine. A computer displays a spectroscopy graph."
• Wikipedia guidelines at WP:ALT demands that "Every visible image should have alt text, unless the image is purely decorative, that is, it has no function and is used only for visual formatting or decoration." And furthermore "Alt text is not the same as a caption. Alt text is meant for those who cannot see the image, whereas the caption is intended for all readers."

As I see it, there are two problems with the disjunct between the purpose of alt in HTML, the way it is generated by Mediawiki software, and Wikipedia's guideline on the usage of alt text:

1. Junk alt text is automatically generated, apparently just to pass validation. In the example above, the alt text on the main page is "Laboratories at Icos", which is also the caption at Icos. This alt text is completely useless to anyone who cannot see the image. "Laboratories at Icos" does not tell me what this picture looks like (except in the vaguest possible terms), which is exactly why WP:ALT requires descriptive alt text, such as the one used on that article, "A woman in a lab coat takes glass vials out of a machine. A computer displays a spectroscopy graph." Furthermore, Wikipedia:Featured article criteria demands alt text, so this had to have been added to this article prior to it reaching the main page. Why would the article itself be required to have useful alt text as a prerequisite to get the to main page, when the main page ends up generating useless junk text? The purpose of alt in HTML is congruent to the guidelines on Wikipedia: alt text gives a description of an image, whereas title and the purpose of captions on Wikipedia is to provide supplemental text that contextualizes an image or gives additional information that enhances the use of that image within a document.
2. Not only is bad alt text being generated—in contravention to the purpose of the tag in HTML, the usability it provides to readers, and the guidelines of Wikipedia—but different alt text is being generated for the same image. File:Icos Laboratories.JPG is the same on both Main Page and Icos (the only difference being its dimensions). Therefore, while a caption at either page might be different—at Icos it might read "Laboratories at Icos" but at Women in science it might read "The 1990s saw a greater influx of women into technical fields."—the alt text should ostensibly be identical. The same image is being described by alt text in either place, so there is no purpose in having it be described in two different, conflicting ways. It also creates an extra burden for Wikipedians to write up alt text repeatedly for the same image. To use another example, File:Flag of the United Kingdom.svg is used in over 50,000 pages. Sometimes, it is included in the page through a template, such as Flag of the United Kingdom, where it is in {{Infobox Flag}}. This template has an Alt field, but it is presently unused. Other times, it is transcluded through a template such as {{UK-poli-stub}}, which has no option for alt text (cf. The Angry Brigade.) Finally, the file can be added simply through the use of [[File:Flag of the United Kingdom.svg]] in an article. In the first example, the alt text reads "See adjacent text." In the second, it says "Stub icon". In instances of the third usage, anything can be the alt text, but it is mostly likely that none is employed, generating the generic, "Flag of the United Kingdom.svg". Furthermore, at File:Flag of the United Kingdom.svg itself, the alt text for this image is the completely useless "File:Flag of the United Kingdom.svg". This serves even less of a function than if it said "Flag of the United Kingdom"; at least in that case a reader could have a good guess at what the image displays if he's ever seen that flag before, but who knows if "File:Flag of the United Kingdom.svg" is accurate or if the flag of the United Kingdom has changed since you last saw it? For that matter, "See adjacent text." and "Stub icon" are completely useless and even confusing. They cause more harm than help to anyone using a screen reader or a browser that doesn't display images.

All that having been said, the solution that I have in mind for the interrelated problems of junk alt text, no alt text, and conflicting alt text used for the same image in different contexts is to make alt text a part of the File namespace and then "transclude" or <includeonly> that text into any article/template/portal/etc. that uses an image file link. This might mean that Special:Upload could have a new field requesting alt text and any image lacking it could be added to a hidden category like, Category:Images without alt text. I have three questions:

1. Is this feasible?
2. Does anyone have any better suggestions on how to fix the problems of no alt, junk alt, and conflicting alt text?
3. Does this belong somewhere else (e.g. mediawiki.org or Meta)?

Thanks for your input. —Justin (koavf)❤T☮C☺M☯ 20:38, 13 March 2010 (UTC)

See bugzilla:19906. Stub icon alt text can be changed by using "imagealt" on the stub template. –xeno^talk 20:42, 13 March 2010 (UTC)

1. absolutely, but who is gonna do the work ? See bugzilla:19906, as xeno said.
2. well i would like to point out that it should always be possible to override the 'default' alt text, because different alt text for different context can certainly be useful. (Think of an image of a president next to the oval office desk. The image description might focus on different details depending on being in the article of the president in question, or being in the article on the oval office desk). Next, i think any auto generated description is probably than NO description, and the File: prefix does provide a function there in making sure that people will realize that it is not human generated I think, though perhaps a different prefix ("File: ")might be something to consider.
3. It has been discussed a few times, and having the ability to add a default alt text to an image is definitely something we would like to have I think. So it's more an issue of who is gonna write the code (see bugzilla). —TheDJ (talk • contribs) 22:08, 13 March 2010 (UTC)

TheDJ's right about the need to override any default. As for the main page, the alt text for its photo should be better, but that's up to the people who maintain the main page. If you edited Wikipedia:Today's featured article/March 16, 2010 right now, for example, you could fix its alt text. Would you like to volunteer to do that? Ideally it'd require only one edit per day. Eubulides (talk) 01:21, 14 March 2010 (UTC)

You might find tools:~dispenser/cgi-bin/steal_alttext.py useful for filling in alt texts until the bug is resolved. — Dispenser 06:45, 14 March 2010 (UTC)

Ahoj lidi. If this is related to the annoying trend of making “decorative” image-links not point to their respective file-description pages, I have a javascript gadget which restores the default link. ―AoV² 09:59, 14 March 2010 (UTC)

Job queue expedite

Is there any way to expedite stuff in the job queue? For example, a widely used but semiprotected template like Template:Sisterlinks was vandalized to prominently display some expletives. The change was unnoticed as it didn't get published until it went through the job queue. After it is published, someone sees it and fixes it. The problem is that this fix will not go live for some time, and in the meanwhile, users of Wikipedia will see that some pages heve a section full of said expletives. Basically, I'm asking if items in the job queue can be modified, deleted, or expedited, and where one can request such a change. ManishEarth^{Talk • Stalk} 01:34, 14 March 2010 (UTC)

If you e-mail a list of affected articles to me at firstinitial surname at wikimedia dot org (my name is Andrew Garrett), I might be able to look into it. — Werdna • talk 06:29, 14 March 2010 (UTC)

Currently, there's no such problem. But I've seen it happen, where there's some template vandalism which has been reverted but the revert hasn't gone live. I was looking for a group (a la Oversighters) who have access to the job queue. The problem with sending a notification about the affected article by email to just you is that there will still be some time until you see the mail, and by that time its probably already fixed. ManishEarth^{Talk • Stalk} 10:50, 14 March 2010 (UTC)

Any known reason Watchlist 7-day view won't work?

Resolved

It's now 14 March, but I can only see 13 and 14 March in my watchlist, even if I click "7 days" or "all". Is this a known issue, perhaps to do with one of the gadgets? Bit of a problem if people are away for a few days. Didn't think it used to be like that. Noticed it a few times recently. PL290 (talk) 14:51, 14 March 2010 (UTC)

It works for me. What happens if you select "show my edits"? There are a few edits of yours that should be displayed in that case. Hans Adler 14:59, 14 March 2010 (UTC)

Thanks for that reply: it wasn't "show/hide my edits" (mine were not hidden) but on learning that full history works for you, I tried deselecting " Expand watchlist to show all changes, not just the most recent" in my preferences, and that turns out to be what was causing the issue. PL290 (talk) 15:17, 14 March 2010 (UTC)

In the preferences, under the Watchlist tab, you can set the "maximum number of changes to show in expanded watchlist" up to the maximum of 1000 changes (default is 250), if you would like to keep the expanded watchlist. Depending on the pages you are watching though, you probably will not see the full seven days, even using the maximum setting. Hint: unwatch busy pages you are not interested in anymore. PleaseStand ^(talk) 15:59, 14 March 2010 (UTC)

Hey!

Resolved

 –  – ukexpat (talk) 04:06, 16 March 2010 (UTC)

Hey! Somebody has gotten into and vandalized my watchlist. This is Not Good. (This could have happened a while back.) Has this been fixed? If not, does anybody know anything about this? Herostratus (talk) 05:24, 15 March 2010 (UTC)

What happened exactly? Sometimes pages you move, edit, or create are added to your watchlist, depending on your prefs (watchlist tab, bottom of page, the 3 checkboxes saying "add pages I * to my watchlist") Your watchlist can't be edited by anyone but you and sysadmins (These are not sysops). Check what pages you're watching here. If there are some pages which you haven't edited at all, then your account might have been hacked. ManishEarth^{Talk • Stalk} 06:13, 15 March 2010 (UTC)

Uh, somebody added "pages" to "watch" with tiles such as "LOL WIKIPEDIA SUCKS", gibberish, and so forth. I am an admin and for this and other reasons hacking of my account is problematic. I'm hoping that there was some known bug which has been fixed which allowed watchlists (only) to be hacked? Anyone know? Herostratus (talk) 06:44, 15 March 2010 (UTC)

DON'T PANIC. Do you have an example title that indicates this hacking? I imagine it's simply the result of page move vandalism. --MZMcBride (talk) 06:46, 15 March 2010 (UTC)

Well, I don't havy any examples any more - I removed them from my watchlist, and I don't know of any way to get a history of the watchlist. Except for one I missed, so-called article HАGGＥR? Ÿ, is still there. Are you saying that these article names are the result of actual articles being moved to names like LOL WIKIPEDIA SUCKS or whatever? But there are so many, wouldn't the normal restoration of the article result in the watchlist being healed? All the bogus links were redlinks. Thank you for your help! Herostratus (talk) 06:53, 15 March 2010 (UTC)

Yes, usually if there's an entry on your watchlist that you really don't recognize, it's the result of page move vandalism. For example, Yo to the NyMpHo is on my watchlist; this was the result of this vandalism. All the unfamiliar titles are red links because the vandalism has been reverted and the redirects were deleted. This issue is bugzilla:13602. --MZMcBride (talk) 07:01, 15 March 2010 (UTC)

Ah, now I understand! Thanks so much for your help! Herostratus (talk) 10:57, 15 March 2010 (UTC)

"sig=yes" not working on uw-ablock

The "sig=yes" parameter is not working on {{uw-ablock}} - it just produces four tildes instead of a proper signature. JohnCD (talk) 13:39, 15 March 2010 (UTC)

You need to subst the template, then it should be fine - Kingpin¹³ (talk) 13:45, 15 March 2010 (UTC)

(Strikes self on head with large trout) Thanks! JohnCD (talk) 14:09, 15 March 2010 (UTC)

Can expr not take a supplied parameter?

If I want to insert one of the parameters as a hard-coded value in another parameter, I am putting

{{subst:#expr:{{{population_total}}}}}

but am getting an error. –xeno^talk 15:58, 15 March 2010 (UTC)

Yes, this is a known issue, see Help:Substitution. I typically just run a script to expand them myself when I need to. Plastikspork _―Œ^(talk) 16:23, 15 March 2010 (UTC)

Thanks for the quick response =) –xeno^talk 16:29, 15 March 2010 (UTC)

Character change count on history pages, just like on watchlists?

I've always wished that page histories would display the number of characters added or deleted from revision to revision, just like watchlists. Is there any reason they don't? AzureFury (talk | contribs) 23:08, 10 March 2010 (UTC)

I agree: that would be very helpful. - Pointillist (talk) 23:17, 10 March 2010 (UTC)

No particular reason they don't. They display the size in bytes instead. We probably don't want to display both the size and change, it would be too cluttered. Actually, the history page is probably already much too cluttered. —Aryeh Gregor (talk • contribs) 23:23, 10 March 2010 (UTC)

I think characters are more informative than bytes, so if given a choice, I'd replace the latter with the former. AzureFury (talk | contribs) 02:44, 11 March 2010 (UTC)

That was my opinion, IIRC, but Brion disagreed. He pointed out that the exact number wasn't important anyway, so being off by a small fixed factor for a given language doesn't really matter. —Aryeh Gregor (talk • contribs) 19:31, 11 March 2010 (UTC)

Guess you're not too good with mental arithmetic? =) I think a script could probably be written to provide this functionality. –xeno^talk 23:58, 10 March 2010 (UTC)

I agree too, this is just what's needed—I was thinking the same thing recently. The total size is almost incidental; it's the number of characters added/deleted that's of immediate interest about each change. The watchlist view (especially aided by the colours) provides this information directly in a single history record in just the right way. ("Guess you're not too good with mental arithmetic" indeed! PL290 (talk) 09:02, 11 March 2010 (UTC)

It sounds like we have a consensus, do we at this point post this somewhere else? AzureFury (talk | contribs) 12:29, 11 March 2010 (UTC)

Consensus to change mediawiki default behaviour cannot be had by a handful of users on the pump. –xeno^talk 19:36, 11 March 2010 (UTC)

I also agree it would be helpful. One side question, if the history shows no page size, how can a page size of an article be found? Regards, SunCreator (talk) 12:36, 11 March 2010 (UTC)

To get total size, not change size, one way is to add Doc PDA's script to your monobook.js. PL290 (talk) 13:20, 11 March 2010 (UTC)

My userhist script does this by the way, for the first 30 revisions on any history page (you can change that number). Ale_Jrb^talk 13:31, 11 March 2010 (UTC)

Thanks, sounds useful--I'll take a look. PL290 (talk) 13:35, 11 March 2010 (UTC)

In English, size in bytes is almost always going to be the same as size in characters (see UTF-8), so I don't see any point in changing it. OrangeDog (τ • ε) 13:05, 12 March 2010 (UTC)

