Wikipedia

Wikipedia:Village pump (technical)/Archive L

< Wikipedia:Village pump (technical)
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This page contains discussions that have been archived from Village pump (technical).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

Remove content warning on talk pages

The warning

Content must not violate any copyright and must be based on verifiable sources. By editing here, you agree to license your contributions under the GFDL.

Is good for article pages, but it should be removed from talk pages. Talk pages are discussion portals, not articles, and discussion should be as free as possible.

The only thing we can remove from that is verifiable sources, as discussion is supposed to be ones own opinion. Copyright violations and GFDL contributions all apply to talk pages as well however. WAvegetarian (talk) (email) (contribs) 20:17, 15 December 2005 (UTC)
Clearly we cannot allow copyright violation, as this is against the law. However, there are some who have made a reasonable case that participants in discussions should, as on newsgroups or mailing lists, retain copyright to their original statements, particularly since we hardly ever edit the comments of others. I like it like it is though. Deco 20:43, 15 December 2005 (UTC)
You retain copyright on any talk page comments you make; you just implicitly agree to license them under the GFDL whenever you post to a Wikipedia talk page. AFAIK, this is non-negotiable. A user who wanted to not license his talk page comments in this way (User:Pioneer-12, I believe) has been banned indefinitely for disagreeing with this policy. android79 21:06, 15 December 2005 (UTC)
Well, then at least remove the references part on talk pages. Elvarg 21:22, 15 December 2005 (UTC)
I think the references part should remain as well. Talk pages are for discussing the article, and so all statements should be based on verifiable sources. Even on the talk page we should be saying "I think we should do X because Y", not "I think X". I do not think we shoukd seek to encourage people to post unverified statements to the talk pages, as they could then become a repository of false information which google can pick up and disseminate further. Wikipedia is not a discussion portal and talk pages are not for general discussion, and there should be an onus on us to source our opinions. Steve block talk 13:27, 16 December 2005 (UTC)
I'd put this differently. On talk pages, people should be clear when they are using unverified sources. For example "I think remember he was the first to land on the moon.. does anyone have a reference for that" hardly needs to be verified. Mozzerati 20:57, 21 December 2005 (UTC)

Watching pages

Why can't I add comments to my watchlist to note down for myself why I marked a page to be watched? --Unforgettableid | talk to me 06:32, 15 December 2005 (UTC)

Why? Because that feature doesn't exist. If you want it, you should request it, either on the proposals page, in mediazilla, or find a friendly developer :) Stevage 03:12, 18 December 2005 (UTC)
Do others agree this would be useful? --Unforgettableid | talk to me 21:22, 20 December 2005 (UTC)
It'd be useful to me if I could use that information in displaying my watchlist. What I'd like is to be able to put tags on pages on my watchlist, like "featured", "vandal-magnet", "discussion", etc., so I can filter the watchlist by those tags (suppose I want to see only discussions where I'm waiting for a reply, or I'm in the mood to fix articles that attract vandals, or whatever). It would be like having sub-watchlists. rspeer 22:19, 20 December 2005 (UTC)
Now on MediaZilla as bug #4354. 86.133.53.111 13:24, 22 December 2005 (UTC)

Huh?

I tried to move a page and was told my account was too new -- but I have been here for a few months now! Has moving broken somehow, are there new restrictions, or something? I haven't heard anything regarding this... --WCQuidditch ☎ ✎ 23:58, 22 December 2005 (UTC)

After seeing a question above -- is this an inadvertent side effect of a recent upgrade? --WCQuidditch ☎ ✎ 00:00, 23 December 2005 (UTC)
You're not the only one who is experiencing this problem. Perhaps it's restarting the count for everyone. -- Dissident (Talk) 01:05, 23 December 2005 (UTC)
That's fixed. --Brion 01:44, 23 December 2005 (UTC)


new protection feature

I don't know if this is intentional, but for pages that are protected from editing by new users, the page protection link still shows as "protect this page" as if it's not protected. For example, see the penis article. --Ixfd64 19:14, 22 December 2005 (UTC)

It's not that different from pages protected against moves having the link show as "unprotect". --cesarb 22:18, 22 December 2005 (UTC)
IMO now we have so many different types of protection changing the text on that button has outlived its usefullness and it should just be marked "protection" and take the admin to a page showing the pages current status and letting them change it. Plugwash 23:11, 22 December 2005 (UTC)
It says 'unprotect' to me. --Brion 23:47, 22 December 2005 (UTC)

No more time delay for page moves either??

User:Guillermo con sus ruedas was created at 17:54 today and started pagemove vandalism immediately. A bunch of other socks followed. What happened to the time delay for doing page moves? -- Curps 19:12, 22 December 2005 (UTC)

  • I was wondering about the same thing as well. It seems that the page-move throttle may have been inadvertently removed during the recent software upgrade. --Ixfd64 19:16, 22 December 2005 (UTC)
It's a four-day delay on new accounts created since the upgrade. --Brion 23:48, 22 December 2005 (UTC)

View source and template list

When viewing source of a protected page, it would be good to be able to show a list of templates that the page includes, in the same way as you currently get when editing a page. Of course, there may be a case for those templates being protected themselves, but that's not a strong reason against having a list of them in a convenient form. TerraGreen 18:31, 22 December 2005 (UTC)

  • I agree. --Ixfd64 19:12, 22 December 2005 (UTC)

Adding CSS for metadata

I would like to add the following CSS to Common.css: 

/* Metadata */

table.metadata {
        border: 1px solid #aaaaaa;
        display: none; 
}
.metadata-label {
        color: #aaaaaa;
}

This will help us to implement metadata schemes like the German Wikipedia does, which will open up a whole new world of possibilities for accessing Wikipedia data. For example, imagine being able to search for all Wikipedia articles on authors who died in 1954, or cities in Italy with more than 100,000 people. For more info on the German effort, see this article. The first step is being able to add hidden metadata to articles, which this CSS will make possible. The second step is to create metadata templates such as Template:Persondata. The third step is to take over the world! Muhahahaha! Kaldari 17:57, 22 December 2005 (UTC)

I've gone ahead and added the necessary CSS so that I can develop a proof of concept for the English Wikipedia. Stay tuned... Kaldari 19:39, 22 December 2005 (UTC)
If you'd like to keep track of my efforts, check here: Wikipedia:Persondata. Kaldari 23:56, 22 December 2005 (UTC)

Semi-protection and George W. Bush

JosephBostwick (talk · contribs · deleted contribs · nuke contribs · logs · filter log · block user · block log) is the first vandal of George W. Bush under semi-protection, and he did so 2 minutes after registering the username. Clearly, the time-delay aspect of semi-protection hasn't been implemented yet.

Probably we should introduce some artificial delays for registering an account... make them do captchas or answer skill-testing questions for about 30 to 60 seconds. Something that can't be automated: random obvious skill-testing questions like "how many legs does a dog have?" that any human can answer but would stump a registration bot. A set of several hundred randomly varying skill-testing questions might be preferable to captcha, because visually-impaired users wouldn't be affected.

This wouldn't be much of an annoyance to legitimate users (who only register one or a couple of accounts, ever), but would raise the hassle factor for large-scale throwaway account creation. It's like spam... if you could charge 1 cent per e-mail message, spam becomes uneconomical. Same concept. -- Curps 11:15, 22 December 2005 (UTC)

Let's just put the 4 days thing into effect...see if we can get the software to accept that...and just do that. --Woohookitty(cat scratches) 12:52, 22 December 2005 (UTC)
Not the whole mess, please. As to the problems - it's quite possible, as Brion appears to mention it's a configuration option, that he made the honest oversight of not enabling that option in the English Wikipedia's configuration. 86.133.53.111 13:19, 22 December 2005 (UTC)
Spelled it wrong in the config. :P Fixed. --Brion 23:56, 22 December 2005 (UTC)

Special:Unwatchedpages

I've enabled Special:Unwatchedpages for users with protect permission (that's sysop and above to those not familiar with the permission system), there 5000 rows generated on that page as opposed to the usual 1000, making that value larger is a bad idea since paging gets exponentially more expensive the further in one goes (that can be fixed though).

Having more people watch pages might help avoid something like the recent incidents, it's restricted so that it won't serve as a "vandalize this" list, currently approximately 27% of pages in the main namespace which are not redirects are watched (252502/906363*100 = 27.85881594901821896900). —Ævar Arnfjörð Bjarmason 06:07, 21 December 2005 (UTC)

It would be useful to know how many of these unwatched pages are categorized and how many appear in lists. It generally more difficult to find a page associated with a particular topic area if it is not included in that topic area by categorization, thus making it more difficult to find for watching. Also, some people monitor large numbers of pages for changes via the related changes function applied to lists; such monitoring can decrease the percentage of pages appearing on watch lists despite their being watched indirectly. User:Ceyockey 05:35, 23 December 2005 (UTC)

Post-Image deletion

Since our policy change, images are being deleted without so much as a word to the uploader, or anyone else. The thing is, Admins often delete an image but then don't go through the articles where it was linked to get rid of the now broken links. Since nobody is contacted prior to the deletion, some articles are left with that gastly red link to the previous location of the image. Just now I found one. The image had been deleted on December 9. That's fourteen days with that link hanging in the article. So I was thinking that perhaps it would be a good idea to create a page similar to the one we get when we move an article, informing us to check and fix double redirects, etc. that might have been created by the page move. Except this one would remind the Admin to check the articles where the image was linked and delete the broken link as well. Naturally, this situation doesn't exist with orphaned images, but other than that...perhaps it could be interesting? Would that be technically possible to create, in any case? Regards, Redux 01:44, 24 December 2005 (UTC)

You mean putting a warning note similar to what is on CAT:CSD?
Always remember to check "What links here" and the page's history before deleting. Also remember to check "File links" when deleting images.
Zzyzx11 (Talk) 07:21, 24 December 2005 (UTC)
If no one objects, I have taken the liberty of adding a reminder to the system message [1]. Zzyzx11 (Talk) 08:09, 24 December 2005 (UTC)
Looks good to me. Thanks Z! Redux 14:29, 24 December 2005 (UTC)


Clickable maps

I've continued my quest to find a good way of creating clickable maps on Wikipedia. Using the new {{click}} I've put together a map of a prominent portion of Ottawa. It looks much better and is more precise than my previous ACSII art maps, and it would be even better if it had been done by someone with a modicum of artistic ability.

My main concern is that this technique, which uses dozens of very small images, will cause the image server to explode. If this process can be used, I'm also wondering if there is anyway to make it easier. The very small map at Wellington Street took many hours of work. - SimonP 20:49, 23 December 2005 (UTC)

This is the absolute antithesis of everything HTML is about. {{click}}, in my opinion, is just a horrible hack. Developer's help needed: what are the security implications of HTML image maps? — Ambush Commander(Talk) 22:14, 23 December 2005 (UTC)
{{Click}} is a bad idea -- it's critical to the GFDL that credit for the images be easily available, and this hack circumvents that. There were long arguments about enabling linked images just for the {{Sisterprojects}} template on the main page. Add comments to bugzilla 1227 if you want developers to work on providing proper image maps, but I don't think this is the way to go about it. — Catherine\talk 00:00, 24 December 2005 (UTC)

vandal edit summaries violating privacy policy

Could a developer please remove the summaries from the following edits? All of these edits are from the Jimmy Wales article:

All the times are in Pacific Standard Time.

I will post the links to the bad edits on the George W. Bush article (also badly-hit) shortly. --Ixfd64 05:23, 23 December 2005 (UTC)

All right, here are the bad edits I found on the George W. Bush article:

Again, all the times are in Pacific Standard Time. I'm pretty sure that similar edits also exist elsewhere. Unfortunately, I do not have the time to look for all of them. If anyone spots any similar edits, please take the time to report them here. Thanks! --Ixfd64 05:57, 23 December 2005 (UTC)

The bush links don't load for me (starts my browser spinning until I close the tab), but the edits to Jimmy Wales giving Jimbo's address are pretty harmless. His address is already readily available to anyone with 30 seconds of time on google; for example it is available here: meta:Articles of Incorporation. Thue | talk 18:13, 23 December 2005 (UTC)

Rollback bug?

I ran across an odd edit at Troy. According to Master Jay this was the result of using "rollback" to revert this edit by 70.188.178.146 to the 20:44, December 18, 2005 revision of the article. But if you compare that version with the version produced by Master Jay's edit, you can see that in fact his edit, in addition to the intended reversion, introduced several other changes. Is there a bug in the "rollback" feature? Paul August ☎ 04:14, 22 December 2005 (UTC)

User:Master Jay isn't a sysop, so could not have used the rollback feature. If he used a local javascript hack, that hack might be buggy. --Brion 04:22, 22 December 2005 (UTC)
Let's cripple such hacks. 86.133.53.111 17:10, 23 December 2005 (UTC)

Watchlist

Is it possible to transclude my watchlist in User:Rschen7754/Tools just like the AFDs are set up? I'm not sure if it is but I thought I'd give it a shot. --Rschen7754 (talk - contribs) 08:27, 21 December 2005 (UTC)

Try it and see. 86.133.53.111 17:18, 23 December 2005 (UTC)

Possible commiting fraud using signatures

If enabling raw signatures, and add to the signature ~~~~ and/or {{subst:template}}, then those will get expanded by the next editor, not the editor whose signature contains given strings →AzaToth 16:17, 19 December 2005 (UTC)

  • Correct. If you know of anybody who does so, please notify us at the admin noticeboard. Disruption of this kind should be nipped in the bud. Radiant_>|< 16:24, 19 December 2005 (UTC)
    • It can be rather difficult, for example if someone combine this with hiding the content (pure css), and using a timed expansion (for example it only expand on fridays). It could be almost a week between the expansion and the original commit. →AzaToth 16:29, 19 December 2005 (UTC)
      • Yes, I realize that. It may be worthwhile to ask the Devs (on Bugzilla) to disallow the ~ character in raw signatures. In the meantime, if anyone is doing this, they should consider that it either violates WP:POINT or is downright vandalistic. Radiant_>|< 16:35, 19 December 2005 (UTC)
        • (Azatoth's comment stricken - WP:BEANS - I'm sure you mean well but please don't create guides to vandalize Wikipedia) Radiant_>|< 16:46, 19 December 2005 (UTC)
          • Oops, sorry. What I ment is that signatures should escaped some chars ({}~), also, I don't understand what WP:BEANS says, is it some internal? →AzaToth 16:48, 19 December 2005 (UTC)
            • In a nutshell, BEANS means: don't give potentially-malicious users ideas on how to do malicious things. android79 16:53, 19 December 2005 (UTC)
              • Hehe, it's tricky to describe a possible problem without give out the idea about the problem →AzaToth 17:02, 19 December 2005 (UTC)
  • Yep. Please notify the developers at Bugzilla about this issue. That should help, given the signature crisis of last month. Radiant_>|< 16:57, 19 December 2005 (UTC)
This is something I thought of separately, although it's good to see my thoughts echoed here. It's now something we're aware of and will likely patch up soon. 86.133.53.111 18:06, 23 December 2005 (UTC)

Seizure-inducing external links

An article I am looking to edit references an external link for relevant information. However, the site that is linked to contains pages that I am concerned may induce photosensitive seizures.

Seeing that the external link is the primary source of credible information for the article, what course of action would be appropriate to address the aforementioned concern? (Folajimi 21:22, 24 December 2005 (UTC))

Add a warning next to the link (for instance: *[http://www.example.com/ My credible information source] - ''WARNING: contains evil pages''). --cesarb 15:23, 25 December 2005 (UTC)


Problem with donations page header?

