This page contains discussions that have been archived from Village pump (technical).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.
User talk:BostonMA has been gone for some time now but his page remains. Today I noticed that the template (guess that is what it is) on his user page has been vandalised by two inserts of User:Essjay/Clickthru/3 into the template formating the top of his page. Since he is not here to fix his page, please would someone look into it? (It must be a vandalised include as I understand these things.)
Sincerely, Mattisse16:07, 4 March 2007 (UTC)
Perhaps a redirect should be instated, if permissible, due to many inclusions of all of Essjay's clickthru templates? Of course, Essjay did request a U1, so this seems murky. See this. That's what my user box would have looked like if I didn't, by pure coincidence, switch to imagemaps a couple of days ago. Of course, we could switch to {{click}} in most cases (or even better, to imagemaps, although this would take much work). GracenotesT § 16:25, 4 March 2007 (UTC)
I don't understand. Why is it in the box at the top of User talk:BostonMA's page (in two different places where previous there were no such links at all)? Could you look at User talk:BostonMA to see what I mean? He never had any links there before (only images) or ever had contact with User:Essjay Sincerely Mattisse16:30, 4 March 2007 (UTC)
I don't know how to do as you are suggesting above. (There was an edit conflict when I posted.) I just think it needs to be fixed. Sincerely, Mattisse16:34, 4 March 2007 (UTC)
Fixed that particular example, but as I said above, it's used in dozens of transclusions--VectorPotentialTalk16:40, 4 March 2007 (UTC)
Someone should probably contact a pywikipediabot owner to change all instances of User:Essjay/Clickthru/3 to User:Llama man/Clickthru/3, the same thing should be done for User:Llama man/Clickthru--VectorPotentialTalk16:44, 4 March 2007 (UTC)
I've gone through all pages that transcluded to User:Essjay/Clickthru and User:Esssjay/Clickthru/3 with AWB and changed them over to the User:Llama man/Clickthru versions. The Special:Whatlinkshere lists seem to be a bit lagged, but I'm pretty sure they're all changed —Krellis (Talk) 00:54, 5 March 2007 (UTC)
Need some assistance
Probably the end of my time here... :)
I just deleted Jimbo's talk page in order to delete a revision (at the request of a user concerned with personally identifying information inadvertantly revealed.) Now I'm unable to undelete it--I just get a blank page instead of a list of the 13,000+ revisions. Any help appreciated. —Doug Belltalk05:53, 5 March 2007 (UTC)
Maybe true, but I saw before I did it that Slim Virgin had done it when there were more than 11,000 edits...so I figured I would be OK. I can email you which version if you think you can help. —Doug Belltalk06:01, 5 March 2007 (UTC)
As it pretty much states above, I want to be able to customise MediaWiki messages just for myself. I get annoyed when I come across userpages that imitate the "You have new messages" message, and I want to change the text so I can easily tell whether it's a joke or whether it's real. Obviously I can't change the style as intelligent people would just use the CSS class for formatting. Harryboyles11:56, 4 March 2007 (UTC)
You could probably tell the difference between a real message and a fake one using JavaScript, and change the formatting that way. But you can't change the HTML before it comes out of the server if that's what you're thinking. -- Tim Starling12:20, 4 March 2007 (UTC)
I would personally love to have my own MediaWiki pages (that is, replacing the text of the messages with my own; in the same vein as having "contributions" instead of "my contributions", or "stalked pages" instead of "my watchlist")> I suppose that this also is not technically feasible. GracenotesT § 13:41, 4 March 2007 (UTC)
Gracenotes: it is possible with Javascript:
addOnloadHook(function(){
var portlet;
if (portlet = document.getElementById('pt-mycontris'))
portlet.firstChild.innerHTML = 'contibutions';
if (portlet = document.getElementById('pt-watchlist'))
portlet.firstChild.innerHTML = 'stalked pages';
})
Unfortunately, MediaWiki messages do not constantly nor consistently have div/span IDs or classes. Also unfortunately, the class on a usermessage div can be emulated in wikicode (as the opening statement lamented). What you can do is specify the css more specifically. For example:
...Would give real usermessage notes a red border (only if they were preceeded specifically by the siteSub and contentSub divs). It could still be spoofed, but not as easily. --Splarka (rant) 08:31, 5 March 2007 (UTC)
Diff function screwing up
See [1]. This edit looked like vandalism, as it was changing comments in the straw poll, so I tried to undo it; it was marked 'conflicting intermediate edits', so I reverted it by hand, preserving the comment in between. This, however, demonstrated that there was no comment-reattribution going on in the original diff after all, so I reverted myself. (I've checked the resulting state of VPR; it seems to be fine in the current revision.) What's confusing me is why the original diff shows reattribution going on in the first place. Does everyone else see this? Does anyone else know what's causing it? --ais523 17:14, 5 March 2007 (UTC)
The diff seems to look fine when viewed through popups, so it must be something wrong with the server-side diff rendering. I'm guessing that maybe an oversight or something has taken place, and something might still have been cached. Interestingly, the URL [2] (note that %28 and %29 have been replaced by brackets and diff=prev has been used) does seem to work.Tra(Talk)17:41, 5 March 2007 (UTC)
No, the diff you gave is the edit after the one I'm talking about. --ais523 17:44, 5 March 2007 (UTC)
I see it. There's definitely something fucked up with the diff engine. Either that, or something's feeding the diff engine broken revisions, but I rather suspect the former. —Ilmari Karonen (talk) 18:20, 5 March 2007 (UTC)
Sig won't link to user page
When I do the 4 tilde thing, my sig won't lionk to my user page
I recently installed the Twinkle script to my monobook.js, and none of it works. I've already asked the script's creator, AzaToth, for help here, and he couldn't figure out what's wrong.