Given any string s, s.length will give you the character-count and encodeURIComponent(s).replace(/\%../g, "_").length will give you the byte-count. ―AoV² 10:33, 14 March 2010 (UTC)

I think that there has been a misunderstanding here. The original question is not about bytes versus characters, it is about showing the change in page size (as per the watchlist) versus the total page size (as per the page history). Consider this edit - in watchlist, it shows as "(+246)", whereas in page history it shows as "(80,847 bytes)". The original poster merely wants consistency. --Redrose64 (talk) 12:36, 16 March 2010 (UTC)

Toolserver.org

Toolserver WikiProject assessment seems to have a problem of some sort at the moment. As it gives a message Bad Gateway. Regards, SunCreator (talk) 12:49, 15 March 2010 (UTC)

Still down or down for me. Can anyone else confirm this behavior? Regards, SunCreator (talk) 01:19, 17 March 2010 (UTC)

The server committed a protocol violation

Does anyone have any idea why I would be getting occasional:

The server committed a protocol violation. Section=ResponseStatusLine

responses from en.wikipedia?
— V = IR (Talk • Contribs) 01:52, 16 March 2010 (UTC)

Using what client? How often? Making an edit, viewing a page, doing what? --MZMcBride (talk) 02:17, 16 March 2010 (UTC)

Saving a page to the server, using C#. I figured it out though. Had to switch down to use HTTP version 1.0 (.NET defaults to HTTP 1.1). I'm not really sure why that matters, and honestly not sure that I care (although I'm curious), but it works fine now.
— V = IR (Talk • Contribs) 03:00, 16 March 2010 (UTC)

Are you giving a user agent with your headers? Peachey88 ^{(Talk Page · Contribs)} 09:45, 16 March 2010 (UTC)

I do know failing to provide a user-agent can cause 403 error:

owner@lappy:~$ wget "http://en.wikipedia.org/wiki/WP:VPT" --user-agent=
--2010-03-16 __:23:16--  http://en.wikipedia.org/wiki/WP:VPT
Resolving en.wikipedia.org... 208.80.152.2
Connecting to en.wikipedia.org|208.80.152.2|:80... connected.
HTTP request sent, awaiting response... 403 Forbidden
2010-03-16 __:23:16 ERROR 403: Forbidden.
owner@lappy:~$
owner@lappy:~$ wget "http://en.wikipedia.org/wiki/WP:VPT" --user-agent=makesomethingup/1.1
--2010-03-16 __:24:58--  http://en.wikipedia.org/wiki/WP:VPT
Resolving en.wikipedia.org... 208.80.152.2
Connecting to en.wikipedia.org|208.80.152.2|:80... connected.
HTTP request sent, awaiting response... 301 Moved Permanently
Location: http://en.wikipedia.org/wiki/Wikipedia:VPT [following]
--2010-03-16 __:24:58--  http://en.wikipedia.org/wiki/Wikipedia:VPT
Connecting to en.wikipedia.org|208.80.152.2|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 178124 (174K) [text/html]
Saving to: `Wikipedia:VPT'

―AoV² 10:37, 16 March 2010 (UTC)

Because if you say you are using HTTP 1.1 (which you probably should be), you have to actually follow the specification (rfc 2616). OrangeDog (τ • ε) 13:08, 16 March 2010 (UTC) Sorry, didn't realise our servers were so '90s. OrangeDog (τ • ε) 23:17, 16 March 2010 (UTC)

Most Wikimedia servers do not support HTTP 1.1, that's why they send HTTP/1.0 in their response headers. I believe C# sends "Expect" headers, which the current version of Squid responds to with "417 Expectation Failed". Squid predates HTTP 1.1, and development work on adding support for it has been proceeding very slowly.-- Tim Starling (talk) 19:31, 16 March 2010 (UTC)

Yup, that's the problem right there. Kind of a pain, but it's an easy fix. It would be nice if the software were updated, but... *shrug*
— V = IR (Talk • Contribs) 00:59, 17 March 2010 (UTC)

on white-space (2)

Consider the following:

input	output	html
foo{{code|int main(void);|cpp}}bar	fooint main(void);bar	foo <code><span class="mw-geshi cpp source-cpp"><span class="kw4">int</span> main<span class="br0">(</span><span class="kw4">void</span><span class="br0">)</span><span class="sy4">;</span></span></code>bar

Note the space after “foo” in the output. This, despite that Template:Code contains no spaces in the “onlyinclude” portion. How might one make this go away? ―AoV² 05:48, 16 March 2010 (UTC)

Not sure if this is a typo in SyntaxHighlight_GeSHi::parserHook():

      if ( $enclose === GESHI_HEADER_NONE ) {
        return '<span class="'.$lang.' source-'.$lang.'"> '.$out . '</span>';
      } else {
        return '<div dir="ltr" style="text-align: left;">' . $out . '</div>';
      }

The code above seems to insert a space after the opening span of inline tags. Taking a wild guess, maybe HtmlTidy is moving the space outside the span, which would then explain the space added in the output before the tags.

— Richardguk (talk) 08:25, 16 March 2010 (UTC)

Based on the spacing convention of the else path, I′d tend to agree this is unintentional. It does appear to render without an extra space when $enclose != "none". This largely concerns me because I had a {{code}}-wrapped word immediately after a left-parenthesis of default font. In the worst case, I could rephrase the sentence. ―AoV² 08:54, 16 March 2010 (UTC)

Yes, i believe one of the things that htmltidy does, is to move spaces inside spans, to the outside. —TheDJ (talk • contribs) 12:23, 16 March 2010 (UTC)

bugzilla:22848 —TheDJ (talk • contribs) 12:42, 16 March 2010 (UTC)

syntaxhighlight (2)

In loosely related news (though I′m sure it′s some kind of bug in the third-party mw:Extension:SyntaxHighlight GeSHi), I noticed that this:

<syntaxhighlight lang="foo">0</syntaxhighlight>

will render as a box containing an empty string (well, actually a single '\n') regardless of the selected language. I suppose I could special-case this in the {{code}} template to force display of the zero, but I′d like to know whether any other known situations exist where this extension has null output (with non-null input). ―AoV² 08:44, 16 March 2010 (UTC)

Note, I was able to fix this (in the {{code}} template at least) by adding padding the parameter with zero-width space. Let me know if anyone has a better way. ―AoV² 09:35, 16 March 2010 (UTC)

leave edit page dialog

In beta, when you try to leave an edit page after making changes (without saving), you get a dialog box which says "Leaving this page may cause you to lose any changes you have made.
If you are logged in, you can disable this warning in the Editing section of your preferences.", with two buttons underneath it, which say "Leave this page", and "stay on this page". How can I make such dialogs using js? It seems to be a modification of confirm(). Thanks, ManishEarth^{Talk • Stalk} 15:26, 16 March 2010 (UTC)

Edit bars

Is there a script or a way to have the "Wiki Markup" option at the bottom of an edit window and the edit bar in the My preferences working at the same time. I can get one or the other but not both : note they will work at the same time during the same session, but once you log out and return the wiki markup is the overridden by the check box in My preferences. Mlpearc MESSAGE 17:37, 16 March 2010 (UTC)

Bug with pipe trick?

There may be a bug with the pipe trick. On my next-to-last edit to checkmate, I tried [[resign (chess)|]] but it put it in there literally instead of linking. I don't know if it failed to work because it is in a note. Bubba73 ^{(You talkin' to me?)}, 18:33, 16 March 2010 (UTC)

You're right, and it's a known issue. –xeno^talk 18:36, 16 March 2010 (UTC)

Generally a good idea to never rely on the "pipe trick." It fails in a lot of cases, as xeno noted. --MZMcBride (talk) 18:38, 16 March 2010 (UTC)

Resolved

Thanks, I didn't know it was already known. The "pipe trick" is so nice when it works. Bubba73 ^{(You talkin' to me?)}, 18:41, 16 March 2010 (UTC)

A lot of things do not work when placed in <ref> markup. There's a bug about it in MediaWiki's Bugzilla, that's been there for five years. The pipe trick works fine elsewhere, just like most other things outside reference markup. Gary King (talk) 18:47, 16 March 2010 (UTC)

Reference statistics bot

I'm interested in gathering statistical information about which sites/sources are most commonly referenced on Wikipedia, in order to gain a better understanding of Wikipedia's systemic bias. The only such statistics I have been able to find are these, which are very outdated (Nov. 2006).

Essentially, the questions I'm looking to answer right now are the following:

1. Which external sites are most commonly referenced on Wikipedia as a whole, both in external links and in inline citations? This could be done by gathering all of the external links from each page, and tallying them by domain name.
2. For each of the categories in Portal:Contents/Categorical index, which external sites are the most commonly linked to in articles in these categories?

I've got a bit of PHP and Javascript experience, but would need technical assistance getting a bot put together to do this, since I've never written a WP bot before. So that I'm not reinventing the wheel, can anyone recommend bots that someone has created that do things similar to the above that I could modify, and/or people who I might go to with questions?

Thanks -- Jrtayloriv (talk) 18:53, 16 March 2010 (UTC)

If you are thinking about a bot that downloads all the articles on Wikipedia one after another and analyzes them, then that is prohibited, because it wastes too much resources. You should instead download the database dump in XML format and analyze that. I think it should be quite easy to do. Also, if you want really precise statistics, templates like {{imdb title}} could cause you problems, because they don't look like an external link in the source code, but render as one. Svick (talk) 19:12, 16 March 2010 (UTC)

Thank you Svick -- I am downloading the dump at the moment, and it looks like it will be rather easy to do with the currently existing Perl modules available on CPAN. Appreciate your help! -- Jrtayloriv (talk) 19:54, 16 March 2010 (UTC)

Thumbnailing of animated GIF's

I'm wondering about the current state of play on this. Animated GIF's don't seem to get thumbnailed, so they load on pages at their raw upload size as far as I can tell. This is a huge issue for dial-up users, a few GIF's I've seen on the Science Reference Desk are 500K-1MB in size regardless of the scaling parameter. (discussion) Looking through the archives and various Bugzillas, I can see discussion on scaling the first frame of the animation and turning the whole thing off because it pooches the scaling server. So just at the moment, does the first frame get thumbnailed or not? And are there any plans afoot to scale every frame, or is that never going to happen? Franamax (talk) 22:49, 16 March 2010 (UTC)

Commons:Graphics village pump ¦ Reisio (talk) 00:39, 17 March 2010 (UTC)

That will be a handy link to bookmark. Thanks! Franamax (talk) 03:36, 17 March 2010 (UTC)

inline NavFrame collapsing?

I was thinking about creating a template for inline collapsing. basically this would involve the NavFrame tools using SPANs rather than DIVs - clicking the 'show' link would expand the NavContent SPAN rightward or leftward inline, rather than up or down on a separate line. It would be useful (IMO) for tucking away material in the middle of a paragraph without disrupting the flow of the paragraph itself (e.g. hiding disruptive, meandering, or off-topic comments while retaining the sense of the text). However, this would require some modification of MediaWiki:Common.css and MediaWiki:Common.js - at a quick glance we'd need new CSS classes for span.NavFrame, span.NavHead, span.NavContent, and we'd need to modify the javascript file, which currently only looks at divs.

My question is whether this be a worthwhile addition, or at least worthwhile enough to justify the effort? what do you all think? --Ludwigs2 00:34, 17 March 2010 (UTC)

Firefox 3.6 - loss of edit window content

Resolved

 – Redrose64 (talk) 16:19, 17 March 2010 (UTC)

I upgraded from Firefox 3.0.18 to 3.6 this morning. I have made eight edits since then, and of those, on two of them I experienced the following problem:

• click edit, the edit window fills with text
• make a change
• click "show preview"
• the upper part of the screen is blank, and the edit window is now entirely empty - both my changes and the original section content have disappeared

Is this a Firefox bug, or MediaWiki? --Redrose64 (talk) 13:31, 14 March 2010 (UTC)

I have used Firefox 3.6 on Windows Vista with hundreds of edits and have never experienced this. You could try reinstalling, or logging out and testing Show preview many times to see if something in your account settings affects it. PrimeHunter (talk) 13:45, 14 March 2010 (UTC)

You aren't using Beta (clicked the "Try Beta" link) or wikEd (Gadgets tab in Preferences), are you? Not sure about now, but those are problematic (at least they used to be). wikEd in particular. I stopped using it weeks ago in favor of the "sans-serif font" Gadget because of copy/paste issues in Firefox 3.6. PleaseStand ^(talk) 16:10, 14 March 2010 (UTC)

No gadgets. No beta. No vista (Windows XP). Tried logging out and logging in again, and it's happened oncetwice again; this timeboth these times on a "Save page" - clicking that gave the same effect as above (didn't save, but behaved as if I'd used "Show preview", and cleared the edit window). Regarding Beta: when it first became available, I tried it for about two hours and then gave up because there was little discernable difference between blue (unvisited) links, purple (visited) links and black text. --Redrose64 (talk) 16:24, 14 March 2010 (UTC)

By "the edit window is now entirely empty", do you mean that you are left staring at a completely white screen, without even the Wikipedia logo? There are various issues that can cause this, including having a particular Firefox extension installed. And of course don't forget the possibility of malware causing problems. PleaseStand ^(talk) 00:44, 15 March 2010 (UTC)

It's not a completely white screen. What I mean is that the edit box itself is completely empty. To simulate the effect that I see: use any graphic-based browser (firefox, IE, chrome you name it), go to any page, edit any section, delete everything in the edit box, click "Show preview". That is what I see approximately one in every four edits that I make. It never happened when I used Firefox 3.0.18, and only began after installing Firefox 3.6 - and it began immediately. It is definitely not my preferences, because yesterday I logged out before shutting down, then today I turned on, started Firefox, went to Wikipedia, picked an article, made changes, went for "show preview", repeated last 2 steps 3 more times and- bang! All gone. That was as an IP user. --Redrose64 (talk) 21:15, 15 March 2010 (UTC)

I was able to reproduce your problem, thanks to Mozilla bug 547525. The fix? Go to the Firefox preferences, Advanced section, Network tab, and Settings button. Make sure that the "no proxy" option is checked ("auto-detect" is the faulty setting). If the correct setting is already selected, do another check. Type about:config into the address bar. Click the button. In the Filter box type network.proxy. Right-click network.proxy.type and click Modify. Change it to 0 (the faulty setting will probably be 4). Wikipedia should load much faster and work much more reliably. And this issue is supposed to be fixed in the next update to Firefox 3.6. PleaseStand ^(talk) 22:13, 15 March 2010 (UTC)

Since the above, I got a worse error: one time only, on going for "Save page", it didn't just blank the edit box, but also threw:

Warning: An administrator deleted this page since you started editing it. Please check the deletion log to see the reasoning.

on a pink background; however on going to the page in a different tab, it definitely wasn't deleted.