Just today, the little box at the top of each page which starts "You can help Wikipedia grow and improve by donating here!" has started appearing to the right of the page title, not above it like it usually does. If this is a bug, it's happening on every page I go to. If it is deliberate, I think it looks terrible :). So which is it? Raven4x4x 04:19, 23 December 2005 (UTC)

I bet some sysops are arguing over the best place to put it, so it's been schizophrenic. Try clearing your cache... if only I could find the page where this stuff gets edited. Personally, I don't like it floated to the right. Looks like some ugly ornament that was lackadaisically slapped on. — Ambush Commander(Talk) 04:21, 23 December 2005 (UTC)
The appropriate place is MediaWiki:Monobook.css. Titoxd(?!? - help us) 04:23, 23 December 2005 (UTC)

I've gotten fed up with the ugly popup box and have taken things into my own hands. Insert this code into your Monobook.css and the fundraiser notice will be waaay less intrusive (but still there). 

div#siteNotice {float:none;border:0;background:inherit;margin:0;}
div#siteNotice * {display:inline;font-size:8pt;}
div#siteNotice img {display:none;}

— Ambush Commander(Talk) 22:11, 23 December 2005 (UTC)


  • My suggestion is to add the following to your monobook.css file and get rid of it entirely:
#siteNotice { display:none; }
#mpbanner { display: none; }

JtkieferT | C | @ ---- 03:09, 25 December 2005 (UTC)

This, kiddies, is exactly what you should not do. =P Remember: you want to support Wikimedia! So leave the banner on! But feel free to get rid of the really annoying parts.
One last thing: setting siteNotice to display:none will prevent you from figuring out when the fundraiser is over, or when there's another important sitenotice to be read. — Ambush Commander(Talk) 03:20, 25 December 2005 (UTC)
oooOOOOooo... the ghosts of Wiki past, present, and future visit Jtkeifer in the night
:) rspeer / ɹəədsɹ 03:47, 25 December 2005 (UTC)

External Link Icon appears on every link

Within the past day or so, my browser has started showing the external link icon (box with an arrow) after every single link within articles, including internal wiki links. Its making WP quite annoying to read. Can somebody explain this? I haven't made any changes to my browser recently. --Mcpusc 12:33, 21 December 2005 (UTC)

Is your browser the Opera 9 beta? This is a known issue; report the bug to Opera. --Brion 23:44, 21 December 2005 (UTC)
Nope, its FireFox 1.0.7, which I've been using for months without this showing up. I can't recall making any config changes, but it this isn't a widespread problem, I guess I'll have to check that. --Mcpusc 00:24, 22 December 2005 (UTC)
I've been seeing this too, but only on links that I clicked to open in a new window or tab (so it's not exactly the same thing but looks related). I'm using Firefox 1.5. -- grm_wnr Esc 04:25, 25 December 2005 (UTC)

Telnet links

It seems that while external links to IRC servers (e.g. - [2]) and FTP servers (e.g. - [3]) link correctly, links to telnet servers (e.g. - [4]) do not get handled properly. I'd like to think this is a simple thing to fix, hence my posting it here. Comments? —Locke Cole 12:13, 18 December 2005 (UTC)

telnet is an obsolete protocol, and should not be encouraged to be used, there shouldn't be any links to telnet-servers from wikipedia at all →AzaToth 13:33, 18 December 2005 (UTC)
IRC and FTP are obsolete as well, so the "it's obsolete" argument doesn't really hold a whole lot of weight for me. As for links to telnet servers, while they continue to disappear, there are still bulletin boards (for example) that operate on the internet that may be worth linking to. —Locke Cole 13:51, 18 December 2005 (UTC)
So we shouldn't be linking to servers that only support HTTP/1.0 either? How is whether or not the protocol is considered deprecated at all relevant to anything? (and AFAIK IRC at least hasn't been deprecated by anything) —Ævar Arnfjörð Bjarmason 14:03, 18 December 2005 (UTC)
Some might argue that IM has largely replaced IRC (especially where the IM protocol supports having groups of people together ala an IRC channel). But that's really besides the point and not something I want to debate: whether or not it's obsolete is irrelevant. Afterall, I note that the old gopher protocol is supported: [5]. So can something be done for poor old telnet? :P —Locke Cole 14:15, 18 December 2005 (UTC)
Define obsolete... Telnet is still is use, remote Linux shells, finantial institutions MUD games etc still use them, as well as remote administration of routers, printers and what not. Now granted there are no longer a huge number of public telnet services available, but still. It's usefull if you want to provide a direct link to for example a MUD like telnet://discworld.imaginary.com or whatnot...--Sherool (talk) 14:43, 18 December 2005 (UTC)
Yeah, in my case I was specifically thinking of MUD servers that are open to the public. And like I mention above, for whatever reason, the Gopher protocol is still supported. So telnet seems like a given. :P —Locke Cole 06:18, 21 December 2005 (UTC)

Since this won't hurt anything (it's not like people are using telnet:// for anything else) and it might be useful I've gone ahead and enabled it in MediaWiki. —Ævar Arnfjörð Bjarmason 06:29, 21 December 2005 (UTC)

Thanks Ævar! —Locke Cole 17:00, 24 December 2005 (UTC)

unblocking long user names

I tried to unblock User:Thiswassostupidihadtocreatethisaccountjusttomakethisonecomment. The Block log indicated that the username was truncated to "Thiswassostupidihadtocreatethisaccountju". The original block seemed to be removed from List of currently blocked IP addresses and usernames. However, the block log for User:Thiswassostupidihadtocreatethisaccountjusttomakethisonecomment does not indicate that the user was ever unblocked. I'm wondering if there is a bug in MediaWiki for when unblock is performed on a long username. --JWSchmidt 21:40, 25 December 2005 (UTC)

I put a new block on User:Thiswassostupidihadtocreatethisaccountjusttomakethisonecomment. The List of currently blocked IP addresses and usernames is now incorrectly showing that User:Thiswassostupidihadtocreatethisaccountju was blocked, but the block log for User:Thiswassostupidihadtocreatethisaccountjusttomakethisonecomment correctly shows that user User:Thiswassostupidihadtocreatethisaccountjusttomakethisonecomment was blocked. --JWSchmidt 21:50, 25 December 2005 (UTC)


Improper Merger

Someone has improperly merged the article entitled, "Gang Stalking" with the article "Gang Stalk Persecution Complex." The result is that, when you attempt to comment on the need to delete the "Persecution" article as being very unprofessional, you are immediately switched over to the "Gang Stalking" article, where your comments in fact are registered.

Please unmerge these two articles. I am not a software expert and am only now beginning to barely grasp your diverse instructions, so, your assistance in this matter will be very greatly appreciated. Thank you.

Julianne McKinney

--70.236.161.247 20:40, 25 December 2005 (UTC)

This is a dup request, responded to at Wikipedia:Village pump (assistance). -- Rick Block (talk) 23:23, 25 December 2005 (UTC)

Date linking

Since the only real reason to link dates is for preferences to work - yet since these links usually just appear as overlinking, couldn't date links be auto-detected & NOT appear as links. Perhaps using 3 brackets could force them to actually link - though there is likely no real need for such anyway --JimWae 19:40, 25 December 2005 (UTC)

Sometimes it is useful to link dates, but generally you are right, it is pointless and missleading. Martin 20:42, 25 December 2005 (UTC)
I fully agree. A sepeerate kind of wikmi-markup to activate date preferences would IMO be best. (This should be done by an auto-detect, becauase in soem cases, such as in quotations, we don't want formats adjusted to fit preferences. Alternatively a markup for "exclude from preference alteration" with all not so marked being subject to prefernces would do the job too, IMO.) DES (talk) 20:52, 25 December 2005 (UTC)
See Wikipedia:Village pump (proposals)#Creating syntax for date preference formatting that isn't linking DES (talk) 01:15, 26 December 2005 (UTC)

Making preferences work is not the only reason to link dates, nor is it even the most important reason. Linking to dates allows users to check on other things happening on that date. Why would you consider unlinking dates? This serves no purpose. User:Zoe|(talk) 01:25, 26 December 2005 (UTC)

  • I think it is an important form of cleanup, a way of reducing overlinking that makes useful ;links less visible, a way of increasing the Signal to noise ratio. That is why i spent several hours working on it today. I feel that Wikipedia:Manual of Style (dates and numbers) supports this view, and has done so for a long time, and this view has consensus. DES (talk) 01:32, 26 December 2005 (UTC)

Autodetect page-blanking

Would it not be fairly easy to detect & override all attempts at page blanking?--JimWae 19:40, 25 December 2005 (UTC)

Yes, it's so easy to detect that the CVU already does it with their anti-vandalism bot. --cesarb 19:55, 25 December 2005 (UTC)

Well, I meant to say auto-detect AND auto-block --JimWae 19:58, 25 December 2005 (UTC)

Some instances of blanking are useful and should not be prevented. It's better to let a human do the determination. Besides, it makes a good and harmless canary. --cesarb 20:19, 25 December 2005 (UTC)
It's easier to fix if we don't stop them; if they knew immediately they'd been prevented from blanking a page, they'd just do something different. Perhaps WP:BEANS is relevant? -- SCZenz 20:35, 25 December 2005 (UTC)

How do you turn on Uploading?

I have a Wiki site and for the life of me I can't figure out how to turn on uploading for pics and such. This is what I have in my LocalSettings.php file and it still doesn't work. 

$wgEnableUploads		= true;
$wgUseImageResize		= true;
$wgUseImageMagick = true;
$wgImageMagickConvertCommand = "/usr/bin/convert";

Any ideas as to how to get it to work? Is this configured properly? Thanks and my email is bucklerchad@comcast.net I'm always online. Sniggity 21:45, 23 December 2005 (UTC)

Those settings should enable uploading. You have ImageMagick installed, correct? And, it would be helpful if you told us what the error message is, as well as your version of MediaWiki. — Ambush Commander(Talk) 22:12, 23 December 2005 (UTC)
that looks right to me. Does the upload link show and if so what happens when you try and use it? Plugwash 22:43, 23 December 2005 (UTC)
Alright, I downloaded ImageMagick, where would I install it? 69.250.106.39 22:54, 23 December 2005 (UTC)

I don't think I have it installed, at least I didn't install it. How/where would I install it if you don't mind me asking? Just upload it to my root folder via FTP? I'm brand new at this as you can probably tell. lol Also, the only thing is tells me is "Sorry, Uploading is Disabled" when one tries to add a picture to my site. The site is http:/www.wikibands.org/ if you'd like to try it out. There is still a lot of work that needs to be done, but it's hard doing it all alone. Thanks very much ! One more thing, I am using the most recent version of Wikimedia. Sniggity 22:46, 23 December 2005 (UTC)

no, the upload link doesn't show. Sniggity 22:46, 23 December 2005 (UTC)

Alright, I downl'd ImageMagick, where would I install it? Thanks Sniggity 22:55, 23 December 2005 (UTC)

Ah... nevermind with the ImageMagick. You probably already have GD installed, so there's no need to add another image editing software. Comment $wgUseImageMagick.
This may sound stupid, but make sure you've copied the configuration changes to the server. And it's MediaWiki (not Wikimedia). — Ambush Commander(Talk) 23:03, 23 December 2005 (UTC)

I scratched the ImageMagick, and now I have the following: 

$wgEnableUploads		= true;
$wgUseImageResize		= true;
$wgUseImageMagick = false;
$wgImageMagickConvertCommand = "/usr/bin/convert";

Also, what does uncomment/comment mean? I uploaded it to the server and this is the page I get when I try to upload a picture. http://www.wikibands.org/index.php?title=Special:Upload Sniggity 23:50, 23 December 2005 (UTC)

There's no need to add a <br> to the start of your message.
Commenting looks like this: //this is a comment. It's the way the line looked like before you put it there.
I'd like to test something: write after that block, put exit;. — Ambush Commander(Talk) 23:57, 23 December 2005 (UTC)

I'm sorry, I don't understand what you mean, after what block? Also, this is what I have now for localsettings.php: 

## To enable image uploads, make sure the 'images' directory
## is writable, then uncomment this:
# $wgEnableUploads		= true;
$wgUseImageResize		= true;
# $wgUseImageMagick = //
# $wgImageMagickConvertCommand = "/usr/bin/convert";

Sorry about all this, I hate being a noob at things. I learn fast though

Sniggity 00:08, 24 December 2005 (UTC)

If by block you mean the code, I now have this uploaded to the server: 

## To enable image uploads, make sure the 'images' directory
## is writable, then uncomment this:
# $wgEnableUploads		= true;
$wgUseImageResize		= true;
# $wgUseImageMagick = exit;
# $wgImageMagickConvertCommand = "/usr/bin/convert";

It's still not uploading. BTW, I apprecaite your help very much. Any other suggestions? Sniggity 00:12, 24 December 2005 (UTC)

I found out what version of GD I have, it's 1.6.2 not 2.0. Does that help in your quest to help me?

Sniggity 00:39, 24 December 2005 (UTC)

I just did this code and it now works: 

## To enable image uploads, make sure the 'images' directory
## is writable, then uncomment this:
$wgEnableUploads           = true;
$wgUseImageResize               = true;
$wgUseImageMagick = "";
$wgImageMagickConvertCommand = "/usr/bin/convert";

But now it's giving me this error message when I try to upload a picture and I CAN NOT FIX IT no matter what I try: 

Upload file
From WikiBands
Upload warning
The file is corrupt or has an incorrect extension. Please check the file and upload again.

Sniggity 00:45, 24 December 2005 (UTC)

What's the name of the file? — Ambush Commander(Talk) 01:49, 24 December 2005 (UTC)

They're just pictures I'm trying to upload from my computer. Here is a screenshot before I try to upload a pic: http://static.flickr.com/38/76739253_930f120362_o.jpg and here is a screnshot after I try and upload: http://static.flickr.com/41/76739254_876d45bc52_o.jpg

I'm just trying to upload regular .jpg's. I've also tried .bmp .png. .gif and none of them will upload. I also have second computer and have tried different pics and different computers and it still will not upload. Does it upload for you? Any idea as to what is happening? this is very frustrating, especially being a noob at this. Thanks again ! Sniggity 02:18, 24 December 2005 (UTC)

Make sure mime magic is on the system... you may have to disable that. Also, make sure that the jpgs are not actually corrupted or named incorrectly. BMP files are not accepted by MediaWiki by default. — Ambush Commander(Talk) 17:22, 24 December 2005 (UTC)

Ok, I went and grabbed this code from defaultsettings and pasted it in my localsettings and uploaded it to my server: 

/** Determines if the mime type of uploaded files should be checked
 * @global boolean $wgVerifyMimeType
*/
$wgVerifyMimeType= false;

/** Sets the mime type definition file to use by MimeMagic.php.
* @global string $wgMimeTypeFile
*/
#$wgMimeTypeFile= "/etc/mime.types";
$wgMimeTypeFile= "includes/mime.types";
#$wgMimeTypeFile= NULL; #use build in defaults only.

/** Sets the mime type info file to use by MimeMagic.php.
* @global string $wgMimeInfoFile
*/
$wgMimeInfoFile= "includes/mime.info";
#$wgMimeInfoFile= NULL; #use build in defaults only.

I changed it $wgVerifyMimeType from true to false and it worked, it let me upload a pic. But when I did, this screen popped up http://www.flickr.com/photo_zoom.gne?id=77073850&size=o so I hit refresh to take a screenshot as it loaded and to be able to read the error code and before it loaded, I got a screenshot of this error code: http://www.flickr.com/photo_zoom.gne?id=77076773&size=o I went to the Exif.php file and changed the print_r file from true to false and it still gives me that error. Everything else seems to be working, it's just that error code is showing up, any ideas on that? BTW, great advice on the Mimemagic being disabled, thanks !!!

This is very strange. It appears debugging is on. The code that's failing:
function debug( $in, $fname, $action = null ) {
		$type = gettype( $in );
		$class = ucfirst( __CLASS__ );
		if ( $type === 'array' )
			$in = print_r( $in, true ); 
	 
		if ( $action === true )
			wfDebug( "$class::$fname: accepted: '$in' (type: $type)\n");
		elseif ( $action === false ) 
			wfDebug( "$class::$fname: rejected: '$in' (type: $type)\n");
		elseif ( $action === null )
			wfDebug( "$class::$fname: input was: '$in' (type: $type)\n");
		else
			wfDebug( "$class::$fname: $action (type: $type; content: '$in')\n");
	}

You seem to be using a PHP version older than 4.3.0, because that second parameter was added in that version. Bug your webhost to get it upgraded. This is likely what was causing all the other problems: the PHP installation was poorly configured and mime.magic was not available. — Ambush Commander(Talk) 20:03, 25 December 2005 (UTC)

Tis very odd. I got it to work though, and here's how: 

function debug( $in, $fname, $action = null ) {
                $type = gettype( $in );
                $class = ucfirst( __CLASS__ );
                if ( $type === 'array' )
                        //$in = print_r( $in, true ); 
         
                if ( $action === true )
                        wfDebug( "$class::$fname: accepted: '$in' (type: $type)\n");
                elseif ( $action === false ) 
                        wfDebug( "$class::$fname: rejected: '$in' (type: $type)\n");
                elseif ( $action === null )
                        wfDebug( "$class::$fname: input was: '$in' (type: $type)\n");
                else
                        wfDebug( "$class::$fname: $action (type: $type; content: '$in')\n");
        }

I commented out the lines with the print_r sentence using "//" and now it works great !Those lines are to store the data passed to the script in a debug log, so, if you are not debugging the PHP script it will not affect you.

Ambush Commander, I can't thank you enough my friend. I helped me out of a bind and you're help won't go unnoticed. If there's anything I can do for ya, don't hesitate to ask. I'm sure I'll have a few more problems, as we all with new sites, so I may be back. Again, thanks ! Have a great holiday season. Sniggity 00:27, 26 December 2005 (UTC)

No problem! Just be careful when upgrading (your patches will be overwritten). And don't forget to bug your host about the PHP version. Happy holidays to you too. — Ambush Commander(Talk) 00:58, 26 December 2005 (UTC)

Google Web Accelerator causing problems?

I recently gave UrineForGas (talk · contribs) a {{usernameblock}} and managed to inadvertently block several other users.

It seems that they are using Google Web Accelerator and this seems to be acting as a IP proxy. As our article says it seems to cause administrative websites such as ours problems.

IP addresses mentioned are 64.233.173.85 and 72.14.194.19 -- both belong to Google web accelerator. See http://ws.arin.net/cgi-bin/whois.pl

Do we have a policy of blocking proxies on site because they cause problems with IP addresses? I know you can't edit through Babel Fish (website). Or not? — Dunc|☺ 17:02, 18 December 2005 (UTC)

If anyone from anywhere on the Internet can use it as a proxy, even if it requires installing a specific program, it should be considered an open proxy, and indefinitely blocked. --cesarb 23:07, 18 December 2005 (UTC)
Open proxies are blocked when they allow editing, not when they simply read pages. It's very common for ISPs and countries to use read caching proxies and those are harmless. The potential for harm from the Google Web Accelerator is that it goes and loads pages which the user hasn't clicked on. It's proably insignificant at present. Jamesday 22:32, 19 December 2005 (UTC)
Yes, that's what I meant; I should be more careful in my phrasing. --cesarb 19:21, 25 December 2005 (UTC)
The official FAQ says it can be disabled for specific sites, which means it might be possible to block it with a message asking people to add wikipedia to the list. But this should probably be discussed at WP:AN first. --cesarb 23:10, 18 December 2005 (UTC)

Prefix index buggy

If you for example type "user" as template prefix, you only get till "user c", then there is a redirect to "user cblah". You can never get "user f*" etc... by this search. →AzaToth 02:35, 27 December 2005 (UTC)


Upload overwrites limited

I've disabled the ability for brand-new accounts to overwrite existing files by uploading a new file with the same name. While sometimes legitimate modifications are made this way, vandalism in image replacements is common and relatively difficult to notice or track (and more importantly, caching of images with our current systems means the bad image versions sometimes still get shown after fixing); so this should cut down on it a bit until things get improved.