I've already purged the server cache and bypassed the browser cache, and as I explained to AzaToth, none of the buttons work. (Note: If it's useful at all, my discussion with AzaToth also has some of the error messages I get when I try to use Twinkle.) PyrospiritFlamesFire02:01, 6 March 2007 (UTC)
Yes, you can use the {{click}} template, but it is VERY discouraged. I would argue the Godfather logo is copyrighted and therefore the image must have a wrong license, though... I heard they were trying to develop a map extension for clicking images. In any case, just leave the template as is, or better yet, change the image for a wikilink to the article. -- ReyBrujo03:49, 5 March 2007 (UTC)
Thanks, but why is it discouraged? I also noticed {{Evanescence}} and its use of Image:Evanescence.svg. If its copyrighted, is it illegal to make The Godfather font? Perhaps its creator had permission, I'm not sure. I honestly think the template looked really bad with just a wikilink, and I can't think of any better font than this. That's just my opinion though. Khoikhoi04:04, 5 March 2007 (UTC)
Clicking on images should normally bring up the licensing information for the image. Using {{click}} and similar may be confusing given those expectations, may have exploits, and hiding the licensing info may not be OK with the GFDL. It also may have accessibility issues. Imagemap is better because it allows a link for the licensing info to remain in the image.
Copyrighted images are only "fair use" for articles, ie in article space, so use in a template means that template should *only* be used in article space. If one were really ornery, the template could be made such that the image only appears in transclusions in article space. Hmmm. Maybe there should be a template to make that easy... Gimmetrow04:16, 5 March 2007 (UTC)
If it's a fair use image, it shouldn't be used in a template at all. In this case, the image was created using a "Godfather" font and is not quite the same as the copyrighted Godfather logo (although it might be close enough that copyright infringement could be claimed). -- Rick Block (talk) 05:07, 5 March 2007 (UTC)
If it's a fair use image, it shouldn't display outside articles where the fair use is claimed. Does that mean it cannot be linked in a template? A template could be written with an image inside an includeonly and conditional on article space, for instance. Gimmetrow17:51, 5 March 2007 (UTC)
Recently, when I go to any wikipedia page my Internet explorer has started locking up. I have the new tabbed version, not sure exactly of which version it is but the only way for me to kill it is to ctrl+alt+delete and kill it throth task manager. When in task manager, it shows that iexplorer is using 99% of the cpu resource. It does not happen every time and after killing it in task manager it works again for an undetermined amount of time before it happens again. Is this a bug with wikipedia or my browser? It is starting to get realy annoying. Thanks! -- Chrislk02 (Chris Kreider)20:59, 1 March 2007 (UTC)
It might be one of your user scripts (maybe popups?). To see if this is the case, blank User:Chrislk02/monobook.js and try browsing the site. If it works fine, then you know it's one of those. You can then add them back individually until the problems come back, in which case, you've found the culprit. Tra(Talk)21:45, 1 March 2007 (UTC)
The owner of the server has apparently followed instructions to add the hash (#) symbol to the front of all the functions.
I discovered (after a looong time) that the pipe (|) symbol doesn't work within conditional statements, so I created a template (Template:!).
But here's the thing: the script I created in the template seems to work ok in Wikipedia, but not on the other server. I created a temporary template, Template:Temptest1 to test with and a page in my user space to check the results, User:Setanta747/temptest.
I also copied Template:Infobox Company to see if that would work, being a relatively simple template (or so I thought!). I had to remove the html code to start with, as the other server just didn't seem to want to know about it, so I replaced the tr and td and th tags with pipe templates and eventually got it to display all the fields.. but the conditional function is still not working - the fields are displayed as {{variable1}}.. {{variable2}} etc if the variables aren't included in the article.
The interpreter or compiler or whatever, seems to be extra-sensitive with regard to spacing - almost like COBOL - if I remove a line break, or add one, it seems to radically change the layout of the template. I had to do a lot of trial and error editing to get anything approaching a template that can be used at all.
Here are links to the relevant pages on the other site, if anyone is inclined to help out.. or if anyone is familiar with what I'm talking about.