Anyway, have now changed setting from "Auto-detect ..." to "No proxy". Bed-time now, see if it behaves tomorow. --Redrose64 (talk) 23:27, 15 March 2010 (UTC)

The above has fixed the problem. No trouble for over 24 hours. --Redrose64 (talk) 16:19, 17 March 2010 (UTC)

Help with javascript-based notice

Can I ask someone to help me prototype a proposed notice for BLP pages? The idea is to put a special "contact us" form at the bottom of such pages. It should be implemented something like Template:BLP editintro, but appear on the article page itself, not the "editing" page.

The code for displaying the BLP editintro is here (and the currently live code for it is in MediaWiki:Common.js). I'm afraid my javascript skills are not up to the task of adapting it to the request I'm making here.--Father Goose (talk) 05:33, 15 March 2010 (UTC)

What's wrong with templates and bots ? —TheDJ (talk • contribs) 12:25, 15 March 2010 (UTC)

It shouldn't go in the article proper. See down at the bottom of this page where it says "Contact us"? We want to replace that text with something a little more verbose when the page has the BLP category on it.--Father Goose (talk) 19:44, 15 March 2010 (UTC)

OK, this is NOT tested for IE or with anything but Vector. But I have created something here, that you can test with this link. It loads the HTML from Template:BLP footer (including "noinclude" elements, so don't add documentation). It blocks on loading this notice, so it will currently slow pageloading (might be fixed later, but makes it more complicated). —TheDJ (talk • contribs) 22:13, 15 March 2010 (UTC)

Outstanding. I think I can tweak it from there. Thanks so much for your help. Looks good so far under Monobook, IE 8, FF 3.6.

The place where this is being developed is at Template talk:BLP help, by the way, with prior discussion at Wikipedia_talk:Biographies_of_living_persons#Helping_non-editors_fix_BLPs.--Father Goose (talk) 23:36, 15 March 2010 (UTC)

Confirmed that it works in Monobook. Link here. —MC10 (T•C•GB•L) 05:03, 16 March 2010 (UTC)

Thanks, TheDJ. That's very helpful. Maurreen (talk) 19:26, 17 March 2010 (UTC)

Any progress on a __NOTIDY__ magic word?

This thread is follow-up to Wikipedia:Village pump (technical)/Archive 71#Template:Blockquote --Redrose64 (talk) 12:26, 16 March 2010 (UTC)

I while back I ran into a problem where div-based templates would break if indented via wikitext. The problem was (apparently) that the HTML tidy function was throwing in spurious closing tags for the divs in cases where a div is placed inside a definition list and contains a carriage return. Someone at that discussion suggested that we could use a NOTIDY magic word to keep tidy from mucking things up.

well, I ran into the problem again, and I'm curious whether any progress has been made on the issue. does anyone know? --Ludwigs2 23:58, 15 March 2010 (UTC)

T24786 was recently opened. ---— Gadget850 (Ed) ^talk 00:09, 16 March 2010 (UTC)

ok, I've added a comment. I find myself wondering if it's just a misconfigured tidy installation, though - tidy just seems to be over-eager about closing tags when it runs into carriage returns. that question is out of my league, though... --Ludwigs2 00:36, 16 March 2010 (UTC)

Quite fresh bug report. From my experience, don't expect a solution any time soon. And, by soon I mean the next two to three years. Things in mediawiki seem to be moving slowwwww... I voted btw. --JokerXtreme (talk) 00:38, 16 March 2010 (UTC)

And, while we're at it. Please vote for this too:[7] --JokerXtreme (talk) 00:39, 16 March 2010 (UTC)

that's because the entire technical side is staffed by beetles; why do you think they call it bugzilla? Beetles are efficient, but never hasty. --Ludwigs2 00:59, 16 March 2010 (UTC)

I prefer ants. Equally efficient and get the job done quicker:P BTW you should use the vote option on those bugs. Bugs that have more votes are prioritized (theoretically). --JokerXtreme (talk) 10:11, 16 March 2010 (UTC)

Let's get these sorted first. OrangeDog (τ • ε) 13:10, 16 March 2010 (UTC)

Putting _NOTIDY__ on a template wouldn't be very helpful; we would still need to tidy the HTML for the articles the template is included in , in case other things on the page were incorrect. The underlying solution is to fix Mediawiki so that the HTML it produces is acceptable to the tidy program, in these cases. — Carl (CBM · talk) 13:47, 16 March 2010 (UTC)

well, as I said above, it may be something as simple as changing the tidy prefs - the problem is that tidy is adding tags where logically it shouldn't, so clearly tidy is parsing the text incorrectly to begin with. I mean, I assume HTML tidy has a 'settings' file where there is at least some control over its behavior. does anyone know where that file would be? I deeply suspect that what's happening here is that tidy is expecting definition data blocks to be single paragraphs, and so is mindlessly retagging things to make sure that definition data blocks are single paragraphs (that seems to be the effect, at any rate). --Ludwigs2 16:27, 16 March 2010 (UTC)

I checked this, it's not tidy, it's the wikicode parser or the sanitizer that simply adds dd or li endtags, regardless of wether it should (basically a side effect of the limited syntax used to create lists in wikicode). It might be that setting tidy's "ignore: auto" in the config might improve it a BIT, but it doesn't solve the core issue. —TheDJ (talk • contribs) 16:32, 16 March 2010 (UTC)

It would be useful to document known issues with HTML Tidy— either a section at Help:Markup validation or its own help page. ---— Gadget850 (Ed) ^talk 16:44, 16 March 2010 (UTC)

Checked out how this works. The problem is in how doBlockLevels( $text, $linestart ) of the parser works. It parses line by line, and when ever the next line doesn't start with any of the wikicode list tokens, the list is ended. Regardless of the HTML that follows such a token (it simply doesn't come into play at all). So this part of the parser is what you could call "dumb" with regard to html. This really is a complicated bug to solve, without breaking anything else, or changing the wikicode syntax. —TheDJ (talk • contribs) 18:31, 16 March 2010 (UTC)

can you provide a link to the code? I'm curious whether there's some way to scam or sidestep it - basically some way of encoding a block of text so that the parser doesn't realize it's reached an end-of-line, or is fooled into thinking a correct token follows, or something like that. or maybe this would be the place for a magic word - __NOBLOCKEND__, say - that would keep the parser from crapping out. heck, mostly I'm just curious about the code. . --Ludwigs2 23:39, 16 March 2010 (UTC)

Search for "function doBlock" in Parser.php. —Aryeh Gregor (talk • contribs) 17:32, 17 March 2010 (UTC)

Got it, thanks. --Ludwigs2 20:22, 17 March 2010 (UTC)

Simple Change Proposal

This is not a major issue, but a minor annoyance that could easily be fixed. When I visit Wikipedia, I have to click in the query box in order to do a search. The functioning of the website could be much more streamlined if the search box were automatically selected, as for most users, the only typing done on the website, especially on the home page, is to search for an article. Thank you for your time, Xyon. —Preceding unsigned comment added by Xyonhurst (talk • contribs) 00:43, 17 March 2010 (UTC)

No, see the second item at the FAQ for this page. Graham87 01:18, 17 March 2010 (UTC)

But there is a gadget that will do just that: Special:Preferences --> gadgets tab --> check the Focus the cursor in the search bar on loading the Main Page. box --> save and bypass your browser cache. – ukexpat (talk) 01:34, 18 March 2010 (UTC)

Help required in creating a code

Hi. The RfC discussions on Vandalism Talk with respect to a specific method of temporarily auto-blocking anonymous IP addresses has reached a general agreement that we could go on to the test phase to see false positive results. Would an administrator/code writer be able to help us in creating a code to implement the proposal and see the test results? Thanks ▒ ♪ ♫ Wifione ♫ ♪ ▒ _―Œ ^{♣Łeave Ξ мessage♣} 06:00, 17 March 2010 (UTC)

We'll need some1 at MediaWiki to make an extension for it... I'm gonna start at MW soon, but not now (I hafta first learn PHP, then learn my way around MW, &so on&so forth). Why not request it there? ManishEarth^{Talk • Stalk} 11:43, 17 March 2010 (UTC)

You don't need an extension to do this, nor do you need any changes to MediaWiki itself. From (what I read on the proposal) you want a bot to block users if N number of rollbackers have reverted them. That's something for a bot or toolserver script to do, not an extension. As a minor sidenote: I don't think that got nearly the feedback it needs, a proposal to allow non-admins to have a direct role in blocking is a pretty far-reaching change. I'm going to drop a note on WP:AN pointing here. ^demon^{[omg plz]} 13:41, 17 March 2010 (UTC)

if then parser functions / magic words

I am attempting to create a template which does the following:

Editor adds only one parameter:

{{User:DASHBot/Wikiproject|WikiProject Syria articles}}

Template creates:

Unreferenced BLPs from [[Category:WikiProject Syria articles]] are provided at [[Wikipedia:WikiProject Syria/Unreferenced BLPs]].

Editor adds two parameters:

{{User:DASHBot/Wikiproject|WikiProject Syria articles|New Blps}}

Template creates:

Unreferenced BLPs from [[Category:WikiProject Syria articles]] are provided at [[Wikipedia:WikiProject Syria/New BLPs]].

Having only one parameter is easy:

Unreferenced BLPs from [[:Category:{{{1}}}]] are provided at [[Wikipedia:{{{1}}}/Unreferenced BLPs]].

But how do I make an if/then parser function which says, if there is a second parameter do this:

Unreferenced BLPs from [[:Category:{{{1}}}]] are provided at [[{{{2}}}]].

I know that the second part of the coding should be something like:

{{#if:{{{2|}}} |<!--then:-->[[:Category:{{{1}}}]] are provided at [[{{{2}}}]]}}.

I think I need some kind of if-then-else coding. Maybe #ifeq?

thanks in advance! Okip 08:50, 17 March 2010 (UTC)

I think you're basically there, aren't you? You can do it like this:

{{#if:{{{2|}}}|text for if there are two parameters|text for if there is one parameter}}

--Kotniski (talk) 09:16, 17 March 2010 (UTC)

(edit conflict) Is User:DASHBot/Wikiproject/sandbox what you're looking for? It's not quite as you asked, since some of the things you said contradicted each other. So instead of {{User:DASHBot/Wikiproject|WikiProject Syria articles}} you use {{User:DASHBot/Wikiproject|WikiProject Syria}}, and instead of {{User:DASHBot/Wikiproject|WikiProject Syria article|New Blps}} you use {{User:DASHBot/Wikiproject|WikiProject Syria|Wikipedia:WikiProject Syria/New Blps}}. I can change the latter to be {{User:DASHBot/Wikiproject|WikiProject Syria|New Blps}}, with the out put of where they are listed still being Wikipedia:WikiProject Syria/New_Blps, but my way provides a little bit more functionality (for example, you could still get it to work if the category and WikiProject have different names. - Kingpin¹³ (talk) 09:17, 17 March 2010 (UTC)

P.S. I've also made so that you can hide the "articles" suffix to the category by entering the parameter |noarticles=yes - Kingpin¹³ (talk) 09:23, 17 March 2010 (UTC)

Weird "Retrieved..." text at bottom of article

I'm seeing "Retrieved from "http://en.wikipedia.org/wiki/Wikipedia_talk:WikiProject_Dungeons_%26_Dragons" Categories: Project-Class Dungeons & Dragons articles | NA-importance Dungeons & Dragons articles | Project-Class Dungeons & Dragons articles of NA-importance | Non-article Dungeons & Dragons pages" at the bottom of aforementioned article, and the cats aren't formatted correctly. Is it just me? - Peregrine Fisher (talk) 00:48, 18 March 2010 (UTC)

Sounds like you are bothered by a non-complete download of the page CSS. Happens sometimes. Just clear your browsercache and you should be fine again. —TheDJ (talk • contribs) 01:01, 18 March 2010 (UTC)

Geo coordinates query

How do I edit geo coords (they aren't on the main edit page and I haven't time to work it out from templates). As far as I can tell the coords given for Brockley, Suffolk are for the wrong Brockley in Suffolk referring to the text (there are at least three around here and the geo coords are not 7 miles south of Bury St Edmunds or on the named road, the text refers to this one http://maps.google.co.uk/maps?f=q&source=s_q&hl=en&geocode=&q=brockley+suffolk&sll=51.398386,-2.760008&sspn=0.03047,0.090895&layer=t&ie=UTF8&hq=&hnear=Brockley,+Bury+Saint+Edmunds,+Suffolk,+United+Kingdom&z=14&lci=org.wikipedia.en whereas the geocords are for http://maps.google.co.uk/maps?f=q&source=s_q&hl=en&geocode=&ie=UTF8&hq=&t=p&ll=52.120631,0.507088&spn=0.059973,0.181789&z=13&iwloc=lyrftr:org.wikipedia.en,10974011200960407957,52.099968,0.517044&layer=t&lci=org.wikipedia.en). --BozMo talk 13:14, 17 March 2010 (UTC)

They are on the edit page; they're the values of "latitude" and "longitude" under "Infobox UK place".--Kotniski (talk) 13:55, 17 March 2010 (UTC)

I've emended the coordinates in the article. Deor (talk) 15:55, 17 March 2010 (UTC)

Thanks both --BozMo talk 22:15, 18 March 2010 (UTC)

Photo upload log

How can I find a log of all the photos I've uploaded? DRosenbach ^{(Talk | Contribs)} 03:00, 18 March 2010 (UTC)

Hows this [8]?--Jac16888^Talk 03:02, 18 March 2010 (UTC)

Or, in more general terms, go to "my contributions" at the top of the page; and at the very bottom of which you will find a link titled "Files uploaded". Note that if you have uploaded to Commons:, you need to check the uploads there as a separate operation. --Redrose64 (talk) 15:07, 18 March 2010 (UTC)

Tool to automate the removal of wikilinked common words?