My general inclination is to make uploaded files immutable, so any new version always would go under a new name, but I'm not going to make such a change unilaterally. --Brion 07:29, 26 December 2005 (UTC)

How do you define "brand new"? Less than X edits? Less than X edits to articles? 'etc Raul654 07:30, 26 December 2005 (UTC)
I imagine it's percentage (newest X% of accounts), similiar to page moves and semiprotection. But that's just an assumption based on current technical measures.--Sean|Black 07:34, 26 December 2005 (UTC)
We dropped the percentage method when semi-protection support came in. :) 86.133.53.111 19:25, 26 December 2005 (UTC)
Currently, less than 4 days since registration. --Brion 07:40, 26 December 2005 (UTC)
I'd really prefer to be able to keep the various revisions of the same image all in one place. Image:Grapes.jpg is illustrative of both the pros and cons of the current approach. Backtracking through various pages, as you must to get to the original image from Image:Pomegranate03 edit.jpg, isn't terribly reliable. I don't suppose it's at all feasible to make overwritten images generate watchlist entries? —Cryptic (talk) 17:40, 26 December 2005 (UTC)
I agree that uploaded files should be immutable. I would also support the ability for admins to soft-delete them; i.e. they could be resurrected if desired. Of course developers could always hard-delete, though I can't really see a circumstance where they would need to. [[Sam Korn]] 19:44, 26 December 2005 (UTC)

How do I get template info boxes on my site?

For instance, I want to put this onto my site, but it comes up as a red link at the very top of the page: 

{{Infobox Guitar model|title=Gibson Les Paul
|image=[[Image:LPSHC.jpg]]
|manufacturer=[[Gibson]]
|period= [[1952]] — [[1960]], [[1968]] — present
|bodytype=Solid
|necktype=Set
|woodbody=[[Mahogany]], [[Maple]]
|woodneck=[[Mahogany]]
|woodfingerboard=[[Ebony]], [[Rosewood]]
|bridge=Fixed
|pickups=2 [[Humbucker]]s (originally [[single-coil]]s)
|colors=Various (often natural-type finishes)
}}

Here is a picture of what happens when I try to add an info box: http://static.flickr.com/38/77463394_d78d26239d_o.jpg

Thanks

It looks like you don't have {{Infobox Guitar model}} on your wiki. You need to copy it over. You may also need to copy over templates that it uses, although some may come as standard with MediaWiki.-gadfium 07:40, 26 December 2005 (UTC)

Know where I can pick up those modules?

It's not a module, it's a template. Just copy and paste the contents from Template:Infobox Guitar model to your corresponding page. Also, don't forget to sign your comments with ~~~~. — Ambush Commander(Talk) 20:23, 26 December 2005 (UTC)

Hey Ambush, sorry about that, but it's me, Old sniggity again. Thanks, it worked !

Sniggity 01:03, 27 December 2005 (UTC)

Talk page diff link

I really like the new diff link that comes up on the "New Message" bar. However, if I could request one small change: The diff link only shows the diff of the last person to edit the message. If it could show all the edits that were made since you last saw the page it would be a lot more useful. — Asbestos | Talk (RFC) 15:47, 21 December 2005 (UTC)

This would be helpful for watchlists, but not very feasible, as the MediaWiki software would have to know which pages you had seen. (huge overhead) — Ambush Commander(Talk) 22:35, 21 December 2005 (UTC)
I believe the proposal is simply to show changes since the last time that user had viewed their talk page (and only that page). This isn't unreasonable at all, only requiring the storage of the version of that page last seen by that user. Deco 22:45, 21 December 2005 (UTC)
Ah... that's right. Well... I'm just some lowly MediaWiki hacker with no arcane knowledge of MediaWiki's internals, so I really can't tell you if it's feasible or not. — Ambush Commander(Talk) 23:01, 21 December 2005 (UTC)
I coded something people like? I must be doing something wrong. Anyway, hmm...well, it's technically feasible, I suppose, but a bit more fiddly. Still, I'll have a play - nothing's impossible, after all. 86.133.53.111 13:37, 22 December 2005 (UTC)
That might seem like a small change to you but in fact, it is not. —Ævar Arnfjörð Bjarmason
Ævar, if it is you who coded the change, then yes I'd like to add that I think it's a great idea, and very helpful. I'd usually open the history right away to see what had changed, and this saves me several steps. Thank you for this excellent addition. — Knowledge Seeker দ 05:21, 26 December 2005 (UTC)
Nah, Avar didn't code it, but he did commit it, however, he was talking to me. I'm not sure what he's on about though, so I'll pester him later. 86.133.53.111 19:17, 26 December 2005 (UTC)

What the hell just happened?

All the links became nonsensical HTML entities. I've just reloaded, and it's OK. What happened? Sceptre (Talk) 14:07, 28 December 2005 (UTC)

Someone pressed the wrong button, I guess. Morwen - Talk 14:07, 28 December 2005 (UTC)
Typo in config, it's fixed. --Brion 14:08, 28 December 2005 (UTC)
Scared me for a second, though that FF had been infected with spyware. Sceptre (Talk) 14:13, 28 December 2005 (UTC)


Please allow Wikipedia users to configure their browsers to block ads

I just figured out why I've not been seeing images on Wikipedia for months now. It first happened when Wikipedia was in the midst of a big upgrade and I figured there were bandwidth problems. Now I see that the trouble is that images are coming from wikimedia URLs (IIRC), and I've configured my browser so that it does not load images that are not from the originating web site. Frankly, I'd recommend this to anybody as it means that I see hardly any ads. I've no desire what so ever to change my browser's configuration, but because the article images are not coming from Wikipedia URLs I see (practically) no images on Wikipedia.

I plead with the Wikipedia techs to create dns CNAME records that point to the wikimedia hosts and re-configure Wikipedia to source images from Wikipedia URLs. It seems a simple enough thing to do and empowers the users of Wikipedia with the simplest, and possibly most effective, ad blocking technique.

Thank you for your attention. (I would like this comment to be more permanent, but this seemed the most appropriate place to post.) --kop 07:15, 28 December 2005 (UTC)

Alas, on reflection it is not quite so simple. The squids would have to be configured to proxy the images as well. Regardless, I would welcome some solution that both lets me block ads in this fashion _and_ see Wikipedia images. Maybe somebody smarter than me can come up with a simple solution. --kop 07:40, 28 December 2005 (UTC)
What browser are you using? In Mozilla Firefox at least there's an option to allow certain sites to load these images, so if you allowed wikipedia.org, wikimedia.org and so on there wouldn't be a problem. Maybe.--Cherry blossom tree 11:33, 28 December 2005 (UTC)
I don't think this is going to get changed as it provides a valuable extra defence against scripting based attacks in the event they slip past the upload sanitizer. Plugwash 12:28, 28 December 2005 (UTC)
There are two things:
  1. Images are served from a separate web server which is optimized for serving static files. As we expand this may become a cache cluster itself, which will be separate from the one that serves wiki pages because it has different needs.
  2. Images are served from a different hostname to protect against JavaScript attacks that might get past our security filters.
If you want to block ads, feel free. But we probably will continue to serve uploads from a separate domain, as we have for over a year, and as many many large sites do, for the forseeable future. --Brion 14:13, 28 December 2005 (UTC)
So, basically what you're saying is, you'd like it so you could use Wikipedia and still be able to view other sites without the horrible burden of glancing at revenue-generating ads. The web thanks you. --Golbez 15:02, 28 December 2005 (UTC)
I, for one, don't see a need to block ordinarly, static ad images in sites. However, I'm getting increasingly annoyed with various sites' attempts to force annoying, intrusive ads on people via popups, popunders, and other techniques, sometimes crafted carefully to get through browser features designed to block these things. I hope Wikipedia never inflicts that sort of thing on us. So far, the most it's done is to get all "PBS" on us with these pledge drives. *Dan T.* 15:12, 28 December 2005 (UTC)
Use Mozilla Firefox, Adblock Plus, Filterset.G and Filterset.G.Whitelist. It will have the same, if not greater, effect of your current setup, and you will be able to see images. — Ambush Commander(Talk) 15:08, 28 December 2005 (UTC)
Except maybe those images that are randomly served from the /ad/ directory. --Golbez 15:09, 28 December 2005 (UTC)
And thus Filterset.G.Whitelist. I was having that problem a while back too. — Ambush Commander(Talk) 15:11, 28 December 2005 (UTC)

Uploaded image not displaying

Hi,

I just uploaded an image (Sergio_Marchionne.jpg) and although it displays on my computer, it only shows the placeholder on Wikipedia.

Primetime 23:09, 27 December 2005 (UTC)

Works for me. -- Finlay McWalter | Talk 00:06, 28 December 2005 (UTC)
Ah, on looking at the image's page, Image:Sergio Marchionne.jpg, user:Dbenbenn reverted your reupload, as the version was defective. Indeed, that version is broken - perhaps it didn't upload correctly. -- Finlay McWalter | Talk 00:08, 28 December 2005 (UTC)

Image upload licensing line broken for IE

If you try to upload an image with IE, the handy-dandy licensing template button doesn't work - or more to the point, only the top three lines can be selected (no license, unknown licence, "image found"). Any useful templates you have to enter by hand. Can this bug be fixed, please? Grutness...wha? 12:33, 16 December 2005 (UTC)

No repro for me with IE 6.0.3790.1830 on Windows Server 2003 SP1 Standard. Deco 02:13, 17 December 2005 (UTC)

Well it's no go with 5.2 on Mac OS 10.2. Grutness...wha? 23:05, 17 December 2005 (UTC)

Please file it as a bug report on BugZilla if it's a reproducible bug, and provide clear information on the conditions under which it occurs, and steps for reproducing, if possible. Thanks. Rob Church Talk 04:00, 18 December 2005 (UTC)
Thanks - now listed as Bug no. 4313. Grutness...wha? 06:11, 19 December 2005 (UTC)
Whoops; the bug it's duped to is one I was aware of. Getting forgetful in my old age. ;-) 86.133.53.111 13:26, 22 December 2005 (UTC)
When talking about IE mac please state that from the start. When you just say IE people will assume IE windows which is build on a totally different rendering engine from IE mac. Plugwash 16:13, 25 December 2005 (UTC)
Whoops - yes, sorry. I did when I started to write this, but it got lost in my copyediting before I pressed save. Grutness...wha? 23:57, 27 December 2005 (UTC)

Problem with old browsers?

I've recently been editing from a 6 year old iMac running IE 5.1. Yes, I know it's ancient. It's a long and painful story as to why I'm not currently on my Tiger powered iBook G4, so don't ask.
The problem I'm having occurs when trying to revert long articles. Most recently it occurred on Julius Caesar. I think my browser might not be loading all the code into the edit window for some reason. Is having an old browser a possible cause? I know that there are special workarounds for it not being unicode compliant, but might there be other issues without workarounds? WAvegetarian (talk) (email) (contribs) 07:45, 29 December 2005 (UTC)

yep older browsers have problems editing longer articles. See Wikipedia:Article size. Just upgrade youyr browser (or edit only sections in long articles). Broken S 08:21, 29 December 2005 (UTC)
IE for Mac can't edit pages longer than 32kb correctly. If you have to use a machine too old to run Safari, try Firefox or a classic Mozilla release... --Brion 10:40, 29 December 2005 (UTC)
Thanks for the input. I just downloaded Mozilla 1.0.2, which is the most recent build that supports OS 9. I hope to be back up and running on my iBook tomorrow as it is due back from data recovery.
Warning: Fuming ahead...Ignore it if you want, that's why it's in small type. Directories are wonderful things until they get deleted. Do not, I repeat do not use AppleCare's provided disk utility program called TechTool if you got AppleCare before upgrading to Tiger. It will crash your machine and erase your directory information. Apple of course maintains that it's third party software and that data loss, even data loss admittedly caused by them, is not covered under warranty. :-t
WAvegetarian (talk) (email) (contribs) 11:28, 29 December 2005 (UTC)


Errors

Got a lot of errors now "Fatal error: Cannot instantiate non-existent class: outputpage in /usr/local/apache/common-local/php-1.5/includes/Setup.php on line 284" →AzaToth 22:33, 28 December 2005 (UTC)

Likely another quick configuration/coding error that slipped past the devs. OutputPage is instantiated for all requests, so it's not surprising that's the one that failed. — Ambush Commander(Talk) 00:09, 29 December 2005 (UTC)
"instantiate "? User:Zoe|(talk) 03:22, 29 December 2005 (UTC)
Instantiate. --cesarb 15:10, 29 December 2005 (UTC)

Deletedpage templates

This is sort of a minor point, but one that bothers me a bit. Is there any way to make deleted articles that have been protected with the "deletedpage" template still appear as redlinks? The illusion created that there are articles under that name is a bit misleading, and throws me off when I'm looking for recreations of deleted pages. I know little about the technical issues of WP, and I suspect that if there is a way to do this it's too complicated for such a minor issue, but I thought I'd throw the idea out there anyway, and also see if anyone has similar feelings. -R. fiend 19:59, 28 December 2005 (UTC)

Set your stub limit to 20 bytes or so in preferences and add a.stub { color:#CC2200; } to your user css? —Cryptic (talk) 21:40, 28 December 2005 (UTC)
A general fix is certainly possible, but would require a software change and would likely require a new option when deleting an article to disallow recreation. Feel free to enter this as an enhancement request at bugzilla:, but don't be surprised if it is not acted on very soon. -- Rick Block (talk) 21:43, 28 December 2005 (UTC)

AOL user trying to create accounts

We keep getting letters to the Help Desk mailing list from AOL users who are saying that when they try to create an account, they are told that they can't because they've already created 10 accounts. Is this new account counter keeping track of all AOL users? How often does it get reset, and is there something that can be done so AOL anons can create accounts without running into this problem? User:Zoe|(talk) 17:49, 28 December 2005 (UTC)

The whole problem is that AOL shares the same IP address for multiple users. Not only that but, from what I've heard, it changes the IP address with each request — but requests to the same page (by different users) often get the same IP address. That would make it look like the 10-account/day limit is being used for the whole AOL (when in fact it's per IP address). --cesarb 20:05, 28 December 2005 (UTC)

MediaWiki:Youhavenewmessages' diff link doesn't work in Classic skin

The message itself is great, and works: you get a link to your talk page, but the diff link doesn't work at all. Instead, you get another link to your talk page. I'm going to go make a bugzilla report, but I figured I'd post here first in case someone had an idea about fizing it. Blackcap (talk) 17:11, 28 December 2005 (UTC)

I've listed the bug at bugzilla:4411. If anyone's had this happen to them or has any information on it, please post there. Blackcap (talk) 17:47, 28 December 2005 (UTC)

Fixed! Thanks to Brion and Rob Church for fixing it so quickly. Blackcap (talk) 22:26, 28 December 2005 (UTC)

Image uploads

Since the implementation of the copyright drop down box, I have noticed a considerable increase in the number of images incorrectly/deceptively tagged as some variant of free use. A more useful inclusion for the uploader to be forced to include would be source information - that way image copyright tags can be verified more simply. Would there be any possibility of including a field on the upload page that prompts the uploader to include image source information?--nixie 16:15, 28 December 2005 (UTC)

No need to worry about images without sources - if they don't provide a source, after being asked, delete the images, no matter what the "license" tag claims. Simple as that. JesseW, the juggling janitor 21:21, 28 December 2005 (UTC)

??? Entrys

???? appear evrywhere instead of english links in mediawiki specific context like on the search button. Language is set to english en in preferences and its the en wikipedia. so there seems to be a problem here. Problem noticed on this page: here


helohe (talk) 14:10, 28 December 2005 (UTC)

It's fixed... everyone had it briefly, I think. Shimgray | talk | 14:13, 28 December 2005 (UTC)
I had this both here and on Commons for about 5 minutes yesterday. I got what looked like Korean characters rather than question marks though. Thryduulf 10:01, 29 December 2005 (UTC)

Renaming images?

any way to quickly rename images or move them to a different name rather than the slow way of creating another image space and listing the old one for deletion? Jarwulf 06:32, 27 December 2005 (UTC)

Not really. I think this has, at times, been done by bots, does anyone know details on that? -- Jmabel | Talk 07:12, 27 December 2005 (UTC)
Is there any particular reason we don't have a "move image" feature though? Even though it would mean to manualy fix the image links afterwards (that could be done by a bot though) moving the image along with all it's revision history to a new name would be better than having to delete the old one and re-upload the image. If it's a security issue we could make it an admin only feature and do it though requested moves or something. --Sherool (talk) 00:19, 28 December 2005 (UTC)
Exactly...we already have a move function for articles...why not images? Jarwulf 07:43, 28 December 2005 (UTC)
I've tried doing it before, and surprisingly, Movepage will display Image:Whatever in the box, but then it will stop you from moving it to another page in the Image namespace. Perhaps it's an architectural problem? — Ambush Commander(Talk) 15:10, 28 December 2005 (UTC)
Well the standard move feature won't work. You need to add some code to "physicaly" rename the image file(s) on the server as they are not stored in the actual database, and I suppose find a way to mark the move event in the file history and such. I suppose a server "hickup" in the middle of a move operation on a image with many revisions might cause some "icky" results too... --Sherool (talk) 01:22, 29 December 2005 (UTC)

Q) Bot usage

Help!! (admin) bot usages!!
1. user:A tagged {{no license|~~~~~}} to "image:sample.jpg" uploaded by user:B
2. user:Cbot tag {{subst:Image copyright|Image:sample.jpg}} --~~~~ to user_talk:B, and change {{no license|~~~~~}} to {{no license notified by bot|~~~~~}} automatically.
What is bot command??
2. user:Cbot insert speedy deletion tag to Image:sample.jpg after 7 days automatically.
What is bot command??
3. user:Dbot(admin) delete all "image:..." in the certain speedy del category automatically.
What is bot command?? --WonYong 13:41, 30 December 2005 (UTC)
Err, what do you mean? There is no spesific command for that, the bot would have to be coded to parse the text and make apropriate edits. If you are looking for a bot to do these things try Wikipedia:Bot requests. Bots are automaticaly delete stuff is generaly not a good idea though, but the tagging can be done, and I believe User:NotificationBot already does most of that (except list for speeedy deletion), but it's not up and running full time yet aparently. --Sherool (talk) 16:47, 30 December 2005 (UTC)

User talk shortcut at image file history

How to add the user_talk shortcut to image file history.
For ex:

(del) (cur) 03:39, 5 August 2005 . . Dtobias . . 182x71 (3396 bytes) ({{Logo}} AFGNIC logo (.af domain registrations: Afghanistan).)

into:

(del) (cur) 03:39, 5 August 2005 . . Dtobias (talk) . . 182x71 (3396 bytes) ({{Logo}} AFGNIC logo (.af domain registrations: Afghanistan).)
--borgx (talk) 06:33, 29 December 2005 (UTC)

Javascript I presume. Try Wikipedia:WikiProject User scripts — Ambush Commander(Talk) 16:38, 29 December 2005 (UTC)
User:Cryptic/addtalklinkstoimagepages.js. There's some not-so-good instructions for installation at m:Help:User style. —Cryptic (talk) 23:37, 29 December 2005 (UTC)

MW now has this feature built in! Thanks to Robchurch (see bugzilla:4426)


White space, but how much?