Mal, it's a complex table, and you might want to test a few simple things down at the bottom of the template, like printing out the variable values to see what got passed when the article invoked it: F1: {{{F1|undefined}}}; F2: {{{F2|undefined}}}; etc.
When using variables that may not be defined, give them a default value, like {{F1|}}}(blank) or {{F1|no}}} -- because undefined variables cause trouble in conditionals.
For instance, where you have: {{#ifeq: {{{F1}}}|yes| define a default value of yes or no, whichever way you want to jump when the variable doesn't get passed by the article: {{#ifeq: {{{F1|no}}}|yes| would default to no, don't invoke the next section, bypass it for the section after that.
P.S. When debugging a template -- rather than save each change and go over to the article to invoke it -- try giving all your variables the default values you'd want the article to provide, then hit the "show preview" button to see the results right away. Something of a timesaver! And then if you later get different results when the article invokes it, you know the problem's got something to do with how the variables got passed. Maybe the whitespace is getting included, and |F1=yes|F2=yes| (without the spaces and linebreaks) would pass the right values. That being a different system, with possibly different flags set, just such an insanely simple problem (and solution) may apply. Again, good luck! -- Ben12:51, 5 March 2007 (UTC)
Ben thanks so much for your insights. I've not got the time to try out your suggestions just now, but it looks like you've given me some fresh direction to go in. Thanks again. :) --Mal13:39, 5 March 2007 (UTC)
I found out that the problem was the PHP version on the server. My friend has PHP version 4.x installed, and apparently the conditionals work properly on version 5.x and above. Change of server for us then. ;) Thanks for your help again Ben - its still useful to bear in mind what you'd said. --Mal20:18, 5 March 2007 (UTC)
Something else to keep in mind: Wikimedia uses HTML Tidy ($wgUseTidy in the settings) which allows html to be broken and later fixed by parserfunctions and transclusions (something the built in HTML sanitizer doesn't allow). On a MediaWiki install without Tidy enabled, many templates from here will break. --Splarka (rant) 08:08, 6 March 2007 (UTC)
Please enable JavaScript
Is anyone else getting a red notice saying "Please enable JavaScript to access the extended special character edit tools" at the bottom below the edit box, even though JavaScript is enabled? I'm using Firefox 2.002 on Windows XP and under Tools>Options...>Content "Enable JavaScript" is checked, yet I'm still getting the message. Any suggestions? —Angr21:12, 6 March 2007 (UTC)
That's this change, this admin probably tested this in only one browser (IE7?). Quite surprising. This will be reverted back quickly and then I hope will be discussed cause the idea itself might be good. — Alex Smotrov21:18, 6 March 2007 (UTC)
Okay, I wondered what was going on. I tried to sign my name with the ~~~~ button but I couldn't, so I had to type it out. Acalamari21:28, 6 March 2007 (UTC)
First of all, have you purged your cache? (WP:CACHE) It may repotedly require a few tries, since this change is a combined effort of .js, .css and a modification of Mediawiki:Edittools. And it should work fine on Firefox. Миша1321:38, 6 March 2007 (UTC)
(edit conflict) I cleared my cache multiple time with no luck. I tried multiple browsers (one with a freshly cleared cache) too. It is possible that there may have been a lag in the software. I reverted your change to MediaWiki:Edittools. Since you are here, I'll let you sort everything out and revert me as necessary. Thanks. --PS2pcGAMER (talk) 21:43, 6 March 2007 (UTC)
Looks like it's happening to everyone. The api.php RSS watchlist is stuck at yesterday for me, but the api.php query watchlist appears to be working (which is how I noticed this post, as it's used by a script I've got). Using the api.php query watchlist may work as a temporary workaround... --ais523 17:56, 6 March 2007 (UTC)
On browsers that can display XML, this link is probably a more readable workaroung for the time being. --ais523 18:01, 6 March 2007 (UTC)
The watchlist hasn't totally stopped; mine seems to be updating every half hour or so as well, and it only registers some edits. Does anyone know exactly what the problem is? Acalamari18:02, 6 March 2007 (UTC)
Glad it's fixed. Maybe it'll start showing edits from the future soon... --ais523 18:21, 6 March 2007 (UTC)
Does anyone have any missing edits because of this? Cometstyles said that they are missing ten minutes of edits. Anyone else have this problem? My edits are all fine. Acalamari18:19, 6 March 2007 (UTC)
Along with the perpetual database locks and error messages recently, I'm guessing servers are suffering from the load. Xiner (talk, email) 19:46, 6 March 2007 (UTC)
The contributions sometimes take a minute to update; I've had that before. I'm more worried about that watchlist delay, and the "server overloaded" warnings. Acalamari19:59, 6 March 2007 (UTC)
I'd like to take a look at the format and source-code of that progress bar / guage that appears when Wikipedia asks for donations - you know, the one that appears at the top of each page to say what percentage of their donations target has been reached.