Out of interest, are there any tools out there that automate or assist the process of de-linking plain English terms (per WP:OVERLINK) from articles? This JS script does it for a slim category of common terms (countries, major cities and simple professions), but it'd be good to have something that drew from a wider pool of words that should very rarely be wikilinked. (I'm balking at the idea of cleaning up an article like QI (G series) by hand, which has dozens of inappropriate wikilinks for a simple words like "red", "shoe" and "eat".) --McGeddon (talk) 11:17, 18 March 2010 (UTC)

I suppose that script could just be copied and modified to use a different list of de-linkable terms, actually, if such a list existed. Does it exist? Or does anyone have a bright idea about how to generate such a list? --McGeddon (talk) 11:19, 18 March 2010 (UTC)

Greetings, R. wikt:Wiktionary:Frequency lists/TV/2006/1-1000 and so on are lists of words most common on television (to speakers other than George Carlin, I think). As a minimal sanity-check I′d suggest that any uh… hypothetical… bot… should skip links containing a capital letter. A sentence like “Owls, unlike other birds of prey, are able to…” I′d leave well enough alone (being the subject of the sentence and all). Likewise proper nouns such as “Jesus” despite being taken in vain numbers which suggest commonality. Best bet? Start with the verbs. ―AoV² 12:27, 18 March 2010 (UTC)

ref desk header oddity

for anyone with sysop status and a little free time, there's a bugaboo or two on the reference desk header - Wikipedia:Reference desk/header - that need sorting out. the easy one is this: there is a link on the top right of the page name header bar that says "WP:RD/?" (the last character will vary according to the desk) this link is (first) senseless, since it always links back to the page you are currently on, and (second) unsightly if you've enabled the first-section edit link option - the edit link overwrites the ref desk link. the line of code that makes this link on lines 3-4 of the page:,

<span style="position:absolute; top: 1.5em; right:20px;"><small>
[[{{{1|}}}]]</small></span>

you can probably just delete it, since it does nothing except make that (pointless) link. or maybe you would want to move it into a {{shortcut}} box?

the more troublesome issue (and the reason why I'm posting this here) is that the talk page for the ref desk header got co-opted as a content page for the reference desk talk pages. there is no place to discuss the ref desk header or the ref desk talk page header. really, the two should be split apart and put on their own separate pages. I'll try do that myself, but I don't know if I'll run into links on other protected pages, so I may need some assistance with the task. I'll post back here if it looks like I'll run into trouble. --Ludwigs2 04:10, 19 March 2010 (UTC)

Second issue is fixed - it turned out to be trivial - and I've requested protection on the new page Wikipedia:Reference desk/talk header. --Ludwigs2 04:28, 19 March 2010 (UTC)

scratch everything. now that the talk page is freed up, I realized I can {{editprotected}} the header properly. sorry for babbling on... --Ludwigs2 05:18, 19 March 2010 (UTC)

Why is autoconfirmed not at Special:ListUsers

Can anybody point me to the (Special:)Page which shows if a user is autoconfirmed or not? I'm pretty sure there is one, but can't find it or be bothered to look hard ;). Also, why is it not just included in Special:ListUsers? Presumably some technical reason :). Cheers, - Kingpin¹³ (talk) 23:51, 16 March 2010 (UTC)

autoconfirmed is not a true group. You could call it a status. On every action of a user, it is determined if at that point the user has "autoconfirmed" or not. You can gain it and loose it automatically (connect with TOR for instance). It is possible to explicitly make people confirmed btw. There are no tools to see if a user is autoconfirmed I think, but the user can look at the source of a page for "wgUserGroups" and it will show if the user is autoconfirmed at the moment he requested the page. —TheDJ (talk • contribs) 00:39, 17 March 2010 (UTC)

Ahh, I see that's true, but it's only for the user account which you're logged into, I'm asking because I was wondering if there would be a way for User:KingpinBot to recognise when confirmed user requesters already have autoconfirmed? - Kingpin¹³ (talk) 09:02, 17 March 2010 (UTC)

No, it's personal and action based, so another user cannot know. It does not exist from the viewpoint of other users, it only exists for actions of your own account. It might be possible to guess, but again, I'm not aware of such tools. —TheDJ (talk • contribs) 09:46, 17 March 2010 (UTC)

See WP:DRACU. –xeno^talk 12:09, 17 March 2010 (UTC)

I think it is possible to determine if the user is autoconfirmed. The Edit Filter can do this, for instance. Autoconfirmed is a true group, but it is an implicit group with automatic promotion. Ruslik_Zero 17:04, 17 March 2010 (UTC)

It would be perfectly possible to tell whether a user is autoconfirmed, just have something expose the info from $user->getEffectiveGroups() somewhere. Of course it might vary based on the time of day or who knows what, but explicit groups can vary too, via explicit granting and revocation. We don't seem to do this anywhere, though. It's possible it would leak information about the user, like whether their e-mail address is confirmed, based on the particular wiki's configuration. Also, as pointed out, enwiki's autoconfirmed configuration works different for TOR users, so that could be confusing.

Looking at this, it seems like the current code for Special:ListUsers goes out of its way to avoid computing groups for most rows. It'd be pretty easy to get it to display all groups, including implicit ones, but I'm not sure what the performance implications are, so I've left it alone for now. —Aryeh Gregor (talk • contribs) 17:28, 17 March 2010 (UTC)

We had discussed this several times, and after the last discussion, I opened a request here to turn it into an implicit usergroup with autopromotion, but those requests take ages to being processed. Cenarium (talk) 23:47, 19 March 2010 (UTC)

Unable to edit Template:Infobox animanga/Print

While trying to update some fields in {{Infobox animanga/Print}} (commenting out some deprecated fields), I found that I am unable to edit the template. The template is semi-protected, however, the semi-protection hasn't been a problem for me in maintaining this template before. I'm not sure what is wrong and why I am unable to edit it. I can still able to edit other template components, such as {{Infobox animanga/Video}} and {{Infobox animanga/Audio}}. —Farix (t | c) 22:21, 18 March 2010 (UTC)

Is it giving an error message or anything? –xeno^talk 22:25, 18 March 2010 (UTC)

Template_talk:Documentation#.2FPrint talks about this. There's some sort of special protection in place, but I'm not familiar with it, nor do I understand why it was placed. Equazcion ^(talk) 22:26, 18 Mar 2010 (UTC)

Based on this, perhaps just use the {{editprotected}} template at the talk page. –xeno^talk 22:32, 18 March 2010 (UTC)

Perhaps a solution might be to move it to another name, say Template:Infobox animanga/print and just leave a redirect behind and then the protection wouldn't be an issue anymore. -- WOSlinker (talk) 22:35, 18 March 2010 (UTC)

I believe the protection is in place for a reason =) –xeno^talk 22:35, 18 March 2010 (UTC)

They are not using it for this purpose though. — Martin (MSGJ · talk) 22:37, 18 March 2010 (UTC)

Pay no attention to the man behind the yellow curtain. –xeno^talk 22:39, 18 March 2010 (UTC)

That will be entirely inconsistent with the naming scheme used the other infobox components and creates confusion. I also find it ridiculous that I now have to go through someone else after maintaining this set of templates for years. Can it be fixed to where I don't have to resort to {{editprotected}}? —Farix (t | c) 23:00, 18 March 2010 (UTC)

Yes, it can be fixed by developers, the issue is bugzilla:22707. Also any admin can lift the ban (it's on the title blacklist), but I and David and a few other admins would object and a discussion would probably have to be opened on the problem. In the mean time, just use {{editprotected}} on the template's talkpage, and any admin can deploy your proposed changes. —TheDJ (talk • contribs) 23:29, 18 March 2010 (UTC)

I completely object to the protection. I find it ridiculous to have to go through someone for a template that I've worked on in one form or another for the last three years. The reason this template was created a year ago was to consolidate two print media related templates that were nearly identical, {{Infobox animanga/Manga}} and {{Infobox animanga/Novel}}. There is no good reason why I shouldn't be able to edit this template directly. —Farix (t | c) 23:41, 18 March 2010 (UTC)

The pages are protected, because the pages never take into account protection of the parent template, which some consider a security risk. You not being able to edit those pages of which the parent is not protected, is collateral damage to this, which is unfortunate. Please request the developers of the extension to fix their extension, so that you can edit unprotected templates again. —TheDJ (talk • contribs) 00:00, 19 March 2010 (UTC)

Oh, I only now see, that you are not using this for template substitution during print. You need to be aware that /Print is a reserved name currently, because any Template/Print will replace the contents of Template when creating a PDF or a Book. Requesting a template move might be the wisest choice here... —TheDJ (talk • contribs) 00:08, 19 March 2010 (UTC)

Why not restore the semi-protection for this template, especially since the "security risk" isn't even an issue for this template? It's far better than waiting forever for something that may never get done. And there really isn't a better name for the template component since it handles print media, and correlates with the template components for video, audio, and games. —Farix (t | c) 00:15, 19 March 2010 (UTC)

The problem is that it is not possible to separate protection and page creation of ALL /Print subpages, yet excepting this one as far as I know. The badtitle list is of a higher level. —TheDJ (talk • contribs) 01:01, 19 March 2010 (UTC)

So in short all further development by WP:ANIME members will come to a complete halt do to a couple of paranoid admins. Thanks for nothing. :rolls eyes: —Farix (t | c) 01:28, 19 March 2010 (UTC)

Things happen. things that ought to be resolved. They just don't resolve on your timetable. I'm sorry if that upsets you. —TheDJ (talk • contribs) 02:22, 19 March 2010 (UTC)

All I'm asking is that I and other editors from WP:ANIME be able to resume editing the template as quickly as possible. I and other project member already have the template, along with the other infobox component templates, on our watch lists. So if something funky happens, we will catch it and fix the problem quickly. I don't see why we are being bared from maintaining this template ourselves without having to jump through any annoying hoops. —Farix (t | c) 02:32, 19 March 2010 (UTC)

An administrator can simply move it to another name so it doesn't conflict with the reserved /Print name which is intended for something else. You objected to the name /print. Is there another acceptable name? I don't think it is reasonable to demand that changes are made to the general /Print scheme in order to allow that this single template can keep a name that will be confusing to others because it isn't used for the purpose /Print pages is now intended for. PrimeHunter (talk) 13:54, 19 March 2010 (UTC)

Template:Infobox animanga/Deadtree ? –xeno^talk 13:57, 19 March 2010 (UTC)

Bug 22707 is only about nonexistent pages, so maybe /Print templates should be protected when created, if necessary, and the "noedit" option removed from the title blacklist (still preventing the creation of new pages). Also maybe /Print was a bad choice of title; maybe something unlikely to conflict with existing pages or with possible titles of new pages, such as "/PrintedTemplate", "/MediaWiki.Print" or "/Print.mw", should have been used for the printed versions of templates. snigbrook (talk) 00:25, 19 March 2010 (UTC)

It is not only about creation, but that is where the biggest problem lies (because you can just pick any template used on the main page, and defile the printouts. Or citation templates would be fun to vandalize in this way as well). And there is no "protect upon creation" rule (though we could muck around with a bot, but I don't think that's really a good solution either). I agree with you that a more distinct name would probably have been better. —TheDJ (talk • contribs) 01:01, 19 March 2010 (UTC)

One way of preventing any edits other than by administrators is by transcluding into a cascade-protected page; I think that this can be used both on existing and non-existing pages. snigbrook (talk) 01:21, 19 March 2010 (UTC)

How do I transclude */Print into a page ? —TheDJ (talk • contribs) 02:22, 19 March 2010 (UTC)

I don't think it is possible to transclude everything that could possibly match that pattern, but it is possible to transclude /Print subpages of all protected templates (it's possible that several pages would be needed for the full list), or all templates with /Print subpages (currently not many). If it is necessary to protect /Print subpages of all templates, maybe it could be done before each page is created (possibly using an editnotice as a reminder), or immediately after (maybe by a bot). snigbrook (talk) 13:36, 19 March 2010 (UTC)

Time warp?

I know that I did some edits to the Citation and citation/core templates, sandboxes and discussion pages recently (after 10 March or so), but they seem to have disappeared; they don't show on my contributions page and the history pages show the last edits as being in Jan & Feb. Is there a problem with the database? Wtmitchell (talk) (earlier Boracay Bill) 05:52, 19 March 2010 (UTC)

Apparently this was my error. I was sure that I had published those changes but I apparently had not. I've just published them. Wtmitchell (talk) (earlier Boracay Bill) 05:04, 20 March 2010 (UTC)

Template help

On another mediawiki wiki, icons are used in great number to annotate things. The icons do not come singly but form a list (with no spaces between the images). I owuld like to make a template to generate the relevant code so that the parameters can be listed in any order yet the icons appear in a standard order. So I would like a list such as [[Image:Green.png|16px]][[Image:Champion.png|16px]][[Image:Pilot.png|16px]] to be generated by {{Icons|16|Pi|Gn|Ch}} where the Pi, Gn and Ch could be in any order and the first parameter 16 specifies the size. There will not be more than, say, 9 image files. How do I do this please? -- SGBailey (talk) 10:58, 19 March 2010 (UTC)

Something like this would work assuming you mean that each image has one possible position, that each position has one possible image, and that your template′s purpose and input are only to decide which ones to display or omit.