I run several times a day each day into articles having too much whitespace between paragraphs, or at the top of the articles. Well, one may say that is no big deal.


I will argue however, that such whitespace makes the articles look unprofessional, especially that in some places the whitespace is wider than others (see above). And most of the time that whitespace is not created on purpose, rather by editor negligence. I have raised this issue before, but now I really want to ask a question:

Does anybody know why the wiki engine does not strip extra whitespace when converting to html, like html itself and LaTeX do? Also, does anybody know why it is convenient to allow users to insert whitespace like that, and why not just use <br> when it is absolutely necessary?

In short, would it be a good idea to file it as a bug at Bugzilla that in future versions of the software the engine strip all the extra whitespace, while still allowing users to deliberately insert whitespace with <br>? Oleg Alexandrov (talk) 05:29, 29 December 2005 (UTC)

Never underestimate the hideous complexity of Parser.php. I see the definite alternative you proposed, so I say go for the Bugzilla and see what the developers think (eg. Don't expect results). — Ambush Commander(Talk) 16:36, 29 December 2005 (UTC)
As to why whitespace is allowed, the wiki editing rules are meant to be extremely simple and not require any particular technical knowledge nor anything other than a text input box in a web browser to use. Knowledge of HTML or use of a WYSIWYG word processor, in particular, are not assumed. Plain text is simply too plain for most people's tastes, so some markup syntax is supported including limited use of HTML. Allowing any markup immediately hits a feature creep slippery slope, but understanding the original intent hopefully helps fight the notion that wikimarkup might as well be an XML. The markup is intended to be directly edited by humans. -- Rick Block (talk) 17:43, 29 December 2005 (UTC)

Cache tells me Access Denied?

When I try to access Wikipedia, I get this error message:

Access control configuration prevents your request from being allowed at this time. Please contact your service provider if you feel this is incorrect.
Your cache administrator is wikidown@bomis.com.
Generated Mon, 26 Dec 2005 22:06:27 GMT by bayle.wikimedia.org (squid/2.5.STABLE12)

The email address is apparently invalid, as emails bounce back. Can anyone point me in the right direction of the person to contact about the block? --Interiot 22:13, 26 December 2005 (UTC)

if it persisits contact noc@wikimedia.org or #wikimedia-tech on freenode. if it just happens from time to time its probablly just poor handling of what happens when the system can't cope. Plugwash 22:29, 26 December 2005 (UTC)
Another explanation is that this one is the message which appears when a "live mirror" is blocked (I've seen it before when checking these kinds of mirror). If so, it will not go away on its own; you should contact the developers. --cesarb 20:24, 28 December 2005 (UTC)
The live mirror message is different, and the developers are insidious enough to link to the actual page and do all the copyright stuff, except they don't give the actual article. I doubt it's a live mirror blocking issue. — Ambush Commander(Talk) 16:39, 29 December 2005 (UTC)

CVU members list help needed with edit sections and ....

CVU members list show on WP:CVU had a problem where when there were numbers and sections for TOC and numbers were not reset editing a section did not have the text of that section. Because of this it was removed. People would like numbers back and edit sections working with text. Could someone fix this? --Adam1213 Talk + 12:51, 14 December 2005 (UTC)

Is that page doing a lot of transcluding? Fancier transclusions (and meta transclusions, the evil things) are known to break section editing and a few other things. 86.133.53.111 19:24, 26 December 2005 (UTC)
That page was causing bugs before, we had the same problem on the stub sorting members list, and there are over 300 CVU's now, as everyone is categorized a current list and count is availble in the category listing, as suggested here, doesn't mean this can't have some repair, but it's a workaround. xaosflux Talk/CVU 04:41, 30 December 2005 (UTC)

New Cite extension (now installed on Wikimedia)

Note: This has now been installed on the cluster, happy hacking;) —Ævar Arnfjörð Bjarmason 01:08, 25 December 2005 (UTC)

Following a discussion on my talk page I've written a followup to my Special:Cite extension, a specification and test page for it are avalible on my test wiki.

It basically adds sofware support for the {{ref|}} and {{note|}} hacks that are being used right now for citations, citations can be specified inline with it as <ref[ id]>Some citation text here</ref>, if you supply id you can reference it again as <ref id/>. There's no need for adding something like {{note|}} for each citation anymore because the citation section will be automatically generated on the page wherever the editors put <references/>. This improves workflow a lot because it's possible to edit an individual section adding <ref> tags in it and it'll automatically be referenced in the section where <references/> is. The output of <references/> is completely customizable through the MediaWiki namespace so editors aren't bound by what I cooked up.

I'd like to get some comments on how it works (and should work), some things that need fixing currently are to make the error output prettier (need some styling, any ideas?) and most of all, make it work with my mortal enemy, HTML Tidy. —Ævar Arnfjörð Bjarmason 01:24, 21 December 2005 (UTC)

Way frickin' cool! Someone built a telepathy machine and figured out what I was hoping for regarding footnotes! Yes, please add this into the code, exactly as you've got it specified, so I can start using it yesterday! Really, that's exactly the way I had envisioned a proper references system working, I just didn't have time to put my thoughts down into a statement more coherent than this spec. B-) Slambo (Speak) 02:20, 21 December 2005 (UTC)
I object to the abuse of XML-like syntax. It would be better to use <ref name="id">...</ref> and <ref name="id"/>. This is more similar to all other uses of XML-style tags on Wikipedia, and the use of "name" has a precedent on HTML forms (use of "id" as the attribute name would be confusing, since IDs are supposed to be used only once, while on HTML forms a "name" value is often used more than once). --cesarb 02:30, 21 December 2005 (UTC)
While I agree with CesarB's objection, I found one part of the extension a bit difficult to understand: the automatic numbering of the list doesn't correspond to the actual citations. So you clicked number "6" and you get sent to number 5. I understand that the way the system is built to handle multiple citations, this must be done, but it is confusing.
While I haven't thought about it much, I might propose keeping all reusable refs the same number. But whatever you do, I think you keep the list number matching the reference number. — Ambush Commander(Talk) 02:34, 21 December 2005 (UTC)
I was a bit unsure about how to handle that, articles like Alchemy currently use a b c d e f ... for that. That's a bit hard to handle in code though (remember that it has to handle every language out there), one way to do that would be to have a MediaWiki message with whitespace delimited tokens that should be used for that, .e.g "a b c d e f g h.." or "á a b é ..." depending on the language. Another way would be as you suggested to use the number of the actual reference, to do that properly however it couldn't use a HTML ordered list (<ol>), that's doable too, and that could easily be supported. —Ævar Arnfjörð Bjarmason 03:21, 21 December 2005 (UTC)
The convention for backlinks became a b c... simply to have symbols which are different for many backlinks. I thought it would be awkward to use numbers for both downlinks and uplinks. Maybe PHP could emit other symbols, although having them memorably sequential is helpful for tasks such as clicking on each one while checking refs. (SEWilco 04:55, 23 December 2005 (UTC))
Yup... it seems the numbers of the links aren't that important to you. Love the idea though. Absolutely awesome actually. — Ambush Commander(Talk) 02:36, 21 December 2005 (UTC)
I like the idea, but is this the 3rd or 4th footnoting system? Already putting urls in square brackets makes big numbers [6]. Then {{ref| whatever makes little numbers. There's some other system I think that forces you to number the references explicitly. And now this. I like it, but please can we kill off the other systems? Also I agree about the faux-XML syntax. It would be nice if it could conform to wiki syntax, like being <<ref id|url>> or ((ref id|url)) or something. Stevage 02:40, 21 December 2005 (UTC)
Discuss the multiple systems at the Talk page for WP:CITE. (SEWilco 04:55, 23 December 2005 (UTC))
Agree, I like the XML though, just use name instead of id. —Locke Cole 02:57, 21 December 2005 (UTC)
I chose to go with <ref[ id]> rather than <ref[ id=id] because it's quicker to type and there isn't a current need (and I can't imagine a need) for extra parameters, but even if extra parameters were to be added it would be easy to support them and the current system. Our extension paramters aren't XML anyway, not even close. —Ævar Arnfjörð Bjarmason 03:12, 21 December 2005 (UTC)
I would strongly suggest using <ref[ name=id]>. —Locke Cole 03:35, 21 December 2005 (UTC)
They aren't XML, but they are XML-style. You'd be having one style (key=value) for almost everything, and another (naked value) just for <ref>. I can also easily imagine a lot of extra parameters: class=, style=, dir=, id= (to turn it into an anchor), etc. Supporting both key and value pairs and a naked value can quickly become confusing. I also think it's better to use <ref [name="id"]> instead of <ref [id="id"]> due to the common implied semantics of the id attribute. --cesarb 03:39, 21 December 2005 (UTC)

Some valid suggestions above (I especially agree with the need for the inline-number to match up with the down-below-number, one way or another, in all cases: otherwise printed output is essentially broken, since you can't figure out which note is for what) but I'm mostly going to say this: PLEASE HURRY! THIS IS GREAT! :-) (Sorry for the shouting.) Oh -- why not dump all the notes at the bottom of the page if a <references/> tag isn't present? That would make it even friendlier to learn. —Bunchofgrapes (talk) 02:48, 21 December 2005 (UTC)

agreed. Despite it's learning curve (I figured it out in 2 minutes), it's better than what we have now. Broken S 02:53, 21 December 2005 (UTC)

There is very little in wikipedia where the code is XML-like and I'm not sure we want to encourage that. To borrow a page from Docuwiki, how about notation like ((footnote text)) and for repeated references ((name | footnote text)) with simply ((name)) being refered back to the defined footnote text. I realize such a configuration would be somewhat harder to write a parser for, but wiki code is meant to be as simple as possible on the writer, not the parser, right? Regardless an effective citation system would be a big step up for wikipedia. Dragons flight 03:20, 21 December 2005 (UTC)

Because for a parser it introduces possible ambiguity. OTOH, XML is a standard, and it should be embraced. —Locke Cole 03:31, 21 December 2005 (UTC)
Another possibility is to use a template as an interface so the template syntax is used. {{ftemplate|text}} or {{ftemplate|name=Time3|text}} could be used, with the template emitting the XML-like wrapper. (SEWilco 04:55, 23 December 2005 (UTC))

This is your brain | This is your brain on Tidy, Tidy completely messes up inline parser extensions for some reason and until that's fixed (or Tidy is disabled) it can't and will not be enabled on Wikimedia sites. I think tidy should be disabled anyway, it encourages sloppy and unportable syntax. It was disabled for a few days not so long ago and people whined to get it back;/ —Ævar Arnfjörð Bjarmason 03:26, 21 December 2005 (UTC)

I'd like to help, but I fail to see any difference between these two links. --cesarb 03:47, 21 December 2005 (UTC)
See these screenshots: without Tidy, with Tidy. —Ævar Arnfjörð Bjarmason 04:05, 21 December 2005 (UTC)
BTW, Avar, your site is consistently giving unknown page format errors on IE for me and wants me to save your content to disk rather than view it. Netscape is well behaved though. Are others getting the same response? Dragons flight 03:49, 21 December 2005 (UTC)
I think this is because the pages have no extensions, and so IE doesn't open them as html pages. I saved the spcs page to my hard disk, remnamed it to cite.html, and then (and only then) I could view it.DES (talk) 03:56, 21 December 2005 (UTC)
That's probably because he's using XHTML with the application/xhtml+xml MIME type, which older browsers do not understand. Using a more recent browser would fix the problem. --cesarb 04:02, 21 December 2005 (UTC)
Yeah, that's it, Internet Explorer doesn't support the reccomended MIME types for XHTML. —Ævar Arnfjörð Bjarmason 04:05, 21 December 2005 (UTC)
Firefox also seems to be happy (not surprisingly). My IE can't get any more current, and so as someone who does web development, I am a little surprised you'd rely on a format not supported by 80+% of the web browsing audience. Dragons flight 04:13, 21 December 2005 (UTC)
He's probably relying on the application/xhtml+xml MIME type because it forces the browser into a strict mode, where even the smallest misclosed tag will cause the whole page to fail to render with an ugly error message. Since it's a test site for his changes to the MediaWiki codebase, being more strict is useful. --cesarb 13:03, 21 December 2005 (UTC)
Exactly, keeps the code I write error free in that area. —Ævar Arnfjörð Bjarmason 13:59, 21 December 2005 (UTC)
If the files have an HTML or an XHTML extension IE will open them. I'm not sure if it then obeys the strict mode or not, but I presume that the browsers which do support this MIME type will still support it even if the extension is present. DES (talk) 22:32, 21 December 2005 (UTC)
Browsers do not use the file extension, unless the file comes from the local filesystem (i.e. a file: URI). They always use the MIME type supplied by the HTTP server. There are some obscure situations in which MSIE ignores the MIME type (thus breaking the specification) and tries to guess the file type, causing much grief to webmasters, but this is not one of them. So, what is happening is that, when you saved the page to your local filesystem, it lost its association to the application/xhtml+xml MIME type, and MSIE had to guess using the file extension (which being .html gives the text/html MIME type). --cesarb 14:50, 22 December 2005 (UTC)

Update: I solved the issues I was having with tidy, this might be enabled soon, brion seemed to like the idea. —Ævar Arnfjörð Bjarmason 04:43, 23 December 2005 (UTC)

Another update: Brion did a review of it and we found some issues with it, in particular it would break the static HTML dumps. Also there were issues with using <ref key> (could only use [A-Za-z0-9_] keys) so I went for <ref name=""> as suggested above. Just as I was about to fix all this stuff, commit & deploy it my hard disk broke down, so I won't be able to do it until sometime after Christmas, hopefully. Currently doing a backup over 802.11b hoping it'll keep spinning (all my MW work is backed up though). —Ævar Arnfjörð Bjarmason 20:18, 23 December 2005 (UTC)

Sorry to hear about your hard disk problems, glad to see this is close to getting implemented though. Nice work. =) —Locke Cole 17:10, 24 December 2005 (UTC)

I managed to get it working anyway, it's now installed, merry winter-holiday-thing;) —Ævar Arnfjörð Bjarmason 01:08, 25 December 2005 (UTC)

I don't think it is working properly, the numbering is all wrong... See Sant Mat. I have converted to the new system, but the numbering is not correct. ≈ jossi ≈ t • @ 23:53, 26 December 2005 (UTC)
The numbering may be correct in the refs, but the numbering in the generated reference section do not match the inline citation numbers, specifically when you use <ref name=Myref /> to note an additional citatiion to a ref previously included. This problem is clearly visible in Sant Mat ≈ jossi ≈ t • @ 00:08, 27 December 2005 (UTC)
Well that's because it wasn't designed to do that, i.e. not a bug. However people didn't like the way it was designed so I changed it, the numbers now line up perfectly. —Ævar Arnfjörð Bjarmason 22:49, 29 December 2005 (UTC)

Special:Undelete

What's the story behind Special:Undelete being restricted? --- Charles Stewart 01:48, 31 December 2005 (UTC)

see #Deleted history permissions above. WAvegetarian (talk) (email) (contribs) 02:26, 31 December 2005 (UTC)
Thanks. --- Charles Stewart 03:06, 31 December 2005 (UTC)