Do you know where I can find that progress bar in a template or something? I've done a bit of searching, looked up template:progress, etc etc and can't seem to find it.
Thanks in advance. 60.227.97.7915:34, 1 March 2007 (UTC)
Great, looks like a php script. Darn I gotta start learning PHP properly cos the syntax isn't singing to me. But I worry that even if it did, I still wouldn't know how to implement it - how do I transclude it in a page or feed it the variables it needs to generate the image? But kudos for the answer. Rfwoolf03:48, 2 March 2007 (UTC)
How to archive entire Wikipedia at once in several languages?
Any suggestions on how this could be done? I'd like to have a Padd
type device using a tablet PC. I realize I could use a spider but maybe someone has experience with this. I'd also like to have archives in several different languages. I realize this would be quite large and I would also like to cache all related images. Thanks --70.106.252.24704:36, 7 March 2007 (UTC)
See WP:SIZE. One way is to search for the article (use "search", not "go") and the last cached size should be listed. If an article is over 30k, the raw size is shown above the edit window. For smaller articles (or to be more precise on "prose size"), you could go to the "printable version" of the article (it's a link in the toolbox on the left), remove the references, and check the resulting file size. (There are some javascript tools which do the same, but if you only need to know for one article, it's probably easier to do it by hand or ask someone rather than install the tool.) Gimmetrow03:35, 6 March 2007 (UTC)
Another method, not necessarily more convenient, is to click the edit tab at the top of the article, select all the wikitext in the edit window, copy to the clipboard, and then paste into an external text editor that has a function to show the file size. This measures the size of the article's wikitext, and does not account for the size of images the article displays, the size of any transcludedtemplates, nor the expanded size of any magic words. If you want to know the size of the downloaded HTML of an article, you could use your Web browser's file information function. If you want to count the words in an article, you could use Unix commands such as lynx and wc. For example, this counts the words currently in the Village pump (technical) page:
I used to have a small tab with a '$' in it that showed a user's groups (i.e. Admin, Bureaucrat, etc) when clicked. I'm sure where I got it or where it went, but does anybody know of a script that does this? John Reaves(talk)01:07, 6 March 2007 (UTC)
That is an interesting idea, the only catch I could think of is if there were two separate articles that differed only by their capitalizations, it might cause a potential conflict. Of course there probably aren't supposed to be article titles that differ by so little, since that would be confusing--VectorPotentialTalk16:45, 3 March 2007 (UTC)
OTOH... there are Some few, which are deliberate-- see WP:WC, and WP:Wc, making for a one character change in the resultant url if one gets a memory fart. The one is a resource page for the newcomers, the other the project, such as it is. I would think system software could be written trying the expressed version first and successive capitalisations sequentially until a match is found whilst ignoring prepositions. Maybe a filled in wrong case could be represented by a violet link color--combining red and blue! <g> // FrankB22:17, 3 March 2007 (UTC)
WP:Wc and WP:WC redirect to the same page for me. I also like the violet idea, if it can be applied to all wikilinks, and if its codable. If it IS the Gobar checking capitalizations for pages, this makes me suspect it is somewhat codable. I wonder who we bring it to then? Logical2uReview me!22:52, 3 March 2007 (UTC)
Bugzilla -- see link at page top, submit as "requested features".
Can post on the village pump at Mediawiki too, but the Bugzilla route has 'procedural handling' in place for such requests, so will at least get 'kicked around' on the email circuit. // FrankB23:33, 3 March 2007 (UTC)
WP:Wc and WP:WC both redirect to the same place. either:
a. frankb is mistaken
b. frankb is correct, but the explanation is hard to understand
c. frankb is correct, and i am not smart, and misunderstood. this seems the most likely.
offhand i can't think of a situation where there should be two articles that differ only by capitalization. --barneca (talk) 22:55, 3 March 2007 (UTC)
Sigh... yet another case of being edited unmercifully, I fear! "Alas! Alack, Oh woe is me!" WP:WC (edit | [[Talk:WP:WC|talk]] | history | protect | delete | links | watch | logs | views)used to go to this page (diff shows being moved)and this diff (shows a shortcut reassignment) both recently, and proves my assertion-- or would have, or 'whatever' and something like that, etc. etc. etc. ad nauseum. <G> Q.E.D. - Both pages existed on Feb 9th, or that last edit made no sense whatever. Sniff. So pick Enone of the above! Oh thou mocking 'Barneca', but be assured when the FrankB speaks there be something behind the statement! (It really doesn't come off very mature either, dude.) The situation in question arose back when they were discouraging subpages except in all but user spaces. So that being settled, moving the page makes some sense. But I'll have a stern talk with The Transhumanist on keeping consistency with shortcut links! Maybe even spank 'em with a wet noodle! Cheers! // FrankB23:33, 3 March 2007 (UTC)
(Edit conflict)
Wikipedia is case sensitive indeed. However, when you click "Go" the search box looks for alternate capitalizations. Thus, if you type Prestige Oil Spill, the search will search for Prestige Oil Spill, Prestige Oil spill, Prestige oil Spill, prestige Oil Spill, etc, and since it finds Prestige oil spill, it will show that page. Using common sense, you should not create two pages that differ only in capitalization (in example, Virtual console and Virtual Console should both redirect to the same page, because the difference between both words is trivial, and not to different pages, as it is currently). Redirects that differ in capital letters should use the {{R from other capitalisation}}. -- ReyBrujo23:23, 3 March 2007 (UTC)
Good point... any editor earning their munificent salaries here needs to back check any redirect link they come across for the proper {{R from ... } tagging. The Category:Redirects links to the reference page so you need not remember that or it's shortcuts. There are maybe five or six common ones. // FrankB23:39, 3 March 2007 (UTC)
So, realistically, we need to check each article on the Wiki for multiple capitalization schemes of the same article before any "Capitalization Fixer" is implemented? Logical2uReview me!01:22, 4 March 2007 (UTC)
the URL are fine when it use in common web browser. do i miss something here?