<!--     reverse-switch technique
-->{{#switch:Gn|{{{2}}}|{{{3}}}|{{{4}}}|{{{5}}}|{{{6}}}|{{{7}}}|{{{8}}}|{{{9}}}|{{{10}}} = [[File:Green.png|{{{1}}}px]]    }}<!--
-->{{#switch:Ch|{{{2}}}|{{{3}}}|{{{4}}}|{{{5}}}|{{{6}}}|{{{7}}}|{{{8}}}|{{{9}}}|{{{10}}} = [[File:Champion.png|{{{1}}}px]] }}<!--
-->{{#switch:Pi|{{{2}}}|{{{3}}}|{{{4}}}|{{{5}}}|{{{6}}}|{{{7}}}|{{{8}}}|{{{9}}}|{{{10}}} = [[File:Pilot.png|{{{1}}}px]]    }}<!--
-->{{#switch:etc|etc|etc}}<!--
-->

The above code compares several parameters to one constant (unlike the traditional exercise of comparing one parameter to several constants). On each line it will display a specific image if one or more comparisons are equal, otherwise do nothing. Should you need to introduce other pictures and letter codes (red–champion–pilot or black–champion–quarterhorse or whatever), the process would become exponentially more complicated and you′d want to use sub-templates to do it sanely. ―AoV² 12:48, 19 March 2010 (UTC)

Seems to work. Thanks. In order to abut them, I have to abut the switches else I get a new paragraph, but that's ok. -- SGBailey (talk) 16:07, 19 March 2010 (UTC)

Quick question in regards to Suppression

Does an edit have to be reverted for it to be suppressed? I.e. if users have made edits after the revision in question can it still be removed? Jeffrey Mall (talk • contribs) - 15:32, 19 March 2010 (UTC)

Sometimes, but not always. If subsequent edits were in different sections or paragraphs, and the order of sections and paragraphs was not changed, then the offending edit may still be revertable. Normally, if you click "undo", it'll tell you either "The edit can be undone" or "The edit could not be undone due to conflicting intermediate edits". Try it, and see which message you get. --Redrose64 (talk) 15:45, 19 March 2010 (UTC)

So am I correct in saying if the edit cannot be undone then the edit cannot be suppressed? Jeffrey Mall (talk • contribs) - 15:47, 19 March 2010 (UTC)

By suppression you mean oversight? Obviously the offending material cannot be in the current revision of the page. If there were many revisions between the offending revision and when it was removed, those edits will have to be suppressed as well - as collateral damage. –xeno^talk 15:50, 19 March 2010 (UTC)

(edit conflict) Sorry, what do you mean by "suppressed"? Is this not the same as "undo"? --Redrose64 (talk) 15:51, 19 March 2010 (UTC)

Here] is an oversight action that occurred about two weeks ago to Wikipedia:Help desk, because the 11:02, 3 March 2010 post included personal details. Of the four edits crossed out, the one shown as "11:07, 3 March 2010 Redrose64" was not, in fact, removed. --Redrose64 (talk) 16:01, 19 March 2010 (UTC)

Oversight yes, so to prevent this collateral damage I suppose any Admin could delete the revision and it would still pretty much remove the information from public view? Jeffrey Mall (talk • contribs) - 16:04, 19 March 2010 (UTC)

Any revisions in between when the offending material was posted and removed need to be suppressed (either through oversight or revision deletion). Regular administrative selective deletion will cause attribution problems in a case like this, so oversighters use WP:RevDel. (Regular admins are still waiting to get this new toy). –xeno^talk 16:06, 19 March 2010 (UTC)

Ok because I've had an email from a new editor who would like to have a post of theirs on one of our discussion pages removed from public view because it contains their IP but this was yesterday morning and as such quite a few other users have edited the page, would it still be possible to remove this revision without causing too many problems? Jeffrey Mall (talk • contribs) - 16:11, 19 March 2010 (UTC)

Yes, just email Special:Emailuser/Oversight with a difflink and they can easily get hide the IP per WP:RD4. –xeno^talk 16:13, 19 March 2010 (UTC)

Cheers Xeno. Jeffrey Mall (talk • contribs) - 16:17, 19 March 2010 (UTC)

Search Feature That Needs To Be Implemented

In the search box, we should be able to type in talk so that we can go to a categories talk page, or a history's talk page, etc.

Also, we should be able to type in something so that we can go to a history page. We can only go to articles and article talk pages. We can't go to wikipedia space talk pages, currently, so this needs to be filed on to bugzilla.174.3.98.20 (talk) 19:14, 19 March 2010 (UTC)

You mean like advanced search? –xeno^talk 19:24, 19 March 2010 (UTC)

Proposal to define an alias for the "User talk" namespace

There is ongoing discussion here of a proposal to define "UT:" as an alias for the User talk namespace. Comments and participation are welcome. -- Black Falcon ^(talk) 20:32, 19 March 2010 (UTC)

Expression error: Unrecognised punctuation character "["

I'm getting a Expression error: Unrecognised punctuation character "[" error on the table at Municipalities of the canton of Zürich. The initial problem was that this table was copied straight over from German Wikipedia. This table used a German template to read a metadata table and fill in the correct population. I replaced the German template with an English template that I created (Template:Swiss populations, which calls the data at Template:Swiss populations data CH-ZH with the CH-ZH coming from the template). This works fine in the table, unless any other operation is done to the data. In the Inhabitants/km² column I've tried applying both the function #expr: and the {{div}} template. Both cause the same error. The div template causes the error outside the table as well, so I think the problem is in my templates. However, I can't find where the "[" character is coming from. Tobyc75 (talk) 22:45, 19 March 2010 (UTC)

A template category was appearing in the article where numbers were expected (I found this by copying the code into Special:ExpandTemplates). I've moved the category into the <noinclude> section. snigbrook (talk) 23:01, 19 March 2010 (UTC)

That's probably it. The only other thing I could see is that you are formating a number before passing it to the expr. But, the page seems to have changed a bit since I first and last looked at it. You probably want also want some rounding in the expr statments. Plastikspork _―Œ^(talk) 23:03, 19 March 2010 (UTC)

Thanks, that seems to have resolved the problem. Tobyc75 (talk) 23:54, 19 March 2010 (UTC)

Transcluding includeonly

I want to create a template which allows an <includeonly> to be transcluded, rather than parsed.

This is so that I can build a series of lists using a standardised template. Those lists will be transcluded into one larger list. One column in the sub-lists is currently tagged with <includeonly></includeonly>, because it is needed only in the combined list ... but I can't figure out how to do this through a template, or even if it is possible.

As a simple illustration of the problem, I have created User:BrownHairedGirl/xtemplate, which contains the words hamster and giraffe. The word giraffe is wrapped in <includeonly>

User:BrownHairedGirl/x1 transcludes the template, and User:BrownHairedGirl/x2 transcludes x1

If everything worked as I want it to, x1 would display "hamster", and x2 would display "hamster giraffe" ... but obviously, that's not what's happening. x1 says "hamster giraffe", because the <includeonly> is being parsed.

Is there any way of getting the template to do what I want? --BrownHairedGirl (talk) • (contribs) 15:23, 19 March 2010 (UTC)

If only one article needs to contain the column(s) in question, the easy approach would be for the template to contain conditionals like these instead:

<table>
<tr>
<td colspan="{{#ifeq:{{FULLPAGENAME}}|That one page|4|3}}">whole-row cell</td>
</tr><tr>
<td>column 1 cell</td>
<td>column 2 cell</td>
<td>column 3 cell</td>
{{#ifeq:{{FULLPAGENAME}}|That one page|<td>column 4 cell</td>}}
</tr>
</table>

I′m pretty sure resolving the include-or-not tags right away (without negotiation or deferral) is what allows /doc-pages to be something other than a fire-hazard (that and the “WP is not paper” meme). ―AoV² 17:24, 19 March 2010 (UTC)

Thank you, AoV²! That's a nice piece of lateral thinking.

It's a slightly inelegant solution, in that relies on the page not being moved ... but it's massively clearer than the {{subst:subst}} stuff that I was experimenting with after reading m:Help:Recursive conversion of wikitext and m:Help:Advanced templates. --BrownHairedGirl (talk) • (contribs) 18:40, 19 March 2010 (UTC)

Another possibility is use of parameters. snigbrook (talk) 19:22, 19 March 2010 (UTC)

Clearly a {{PAGEID}} variable would solve that, if it existed. I don′t see why it shouldn′t, as the number is available for Javascript purposes as wgArticleId, and as a cgi parameter for page look-up—This is how you bookmark a page which may move. ―AoV² 21:53, 19 March 2010 (UTC)

It looks like you could use empty pairs of includeonly tags for this, e.g. <inc<includeonly></includeonly>ludeonly> for the opening tag. This is done in the {{afd2}} template, although that template should be substituted and not transcluded. snigbrook (talk) 18:21, 19 March 2010 (UTC)

It appears to be different when transcluded, this is possibly a bug. snigbrook (talk) 18:28, 19 March 2010 (UTC)

well, the thought that occurs to me is to play games with tags - e.g. wrap your INCLUDEONLY in NOWIKI tags, but wrap the NOWIKI tags in INCLUDEONLY tags in such a way that they only appear in the right place - but I'll confess I can't quite see what the result you want is. can you point to the nested templates you're working on so we have an example to play with? --Ludwigs2 18:53, 19 March 2010 (UTC)

Sorry, but no. I haven't bothered trying to create the templates until I know whether what I want to achieve is doable. Apart from this issue, they will be very simple, but they will be no use without this feature. An example of where the template would be used is in User:BrownHairedGirl/nm2/Dublin: I would like to replace each row of the table with a template to simplify maintenance of the list, but as you can see there, each row needs one column to be wrapped with <includeonly></includeonly>, so that it only appears when transcluded into User:BrownHairedGirl/nm2/Fully sortable list. --BrownHairedGirl (talk) • (contribs) 01:17, 20 March 2010 (UTC)

well, you could futz with the INCLUDEONLY stuff, but the little bit I've played with it, it doesn't do what's expected. I'd suggest you use a trickle-down parameter: in the row-level template have a conditional table cell like so: {{#if:{{{x|}}}||{{!}} cell contents }}. This cell will show up in the row template. when you call the row template from the county template, call it like so: {{row template|x={{{x|}}} }}. that will pass a blank to the row template and the cell will get transcluded into the county template. When you call it from the full template, however, call it like {{county template|x=1}} the x=1 will percolate down and keep the undesired cell from being transcluded up into the full table. will that work for you? --Ludwigs2 02:06, 20 March 2010 (UTC)

Thanks Ludwigs2, but I think you misunderstand how the template would be used. The template call would only ever be coded in User:BrownHairedGirl/nm2/Dublin. The combined list User:BrownHairedGirl/nm2/Fully sortable list does not explicitly call the template; it transcludes User:BrownHairedGirl/nm2/Dublin, and so there is no opportunity to alter the parameters with which the template is called. The only way of making the template do what you suggest is to build a test intoi the template, as suggested above by AoV². --BrownHairedGirl (talk) • (contribs) 11:20, 20 March 2010 (UTC)

Are you subst:ing '/nm2/Dublin'? Because if you're not, I don't think what you said here is correct. When '/nm2/Fully sortable list' calls the template '/nm2/Dublin' it parses the wikitext again with any new parameters. for example, I've made test pages at User:Ludwigs2/test-BHG1 and User:Ludwigs2/test-BHG2 (I've had to add noincludes on the table definition elements - maybe that's where the confusion lies). At any rate, look at those two templates, and look at this result:

{| class="wikitable" style="width:30%"
|+ This is a full table made from transcluded subtables
|-
|{{User:Ludwigs2/test-BHG2|x=1}}
|-
|{{User:Ludwigs2/test-BHG2|x=1}}
|}

gives:

This is a full table made from transcluded subtables

User:Ludwigs2/test-BHG2

User:Ludwigs2/test-BHG2

Note that the x=1 propagates down to prevent the first column cells from showing, even though it shows in the two other templates. that is the effect you're looking for, right? --Ludwigs2 16:13, 20 March 2010 (UTC)

Ok, I see what you are doing. (Sorry, I had slightly misunderstood you first time around).

Somewhere along he way I had considered a pass-through technique like that, but rejected it because it would complicate the markup of the entries in the list, which would have to include something like county = {{{county|}}}. Not a big deal to you or me, but the idea of using the templates is to simplify the markup for non-technical editors, and entering county = {{{county|}}} will seem like gobbledygook to many of them. --BrownHairedGirl (talk) • (contribs) 02:27, 21 March 2010 (UTC)

This reminds me of something. Vote and add some comments here:[9] to promote this. As for the temp workaround, I believe the prod template has managed to do that. I think that it cannot be trancluded though, it can only work through a substitution. --JokerXtreme (talk) 19:23, 19 March 2010 (UTC)

That is correct. If an <includeonly> tag from page A is visible on page B, it is because page A escaped it as <includeonly> (for literal rendering) before forwarding it to B for transclusion. Using “subst” is different from that, doing pre-save replacements with raw (un-sanitized) wiki-text.