New picture uploading bug problem

Hello, I figured out how to upload pics to my site with help from Ambush Commander. Now I have a new problem. When I upload pics to the server and then create a new page/infobox about the band I'm working on and hit submit, the picture won't show up within the info box. The infobox template is also uploaded to my server. Here is the message I get when the page loads: http://static.flickr.com/36/79502535_9bf0ba3a4e_o.jpg I captured a screenshot of the error message I get when I try to put up the picture. Here is the message I get before the page loads:: http://static.flickr.com/36/79502533_0a0c5faf62_o.jpg Any ideas? I only have GD 1 installed on my server and I don't have any wayt to change that at the moment. But I think the error is more than that. Is there a way to turn off imagecreatetruecolor ? Any ideas on why this shows up? Thanks

Sniggity 21:54, 30 December 2005 (UTC)

All your problems will be solved if you get your host to upgrade the software. Sorry. MediaWiki needs GD in order to resize images, and GD 2.0 added support for true color images. Unless you're willing to rewrite all image related code in MediaWiki to be compatible with GD, you must upgrade. — Ambush Commander(Talk) 04:20, 31 December 2005 (UTC)

Right on, I just wasn't sure. I think I may just switch hosting, because godaddy is inexpensive, but very unfriendly, not up to date and unhelpful. Thanks, Ambush Commander. Sniggity 05:04, 31 December 2005 (UTC)

Severe bug in Windows

http://www.washingtonpost.com/wp-dyn/content/article/2005/12/29/AR2005122901456.html

A previously unknown flaw in Microsoft Corp.'s Windows operating system is leaving computer users vulnerable to spyware, viruses and other programs that could overtake their machines and has sent the company scrambling to come up with a fix.

normxxx 18:08, 30 December 2005 (UTC)

Related slashdot story [7]
The temporary workaround would be to unregister the affected dll, like this:
regsvr32 /u shimgvw.dll
And then, once the dust has cleared, reregister it like this:
regsvr32 shimgvw.dll
— Ambush Commander(Talk) 18:23, 30 December 2005 (UTC)
Related article: Windows Metafile. --cesarb 20:53, 30 December 2005 (UTC)

Right-to-left edit summary bug

File:RTL bug.png

I'm not sure if this bug has been reported to bugzilla yet. It appears that the reason for the bug is that one of the edit summaries is in a right-to-left language. The next entry on the watchlist is right-to-left for the bullet, time, and first word of the link. The rest is left-to-right. — 0918BRIAN • 2005-12-30 16:39

Red links

I don't know where to ask this at, but what do the red links mean? — Preceding unsigned comment added by TenPoundHammer (talk • contribs)

Red links are for pages that don't exist. They lead to an edit box, just like you get when you edit any other page, except it's blank.--Sean|Black 03:21, 30 December 2005 (UTC)
more to the point, red links are for pages that don't exist yet. They're just waiting for input from someone who knows what to put there. Grutness...wha? 03:25, 31 December 2005 (UTC)

What happened to Create new page in searches?

I'm sure there used to be a set of options that came with a search that didn't find exactly what was typed - in particular the option to create a new page. Why has this disappeared (or why can't I still see it...). -- SGBailey 21:11, 29 December 2005 (UTC)

Do you see this if you enter something in the search box and click "Go", rather than "Search" (I do). -- Rick Block (talk) 23:10, 29 December 2005 (UTC)
"Search" doesn't offer a create new page option. As you suggest, "Go" does. There is also some other variant which I occasionally stumble across where there are half a dozen options listed - but I've no idea how I get there. -- SGBailey 23:09, 30 December 2005 (UTC)

MediaWiki 1.6 Release Date?

Is there any estimated release date for the first beta release of MediaWiki 1.6? Thanks, Deyyaz 20:35, 29 December 2005 (UTC)

If you want to use the same cutting edge software that runs Wikipedia, you can always check out the code from the CVS repository... — Ambush Commander(Talk) 04:23, 31 December 2005 (UTC)
Sometime Januaryish. --Brion 07:20, 31 December 2005 (UTC)

Confusion about renamed page and double redirects

Hi. A few minutes ago I renamed the List of SIRIUS Satellite Radio stations page to List of Sirius Satellite Radio stations, per the Manual of style's guidance to "Follow standard English text formatting and capitalization rules even if the trademark owner encourages special treatment." But just before doing that, I first created the List of Sirius Satellite Radio stations page as a pure redirect to the original "SIRIUS" page. (Why I did this, I'm not entirely sure.) All is well, except that after I updated all the "What Links Here" pages to link to the newly-renamed page, I've got this one redirect I can't fix on the old "what links here" page (which also shows up in the "what links here" list on the new page as well). It looks like a double redirect, but it doesn't seem to have any of the problems that a double redirect usually causes. Is this something I need to worry about? How can I fix it? --Aaron 00:00, 1 January 2006 (UTC)

It is a problem and you simply need to edit it to point to the correct target. The redirect in question is List_of_Sirius_Satellite_Radio_channels (note that it's "channels", not "stations"). Click this link and you end up at List of SIRIUS Satellite Radio stations which is redirect (not the ultimate target article). If you click the link following "redirected from ....", you'll be at the redirect you need to fix. -- Rick Block (talk) 01:36, 1 January 2006 (UTC)
*slaps self on forehead* That's what it was; I didn't notice that one said "stations" and the other said "channels". I've fixed it. Thank you Rick ... have a happy new year! --Aaron 02:51, 1 January 2006 (UTC)


Category:Cleanup by month

Any idea why Category:Cleanup by month is no longer showing the categories for cleanup in each month that should be under it? 68.109.18.226 02:22, 2 January 2006 (UTC)

What do you mean? -- Tim Starling 02:30, 2 January 2006 (UTC)
OK, Now it's back (and something logged me out and I didn't notice it). Might be part of categories not working above. (which I couldn't find when I looked at first because it was general categories instead of this category). Sorry about that. Apparently its been fixed. RJFJR 02:39, 2 January 2006 (UTC)


Template links

Four bug reports above were due to this. A new table has been introduced into the schema to track template links. Previously we just assumed that a link to the template namespace was a template inclusion. Now this assumption is gone and you can put templates in any namespace without subtle problems like them being missing from the list of used templates on the edit page. It'll take a while for the updates to finish though, in the meantime we'll have a mixture of the old way to identify template inclusions and the new way. -- Tim Starling 02:20, 2 January 2006 (UTC)

Preview error

Erroring out when hitting preview with message: Fatal error: Call to undefined function: openshowcategory() in /usr/local/apache/common-local/php-1.5/includes/EditPage.php on line 975 Perhaps related to category problems listed above? xaosflux Talk/CVU 21:04, 1 January 2006 (UTC)

This error appears to be resolved. xaosflux Talk/CVU 02:18, 2 January 2006 (UTC)

Where did the category listings of articles go?

I have seen several category pages, and they are NOT showing anything except any text related to the category that is already there. It is not a good thing, as leaving it alone will lead to CSD backlogs. What's going on? --WCQuidditch ☎ ✎ 20:47, 1 January 2006 (UTC)

Already asked, it looks like. Can't be good... --WCQuidditch ☎ ✎ 20:48, 1 January 2006 (UTC)

Question of Grammatical Technicailities

this is a critical (and possibly controversial) question!!!!!

Technically, other than being a free, open-content, online encyclopedia, isn't Wikipedia a general encyclopedia?

My school's MLA guide says that I'm not allowed to have general encyclopedias as sources.

Help put an end to this madness! --anonymous

Wikipedia is a general encyclopedia and your school is right you shouldn't be using general encyclopedias as sources for anything important. Wikipedia is great as a first step in learning from nothing but if you wan't you information to be solid and in depth then you should go elsewhere. (which is why wikipedia articles are supposed to cite sources although many unfortunately don't). Plugwash 20:47, 1 January 2006 (UTC)
Wikipedia is great as a first step in learning from nothing: yes, but it's also a great first step in researching. Using Wikipedia as a first step for research is fully in line with Wikipedia's being an encyclopedia. [[Sam Korn]] 21:07, 1 January 2006 (UTC)

ok. BUMMER! i wanted to use it! oh well. thanks anyways! --anonymous

You ought to read Wikipedia:Researching with Wikipedia. — Ambush Commander(Talk) 22:21, 1 January 2006 (UTC)

i just wanted to know because I have this project that I desperately needed to work on. So, thank you. I hope you have an awesome new year! --anonymous

Repeated error message

Today, every time I go to a new page, I get "Fatal error: Unknown function: cefined() in /usr/local/apache/common-local/php-1.5/extensions/BoardVote/BoardVote.php on line 6". If I refresh, it goes away. User:Zoe|(talk) 20:16, 1 January 2006 (UTC)

Me also. Martin 20:18, 1 January 2006 (UTC)
It seems to have gone away. User:Zoe|(talk) 01:38, 2 January 2006 (UTC)

categories not working

Am I the only one For which the ategorys won,t work in the Main namespace??? Circeus 19:45, 1 January 2006 (UTC)

Nope. Me, too. Sparkit 19:51, 1 January 2006 (UTC)
They are indeed a little bit broken. Trying to create one also throws up this error: "Fatal error: Call to undefined function: openshowcategory() in /usr/local/apache/common-local/php-1.5/includes/EditPage.php on line 975". --TheParanoidOne 19:56, 1 January 2006 (UTC)
A little??? Circeus 20:03, 1 January 2006 (UTC)
Me too. Seems to be a widespread issue. Lbbzman 20:27, 1 January 2006 (UTC)

Now we have the categories, but no listings...Circeus 20:58, 1 January 2006 (UTC)

And now the listings are back! --WCQuidditch ☎ ✎ 22:36, 1 January 2006 (UTC)

Clarification

Hi. Something just happened while I was editing an article, and it got me somewhat confused. in the article Saint Silvester Marathon (edit | talk | history | protect | delete | links | watch | logs | views), while I was adding the results for the current year, an anon edited a section of the article. I was editing the whole article, as I was reverting the edit from another anon and taking the opportunity to add the latest result. The second anon added the same result. In the history, his edit shows as happening one minute sooner than mine (which was most likely a matter of seconds, really). Well, the reverted part of the article was in a completely different section, but since I was also inputing data in the same place where the other editor was working, shouldn't I have gotten an edit conflict?? Especifically, I was editing an "outdated version of the article", since I clicked the oldedit link from the history page in order to revert the other anon's edit. So, since I saved that outdated version with my own modifications, shouldn't the edit from the second anon have been overwritten once I saved my edit?? What actually happened was that both edits went through, and the specific results board ended up with a double entry, which I then corrected by editing the article again. I'd like to know how this works exactly... Thanks, Redux 20:34, 31 December 2005 (UTC)

Oh, and it's important to mention that, when I started my editing, the second anon had not yet done anything (as I said, I went to the history page in order to revert another anon's edit). So, in all likelihood, he started his editing and saved it while I was still editing it — so, naturally he would not have had any problems, since his edit went through undisturbed by mine, but since someone else edited while I was editing, maybe I should have gotten an edit conflict. Maybe the fact that I was editing the whole article and the anon edited only a section played some part in this? Redux 20:39, 31 December 2005 (UTC)
If you edit an older version of an article, the edit conflict rules are different. I think your edit in that situation is supposed to override any subsequent edits without an edit conflict message being shown (but I think at some point I have seen edit conflicts in this situation, so I might be wrong). Editing an old version of an article is generally done in revert situations, and if you want to clean up the article at the same time you might be better making that a separate edit.-gadfium 02:06, 1 January 2006 (UTC)
Naturally. I just got "overly frugal", since, other than the revert, the new info to be added was really small, and 90% of it consisted of a copy and paste from a previous line in the board. I must have had the edit box opened for 45 seconds, but the anon probably did his edit in 30 seconds. In fact, the only reason why it took me longer was because I was editing the whole article, which is quite long, and that made me spend slightly more time while sliding down the sidebar from the top of the article, where it opens, all the way to the section where the addition was to be made. Redux 17:02, 1 January 2006 (UTC)

Rendering subsection edit buttons besides picture inserts

If you look at http://en.wikipedia.org/wiki/Photosynthesis#Z_scheme, you'll find that the "edit" button for that section overlays the main text. - Samsara 18:41, 31 December 2005 (UTC)

This is using Firefox 1.0.7 on Windows, IE seems to work fine on this occasion. - Samsara 04:28, 1 January 2006 (UTC)
It's the other way around on Safari 2.0.2 running on Mac OS X 10.4.3 -- main text overlays edit button. --WCQuidditch ☎ ✎ 20:52, 1 January 2006 (UTC)

Image:DEFCON_Nothing.png

May an admin delete this image as now in commons??

thanks,

de:User:Klever

Done -- Chris 73 | Talk 10:23, 2 January 2006 (UTC)

Section saves

I was recently reminded to use Preview on the grounds that a Save is more costly in resources, which makes me ask "Isn't it less costly to do a Save page on an article section, than on an entire article?". My mental model is that Rendering the article is done probably on the Apache servers only, and that Save hits the Database servers as well as the Apache servers. --Ancheta Wis 13:44, 2 January 2006 (UTC)

Previews are cheaper than saves (Apache only, no database writes), a single save is cheaper than multiple saves (one database hit and cache invalidation rather than several). I'm not entirely sure, but I'd suspect section save rather than entire article save is roughly the same (the whole article must be saved and regenerated in either case). On a slow link, editing large articles in their entirety might take more clock time than multiple section edits and if you're editing an often edited article you're more likely to encounter an edit conflict, so there certainly may be reasons to edit by section. However, from the Wikipedia server perspective, it's almost certainly cheaper to make a single "whole article" edit than multiple "by section" edits. -- Rick Block (talk) 17:13, 2 January 2006 (UTC)
The real problem with not using preview is that it clutters up page history, not that it places more of a load on the servers. -- Cyrius|✎ 00:21, 3 January 2006 (UTC)
You can also put less load on the servers by writing really good articles and citing your sources. -- Tim Starling 05:18, 3 January 2006 (UTC)
Wouldn't that cause more server load? If we get better, people will visit us more... :) Titoxd(?!? - help us) 05:19, 3 January 2006 (UTC)


Any easier instructions on how to use importdump.php?

I'm pretty lost with this issue. I also looked into bzip2 and mwdumper and for a beginner these programs are not user friendly at all, nor are there any step by step instructions. That is nobody's problem but my own, granted. And these are wonderful programs, but that still doesn't help me learn from a beginner's standpoint, which is what I want to do. So I want to use my importdump.php. I'm trying to transfer the most recent wikipedia pages_articles.xml.bz2 to my site. On the database download site it says [[8]] you might run the following command: gzip -dc dump.xml.gz | php maintenance/importDump.php and I don't know what this means. How do I do this? I opened the importdump.php file in my notebook, and have the .xml file on my desktop. now what? Thanks for the help. Sniggity 06:01, 2 January 2006 (UTC)

  • I'm not familiar with the program, but I may have some general knowledge that will help, so here goes: The command looks to be intended for a *nix. Basically, it says to decompress dump.xml.gz and feed the output into importDump.php. importDump.php is a PHP program, and since PHP is not a compiled language, it needs a separate program to run it, hence "php maintenance/importDump.php". You can use anything you like to decompress the file, personally I use 7-Zip in Windows, it's reasonably user-friendly. In any case, you'll have dump.xml decompressed afterwards. Feeding that into importDump.php could be a bit harder. On a *nix, piping (such as the "|" above) makes it easy to pass files and data around from the command line, in Windows there's probably an argument you can pass to PHP to make it pull data from a file. You will need to have PHP on the server. Hope that helps! -FunnyMan 19:48, 2 January 2006 (UTC)
    • On Windows, it probably would be "php maintenance/importDump.php < dump.xml" (after uncompressing dump.xml.gz). You can use "gzip -dc dump.xml.gz | php maintenance/importDump.php" on Windows (if you have gzip), but it would gain nothing, since (AFAIK) it would first uncompress into a temporary file and then read from the temporary file. --cesarb 20:09, 2 January 2006 (UTC)
  • The instructions at m:Data dumps seem more up-to-date and useful. --cesarb 20:14, 2 January 2006 (UTC)

partial protection implementation

I run a website using MediaWiki (Wikireason.org), and I just installed the lastest software update (1.5.4, released 21 Dec 2005) hoping that it would include the code for the new Semi-protection policy described in the announcement of 22 Dec 2005[9]. However, nothing seems any different when I click on the "protect" tab at the top of a page. I'm not an admin on Wikipedia, so I can't compare the behavior of Wikireason to that of Wikipedia. I don't know if I'm doing something wrong as a user, as a site-admin, or if I misunderstood the nature of the software upgrade. I found no mention of the semi-protection policy in the release notes for 1.5.4. Any clarification will be greatly appreciated. AdamRetchless 04:03, 2 January 2006 (UTC)

It's in 1.6. Only bugfixes are going into the 1.5 branch at the moment, not new features. So you'll have to wait until 1.6 is released. -- Tim Starling 04:44, 2 January 2006 (UTC)
Thanks for the clarification. AdamRetchless 17:49, 2 January 2006 (UTC)

Categories

In my opinion, the two greatest software modifications to Wikipedia were categories and redirects. My concern is that you can't redirect categories. It should work so that if category:Physical property redirects to category:physical quality, all the pages in category:physical property get moved to category physical quality instead. Can this be done? Bensaccount 03:02, 2 January 2006 (UTC)

In cases where this is necessary, soft redirects can be created using template:categoryredirect. The developers are aware of the request. -- Rick Block (talk) 03:42, 2 January 2006 (UTC)
Isn't the "moving" part done by one of the many bots? --cesarb 20:17, 2 January 2006 (UTC)
user:NekoDaemon (a bot) recategorizes articles from categories redirected this way, but currently only categories last edited by an administrator. See discussion at Wikipedia talk:Categories for deletion#NekoDaemon upgraded. -- Rick Block (talk) 21:08, 2 January 2006 (UTC)

Grabbing the backbutton redux

See Wikipedia:Village_pump_(technical)#.22grabbing_the_back_button.22 above. The person who was compalining has responded:

XP with service pack 2. Internet explorer 6.0
I went to Goggle and looked up the sled dog statue in Nome Alaska. I couldn't use the back arrow to get back to Goggle. so I closed the browser window and tried again. I didn't remember "Wikipedia" so I stumbled into the same site again.
I'll leave it to a technical guru to try to explain this. User:Zoe|(talk) 02:12, 2 January 2006 (UTC)
This is actually probably on Google's end. Occasionally I've noticed them piping outbound search links through a redirect page on the Google side, presumably to sample which links people go to for a given query. This person likely hit 'back' and landed on the redirect page, which sent them right back to us. There's nothing we can do about this problem, if it is the case. -- Cyrius|✎ 00:27, 3 January 2006 (UTC)

Underlined links?