i know that wikipedia article are under GNU Free Documentation License, so didn't think i violate any rules, did i?
The servers may be rejecting your request due to the user agent string of your request; try customizing it to something unique to your own mobile browser. --ais523 10:42, 7 March 2007 (UTC)
ais523 is correct; the server has banned certain user agents, so you will have to change the user agent to something else (e.g. MyMobileBrowser/0.1). Jayden5412:42, 7 March 2007 (UTC)
Dear all, thank you so much for your kind suggestion. What you have say is right, indeed i have to specify a unique user agent on my request header. By the way, i did some mistyping on the code i write in this post, my real code is truly passing urlstr to URL(). regardsDidito13204:09, 8 March 2007 (UTC)
Perhaps the new version of the software is not yet running on this wiki. When I delete a page that has been deleted before, I don't get the old deletion log entries either. Kusma(討論)07:38, 6 March 2007 (UTC)
Soon, it was reverted due to an SQL error for newbie contribs due to a mistake. I've committed it again with that issue fixed. Voice-of-All00:48, 7 March 2007 (UTC)
To solve the problem we encountered here regarding the PHP version on our server, the server owner is wanting to migrate to a different server.
I'm unfamiliar with the installation process, and what he needs to know is, basically, the best order in which to do things. We are running Mediawiki 1.69 at the moment and we want change to another server and upgrade it to the most recent version at the same time.
As we understand it, the MySQL databases can just be backed up and rebuilt on the new server, and other files can be FTPed, but what order is best for upgrading. Does he have to install most recent version of mediawiki to the new server first. or what..?
You might also try experimenting first with the move and upgrade on a workstation, using the m:Wiki on a stick method. That may be more convenient for working out your exact procedure, verifying that all your extensions work, that you have preserved all your image uploads, customizations, and so on. Be sure to document all the steps in your procedure on a wiki page, because you will need to upgrade MediaWiki repeatedly in the future, and because the available documents (mw:Manual:Upgrading, mw:Manual:Moving a wiki) are a bit vague and general (they don't list all the specific commands you will need for your setup). Once you have seen how the move and upgrade work on your workstation, you can be more confident when you repeat them on a full-blown server. --Teratornis16:39, 7 March 2007 (UTC)
What an excellent resource, and a helpful bunch you lot are! Thanks yet again for going above and beyond the call of duty to help others out with their problems. :) -- Mal19:27, 7 March 2007 (UTC)
Thanks for being one of the few to say thanks. (I think I forgot to do that when I asked some questions, and now my ingratitude is archived forever. I swear, I felt grateful.) Anyway, I figure we're getting this fantastic software to use for free, we might as well give something back. --Teratornis08:14, 8 March 2007 (UTC)
Serbian-style transliteration
Hi, not sure if this appropriate place to ask, but we would like to create a Syllabics<->Latin converter for the Inuktitut wikipedia in the style of the Serbian Cyrillic<->Latin converter, but I've no experience with the actual MediaWiki software. Can someone point me to a guide on how the Serbian wiki accomplished this, or at least point me to the relevant source code? Thanks, Moszczynski17:23, 8 March 2007 (UTC)
How do you keep history?
I am curious about the feature which allows for people to be able to see the history of changes.
What do wiki folks do regarding that? There can be two ways of doing this:
1) store the ENTIRE article
This is trivial and changes can be reflected easily but it is highly inefficient because you are storing tons of repeated data. So there has to be a better way ... maybe ...
2) store the incremental changes?
If you store the incremental changes how do you store these changes? How do you remember that a word in certain paragraph was modified? Is there an algorithm for this? Someone told me about "Levenshtein distance", is this used by wikipedia?
would someone be kind enough to email a response to "atharshiraz" user id for a yahoo email account. Appreciate it! May wikipedia continue to grow and prosper.