While I can think of potential uses for an “includeonce” tag, for this she actually′d want one that says something like “includeonlyafterthefirst” and does the opposite thing. ―AoV² 21:53, 19 March 2010 (UTC)

It would probably be more useful than <includeonce> as there does not appear to be a consistent workaround; an "expand" tag would also be useful. It's understandable for < and > to have the effect it does, what is less understandable is that <noinclude> tags have an effect on pages they are transcluded in. snigbrook (talk) 22:22, 19 March 2010 (UTC)

A <noincludefirst> then. All these tags should be created instead of us trying to do it the alchemist way. --JokerXtreme (talk) 22:44, 19 March 2010 (UTC)

Yes, <noincludefirst> would be just the ticket. But I get the impression that the developers aren't exactly racing away to create new features. --BrownHairedGirl (talk) • (contribs) 01:19, 20 March 2010 (UTC)

It would be better to give includeonly, noinclude, and onlyinclude some parameters, for example :<includeonly wait=3>Wait 3 pages before transcluding</includeonly> It would have worked if {{#tag:includeonlyworked. #tag should support the include tags and nowiki. I know includeonly's a core tag and not an extension (#tag sanitizes XML extension tags, like <syntaxhighlight lang="">), so it doesn't fall under #tag's scope, but still... I've asManishEarth^{Talk • Stalk} 02:22, 20 March 2010 (UTC)

Many thanks to everyone for their help and suggestions. I have now finished the lists, using the {{#ifeq: {{PAGENAME}} technique suggested by AoV². The lists are at Category:Lists of National Monuments of Ireland, templates at Category:National Monuments of Ireland templates. --BrownHairedGirl (talk) • (contribs) 02:28, 21 March 2010 (UTC)

Stacking images

I was attempting to fix a bunching problem on Cisco, Utah, and I noticed that if I use {{stack}}, that the images stack to the left of the infobox, rather than below the infobox. To do a bit of experimentation, and created {{stack images}}, which uses the "infobox class" in the table definition. The result is something that stacks the images below the infobox. My question is three-fold, (1) Is this the best way to achieve this? (2) Is there a reason why not to make this default behavior of {{stack}}? or (3) Should I just add this as an option to {{stack}}? The default behavior of {{FixBunching}} is to stack them to the left of the infobox as well, so I suppose there is some consistency. To see examples of the behavior, see stack image, stack, FixBunching. I suppose it might also be achieved by stacking the infobox with the images, but that seems less desirable since the start and the end of the template are quite spread apart in the wikitext. Thanks! Plastikspork _―Œ^(talk) 22:43, 19 March 2010 (UTC)

The normal solution would be to put each image in the section to which it is relevant, rather than stacking them all in section 0. I believe to use FixBunching with any success, you′d start the invisible floating table at the beginning of the page so that the infobox is in the first cell, but I doubt this is ideal. Honestly, given the dearth of text in this case I′d just use a <gallery> tag to arrange the photos horizontally. You′ll notice also that the interwiki-linked articles fr:Cisco (Utah), it:Cisco (Utah) avoid this problem completely (by using an ugly javascript/css work-around for bugzilla:11555). ―AoV² 00:08, 20 March 2010 (UTC)

Yes, I agree that in this case, a gallery is the best solution. I will probably just add this alternative stacking option to {{stack}}, just in case it is of some use in another situation. Thanks! Plastikspork (talk) 02:44, 20 March 2010 (UTC)

Okay, I added a "clear" option to {{stack}}, just in case someone wants to force the stacking to the far right. My experimental template is now redundant, so I redirected it to the main {{stack}}, just in case anyone is interested. Thanks! Plastikspork _―Œ^(talk) 17:07, 20 March 2010 (UTC)

Lupin's tool doesn't work properly?

When I was in the "Filter recent changes" section of Lupin's anti-vandal tool, I saw an edit by Bottleofjag (talk · contribs) appear as an "Admin" edit. True? No way, that edit doesn't look like something an admin would do at all, and a quick glance at my User:PleaseStand/userinfo.js script reveals that he is not even a rollbacker, only "A user, 13 hours old, with 5 edits." Why is this? I am using Firefox 3.6. PleaseStand ^(talk) 02:58, 20 March 2010 (UTC)

P.S. See User talk:Lupin/Anti-vandal tool#Issues under Firefox: another editor has noticed the same, but the issue is far worse than he thinks (see above). PleaseStand ^(talk) 03:09, 20 March 2010 (UTC)

The tool builds a regex of all admin names and tests usernames against it. However, the regex doesn't have boundaries on it, so if your username contains the username of an admin, it will be marked as admin. As we have an admin called 'B', any username with a (capital, the regex is case sensitive) B in should be marked as an admin etc. I'll fix it at some point, unless someone beats me to it. Ale_Jrb^talk 13:37, 20 March 2010 (UTC)

Working on that fix right now, as well as changing/removing some unnecessary code that is now in site-wide JS. The script dates back to November 2005. PleaseStand ^(talk) 16:50, 20 March 2010 (UTC)

Oops - I just came here to say that I've fixed this, and then I saw your comment. I have in fact fixed it, but if you think you can make any significant improvements, I say go for it; Lupin is no longer active (if you're making big changes, though, I suggest forking it). Ale_Jrb^talk 17:16, 20 March 2010 (UTC)

OK, here is a a small but significant update, update to Lupin's tool. You can see a diff of the changes as of this post. Changes include changing the words "non-admin rollback" to "'fake' rollback" in the settings section, checking the Huggle whitelist, adding an option to hide rollbackers' and whitelisted users' edits, and removing the user warning links (in favor of a new link that is made to work with Twinkle). As there are many rollbackers, though, this version takes longer to load (four times as long, at least the first time, since there are three times as many rollbackers as admins). It also checks the Huggle whitelist. Many bugs are still left, including that the word filter matches text that is not in red. PleaseStand ^(talk) 00:05, 21 March 2010 (UTC)

air dye technology

Resolved

 – PleaseStand ^(talk) 12:45, 20 March 2010 (UTC)

hello sir, can anyone assist me to know the principle used in the air dyeing technology? —Preceding unsigned comment added by Rameshkaruppiah (talk • contribs) 09:10, 20 March 2010 (UTC)

Welcome to Wikipedia. The information is in the article AirDye. Any further questions should be directed to the reference desk, as the page you are on is for discussion about technical issues of Wikipedia itself. PleaseStand ^(talk) 12:45, 20 March 2010 (UTC)

Flagged Protection: ready for more testing

Wikipedia:Village pump (policy)#Flagged Protection: ready for more testing --MZMcBride (talk) 00:45, 21 March 2010 (UTC)

Maximum references

Is there a maximum number of references that can go on a page? I'm trying to add lots of references to the page List_of_chess_grandmasters, but it just times out when attempting to preview/submit, giving the Wikimedia Foundation Error message. Regards, SunCreator (talk) 14:40, 19 March 2010 (UTC)

The answer is yes. Going to split it up. Regards, SunCreator (talk) 14:55, 19 March 2010 (UTC)

The flag icon templates and sort templates don't help either. I suggest you use {{scref}} and {{scnote}} in lists where the same reference is used multiple times. ---— Gadget850 (Ed) ^talk 16:17, 19 March 2010 (UTC)

Thank you. Same reference being re-used multiple times is the issue. Regards, SunCreator (talk) 16:09, 20 March 2010 (UTC)

There is no hard limit on the number of <ref> tags that you can have on a page. However, there are length limits on pages, while you'll hit if you use too many complicated templates like {{cite}}. And even if you don't actually hit the length limit, you might make the page so complicated that the rendering times out on some level. —Aryeh Gregor (talk • contribs) 17:30, 19 March 2010 (UTC)

I found once you exceed reference zz after 702 references(27*26; 26 letters combined together) of the same occurrence using cite then errors and failure to save occur. It's not appropriate to use so many references of cite, so the technical side is fine. Regards, SunCreator (talk) 16:09, 20 March 2010 (UTC)

Wikipedia:Citing_sources#Parenthetical_reference doesn't have those technical limitations. --Enric Naval (talk) 07:17, 21 March 2010 (UTC)

Transclusion History

See this discussion

Sometimes editors involved in disputes want to see an article as it was a while ago, and then look at the changes made. If the article transcludes content from another article, they will see the old version of the article but the current version of the transcluded content, which is confusing. Is there a way in which they could see the old version of the article with the transcluded content as of the date/time that version was created? This could be a big improvement. Thanks, Aymatth2 (talk) 23:28, 20 March 2010 (UTC)

It′s not possible as the software is currently written, though I agree it would be a useful feature. Designing a “gadget” which re-parses from old template revisions using javascript and the API might be (remotely) feasible (albeit madness—yet I′m not above attempting a proof-of-concept if/when I find the time). ―AoV² 14:21, 21 March 2010 (UTC)

It doesn't seem that difficult (easy for me to say). The change would be: "when transcluding content into a historical view of a page, select the most recent version of the source page that was created no later than the date of the target page". That way the historical view would show the page as it really looked at that time, rather than a jumble of past and present. Many editors would be deeply grateful for the improvement! Aymatth2 (talk) 16:13, 21 March 2010 (UTC)

• Posting a summary of this and links at the wikitech-l at lists.wikimedia.org mailing list, see what interest appears from those with their finger on the wiki software pulse. Trev M ~  19:47, 21 March 2010 (UTC)

Show/hide

Is there a script (or gadget, or whatever) to automatically uncollapse everything collapsed in show/hide tables, like in {{Navbox}} and {{ArticleHistory}}? Ucucha 15:28, 21 March 2010 (UTC)

tr[style*="display: none;"] { display: table-row !important; }* ¦ Reisio (talk) 17:55, 21 March 2010 (UTC)

Thanks, that works. Ucucha 18:02, 21 March 2010 (UTC)

Do I live near San Francisco? No I bloody well don't!

There's one of those annoying "help us" messages popping up occasionally at the top of pages, no option to hide or dismiss it, doesn't shew up on every page, and doesn't always reappear when I return to a page it was previously on. How can I get rid of it, and where is it coming from? DuncanHill (talk) 22:12, 17 March 2010 (UTC)

Adware? ¦ Reisio (talk) 22:17, 17 March 2010 (UTC)

No, it's a Wikimedia usability survey thing or something. And apparently nobody can be fucked to include some geo IP code, so it shows to a lot of people unnecessarily. Such is life. Such is Wikimedia. --MZMcBride (talk) 22:19, 17 March 2010 (UTC)

Yep, and even if you click on it there's no option to contact the muppets responsible to complain. DuncanHill (talk) 22:22, 17 March 2010 (UTC)

Haha, I just got it https://ethnio.com/remotes/56415/edit? ¦ Reisio (talk) 22:25, 17 March 2010 (UTC)

I've emailed Ethnio, the company responsible, to complain. DuncanHill (talk) 22:30, 17 March 2010 (UTC)

E-mailing Ethnio is utterly useless; it's just a platform we use to recruit participants for a usability study. guillom 22:34, 17 March 2010 (UTC)

Who is this "we" of whom you speak, why do you subject people thousands of miles from San Francisco to this advert, why is there no "hide" or "dismiss" option, and why is the identity of those responsible hidden? DuncanHill (talk) 22:37, 17 March 2010 (UTC)

This bit, merged from my rant at the pump misc, now redirected here  Chzz  ►  22:35, 17 March 2010 (UTC)

"Live near San Francisco, or in the US, and have an hour to help Wikipedia?" - banner has just appeared.

No, I don't live in the USA; believe it or not, some Wikipedians are not in America. You can tell by our IP address.

Why is this spam permitted, and why is there no 'dismiss' option? Grr.  Chzz  ►  22:29, 17 March 2010 (UTC)

• If it looks like spam and it smells like spam then..... When did the foundation go into the spam business anyway. –droll [chat] 23:20, 17 March 2010 (UTC)

Technically, it isn't spam - it's part of the usability initiative, like the one mentioned here. I had a look at the code, and it's being loaded through the central notice system. It should only ever display once, unless your cookie settings are overly aggressive and are being cleared at the end of each session. If so, or if the idea or ever seeing it really offends you, you should be able to disable it by adding this code to your javascript page:

hookEvent ( 'load', function () { if ( typeof Ethnio !== 'undefined' ) { Ethnio.set_cookie (); } });

Hope that helps. Ale_Jrb^talk 23:52, 17 March 2010 (UTC)

I've had it five or six times, without leaving Wikipedia in between. Thanks for the code, but editors really shouldn't have to resort to fiddling about with javascript (which most of us don't understand) to get rid of irrelevant and annoying banners. DuncanHill (talk) 23:56, 17 March 2010 (UTC)

When you see it for the first time, it tries to set a cookie for like, several years. If it succeeds, it won't display again. Are your cookie settings or a third-party security program preventing Javascript set cookies? If so, my code won't help you... What browser/security software are you using? Ale_Jrb^talk 00:00, 18 March 2010 (UTC)

Chrome on WinXP, and AVG. No special settings AFAIK. But that is irrelevant to my point that there should be a dismiss or hide option, and to the point made above that it shouldn't be randomly targetting people from all over the world. DuncanHill (talk) 00:09, 18 March 2010 (UTC)

Well, yes there should, and no it shouldn't be... but it does sort of have a dismiss. If you see it once, you shouldn't see it again - you don't even need to click a dismiss link (when it's working properly heh). I'm not sure when or why this one was implemented though, and I had a little look. Hmm... Ale_Jrb^talk 00:12, 18 March 2010 (UTC)

Maybe I shouldn't see it again, but I have seen it again, about 20 times (Vista, Firefox 3.6, AVG). Please just make the wretched thing hideable, because your javascript thingy clearly is not doing the biz. --BrownHairedGirl (talk) • (contribs) 00:18, 18 March 2010 (UTC)

The tagline is inaccurate; the Mar 30th date is for San Francisco views, while the Mar 31st is for online interviews. This is actually an issue for the interviewing company to fix. --Ancheta Wis (talk) 00:25, 18 March 2010 (UTC)

Cookies/javascript unblocked for whom? Ethnio.com?