All of a sudden all of the links in articles are underlined... previously they would only be underlined when the mouse hovered over that particular link. monobook.css, which is the skin I'm using, hasn't changed. I'm pretty sure I haven't made any changes on my side either... any ideas? ~MDD4696 00:22, 2 January 2006 (UTC)

It's a preference: the CSS that creates the change is generated here. Try refreshing your cache amd check the preference for link underlining. — Ambush Commander(Talk) 00:26, 2 January 2006 (UTC)
If you are using Firefox, Shift-Reload almost always works. --cesarb 20:18, 2 January 2006 (UTC)

Interesting week calculation byproduct

According to {{CURRENTWEEK}}, we're still in week number 52, but {{CURRENTYEAR}} now shows 2006. Where this comes out to be a problem is where a page uses these parameters to create links to things like [[Portal:Trains/Featured article/Week {{CURRENTWEEK}}, {{CURRENTYEAR}}]]. Is there a parameter that we can use that performs the full calculation showing both week number and year? Slambo (Speak) 17:41, 1 January 2006 (UTC)

At worst in this case is that the "Selected article" and "Selected image" for the week aren't displayed until tomorrow, but it'd still be nice to have a real solution. Thanks. Slambo (Speak) 17:50, 1 January 2006 (UTC)
There's no "solution", it's by design. The week number is associated with the year in which Thursday occurs (see Week#Facts and figures and ISO 8601#Week dates). --cesarb 20:22, 2 January 2006 (UTC)

Archiving Current events page

I tried to follow the instructions in Wikipedia:How_to_archive_Current_Events, but failed because (a) December 2005 already exists, so I couldn't do a move to it and (b) Current events has its "move this page" link suppressed. Does this mean that only an administrator can do the archiving? -- SGBailey 08:54, 1 January 2006 (UTC)

Not that I want to do it, but now that Current events has been moved to December 2005, I observe that December 2005 has no Move this page option, whilst Current events does. How, what, why? -- 09:07, 2 January 2006 (UTC)
Probably because the page protection follows the page when it's moved. --cesarb 20:27, 2 January 2006 (UTC)

Page View Counter

How can I look up how many page views an article has had? Thankyou. --Rachel Cakes 04:25, 1 January 2006 (UTC)

I believe that feature has been disabled by developer due to high cost query. borgx (talk) 05:04, 1 January 2006 (UTC)
Yup. Please see Wikipedia:Technical FAQ#Can I add a page hit counter to a Wikipedia page?. -- Rick Block (talk) 05:28, 1 January 2006 (UTC)
Uh, I am pretty sure I saw a website where you could look up a page and see how many visits it has had? Does anyone know the URL? --Rachel Cakes 04:43, 3 January 2006 (UTC
Maybe Alexa? http://www.alexa.com/data/details/traffic_details?q=&url=wikipedia.org --AMorris (talk)●(contribs) 09:48, 3 January 2006 (UTC)

User subpages?

Is there a way to keep track of all my user subpages? --Revolución (talk) 18:33, 3 January 2006 (UTC)

You can see them all from Special:Allpages (enter your user name). Is this what you're looking for? -- Rick Block (talk) 19:42, 3 January 2006 (UTC)
Yes, thanks. --Revolución (talk) 21:20, 3 January 2006 (UTC)


WMF safety of MediaWiki

I know it's really Microsoft's responsibilty to fix the windows metafile problem, but as one of the most popular places on the internet, is Wikipedia safe from people uploading WMFs?

If, for instance, someone renames a malicious WMF as a JPG, will wikipedia try to display it?

What about uploading a WMF, and using [[image:evil.wmf]] to infect a computer?

Thanks for info from anyone, happy new year! :-D --Tristanb 04:49, 2 January 2006 (UTC)

Not that I could ever condone such evil and probably illegal behavior, but... cool! :-) Windows sucks, you knew it, you paid Bill anyway, and now you get infected. It's not as if you don't reinstall all the time anyway though, so this isn't all that big of a deal actually. Next time you reinstall, give Fedora Core a shot. 24.110.60.225 05:04, 2 January 2006 (UTC)
Well, most people actually get Windows preinstalled on the computer they buy, and don't really understand the concept of an operating system. Many don't actually reinstall their software, they buy a new computer because their old one is too slow (bogged down in spyware).
But no-one deserves to be infected just by looking at a file!
Anyway, my point is whether Wikipedia is vulnerable as a disseminator of malicious files. Tristanb 05:17, 2 January 2006 (UTC)

This is briefly mentioned in the Wikinews article. MediaWiki running on Unix does not allow the upload of .wmf files. It also preforms some basic sanity checks of some file formats, including .jpg. To exploit users of wikipedia, you would need to experiment with all the various file formats wikipedia accepts to find one where MediaWiki could be tricked into miss-recognizing a .wmf file as a true image file. It is not impossible that MediaWiki is vulnerable, but its not trivial to find the vulnerability. It is possible that MediaWiki running udner Microsoft Windows is itself moree vulnerable, if system calls are used to identify file type. JeffBurdges 20:03, 2 January 2006 (UTC)

Here is the list of file types supported at commons, seems fairly restrictive.
According to a recent edit on the above linked Wikinews article, at least one format supported by MediaWiki is vulnerable. -- KTC 21:27, 2 January 2006 (UTC)
Only if no sanity check is done (if you rename a .wmf to .jpg, it probably will not be a valid JFIF file, which a sanity check could catch). The question being asked here is, "just how strong is MediaWiki's sanity checker?".
Alternatively, one might ask how corrupt a file can be before Microsoft stops recognizing it as a .wmf file. If Microsoft stays fairly close to the spec when detecting wmf files, it should be trivial to add code that specifically zaps the header for wmf files. — Ambush Commander(Talk) 22:05, 2 January 2006 (UTC)
Please read linked to Wikinews article (more specificly, its talk page). Mediwiki does do sanity check, but it is still vulnerable. As far as the tester commented / discovered, none of the natively displayed format is vulnerable (due to the sanity checks), but a non-natively displayed format that's accepted by MediWiki is. -- KTC 22:14, 2 January 2006 (UTC)
Dirrect linky for the lazy. Wikinews:Talk:Microsoft_Windows_metafiles_are_a_vector_for_computer_viruses#MediaWiki.27s_vulnerability

Bawolff 22:36, 2 January 2006 (UTC)

This was fixed on Wikimedia websites a few hours ago, but the fix hasn't been released yet for use in other installations. -- Tim Starling 05:08, 3 January 2006 (UTC)

A patch, however, is available here: http://mail.wikipedia.org/pipermail/mediawiki-cvs/2006-January/013086.html . There is not fix for 1.4, so you should upgrade. — Ambush Commander(Talk) 20:48, 3 January 2006 (UTC)

Ampersands in wikilinks

As part of the move towards a stable version of articles, we would like to create wikilinks to particular versions of the article. (e.g. Article/Stable would be a redirect to a particular article version.

Is there a way to use the Permanent link from the link on the left side toolbox to do this? I.e. #REDIRECT permanent link from the /Stable ... /Stable would this be a redirect to permanent link?

I tried this with Common Unix Printing System/Stable as #REDIRECT Common_Unix_Printing_System&oldid=32844867 ... but that doesn't work, it gives this address: http://en.wikipedia.org/w/index.php?title=Common_Unix_Printing_System%26oldid%3D32844867&action=edit

which implies the article isn't there, but replace %26 with & in the address bar (and delete the "&action=edit) and it works, giving what I want: http://en.wikipedia.org/w/index.php?title=Common_Unix_Printing_System&oldid%3D32844867

i.e. there is a problem with the ampersand in the wikilink or in REDIRECT.

Alternatively, I could redirect to a link rather than a wikilink, but that won't take me directly where I want to go, just give me a redirect page (e.g. Common Unix Printing System/Stable2 )

Thanks in advance dml 02:36, 5 January 2006 (UTC)

As far as I know, you have to specify the full url to get a link to a specific revision. Lupin|talk|popups 03:22, 5 January 2006 (UTC)


Search features in WP

People should strongly appreciate to find something when using the search button.

It is bad advertising to us to get nothing. I would suggest, at least :

  • a redirect to Wiktionary,
  • a possibility of searching Wikipedia in other languages (using latin alphabet),

when a search returns nought. This feature is more acutely needed in other languages, where many internationally spread words exist without having a link.


    • Also, could the search results return more lines of text (see Google for example) ?

-- Harvestman Post a comment 21:51, 4 January 2006 (UTC)

onsubst

posted bug 4484 for an enhancement for subst. →AzaToth 20:12, 4 January 2006 (UTC)

Access to copy of deleted image?

  was deleted on 03 January. It was being used with permission, but I don't have a copy of the image (which was scanned from an old photo at my parents' rural property) so I can't re-upload it with the appropriate tags. Could someone please place a copy somewhere I can download it from? — JEREMY 19:29, 4 January 2006 (UTC)

I don't think that is possible. I believe that images are stored in a file system, not in the database. When they are deleted, they are gone permanently. --GraemeL (talk) 19:33, 4 January 2006 (UTC)
Yes. Unlike articles, admins don't have access to deleted images, and they can't be undeleted. -- Finlay McWalter | Talk 19:33, 4 January 2006 (UTC)
See Wikipedia:Lost images. answers.com usually still has recently deleted images. -- Rick Block (talk) 20:14, 4 January 2006 (UTC)

Rollback text changed

I recently noticed that the text for the rollback button has been changed to "rollback vandalism". This is an unfortunate change. There is no prohibition against using rollback for non-vandalism reverts described in Wikipedia:Revert, which is the only WP-space page I could quickly find that describes the button. (Incidentally, the status of that page is not noted – Guideline? Policy? Neither?) If I revert erroneous changes I make myself, I'm certainly not reverting vandalism, and there have been times when I used rollback to quickly and efficiently revert large numbers of good-faith but bad-result changes (with a polite note on the rollbacked user's talk page first, of course). IMO, it should be changed back. android79 22:34, 3 January 2006 (UTC)

(Sort of) agreed. The motivation for this interpretation of the rollback button is here, but I don't really agree with it. — Omegatron 22:36, 3 January 2006 (UTC)
It's an excellent change. It should remind admins that the tool may not be used in an edit war or other non-vandalism points. In the circumstance you mention, you can still do the same thing, as the output edit summary is the same. [[Sam Korn]] 22:41, 3 January 2006 (UTC)
Anyone with access to the button ought to know better and not use it in an edit war. If they don't, they ought not be an admin. Changing the button text will give the impression to new admins (and others unfamiliar with rollback) that its only use is reverting vandalism, which is not the case. android79 22:46, 3 January 2006 (UTC)
If they don't know how to use it properly, they shouldn't. Better to be conservative. [[Sam Korn]] 22:51, 3 January 2006 (UTC)
I see no change in the message since November. --cesarb 22:42, 3 January 2006 (UTC)
See MediaWiki:Rollbacklink. android79 22:43, 3 January 2006 (UTC)
  • I've undone this change. A proposal was made a month ago to restrict rollbacking to vandalism only (or rather, to change the "reverted edits by foo to last version by bar" message to "reverted obvious vandalism by foo..." on the incorrect assumption that rollbacking was already restricted to vandalism only). This proposal was not approved by consensus. If the result AND the summary can be the same (as Sam suggests) it is irrelevant whether it was done manually, by Godmode Lite, or with the rollback button. Radiant_>|< 22:44, 3 January 2006 (UTC)
It has always been the case that rollback has been for vandalism only. User tests are counted as vandalism for this. Reminding admins of this is a good thing. There is no difference in the edit-summary output due to this change, so it should stay. [[Sam Korn]] 22:51, 3 January 2006 (UTC)
Do you have a policy page to back that statement up? The status of Wikipedia:Revert is unclear, but it implies that non-vandalism rollbacks are okay in rare circumstances. android79 22:53, 3 January 2006 (UTC)
  • Good point, time to clarify it once more. However I wish to claim precedent on this one, because checking usage shows the obvious fact that the button is used in cases that aren't vandalism, and this usage isn't usually contested. Per m:instruction creep, we don't need strict rules telling us when to rollback and when to manually do the exact same. Radiant_>|< 22:58, 3 January 2006 (UTC)
Per Wikipedia:Revert - "(Rollback's) intent is solely to be a timesaving shortcut for reverting mass vandalism... No one should ever be in an edit war, sysops in particular should be aware that that's not cool, so there's no need to think about whether or not 'rollback' should be used in an edit war. It shouldn't, because we shouldn't be in that position in the first place.". -- Netoholic @ 22:47, 3 January 2006 (UTC)
Also per Wikipedia:Revert: If you insist on using the rollback feature for non-vandalism edits, be sure to explain on the talk page of the user whose edit(s) you reverted. Which is just what I described above. android79 22:48, 3 January 2006 (UTC)
So are you saying that because several people misuse it, or don't realize it's intended purpose, that we should ignore why it really exists, directly from the words of the developer that put it in? Changing the rollback text is exactly what is needed so people can realize what it's for. -- Netoholic @ 23:12, 3 January 2006 (UTC)
It's not clear that the use I describe is actually a misuse. It doesn't matter what its intended purpose was; it is useful for things other than mass reversion of vandalism. android79 23:16, 3 January 2006 (UTC)
  • It absolutely should never be used in a revert war. But there are circumstances in between a revert war and vandalism where rollbacking is appropriate. For instance if a newbie in honest mistake adds several dozen articles to the wrong cat. Radiant_>|< 22:49, 3 January 2006 (UTC)

I'm the person who was so bold to make the change a few minutes ago, which was thankfully not reverted using the rollback tool ;-). First, as has been pointed out, changing the link text does not change the text that is inserted into the edit summary. The earlier proposal Radiant refers to above was a change of the edit summary text, which makes a big difference. If we change the link label to "rollback vandalism", and we make a very short list of exceptions to this rule, then the text "vandalism" in the UI doesn't really matter except that it scares away new admins from using rollback when they haven't studied the policy. I think it could be argued that this is actually a good thing.

For this reason, I would be in favor of changing the text back to "rollback vandalism", but I'll let other people work out whether we want to do this. The reason is that the label change is basically a hack to cover ~95% of use cases and discourage misuse -- and I don't like hacks. Since we seem to essentially agree that there are a small number of cases where we want to allow the use of rollback, it would make sense to have a single-click interface that allows the selection of these reasons (e.g. "vandalism", "own mistake", "multi-page mistake, see user talk"), perhaps via radiobuttons and a simple JavaScript to submit when a radiobutton is selected. I'll jot it down in my loooong list of things to hack on.--Eloquence* 23:20, 3 January 2006 (UTC)

  • The point is really that as indicated in the ANI discussion, rollbacking isn't limited to vandalism - and if we write somewhere that it is, then that's just more food for rules lawyers. I don't like hacks either. I'm afraid that requiring extra clicks for a rollback kind of defies the point of rollbacking. It should only be used in those circumstances where an explanation isn't necessary. That generally, but not always, means vandalism. Radiant_>|< 23:25, 3 January 2006 (UTC)
I can be fairly rules-lawyerish myself, but having a link label that says "rollback vandalism" does not, in my opinion, contradict having a policy that explains that it may be OK to use that link in a small number of other cases if the admin is wary of the message it may send (particularly to non-newbies). As for alternatives, notice that I said "single-click interface" ;-). The only difference is that you would have to position your mouse to a different location on your screen before clicking, depending on the reason for the rollback. And it could have the added advantage of inserting a better summary, based on the reason you chose (we should still avoid the v-word in the summary, though).--Eloquence* 23:29, 3 January 2006 (UTC)
As I suggested in October: add a text field (that defaults to a blank text), in the case we want to specify a reason why we reverted. That way, those who want to abuse the rollback feature by taking out PENISPENISPENIS from an article can keep doing it the way they've always done... Titoxd(?!? - help us) 23:30, 3 January 2006 (UTC)
The effect of a text field would probably be the opposite of the desired -- the amount of "machinelike" summaries complemented by short texts would increase, reinforcing the impression of a distinction between sysops and regular users. Given the philosophy of sysops being housekeepers and not moderators or editors, I do like the notion that a sysop has to go through the same processes as everyone else in order to deal with editorial issues.--Eloquence* 23:35, 3 January 2006 (UTC)
Having a link label that says rollback vandalism does contradict what the button actually does. It does a rollback, and whether it actually rolls back vandalism depends on the situation. I'd like my UI to be clearly labeled. If this spurs development of some new rollback features, hey, great. :-) android79 23:33, 3 January 2006 (UTC)
Empirically speaking, I wonder if it is a contradiction in more than 5% of cases. And if it isn't, one could argue that even in those cases, it is a beneficial contradiction (deterrent).--Eloquence* 23:36, 3 January 2006 (UTC)
I should tell that to the human factors guy at work. "Hey, design some cognitive dissonance into your next UI!" ;-) In all seriousness, though, you're probably right; the amount of cases is probably under 1%. But IMO the solution is not to obfuscate the UI but to educate the users. android79 23:39, 3 January 2006 (UTC)
I think Microsoft did that the moment they made you go through the "Start" button to shut down. ;-) In all seriousness, ambiguity and labeling via predominant meanings are very, very common in user interfaces, good and bad ones.--Eloquence* 23:47, 3 January 2006 (UTC)
Well, a good rollback feature would be a "revert to last good version" link, combined with Tim's last good edit thingie... Titoxd(?!? - help us) 23:38, 3 January 2006 (UTC)
  • Occam's razor, really. Given the choice between a somewhat more complicated text that is correct 98% of the time, and a simpler text that is correct 100% of the time, I prefer the latter. Radiant_>|< 23:42, 3 January 2006 (UTC)
    • I would counter by saying that the shorter text is more complex because it leaves things open to interpretation. A simpler solution to admin misusing this ability would be to change the text. -- Netoholic @ 03:40, 5 January 2006 (UTC)

Certainly, it isn't only used for vandalism. For eg: "FeloniousMonk (Talk | contribs) Reverted edits by FeloniousMonk (talk) to last version ..." [10] Give people a tool, and they'll use it. Regards, Ben Aveling 07:31, 5 January 2006 (UTC)

Kate's tools down?