Pretty sure it keeps the full version of each revision, then uses a difference engine to tell you what changed in the diffs. --Golbez21:35, 7 March 2007 (UTC)
This is not an entirely technical question, but I can't see a more suitable section.
If a user uploads an image which contains metadata that indicates that it is copyrighted and licensed under the GFDL, the first thing WP does is create and redistribute a copy of the image with this metadata stripped out. That would appear to be a violation of the GFDL by WP itself -- licensees are not permitted to remove copyright notices or license notices from copies. I'd argue that the image scaler should keep such metadata intact.
I asked about this elsewhere and another user pointed out that a browser displaying the image file on a user's screen also constitutes making a copy without the metadata. That's true, but until the image file leaves WP's servers it remains WP's problem. If the user then does a screen capture and redistributes the captured image, it is the user violating the GFDL, not WP. Rling21:00, 6 March 2007 (UTC)
It is, but I still think this is not enough. The purpose of having embedded metadata is so that it travels with the file. IPTC metadata was created for exactly this purpose, EXIF also. They would not exist if having the metadata in, say, an accompanying text file, was good enough. It's too easy for the two to be inadvertently separated. How many digital images do you have copies of? How many of those have (non-embedded) information saying where they came from? Rling20:41, 8 March 2007 (UTC)
The image scaler (ImageMagick) probably does not know anything about the metadata, and thus fails to copy it (it's not stripped out, it's simply not copied). How is that metadata implemented? If it's not a critical chunk, programs are supposed to be allowed to completely ignore it when processing the image. --cesarb22:34, 6 March 2007 (UTC)
This thread [3] suggests ImageMagick at least attempts to copy metadata, although this one [4] suggests it doesn't. I'll have another look. Note that the PNG specification includes the text "we strongly encourage programs handling PNG files to preserve ancillary information whenever possible"[5]. Rling20:41, 8 March 2007 (UTC)
DEFAULTSORT question
Before reporting a bug, I want to make sure I am doing this right. According to DEFAULTSORT's help, you can have multiple defaultsort statements. However, Oliver Twist didn't do as I expected in Category:Big Read Books - it sorted under "Twist, Oliver" instead of "Oliver Twist."
{{DEFAULTSORT:Oliver Twist}}
[[Category:1838 novels]]
[[Category:Novels by Charles Dickens]]
[[Category:Serialized novels]]
[[Category:London in fiction]]
[[Category:Black and white films]]
[[Category:1909 films]]
[[Category:Big Read Books]]
{{DEFAULTSORT:Twist, Oliver}}
[[Category:Charles Dickens characters]]
[[Category:Fictional orphans]]
[[Category:Fictional thieves]]
The help says you can, which is why I am asking: Once this has been included in an article, the new default sort key will remain in force until the end of the article, or until a fresh {{DEFAULTSORT}} is used. — RevRagnarokTalkContrib21:09, 24 February 2007 (UTC)
The help page is ambiguous and misleading. What is actually means is that the last {{DEFAULTSORT}} will be used, no matter what previous ones might contain. You may want to make a feature request to do what you'd like to do. --MZMcBride23:27, 24 February 2007 (UTC)
Agree strongly that it should be doing what the documentation says it can do, and the Oliver Twist example above is a perfect demonstration of why. :-) — SMcCandlish [talk] [contrib]ツ02:29, 1 March 2007 (UTC)
into {{onlinesource}} - can anyone help me by pointing out any glitches caused by this? (I already caught some whitespace problems; purge the cache on the merged template and the talk page it appears on, in that order, before reporting problems) --Random83218:40, 8 March 2007 (UTC)
You just need to make the parameters in the former 9 templates optional, so that if "section" (for example) is not defined, its value will be "" and not the text "{{{section}}}", and so that in {{onlinesource}} the #if statement will work. Great job merging them :) I'll see if I can help fix 'em. GracenotesT § 01:48, 9 March 2007 (UTC)
Ok, reading that page didn't help me, though maybe it should have. I fixed a link in the template to Harana/Valle de Arana after I moved it to the proper name, but What links here still shows all the articles that have the template transcluded as linking to the wrong title, but I can't see where they do. Is this just a caching issue? - TaxmanTalk17:58, 8 March 2007 (UTC)
The link table from an article isn't updated until that article is edited, I think (or it might be updated via the job queue (sort of like a backlog for the servers), which is running at over 100000 at the moment). If you don't mind wasting a crazy amount of time, you could null-edit each of the linking articles; normally, though, it would just make more sense to let the links disappear over the course of normal editing. --ais523 18:01, 8 March 2007 (UTC)
Makes sense. I just wanted to make sure I didn't mess something up. With that information I'll just wait it out, since the actual articles are correct. Thank you. - TaxmanTalk18:29, 8 March 2007 (UTC)
Apparently it was the job queue, as it's fine now. Mind writing a sentence or two about that up for the help page? - TaxmanTalk03:59, 9 March 2007 (UTC)
I've updated it on Meta; with the next update, it will be updated on Wikipedia too. --ais523 13:34, 9 March 2007 (UTC)
Included JavaScripts
Some people have included JavaScripts for signatures. I was wondering if anybody analyzed the security implications of this. In theory I could have a JavaScript that does the equivalent of a cross-site scripting without actually crossing any sites. Is this code executed? — RevRagnarokTalkContrib02:06, 9 March 2007 (UTC)
A link which you forgot to mention: User_talk:Netscott/s1.js. No, there are no security implications because .js extension has nothing to do with script execution. As far as I understand, .js extension only affects the way the article is formatted and protects it from editing by others. Only some special .js files (monobook.js etc. depending on your chosen skin) are executed only for the owner of the file. — Alex Smotrov03:10, 9 March 2007 (UTC)
For all practical purposes, it is just a regular page that only Netscott and admins can edit. No code is executed for anyone. Prodegotalk03:14, 9 March 2007 (UTC)
OK, I just wanted to make sure. I didn't want to name Netscott specifically because it's not like I am accusing him of doing anything wrong. — RevRagnarokTalkContrib11:47, 9 March 2007 (UTC)
Makesysop question
If anyone can help me with this, I would appreciate it.