And, my IP range is very firmly and consistently in Canada. Someone needs to update their geoIP listings. [pile-on-comment]

And, around 6-16 March 2009 we were also discussing ethnio! ;) -- Quiddity (talk) 00:48, 18 March 2010 (UTC)

Adblocking */centralnotice.js* should solve it for good. ―AoV² 00:30, 18 March 2010 (UTC)

I'll point out that I am in no way related to the usability initiative, I did not write the code, and I have no control over its implementation. Did you try the code I posted above? If so, and it doesn't help and you don't want to block the whole central notice system, this should be a little more aggressive:

hookEvent ( 'load', function () { if ( document.getElementById ( 'ethnio_pitch' ) ) { document.getElementById ( 'ethnio_pitch' ).parentNode.removeChild ( document.getElementById ( 'ethnio_pitch' ) ); } });

Ale_Jrb^talk 00:33, 18 March 2010 (UTC)

Sigh. Perhaps people should demand less, and request more..... —TheDJ (talk • contribs) 01:02, 18 March 2010 (UTC)

OK, first things first. What are people thinking directly emailing ethnio ? This is Wikipedia and any technical issues should be filed in bugzilla:, any other issues should probably be communicated through the foundation, with which this company has a contract.

Secondly, I asked online. The banner should be online only for 3 days or so (1,5 have already passed). Also, there are currently requests for 2 studies (multimedia upload, and wikipedia usability) in 2 modes going on (San Francisco interviews and online interviews). The exposure is a bit higher than usual because of this.

• bugzilla:22873
• bugzilla:22874

And the bugreports. Please add your information there. Some browser config info might be useful to add to the 2nd bug. —TheDJ (talk • contribs) 02:01, 18 March 2010 (UTC)

Why did I email Ethnio direct? Because there was no indication whatsoever in the notice that anyone else was involved. If the Foundation was responsible for the banner, then they should have said so in it. As for bugzilla - user friendly it isn't. I came to VPT because it's a very helpful place full of very helpful people. DuncanHill (talk) 02:15, 18 March 2010 (UTC)

And further, re Bugzilla, I'm buggered if I'm going to post my email address publically on it, so I can't have a log in, so I can't report things there. DuncanHill (talk) 02:20, 18 March 2010 (UTC)

If just any organization could hijack Wikipedia pages, then we would have a security issue and then you should definetly turn to the foundation. —TheDJ (talk • contribs) 02:43, 18 March 2010 (UTC)

But we've got no way of knowing if it's the foundation, some wikichapter thingy or just a mad admin plating with sitenotices. DuncanHill (talk) 02:50, 18 March 2010 (UTC)

We ? You might not know. But others clearly do. Don't make rash actions if you don't understand what is going on, just hang out and sit tight and wait for people that do understand what is going on. —TheDJ (talk • contribs) 03:17, 18 March 2010 (UTC)

Rash actions? You mean like asking what it is and who is responsible? That's hardly rash in my book. DuncanHill (talk) 03:19, 18 March 2010 (UTC)

Or you can add some CSS and never see stupid sitenotices again:

#siteNotice { display: none; }

Done it and never looked back. Q ^{T C} 01:57, 18 March 2010 (UTC)

Indeed. The ratio of useful information to rancid shit in the sitenotice dropped to near zero years ago; I've had it blocked for years and never missed anything that needed my attention. — Gavia immer (talk) 02:10, 18 March 2010 (UTC)

I was wondering what this thread was referring to for a while, then started seeing the sitenotice, only within the last 5 minutes for some reason. No hide link? Since when? Not good. Not good at all. Yeah I know about CSS. Still, not good. Equazcion ^(talk) 02:24, 18 Mar 2010 (UTC)

It pops up a lot for me too (with Epiphany running under Ubuntu 8.04, and no bizarre settings for cookies). If I add

#siteNotice { display: none; }

as recommended above, will this also kill notices of arb elections and the like? (I don't think I've ever yet been bothered to participate, but one day I might surprise myself by doing so.) ¶ It's a quaintly pre-interwebs project, telling us: Remote interviews will be held online on March 31st with users throughout the US. Why not users outside the US? Email is email, no? (But then again perhaps only the opinions of US Americans are worth listening to.) -- Hoary (talk) 03:15, 18 March 2010 (UTC)

I believe the remote interviews are being done by phone. Mr.Z-man 03:37, 18 March 2010 (UTC)

Just a minor point, but for those who want to block the sitenotice, the CSS code you really want is

#siteNotice { display: none !important; }

Which is almost identical in the usual case but gives you at least some protection against the inevitable "lots of people block the sitenotice, but this usability survey is really important, so we'll override blocking in the sitewide CSS" banner code that will probably get rolled out eventually. — Gavia immer (talk) 03:40, 18 March 2010 (UTC)

These merely will hide it after it already has loaded. Barring a user-preferences option to disable it completely (uh-huh), the best approach is ABP or some other way for your browser to ignore “centralnotice.js” outright. ―AoV² 05:00, 18 March 2010 (UTC)

I can confirm that the notice appears repeatedly using Firefox 3.6 on Windows XP. I never did anything about the cookie settings, so I am confident I have either the FF 3.6 factory defaults or those of an earlier version (max. about a year old) over which I installed. Hans Adler 22:24, 18 March 2010 (UTC)

Yeah, it was my bad - I was tired and the code is sh-... not very tidy. It only disappears permanently if you click on it. That does mean the first code I put up will hide it for you (without you having to click on it), AFAIK. Ale_Jrb^talk 23:13, 18 March 2010 (UTC)

So far, every single thing I've seen that the foundation has paid for has been completely unusable. Thank goodness I perma-blocked the sitenotice after the fundraising banner fiasco. Glad to see the money's being well spent on these web-designers from 1994. OrangeDog (τ • ε) 23:03, 18 March 2010 (UTC)

Why not use Wikipedia:Geonotice?--Pharos (talk) 03:06, 19 March 2010 (UTC)

Because we needed to recruit logged-out users as well; the real solution would be to fix bugzilla:21295. But honestly, I think the community of the English Wikipedia is way too touchy, and I would like to ask you (in the general sense) to give us some leeway. I mean, the banner was up for only ~6 hours, and that's all we needed. Is it too much to ask, when our goal is to make MediaWiki better and to perennially improve the user experience for everyone (including experienced users like those posting here)? guillom 14:38, 20 March 2010 (UTC)

Since the "we" behind the message was not identified, it is too much to ask for any leeway. Jc3s5h (talk) 14:49, 20 March 2010 (UTC)

There is this very useful thing called a "user page". My sincere apologies for assuming you knew what it was and how to use it. Just click on my signature and you should be able to identify who the "we" is. guillom 16:15, 20 March 2010 (UTC)

Also nice to see WMF employees resorting to childish personal attacks so quickly. OrangeDog (τ • ε) 19:05, 20 March 2010 (UTC)

Guillom, what Jc3s5h meant was it was difficult to tell who was behind the notice, from the notice itself. Not that it was difficult to identify who you were, or who you were referring to (I believe) - Kingpin¹³ (talk) 20:09, 20 March 2010 (UTC)

The advert wasn't signed by you. It was only after I complained here and you made a comment that anyone would have been able to deduce that you were responsible for adding the advert. You still haven't explained (as asked above) why there was no dismiss option. DuncanHill (talk) 16:19, 20 March 2010 (UTC)

I'm all for giving you good folks leeway on this, but please work on bugzilla:21295 too, because if implemented it would really be a tremendous boon to the Wikimedia movement in so many ways.--Pharos (talk) 15:47, 20 March 2010 (UTC)

Since when does the WMF do these things properly? And, yes, Adblock Plus and/or NoScript (it's the only script I know of hosted on upload.wikimedia.org) is the best way to permanently kill off crappy central notices, including awful fundraising banners. I blocked the central notice after the marquee fundraiser and I've never looked back. MER-C 05:49, 20 March 2010 (UTC)

Strongly Agree I dont live anywhere near 'Cisco (im on the east coast) so these messages sometimes do nothing for me. But.. the latest one had a opt in for an online interview as well.. so.... but i see what your point is Evenios (talk) 07:40, 22 March 2010 (UTC)

title-only search?

This seems like an obvious thing, but I can't see how to do it. is there a way to search only page titles? for instance, if I'm looking for every page that has the word 'dog' in the title, how do I find that? the search box tooltips will give me pages that start with the word dog, and the main 'special pages' search will give me every page that has the word 'dog' somewhere in the content, but the first is too restrictive and the second is way too expansive. what am I missing? --Ludwigs2 21:04, 19 March 2010 (UTC)

You would search the term preceded by "intitle:", for example, if the term were "airport", the search intitle:airport would give all pages with "airport" in the title. Cheers, Intelligentsium 21:11, 19 March 2010 (UTC)

Ah! I knew I was missing something. thanks. --Ludwigs2 21:13, 19 March 2010 (UTC)

Just out of curiosity, would it be possible to get a link to help:searching (which I suppose I should have looked at first) from the Special:search page? I think that's normal for search sites, but there doesn't seem to be anything like it here. --Ludwigs2 21:47, 19 March 2010 (UTC)

After an unsuccessful search (i.e., no matches to the criteria given) there is a statement, "For search help, please visit Help:Searching.". Is this what you were looking for? Intelligentsium 21:53, 19 March 2010 (UTC)

actually, no. I was thinking about on the main search page (what you get initially if you click the sidebar search button with an empty field). or at least in the modified page you get when you click on the Advanced link. --Ludwigs2 22:50, 19 March 2010 (UTC)

We have requested this to developers but as usual bugs take ages to be processed. Cenarium (talk) 17:09, 20 March 2010 (UTC)

• There is also http://toolserver.org/~nikola/grep.php?lang=en <- pagename search (accepts regex input). –xeno^talk 17:57, 22 March 2010 (UTC)

Bengali wikipedia [[show/hide]] at collapsible table template

I know that this is not a right place for this issue. I added in bn:MediaWiki:Common.js Default skin for navigation boxes of .collapseButton and bn:MediaWiki:Common.js Collapsible tables javascript properly from just copy/paste from this wiki. But bn:Tempalte:Collapsible list and other template not shown [[show/hide]] button. Are there any mistake? or add any js to create that or it is a bug. I don't know. Could you help us for this matter? Advance thanks.- Jayanta Nath ^{(Talk|Contrb)} 07:50, 21 March 2010 (UTC)

I think it's because bn:MediaWiki:Monobook.js already contains old, incompatible, version of the show/hide functionality, that overrides the new version in bn:MediaWiki:Common.js. If you use the Vector skin, it works. Removing the “Dynamic Navigation Bars” from Monobook.js should fix the issue. Svick (talk) 15:56, 21 March 2010 (UTC)

Thank you dear. Its working fine after removing “Dynamic Navigation Bars”. And another small questions about our bn:MediaWiki:Monobook.js which contents /* tooltips and access keys */ may be not necessary now. (this /* tooltips and access keys */ section is currently use in all Indic language wikipedia). If I remove this. Is it effect the system?- Jayanta Nath ^{(Talk|Contrb)} 16:39, 21 March 2010 (UTC)

Tooltips and accesskeys have been defined in the software for over 4 years now I believe. What you are describing is the old system that was used before and it can be removed. —TheDJ (talk • contribs) 20:36, 22 March 2010 (UTC)

toolserver is down

Is it just me or is toolserver.org down? I'm getting a "page not found/broken link?" error ManishEarth^{Talk • Stalk} 14:53, 21 March 2010 (UTC)

It seems to be working for me. --JokerXtreme (talk) 14:58, 21 March 2010 (UTC)

Toolserver is alive and well. There's a bunch of lag happening on the Wikipedia servers right now. Josh Parris 15:00, 21 March 2010 (UTC)

Why is http://stats.grok.se/ not working though? --JokerXtreme (talk) 19:48, 21 March 2010 (UTC)

Works for me. I'm guessing it was the lag that Josh mentioned. That's a different tool by the way, separate independent of the toolserver. Reach Out to the Truth 17:51, 22 March 2010 (UTC)

SVG renderer progress

Where can I go to be kept up-to-date on the development progress of the SVG renderer Wikipedia uses? Thanks. SharkD  Talk  02:55, 22 March 2010 (UTC)

http://meta.wikimedia.org/wiki/SVG_image_support ¦ Reisio (talk) 09:15, 22 March 2010 (UTC)

That seems mostly unrelated and/or outdated. Wikipedia uses rsvg last I heard. It's open-source, so probably has development fora you can follow. —Aryeh Gregor (talk • contribs) 18:15, 22 March 2010 (UTC)

Category depopulation

Depopulation over at this category has stalled. Two and a half weeks ago, it was having issues depopulating, but it hasn't moved in two days. I'm all for removing the stub template and adding it back in, but if anyone else has any other ideas, I'm all for it. Thanks. Kevin Rutherford (talk) 02:23, 19 March 2010 (UTC)

Question was originally raised here. I left other messages on this page which were fruitless, such as here (I did not raise a bugzilla ticket, because to do so requires me to publish my email address, which I object to), and also at Help talk:Job queue#Stuck queue, where I periodically tallied the progress.

Category:Skiing stubs appears to be clear now. --Redrose64 (talk) 14:12, 19 March 2010 (UTC)

I suspect someone nulledited all the pages to force the category to clear ? If someone can confirm that. —TheDJ (talk • contribs) 22:05, 23 March 2010 (UTC)

Geotagged EXIF information Wikipedia understands, { { Location } } for Wikipedia?

G'day guys!

I'm trying to find out information on what way (if at all) Wikipedia is able to automatically understand and translate information geotagged into the EXIF data of a JPEG image so that a) it displays properly in "Show Extended Details" and b) Wikipedia is able to automatically add this to the metadata of the image (i.e. where the summary et al is displayed).