Are Kate's tools down, or has the address been changed? Grutness...wha? 03:23, 31 December 2005 (UTC)

Try http://tools.wikimedia.de/~kate/pages/projects — Ambush Commander(Talk) 04:21, 31 December 2005 (UTC)
Which isn't working as of now. If you're looking for edit counts, try the Kate wannabe counter (sorry, I don't have a link). — Ambush Commander(Talk) 04:22, 31 December 2005 (UTC)
The Kate wannabe edit counter (Interiot's Tool) is blocked too... but there's still Flcelloguy's Tool, but you have to compile it yourself. Titoxd(?!? - help us) 04:25, 31 December 2005 (UTC)
Likely too much trouble for most users... ah well, you can always do it the old way: counting by hand... I wonder why they're down... — Ambush Commander(Talk) 04:35, 31 December 2005 (UTC)
Has this tool been taken down purposefuly, due to it's integration with deleted edits? xaosflux Talk/CVU 01:30, 1 January 2006 (UTC)
  • Please note: This is again functional. xaosflux Talk/CVU 04:44, 3 January 2006 (UTC)

Minor edit link under edit box

Hi. Recently, the term minor edit, in the sentence This is a minor edit, right next to the checkbox under the edit box, has been made into a link. That's great and all, but have you guys noticed that the link is right over the Save page button, as well as the Show preview button (two of the most used buttons when we are editing an article), and that it's awfully "on top" of those buttons? Don't you think that creates a risk of misclicking, that is, people accidentally clicking on the link to the project page when they wanted to either save the page or preview their changes? Although usually it would be possible to "back up" to the edit box, some might end up loosing their work because they...miscalculated the move of their mouses.
Has there been any complaints about this? Not that it happened to me, but it got me thinking. A way to prevent any problems would be making the link one of those that forcibly opens in a different window (which it does not at this point). Sorry for the lack of technical terms on this... Regards, Redux 00:42, 6 January 2006 (UTC)

Humm, "minor edit" appears to have been delinked. Redux 03:24, 6 January 2006 (UTC)
The changes being attempted to be put in to that section were not working properly, and have all been backed out. xaosflux Talk/CVU 03:31, 6 January 2006 (UTC)


Plus now allowed in article titles

Plus signs are now allowed in article titles, e.g. C++ -- Tim Starling 05:47, 5 January 2006 (UTC)

See Category:Articles with a plus in their title for articles which can be converted. -- Netoholic @ 06:23, 5 January 2006 (UTC)
So is "++ungood" allowed? However, I believe Orwell never used that spelling. (SEWilco 06:28, 5 January 2006 (UTC))
Yes, wikified above for your convenience -- Tim Starling 07:04, 5 January 2006 (UTC)
Hooray! Morwen - Talk 12:05, 5 January 2006 (UTC)
Sweet. --Golbez 22:21, 5 January 2006 (UTC)

Edit dating

As I was browsing my contributions, I noticed that an edit to LASEK which I distinctly remember making [11] was dated as 19:03, January 3, 2003. Unfortunately, I wasn't around in 2003; I actually made my first edit [12] at Viola at 22:54, January 10, 2005. To compare the LASEK article edits, you can see that I also edited [13] the article at 10:03, March 10, 2005. Both these edits were during a merge I was orchestrating. Does anyone have an idea what happened? Thanks, Bratschetalk | Esperanza 02:55, 5 January 2006 (UTC)

I'd guess it was one of these times where a few of the Apache servers have their clocks out of sync (it has happened before, and causes exactly that kind of effect). You could try asking a developer to see if they can edit the date directly on the database (or even run a script to find and fix all such situations). --cesarb 12:03, 5 January 2006 (UTC)
Not sure; it might be more. See #Contribution list problem above. Lupo 12:12, 5 January 2006 (UTC)
That's a different case. In this case, the user not only remembers doing the edits, he also provides the date it happened, which is after he registered. --cesarb 18:31, 5 January 2006 (UTC)
Trust me: I don't think I even knew what Wikipedia was in 2003. In case you didn't know "Bratsche" means "viola" in German: that was the first thing I was looking for. Saw the article, created an account, and started becoming hooked on Wikipedia. Thanks for the help. Bratschetalk | Esperanza 03:53, 6 January 2006 (UTC)

homopolar + dynamotor

Dear Editor(s) Wikipedia must update its pages devoted to homopolar motor/generator. The truly relativistic physics underlying such engines has been disclosed by the Argentinien physicist Jorge Guala-Valverde et al (Apeiron, 2001; Am. J. Phys. 2002; Royal Swedish Ac. Sc. 2002, Apeiron 2004, Apeiron 2005). By "googlin" for "Guala + Valverde + Physics", you can find more references. In <www.fjp.org.ar> there is enclosed a movie concerning actual experimentation on homopolar motor. With best regards, J.A. Tramaglia.-

Thank you for your suggestion! When you feel an article needs improvement, please feel free to make whatever changes you feel are needed. Wikipedia is a wiki, so anyone can edit any article by simply following the Edit this page link at the top. You don't even need to log in! (Although there are some reasons why you might like to...) The Wikipedia community encourages you to be bold. Don't worry too much about making honest mistakes—they're likely to be found and corrected quickly. If you're not sure how editing works, check out how to edit a page, or use the sandbox to try out your editing skills. New contributors are always welcome. --Golbez 15:10, 3 January 2006 (UTC)
Instead of providing a boilerplate reply, if you think about what the user above is actually saying, the more relevant policy to apply would be Wikipedia:No original research, since we don't cover "breaking" papers etc. Dysprosia 09:55, 5 January 2006 (UTC)
But if it has already been published in peer-reviewed journals, it wouldn't be original research on Wikipedia. At least that is how I understand the WP:NOR policy to work in a case such as this. Tupsharru 10:35, 5 January 2006 (UTC)
We don't cover "just-published" work, as this appears to be. We cover established work. This is longstanding practice. Once this becomes more mainstream we may. Dysprosia 23:25, 5 January 2006 (UTC)

"grabbing the back button"

We have had two people write to the Help Desk mailing list complaining that Wikipedia grabs their browser's back button and doesn't let them leave. I know it's not a function of Wikipedia, could it be some combination of browser and operating system? I've asked the most recent poster to let us know what they're working with. User:Zoe|(talk) 16:22, 29 December 2005 (UTC)

Hm. Do you know where they're coming to the site from? It could be they're directed to us via a mirror that either opens in a new window, or does something stupid with frames... Shimgray | talk | 16:33, 29 December 2005 (UTC)

"grabs their browser's back button"... never heard it worded that way before... — Ambush Commander(Talk) 16:35, 29 December 2005 (UTC)
I'll post more detail when the person reports back. If they do. User:Zoe|(talk) 19:33, 29 December 2005 (UTC)
If a site does a http redirect (is that the correct term?) from page A to page B, and then from page B they hit the back button they will go back to page A (which redirects them back to page B). In IE you can click the little down arrow beside the back button to bring up a recent history and "hop over" the redirecting page. hope this helps. Matt 04:07, 2 January 2006 (UTC)
A real HTTP redirect won't cause this problem. Using javascript or meta refresh tags to emulate one can do so, though. —Ilmari Karonen (talk) 20:49, 5 January 2006 (UTC)

Remove / increase recommended article size limit

Right now WP gives a recommendation to keep article sizes under 32kb every time you edit an article larger than that.

I know it once was a technical issue, but that is no longer so. While it is also true that large articles are better off broken into different subsections with the main article only listing summaries, still, even then 32kb is too little for good articles.

Another former reason reasonably valid 3-4 years ago when WP started was that not everyone has fast internet (many people still had 14,400 baud and 56k was considered good), but internet speed (like other computer-related measurable quantities) has increased exponentially in the last years, and now over half of internet users have cable/dsl for which downloading 100kb is nothing, and 99% of the rest have 56k which can also load a 50-100kb page reasonably fast.

About 90% of featured articles are WAY over 32kb. Except for really specific topics where one can cover it all in a smaller article (such as the recent shoe wax article), an article of 50 to 100kb is pretty much required to have enough material for featured status. Whereas before a good broad article was pretty much a collection of subpages with short summaries, the present widely accepted definition is an article which is strong and sound by itself, which can be informative on its own, and subpages would further expand on the subject for those interested in the particular subtopic, rather than being forced to go there just to understand the material of the main article.

And large article sizes are even more essential now when there is a big initiative to include lots of references and links.

You may say that its still not a big deal to have the reminder, but I'd argue that it may actually discourage editors (especially new editors) to contribute new material to an article which already contains quite a bit, but is still far off from FA.

In light of that, I propose to, either,

- Remove the warning at all, as irrelevant on both technical and editorial grounds, or

- Increase the article size reccomendation from 32kb to at least 100kb. Elvarg 04:19, 15 December 2005 (UTC)

I agree that instead of recommending any particular size, we should instead recommend that articles that grow to unwieldly lengths be broken up, where this term is subjectively determined on a per-article basis. Deco 04:26, 15 December 2005 (UTC)
I was under the impression the limit was first enforced because of a hard limit on the amount of text allowed in a editable text box in certain ld browsers. I think that that reason is now gone. I would support getting rid of the warning, and enforce common sense at the FAC stage. --Martyman-(talk) 09:30, 15 December 2005 (UTC)
Somebody mentioned the other day on the extra-long AfD that this is still an issue on handheld devices. Even there it's probably not a big problem, since you can edit individual sections (until you hit an edit conflict). Zocky 09:49, 15 December 2005 (UTC)

This is not only a problem with handheld devices, but also article style. Long articles tend to be daunting to edit and discourage further expansion. I think the 32k limit for a plain text simple article is already quite large. My suggestion is; create a new "Wikifeature" where long articles can be (like in most encyclopedias) and break up all articles longer than 40k into bits smaller than 20k. The only exception I can think of is the situation where the article is clearly too detailed already. Mozzerati 20:53, 21 December 2005 (UTC)

Are the figures of "over half of internet users have cable/dsl, and 99% of the rest have 56k" valid worldwide, or for a particular country? They sound a bit suspicious to me wrt many developing countries where poor quality telecoms are an ongoing beef. Given the potential value of WP in the Less Connected World, let's go easy on article length where feasible. JackyR 18:01, 28 December 2005 (UTC)

An article of substantial length, when properly organized, should have multiple sections, and thus each section is separately editable. I support removing the length recommendation/warning, or increasing greatly increasing it.--StanZegel (talk) 18:50, 28 December 2005 (UTC)
Knowing article size is still very useful. The current notice does not mandate sizes of 32KB or less, it just mentions that articles that size of readable prose (all non-readable stuff excluded) may start to be longer than necessary to efficiently cover its topic. See Wikipedia:Summary style (which is part of the featured article criteria). --mav 16:22, 1 January 2006 (UTC)

I support the proposal to remove or increase the recommended maximum. --Anonymous, 09:12 UTC, January 2

I think that knowing the article size is helpful, so perhaps the warning should stay. However, it should be reworded to indicate that when using 32kb as a style guide (as opposed to a technical one), "extras" are not included in this limit. It's ludicrous to think that a well referenced article with lots of images should be shorter in prose than one without these things, but that's how some people are interpreting this (on Wikipedia:Featured article candidates at least). — BrianSmithson 13:50, 6 January 2006 (UTC)

Random article

Can we get the random article button to exclude stubs and disambiguation pages? If people press random article, they want to read something. We can always have a seperate link to stubs, or even a random stub, button for people who want to write about a random topic. JeffBurdges 23:00, 7 January 2006 (UTC)

I suggest reading User talk:Rambot/Random page. I was under the impression that "random" = "random", and even random stubs can contain random topics. — Ambush Commander(Talk) 23:35, 7 January 2006 (UTC)

MediaWiki custom interwiki links

I set up two custom wikis at wiki.ddowiki.net/wiki/index.php/$1 and wiki.dndwiki.net/wiki/index.php/$1

I want, when I write a wikilink to DDO:Page at the dndwiki.net wiki for it to link to the "Page" article in the ddowiki.net wiki; and when I write a wikilink to DND:Page in the ddowiki.net wiki for it to link to the "Page" article in the dndwiki.net wiki - how would I set this up? Something to do with custom namespaces or somesuch? -- ℑilver§ℑide 10:13, 6 January 2006 (UTC)

No, there is a MySQL table that contains interwiki links. It should be fairly self-explanatory. Use PHPMyAdmin. — Ambush Commander(Talk) 23:39, 7 January 2006 (UTC)

Deleted pages

Given that deleted pages no longer show their deleted edits to non-admins, would it be possible to include a link to the deletion log entry if there is one? Currently, there's no way to know that a page has ever existed before (or, if it's been recreated, that it was ever deleted). Thanks. Chick Bowen 17:20, 7 January 2006 (UTC)


Problems with new citation style

Well, I'm trying to adopt the new citation style described here, and I really like it. It works great for simple references, and it saves on article size to boot. However, I'm having problems when a single reference is used more than once. The article I'm trying to fix is Dixie (song). As you can see, there are six page citations that repeat at least once. However, when you click on the reference links by these repeaters, nothing happens (the page does not pop up or down properly like it does with the page citations that are used only once). Secondly, Note 17 should read "Quoted in Abel 36.", but currently it says nothing. Likewise, Note 72 should read "Abel 49.", but it too is blank. Can someone tell me if I'm doing something wrong? Thanks . . . . — BrianSmithson 14:01, 6 January 2006 (UTC)

Looks like two things:
  1. The <ref> with the content needs to come first. Working as designed, I think.
  2. ref names with spaces appear to be broken. I think that's a bug.

I fixed up #17, but didn't fix 72. —Bunchofgrapes (talk) 02:49, 7 January 2006 (UTC)

  • I believe I have resolved the issue with the "Abel 49." reference.[14] The issue appears to be that the reference name was not defined at the same time the text of the reference was defined. Adding the name to the first use of the reference (were the text of the reference is also defined) allowed the second use of the reference to find the previously defined text. Please double check that the corrections made actually do what you wanted them to do. --Allen3 talk 14:17, 7 January 2006 (UTC)
    • It seems to be working fine; thanks to both of you. The page is even popping up and down correctly on the repeated references, which I'm assuming was a bug that got fixed. Thanks again! — BrianSmithson 16:54, 7 January 2006 (UTC)

Contribution list problem

Until recently, I was absolutely sure that my first edit on Wikipedia was on December 18, 2003, part of cleaning up some vandalism at Spartan Alphabet. Now it appears the software has miraculously granted me with three even earlier edits I allegedly have done on December 29, 2001: [15], [16], and [17]. I know for sure that I've never touched these articles. Clearly something is amiss, check this alternate diff link for the first edit mentioned before. What gives? Lupo 15:09, 4 January 2006 (UTC)

Were they made by some other person using the username Lupo on the old software in 2001, who never got around to registering their account on the new software in 2002? --Brion 18:00, 5 January 2006 (UTC)
I wouldn't know. The diff link doesn't show any user at all! Lupo 19:38, 5 January 2006 (UTC)
Did you ever edit the sandbox in your early days? I've got similar phantom edits, and can only surmise that whatever edits I made to the sandbox were moved to different titles. For example, a silly edit I think I recall making in the sandbox somehow ended up at Matt Couper, whoever that is. android79 19:45, 5 January 2006 (UTC)
No, I don't think so, although I'm not absolutely sure as it's been so long ago. But as I had previous Wiki experience from elsewhere, I believe I never used the sandbox—as I remember, I jumped right in (and used preview a lot until I got used to MediaWiki syntax). Lupo 07:54, 6 January 2006 (UTC)

I've got a similar phantom entry for my name in 2002 [18]. I have no idea where it comes from, and wasn't always there (i would have noticed when i first joined wikipedia). I think it happened after a software or database upgrade/maintenance. Don't have a clue what causes it. Tristanb 05:15, 7 January 2006 (UTC)

Yes, it must be recent. When the "earliest" link was introduced on the contributions page, I tried it out and these three phantom contribs definitely weren't there. Lupo 15:01, 7 January 2006 (UTC)

Logged in differentiation

Is there any general way to make pages and items appear differently depending on whether one is logged in or not? One application I have in mind is having the donations request box go away for logged in accounts but stay longer for visitors. I know it is possible to do this on an individual basis with personal style sheets, but is there any way to vary things on a global level for all logged in users? Dragons flight 14:05, 8 January 2006 (UTC)


Browser stats

Is there somewhere I can find updated browser stats for Wikipedia? The most recent I have found is Wikipedia:Browsers, dated from October 2004, as is http://wikimedia.org/stats/en.wikipedia.org/agent_200410.html. Thanks. —Kmf164 (talk | contribs) 16:10, 9 January 2006 (UTC)

Color on Links

I want to edit an internal link to change the color from blue to white. How do I do this? I tried <font color="white">[[Link title]]</font>, but the link was still blue.--God of War 18:16, 9 January 2006 (UTC)

Don't do that. It's possible, but confuses the readers. --cesarb 18:44, 9 January 2006 (UTC)
Tell me how, it's for a userbox so it's not like a big deal or anything.--God of War 18:46, 9 January 2006 (UTC)
OK then. You just need to put the <font> tag inside the link, for instance [[Example|<font color="blue">Example</font>]]. --cesarb 19:08, 9 January 2006 (UTC)
The idea might be css color declaration. Take a look at "css tutorial". plenty of pages returned by goodoldgle. --Harvestman 19:54, 9 January 2006 (UTC)
I would prefer CSS. It would look like this: [[Link target|<span style="color:#FFF;">Link name</span>]]. The underline, however, will not change color on Firefox.