I've got three installations of MediaWiki on my WAMP server, and I have got Makesysop installed on one of them.
What I would like to do would be to set user rights cross-wiki, like Meta does here.
I have tried to get it to set user rights for "User:Testuser@wikidb1" but it says no such user exists.
I'm completely guessing here as I have no knowledge at all on how meta works, but it's likely custom code, and not the standard makesysop function. Each MediaWiki installation has its own database, which means a global makesysop function will have to access multiple databases. Maybe you can find the code for this in the SVN repository available at svn.wikimedia.org or wait until one of the dev's reply. Cheers, Jayden5418:29, 6 March 2007 (UTC)
What would I need to fix to allow access to other MediaWiki databases on the same server?? On my MediaWiki installations I am considering using stewards to assign checkuser, oversight permissions to users, and to de-sysop them. I have tried to get this to work, but to no avail. --sunstar nettalk13:27, 9 March 2007 (UTC)
Time
For the last few hours, the time and date displayed precedent to my user options have been one full day early (at the moment, that is, the time is given as 23:15 UTC on 10 March). Whilst one may, IIRC, edit his/her monobook in order that it might display the time other than in UTC, for instance, I don't think myself so to have edited, and so I wonder whether anyone else is similarly affected at present. It's altogether significant, to be sure, but just rather odd (unless I'm wholly crazy and it has always been this way). Joe23:16, 9 March 2007 (UTC)
Transparent backgrounds
A recent change to the software caused images with transparent backgrounds to display the backgrounds as a checkerboard pattern. This is great for image description pages. However, it looks really bad on image galleries. A couple of examples:
I think it is very helpful on galleries (which should be about the images themselves, not so much about their content, if that makes sense), you can easily identify transparency on images that way. --Splarka (rant) 08:11, 9 March 2007 (UTC)
Actually, galleries are used to great effect in some articles, where showing a checkered background detracts from showcasing their content. —Remember the dot (t) 20:32, 9 March 2007 (UTC)
I cannot access my pref.
I get in but am not able to do anything whatsoever. Not even change my sig. Please help me. -- Darkest Hour 20:12, 6 March 2007 (UTC)
I get in but thats where it ends. I keep trying to access this linky but nothing happens. I hate this!!!! It bugs me!!!! PS; Your link did nothing for me. Darkest Hour 01:25, 7 March 2007 (UTC)
Working! Yay! Dunno what it was but it wors! -- Darkest Hour 19:53, 9 March 2007 (UTC)
External link question
What's the wikicode for hiding that little arrow at the end of an external link?
<span class="plainlinksneverexpand">[http://google.com Google]</span> creates Google. The CSS class can be used in tables and <div> tags as well. --MZMcBride22:13, 9 March 2007 (UTC)
Or simply class=plainlinks. See: example.com. I suggest you save all Wikipedia CSS files locally and inspect them, for example in main.css you could find
/* this can be used in the content area to switch off
special external link styling */
#bodyContent .plainlinks a {
background: none !important;
padding: 0 !important;
}
/* Don't display some stuff on the
user page */
body.page-Main_Page #lastmod,
body.page-Main_Page #siteSub,
body.page-Main_Page #contentSub,
body.page-Main_Page h1.firstHeading {
display: none !important;
}
What is the syntax for specifying the selector for a particular page, in the various namespaces?
And more importantly, where is this documented?
Also, I can't find any documentation on the body.page selector, nor any of the selectors of the set that it belongs to. Please point me to them if you know.