Take this image for example, one of my own. Originally it was on Wikipedia but I've since decided to move it to the Commons for the benefit of other projects. Still needs some fixing though. If you'll notice in "Show Extended Details" where the EXIF data is displayed, The last four entries are the geotagging details, though the actual coordinates don't show, just 'South latitude' and 'East longitude' respectively.

I've used Adobe Lightroom and Photoshop CS4 to process it and used Jeffrey’s “GPS-Support” Geoencoding Plugin for Lightroom to geotag it. While I used Google Maps to geotag this image, I usually use my Nokia N78 and Nokia Sports Tracker for the GPS data. Maybe this will shed some light on why the EXIF data displays so? I have a feeling it might be Jeffery's plugin. I'll drop him an email and see.

Also, is there a Wikipedia version of { { Location } } (apart from coord) so that GPS data (if manually added) displays in a much more user friendly fashion in the image summary? Coord doesn't seem to work, only adds 31°57′12″S 115°48′25″E and not the self contained box.

Thanks heaps!!

DeltaFalcon ^{talk / contribs} 16:44, 20 March 2010 (UTC)

There is a bug in the way the "Extended details" parser displays EXIF information, that is causing the location info not to be visible. This is detailed in bugzilla:13172 —TheDJ (talk • contribs) 15:19, 23 March 2010 (UTC)

I operate a bot [10] on commons that analyzes the EXIF data of newly uploaded images and automatically adds a Location template. --Dschwen 15:36, 23 March 2010 (UTC)

wikitech bad certificate

The interwiki link (wikitech:) to Our backstage tech site wikitech is linking to the https: version, while the link under the list of backstage projects leads to the http: version. That would be fine with me, except that the https: version is giving me an "invalid certificate error". I checked the certificate, and it seems that wikitech has issued itself a certificate, without being a certificate authority. THis means that a first-time user, when viewing it, will get an error:

• "SSL Error-The site's security certificate is not trusted!" (in chrome)
• "Certificate Error:Navigation blocked--There is a problem with this website's security certificate." (in IE).

Shouldn't WMF get it signed by equifax like the secure server? Or change the interwiki mechanism so wikitech: leads to the http: version. Thoughts? ManishEarth^{Talk • Stalk} 07:43, 23 March 2010 (UTC)

Filed bug report. Please vote there. ManishEarth^{Talk • Stalk} 08:04, 23 March 2010 (UTC)

meta:Interwiki map is the place to get the interwiki link changed. The bugzilla report will only consider the signing issue. Happy‑melon 13:51, 23 March 2010 (UTC)

Prefs submit button accesskey

I'm quite used to using keyboard shortcuts, so when I pressed alt-s on preferences and nothing happened, I was surprised. I decided to add keyboard shortcuts to prefs and watchlist edit pages. The js code for prefs is:

document.getElementById('prefcontrol').accessKey="s" //prefs

Somehow, the watchlist edit pages don't have ids, so I can't give them accesskeys. Could someone give me the code for accesskeying those pages? Thx, ManishEarth^{Talk • Stalk} 09:00, 23 March 2010 (UTC)

I know the Special:Preferences page imports javascript only from bits.wikimedia.org, none local. A sanest thing would be to disable js completely on pages where a user may enter a password but I guess people like to have fancy tabs and dancing pigs on the prefs-screen.

Anywho the consequence is that one cannot customize the prefs page to add this access-key without using greasemonkey or something similar, because it will not load any js stored on a wiki page. The other option is to add the access-key for everyone through a software change, but then it would be immune to the script which removes access keys for those who have tired of invoking them by accident, again because the latter will not load on the prefs page.

Do you have greasemonkey? ―AoV² 13:47, 23 March 2010 (UTC)

Any way to not return to Special:IPblocklist after unblocking a user?

^^^ ? –xeno^talk 16:29, 23 March 2010 (UTC)

Could write a script in a few seconds that does it in the background. Do you mean that you stay on the unblock form after unblocking, or did you want an option from somewhere else? Ale_Jrb^talk 17:39, 23 March 2010 (UTC)

Maybe back to the user's talk page? Not sure why it sends us to the ipblocklist. –xeno^talk 20:26, 23 March 2010 (UTC)

It would be nice if this were a standard feature. Until then, though, I did write a small script - use importScript ( 'User:Ale_jrb/Scripts/unblock.js' ); // [[User:Ale_jrb/Scripts]] - it adds a new button to the unblock interface. It doesn't redirect you at all at the moment - it does the unblock in the background and provides you with some links, including to their talk page. I don't like redirects much, but I could make it if you'd especially prefer. Ale_Jrb^talk 21:56, 23 March 2010 (UTC)

Works great - thanks. –xeno^talk 22:02, 23 March 2010 (UTC)

This is bugzilla:16835. --MZMcBride (talk) 20:45, 23 March 2010 (UTC)

Tag as a library

Not sure if we tag public libraries as we do schools, and I'm not sure what the template is, but this IP address is Toronto Public Library - 205.189.194.208 (talk) 21:10, 23 March 2010 (UTC)

Template:SharedIP is typically only added when abuse is originating from the IP. –xeno^talk 21:19, 23 March 2010 (UTC)

And for libraries, it's actually Template:SharedIPPublic. —TheDJ (talk • contribs) 21:44, 23 March 2010 (UTC)

Problem with DYK sections in multiple portals

Does anyone that knows anything about portals know what is up with the Did You Know sections of various portals? See Portal:Film and Portal:Tanzania for example. They are dynamic sections, using {{Random subpage}} and the "seed" parameter. For some reason, eg. Portal:Film/DYK is trying to display pages of huge numbers like 627, even though it's set to a maximum of 28. As a result, red linked sections are showing up on the portal page. I'm hoping someone has a clue what's going on. {{Random subpage}} hasn't been edited since January, and this is a very recent problem, so I can't work out what's changed.--BelovedFreak 22:03, 23 March 2010 (UTC)

There were some changes to {{Random number}} and {{mod}} that were causing this. I have reverted the changes and will ask MSGJ to look into the problem —TheDJ (talk • contribs) 22:34, 23 March 2010 (UTC)

Thanks.--BelovedFreak 22:43, 23 March 2010 (UTC)

What happened to the tokipona wiki

Sometimes I go to Special:MergeAccount just to see which wikis I've been on. This time, I noticed that there was a link: tokipona.wikipedia.org, which gave me a "wiki does not exist" page. If the wiki does not exist, how did I visit it and how did it become part of my global account. (toki pona seems to be a constructed language) ManishEarth^{Talk • Stalk} 03:06, 24 March 2010 (UTC)

Just a note: I never recall visiting the tokipona wiki either. ManishEarth^{Talk • Stalk} 03:09, 24 March 2010 (UTC)

m:Proposals_for_closing_projects/Closure_of_Toki_Pona_Wikipedia. I believe it might have been deleted from the Wikimedia Servers, but I cannot recall. It was closed and moved to a wikia server. User:Zscout370 ^{(Return Fire)} 03:10, 24 March 2010 (UTC)

Been nuked at m:Proposals_for_closing_projects/Deletion_of_Toki_Pona_Wikimedia_projects User:Zscout370 ^{(Return Fire)} 03:12, 24 March 2010 (UTC)

While Special:MergeAccount is nice, I'd recommend checking out sulutil:Manishearth; a lot more information there than just a straight listing. Enjoy. :) EVula // talk // ☯ // 03:32, 24 March 2010 (UTC)

Nice tool. Thanks! ManishEarth^{Talk • Stalk} 03:38, 24 March 2010 (UTC)

Gallery tags

When galleries are automatically generated in categories of images, a text link to the file is automatically generated and displayed below the image. Is it possible to trigger this behavior outside category space, such as in user space? If not, is it possible to pipe code into the gallery tag via templates? SharkD  Talk  04:44, 24 March 2010 (UTC)

You could create a template which renders a gallery/table of images up to an arbitrary maximum. However, I think using inline div elements would be better because unlike table cells they would wrap intelligently to the available screen-width. Examples available upon request. ―AoV² 16:41, 24 March 2010 (UTC)

mysterious csd

User:NativeForeigner/Development is showing up in Category:Candidates for speedy deletion by user. Page does not seem to have been edited in three months, and I can't find any tags on the page. I cleared out this cat yesterday so I know it wasn't in it then. So, why is it showing up in this category? Beeblebrox (talk) 17:25, 24 March 2010 (UTC)

• I added noinclude tags to the page and that seems to have fixed it, but I still can't figure out why this happened to begin with... Beeblebrox (talk) 17:29, 24 March 2010 (UTC)

I'm betting one of the userboxes was put up for CSD, and it carried over. Several on the right have been deleted, though who knows when without checking the log. UltraExactZZ ^Said ~ Did 19:32, 24 March 2010 (UTC)

More CSS help needed...

So, I'm making a skin for Vector in addition to the one for Monobook, and I'm trying to change the image links for the tabs, etc. My code so far (or the relevant part) is here, and I'm trying to change the following things ATM:

1. The Tab border color.

2. The Logo.

However, only the leftmost tab border is changed using the theme, and the logo-changing bit only works if it's used in Stylish. It work in Monobook, It works in Stylish, but not in the Vector css. I'm practically going insane trying to figure out what's wrong with the thing. Any help would be greatly appreciated. Hmmwhatsthisdo (talk) 00:10, 24 March 2010 (UTC)

There are a couple of problems in your css.

1. The .vectorTabs line is missing a closing }
2. It appears the background image is actually attached to .vectorTabs li a not to the parent div vectorTabs
3. For the logo it looks like you meant to be using the shorthand background: declaration, but instead have used background-image:

Dr pda (talk) 03:20, 24 March 2010 (UTC)

I posted the fix here. DON'T reply there, just copy the code and tag it with {{U1}} to delete the talk page. ManishEarth^{Talk • Stalk} 03:44, 24 March 2010 (UTC)

Oh. OK. May I ask how you found this? I have probably less CSS knowledge than a small green fish, and I have no bloody clue what I'm doing. Essentially, I'm trying to modify the Vector skin to make it a red hue, as opposed to blue. However, the coding part is a lot more difficult than I thought. So far, I've tried just copying the lines of code containing the image paths I'm looking for, and just replacing the image names. I'm here now, so we can all guess how that turned out! So, would either of you be willing to give me some help understanding this? Hmmwhatsthisdo (talk) 06:00, 24 March 2010 (UTC)

I made the changes according to Dr pda. Explanantion of the points:

1. Syntax error
2. The class .vectorTabs applies to sets of tabs. For example, The "Article" and "discussion" tabs are in one div with class .vectorTabs. Giving it a border would have worked, except that it would not have overridden the border of the individual tabs, as they cover the div. So only a part of the colored div was showing. The rest was masked by the tab li's. So, instead of applying the code to .vectorTabs, it should be applied to the li inside .vectorTabs (CSS code: .vectorTabs li {.....})
3. You were applying the percentage parameters to background-image, which takes only one parameter. You must have meant background:

Why not use Chrome? It has some developer tools (ctrl-shift-i) which give you the CSS properties and allows you to do js stuff, too. Hope this helps! ManishEarth^{Talk • Stalk} 07:36, 24 March 2010 (UTC)

So does FF, and Opera. (other browsers are available) OrangeDog (τ • ε) 13:26, 24 March 2010 (UTC)

FF requires FireBug... But yes, others are availible. I like chrome best for dev'ing, though. For a multitude of reasons. ManishEarth^{Talk • Stalk} 13:46, 24 March 2010 (UTC)

I got Firebug, but it's still all Greek to me. I'm trying to figure out which class the 'borders' of the tabs are. I tried #p-namespaces h5, p-namespaces ul, .vectorTabs ul, & .vectorTabs h5. None of them seem to work. Could somebody perhaps, explain where my thinking is going wrong here?

EDIT: Massive breakthrough. I've been using Firebug, and have gotten almost everything done, with the exception of the border, text color, and tab dividers. It turns out that maintaining proper order is imperative to the script working. Hmmwhatsthisdo (talk) 01:34, 25 March 2010 (UTC)

Another Edit: I have gotten all of the images replaced, spare the tab separators. It turns out that they should be modified with div.vectorTabs li a, but when I change them, the tabs seem to sink down into the page. If needed, I can post a screenshot. Hmmwhatsthisdo (talk) 02:45, 25 March 2010 (UTC)

Side note: I really like the visual efect of the CSS. Let me know when you're done! ManishEarth^{Talk • Stalk} 03:40, 25 March 2010 (UTC)

Well, I'm almost done. I still need to tweak the edit box and fix those darned tab breaks. Here's the (non-working) vector.css with div.vectorTabs li a, and div.vectorMenu h5 a. —Preceding undated comment added 04:07, 25 March 2010 (UTC).

(edit conflict) Comments on script:

• To make the lines underneath section headings red, add h1{border-bottom-color: red;}
• To make the selected tab look as it its selected (like in normal vector) use : div.vectorTabs ul li.selected {background-image:url('http://bits.wikimedia.org/skins-1.5/vector/images/tab-current-fade.png') !important}
• Looking at the links in the tab bar, its impossible to tell if a page exists or not. Add div.vectorTabs li.new a span{color:#FFAAAA !important}
• You're missing a closing brace at the end. ManishEarth^{Talk • Stalk} 04:13, 25 March 2010 (UTC)

Well, I'm done with the 'main' part, I have only the edit box left to do. However, something appears to be wrong with the tabs. It's kinda hard to explain, but looking at the actual CSS would explain it. Also, if anyone reading this has access to /skins-1.5/vector/images (readonly+view should do it) on bits.wikimedia.org, a readout of what images are stored would help a lot with this.Hmmwhatsthisdo (talk) 00:55, 26 March 2010 (UTC)