==> FWIW, the "E-mail new password" function is very slow. If you use it; take a coffee break! normxxxtalk 19:15, 9 January 2006 (UTC)

E-mail new password

The "E-mail new password" button seems not to send any passwords. Or at least not within minutes. With regards, Ilse (01:23, 9 January 2006 (UTC))

Why are you trying to email yourself a new password? You are logged in and can change it in your preferences. Or is this on behalf of someone else? WAvegetarian (talk) (email) (contribs) 04:19, 9 January 2006 (UTC)
You can't change your password if you don't remember what the old password was. So if you logged in a month ago, checked the "remember me" box and then forgot your password, it's quite possible that you can still be logged in but can't relogin. —Ilmari Karonen (talk) 06:21, 9 January 2006 (UTC)
Good point. I wasn't thinking. WAvegetarian (talk) (email) (contribs) 07:20, 9 January 2006 (UTC)
Isn't that "remember me function" implemented by setting a cookie (client side)? At least I would search for that and kill any related cookies. Adrian Buehlmann 12:21, 9 January 2006 (UTC)
Surely killing the cookies is the last thing you'd want to do in that situation. You could maybe write down the cookie data, just in case it gets lost somehow... —Ilmari Karonen (talk) 13:42, 9 January 2006 (UTC)
AFAIK, you can change your email address without having to use your old password. Just make sure your email address is correct (you could also try using an alternate email address, if you have one). --cesarb 14:58, 9 January 2006 (UTC)

User creation throttle

I've put a temporary user registration rate throttle on the site since there was a registration flood attack recently with a few hundred accounts registered in a few minutes, with various ugly chars etc (see Special:Log/newusers). Most of the flood items I've manually removed, and the network they came from has been blocked.

Between other things I'm working on updating some old experimental captcha code, which will probably be put on registrations later this weekend (possibly regularly, possibly just in response to floods). --Brion 01:53, 7 January 2006 (UTC)

It's funny how apathetic the response is. Sounds good. — Ambush Commander(Talk) 23:11, 9 January 2006 (UTC)

Infobox Philosopher

The template Template:Infobox Philosopher is being used to link the biographies of philosophers on the Wiki. As the template stands, it cannot deal with cases in which we do not have a picture. This could be fixed with a single-point GIF, but is there a More elegant fix? failing that, is there a single-point GIF somewhere on the Wiki we could use? Banno 00:05, 9 January 2006 (UTC)

not easilly because of the way the box is structured. if you made the entire image tag a parameter then you could make it an optional parameter but that would mean editing all the existing articles with the template (lukilly there don't appear to be very many) Plugwash 00:24, 9 January 2006 (UTC)
on second thoughts you could do it by making an optional parameter round the image tag and setting that parameter to a null value on pages you had no image for.

I've added the standard hiddenStructure CSS hack to the template. It should work now. —Ilmari Karonen (talk) 00:37, 9 January 2006 (UTC)

Ripperbeudy. Works like a bought on. Thanks! Banno 00:57, 9 January 2006 (UTC)


France

I'm having some trouble accessing the article France, which I came across being edited by a vandal on RC patrol. I can get to any other article, but not that one. My browser just hangs if I try, and the page refuses to load. Any idea why? Crotalus horridus (TALK • CONTRIBS) 18:17, 10 January 2006 (UTC)

It's not just you. I can't get to the article either. Not sure why. —Kmf164 (talk | contribs) 18:22, 10 January 2006 (UTC)
Works now. —Kmf164 (talk | contribs) 18:30, 10 January 2006 (UTC)
The vandal put in an edit too long for the servers to handle. You have to revert it via history, or for admins by the rollback link on their contributions page. [[Sam Korn]] 18:39, 10 January 2006 (UTC)
Good to know that. Thanks. —Kmf164 (talk | contribs) 20:00, 10 January 2006 (UTC)


The Search Results page is broken

When you do a search and cannot find the results, you get a broken HTML line. I just tried looking for the movie "The Fabulous Dorseys", and got the following:

You searched for "<a href="/wiki/The Fabulous Dorseys">The Fabulous Dorseys</a>" <a href="/wiki/Special:Allpages/The Fabulous Dorseys">[Index]</a> User:Zoe|(talk) 04:45, 10 January 2006 (UTC)

Works fine for me, using the exact search term you used. --TheParanoidOne 12:55, 10 January 2006 (UTC)
It was happening consistantly when I reported it last night, at the same time that Search was disabled, but now seems to be all right. User:Zoe|(talk) 18:46, 10 January 2006 (UTC)

image overlap of text

on Thirty Years' War, the image of Rudolf II, Holy Roman Emperor, overlaps some text in the article. →AzaToth 22:07, 7 January 2006 (UTC)

It seems there's a div.thumb {margin-top: .5ex} missing in monobook. It doesn't show in Opera, though...--Joris Gillis 20:46, 10 January 2006 (UTC)

Bug in Monobook.css

Someone incorrectly commented out this chunka code in monobook.css: 

/#fundraising {
/ text-align: center;
/ border: 1px solid gray;
/ padding: 5px;
/}

I'd just fix it, but the file warns to check here first before editing. (Doesn't anyone run their changes with a javascript console running? I get three errors every page.) --jpgordon∇∆∇∆ 18:36, 11 January 2006 (UTC)

Thanks for reporting the problem. Fixed. You probably will have to clear your cache for it to have any effect. --cesarb 18:49, 11 January 2006 (UTC)
Great. Now, what about these two?
Error: Unknown property 'filter'.  Declaration dropped.
Source File: http://en.wikipedia.org/w/index.php?title=MediaWiki:Common.css&action=raw&ctype=text/css&smaxage=2678400
Line: 243

Error: Error in parsing value for property 'line-height'.  Declaration dropped.
Source File: http://en.wikipedia.org/w/index.php?title=MediaWiki:Common.css&action=raw&ctype=text/css&smaxage=2678400
Line: 234

--jpgordon∇∆∇∆ 19:01, 11 January 2006 (UTC)

The first one is an IE-only feature; its removal was requested at least twice already. The second one I asked the user who added it to fix, because I do not know what's its intended meaning, but I didn't receive any answer. It would probably be better to ask on Mediawiki talk:Common.css before changing them. --cesarb 19:34, 11 January 2006 (UTC)

Since we are chatting about CSS, is there any reason no to remove the AD/BC CE/BCE stuff from Mediawiki:Common.css? I believe it was created as part of a failed proposal. Dragons flight 22:58, 11 January 2006 (UTC)

Move Brugia malayi stub out of Bacteria -stubs

Brugia Malayi is a helminth parasite of human beings. It has been listed as a bacteria stub. This is potentially misleading as it is not infact a bacterium. Is there a way to move it out of the bacteria stubs category? I propose creating a new category of stubs called helminths. Sub categories Nematodes, Cestodes and Trematodes can be added to this. This would be in keeping with current scientific knowledge. I am new to posting on Wikipedia so all information will be helpful. Thanks
this unsigned posting was added by user:R manish at 01:43, January 11, 2006 UTC

I've answered on your talk page. --Adrian Buehlmann 19:20, 11 January 2006 (UTC)

God Mode Light

I realize that this is a javascript extension and not part of the wiki proper, but I am hoping that somebody can help me with it anyway. When I try to use the rollback link it goes to history and parses it like it should, but just before what I assume is the actual edit occurs ti crashes Safari back to Finder, i.e. quits without warning. Is this a known problem that I missed the mention of? WAvegetarian (talk) (email) (contribs) 17:54, 6 January 2006 (UTC)

I'm running Mac OS 10.4.3 (the latest), Safari 2.0.2 (the latest) and have installed scripts at monobook.js. WAvegetarian (talk) (email) (contribs) 17:59, 6 January 2006 (UTC)
Is this on wikipedia or another mediawiki installation? -- ℑilver§ℑide 13:35, 9 January 2006 (UTC)
Sorry I abandoned this posting. It is on Wikipedia. I realized that it would probably be best to talk to Sam Hocevar as it is his script that is acting up. In case anyone else is experiencing this problem, it seems to be related to the fact that God Mode Light is broken as of deleted edits no longer being viewable to non-admins. No one's really quite sure, though. There are a couple of discussions underway at Sam's talk page. WAvegetarian (talk) (email) (contribs) 08:57, 11 January 2006 (UTC)


What links here of template:ll

I'm substing template:ll in preparation for deletion after TfD. The What links here of that spuriously presents an empty list. Adrian Buehlmann 16:37, 6 January 2006 (UTC)

Hmm, it doesn't to me. Maybe it was a temporary problem? I've posted what I see on your talk page. —Ilmari Karonen (talk) 19:08, 6 January 2006 (UTC)
Yes that's what I wanted to say. It seems to be a temporary problem (Took a wrong word, sorry). And not stable too (See my message on your talk) Adrian Buehlmann 21:45, 6 January 2006 (UTC)

Not the only template with a problem. For example, the article space group has on it the {{attention}} template. However, if you visit that template and do a "what links here", here are all the results, and that article does not show up. Strange. This is confusing my bot who adds articles to Wikipedia:Pages needing attention/Mathematics based on this "what links here feature". Oleg Alexandrov (talk) 22:27, 6 January 2006 (UTC)

Yup, I'm here to report the same problem. The whatlinkshere of Template:Infobox Philippine city isn't showing articles that include it that haven't been edited in the past few days. Coffee 06:15, 7 January 2006 (UTC)

Khoikhoi wrote on my talk page:

"Hi. I've been watching what links to Template:ll, and it seems that about 6 or 10 pop up every day or so. When I check the history tab of those pages, it seems that the template was added a long time ago. What's going on? --Khoikhoi 05:35, 8 January 2006 (UTC)"

For me it smells like a bug. Adrian Buehlmann 08:09, 8 January 2006 (UTC)

I have noticed this too; {{WikiProject Wisconsin}} should link to a few hundred pages, but currently shows only a handful. HollyAm 03:30, 11 January 2006 (UTC)

I also brought this up at the Admin's Noticenoard and User:Splash replied saying "There was something on wikitech-l about template whatlinksheres being broken at the moment and the devs being aware of it". So I guess we'll just have to wait. Coffee 08:13, 11 January 2006 (UTC)

Problem with My Skins

I hope someone can assist. I changed my skin from Cologne blue to (I think) My Chick, or something like that. Now I can't change back or change skins at all. When logged in, I'm allowed only limited functions and can't change some of My Preferences. I think I've followed all instructions, but I'm not into the technical stuff. Assistance much appreciated! --Robert Fraser 02:44, 6 January 2006 (UTC)

You can temporarily change your current skin for a single page using ?useskin=, for instance http://en.wikipedia.org/wiki/Special:Preferences?useskin=monobook. With it, you can simply change your current skin to Monobook when opening the Preferences page and change everything you need (including the skin itself). --cesarb 18:44, 6 January 2006 (UTC)

Works like a dream. Many thanks. --Robert Fraser 10:03, 8 January 2006 (UTC)

This was a reported bug back in November or December of 2005. The problem was that users who'd selected the MySkin custom skin couldn't change back. I don't recall why, but I'll look in the tracker and what happened. It's obviously not fixed. Rob Church Talk 19:20, 12 January 2006 (UTC)

Session data

I'm having a lot of losses of session data. Now, in the past, this impacted only rollbacks, which was annoying enough. But now it's impacting edits, and I am having to log in and out for simple edits. This needs to be fixed, please. --Golbez 22:22, 5 January 2006 (UTC)

We need contextual information! Conditions, configurations...etc. Rob Church Talk 19:19, 12 January 2006 (UTC)
It resolved itself a day afterwards. --Golbez 19:29, 12 January 2006 (UTC)

Resolver talk page cross linked

When I go to resolver then its Talk page, I get taken to the Talk page for Resolver (album). Nurg 02:25, 12 January 2006 (UTC)

Fixed. When you get redirected, it will give you a little link to the page you were redirected to: click that link, then click edit. — Ambush Commander(Talk) 02:42, 12 January 2006 (UTC)


Why is this deleted image stil trying to load?

Yesterday, File:Green Shahada crescent.png was deleted, but the image's page still exists along with the data for it. What causes this to happen? --Spring Rubber 22:23, 13 January 2006 (UTC)

Happens sometimes (not sure why), I restored one revision and deleted it again. That usualy does the trick and indeed it seems to be fixed now. --Sherool (talk) 22:37, 13 January 2006 (UTC)


Page Differences

I would like to know if this is just me or not... Take a look at the difference links (last) on this page. For me, all the difference links are coming up with #header added on to the end of the URL, for example http://en.wikipedia.org/w/index.php?title=Wikipedia:Community_Justice&diff=34934305&oldid=34933734#Tasks - now, this doesn't only occur when the page has been edited under a specific question, it seems as if it happens all the time. If you look down the page, each difference link below a section-edit (i.e. (→Governance)) has that link added to each of the URLs (i.e. #Governance), down until the next section edited (i.e. (→Tasks), where it then adds #Tasks on to the URL etc). I hope that makes sense, I'm not a very good explainer. Is this a fault, is it new, it is something that's happening to me only, or is it something that's old and I just haven't noticed before? Thanks, →FireFox 19:34, 13 January 2006 (UTC)

Looks intermittent; perhaps a bug under PHP 5. Checking it out. --Brion 19:52, 13 January 2006 (UTC)
Ok, should be working correctly now. --Brion 20:15, 13 January 2006 (UTC)
Thanks. →FireFox 20:55, 13 January 2006 (UTC)

Gregory Lauder-Frost

Gregory Lauder-Frost won't load Fred Bauder 19:27, 13 January 2006 (UTC)

Special: New pages

Where is one menat to post comments on special pages? They have no talk page. Anyway, is it the case that #REDIRECT [[articlename]] pages are not listed in Special new pages? I ask because I thought they were / should be and I created 2 pages earlier today which were not listed. Puzzled. -- SGBailey 15:37, 13 January 2006 (UTC)

I made an extension that adds a talk tab to special pages or to Project talk:Page name for Special:Page name) (testwiki). I might install it if people are interested. —Ævar Arnfjörð Bjarmason 22:14, 13 January 2006 (UTC)
Nearly good, but then the view project page button to return to special newpages doesn't work. Also I use the Classic skin - will it still work in that? -- SGBailey 23:21, 13 January 2006 (UTC)
Anyone know the answer to the redirect no show part of the question? -- SGBailey 23:21, 13 January 2006 (UTC)

Week number changed

Looks like the number used for the "week in the year" was changed to now contain a preceding zero. This makes the Portal:Astronomy picture of the week not work. Is this a permanent change or something that can be reverted back? Awolf002 15:05, 13 January 2006 (UTC)

I noticed it too on Portal:Trains today. A quick move of the subpage to include the leading zero got the portal showing the right data again, and if CURRENTWEEK is changed back to exclude the leading zero, it'll still work. However, it'd be nice to know about changes like this beforehand so we don't have to go back and cleanup newly formed red links. Slambo (Speak) 15:10, 13 January 2006 (UTC)

Right, that's why I'm asking. Should I do the "move" or wait for a "fix"? Awolf002 15:15, 13 January 2006 (UTC) Okay, I moved the pages in question. Awolf002 16:03, 13 January 2006 (UTC)

It seems to be bouncing back and forth between the two when the page cache is purged. —Kirill Lokshin 17:10, 13 January 2006 (UTC)

Apparently behavior of the function that generated this changed in PHP 5; we've upgraded some of our servers to PHP 5.1 and so it was a little wacky depending on which server you've hit. It should now be consistent once re-rendered; if you see an odd page, purge it to make sure it's cleared. --Brion 18:32, 13 January 2006 (UTC)

Thanks for the info! Awolf002 18:39, 13 January 2006 (UTC)

Images are not showing up

Images uploaded to Commons are not showing up (on all articles). Any idea what is happening? - GaneshkT/C\@ 04:56, 13 January 2006 (UTC)

Trouble with the Commons image server, most likely - it might be temporarily down. Commons is showing no images even on Commons. Deco 04:59, 13 January 2006 (UTC)
It's back up now. The images are showing. - GaneshkT/C\@ 05:39, 13 January 2006 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(technical)/Archive_L&oldid=1091939117"