In my monobook.css, the above syntax works for pages in the main namespace, but doesn't work on my user page, for instance. (I swapped out "Main_Page" with "User_The_Transhumanist_(AWB)", but nothing happened. And yes, I cleared the cache). I also tried a colon after "User". Nuthin'. The Transhumanist (AWB)04:05, 9 March 2007 (UTC)
I'm not sure if these CSS classes are documented anywhere but you can always check HTML source code of the page. Go to your user page, view source code, you'll see <body class="mediawiki ns-2 ltr page-User_The_Transhumanist_AWB"> — so no parentheses in your case. — Alex Smotrov06:22, 9 March 2007 (UTC)
I don't understand the syntax of <body class="mediawiki ns-2 ltr page-User_The_Transhumanist_AWB">. Please point out the CSS selector from that string of code, and show me what it's supposed to look like inserted into the code block presented above. Forgive me, I'm clueless. The Transhumanist20:04, 9 March 2007 (UTC)
The syntax is a space-separated list of classes. In this case, it has the classes mediawiki, ns-2, ltr, and page-User_The_Transhumanist_AWB. The CSS selector is a class selector, so it would be either .page-User_The_Transhumanist_AWB or body.page-User_The_Transhumanist_AWB. --cesarb20:19, 9 March 2007 (UTC)
I've tried the following, and none of them work:
/* Don't display some stuff on the user page */
body.page-User_The_Transhumanist_AWB #lastmod,
body.page-User_The_Transhumanist_AWB #siteSub,
body.page-User_The_Transhumanist_AWB #contentSub,
body.page-User_The_Transhumanist_AWB h1.firstHeading {
display: none !important;
}
/* Don't display some stuff on the user page */
page-User_The_Transhumanist_AWB #lastmod,
page-User_The_Transhumanist_AWB #siteSub,
page-User_The_Transhumanist_AWB #contentSub,
page-User_The_Transhumanist_AWB h1.firstHeading {
display: none !important;
}
/* Don't display some stuff on the user page */
mediawiki.ns-2.ltr.page-User_The_Transhumanist_AWB #lastmod,
mediawiki.ns-2.ltr.page-User_The_Transhumanist_AWB #siteSub,
mediawiki.ns-2.ltr.page-User_The_Transhumanist_AWB #contentSub,
mediawiki.ns-2.ltr.page-User_The_Transhumanist_AWB h1.firstHeading {
display: none !important;
}
/* Don't display some stuff on the user page */
mediawiki ns-2 ltr page-User_The_Transhumanist_AWB #lastmod,
mediawiki ns-2 ltr page-User_The_Transhumanist_AWB #siteSub,
mediawiki ns-2 ltr page-User_The_Transhumanist_AWB #contentSub,
mediawiki ns-2 ltr page-User_The_Transhumanist_AWB h1.firstHeading {
display: none !important;
}
Hi! I'm putting together a schedule table for an event for a club that I am in that has it's own wiki. I'd like the table to be something like:
Date
Joe
Tom
etc...
April 14 8am
X
P
...
April 14 9am
B
B
...
many more...
...
...
...
B = Booth
P = Presentation
X = Not at the conference
My questions are:
Can text in a table be written at an angle (i.e. if Tom's name is longer, can I orient his name 90 degrees)?
If I have dozens of people in the table, can I break the table up into smaller pieces, so the user can edit a smaller and more understandable section?
Possibly like a pivot table in Excel?
Possibly like transclusion for composite wiki pages?
Possibly like templates, that could be merged together for the complete schedule?
Possibly like article sections?
Can I set the domain of cell values (B, P, X, etc...) with a drop down list or something, so the attendees can change their schedule on the fly and the rest of us will know where they are?
Yes, see here but it only works in Internet Explorer.
It would be quite hard to split up the columns but you could quite easily split up the rows by putting a section header between them which will allow only a few rows to be edited at a time.
There are two main ways of doing this. One is to have several separate tables above each other, each separated by headings, e.g. in Wikipedia:Bots/Status and another way is to separate each group of rows with a cell that spans accross the columns and contains a heading, e.g. in Wikipedia:WikiProject User scripts/Scripts#Modules. You could at each way and see which would work best, then use the wiki-code on these pages to show you how to do it. Tra(Talk)21:58, 5 March 2007 (UTC)
Tra - thanks for the examples. Very clear. I was actually able to get templates to create the table, by having {+begin the table { table... however, it took a lot of work for a very un-friendly table. Thanks, I'm going to try out some of your examples! --Shelton123418:47, 6 March 2007 (UTC)
No, you can't set up a dropdown list but since the values you are putting in are quite short, it would probably be easiest to just put them in directly. Tra(Talk)12:56, 3 March 2007 (UTC)
You could also put the long name into its own nested table, and display each character in its own row, so as to stack up the letters vertically, but with each letter in a normal orientation. That should work in any browser:
A comment about question #2: I had a similar problem on a corporate wiki I administer. We have a telephone directory/contact information page with about 50 people on it. The resulting table is pretty complex, and evidently nobody at my company other than me has any prior experience editing on a MediaWiki wiki. So I wrote a section of detailed instructions at the bottom of the page that tell people how to find the row in the table with their personal data, and edit it. At the top of the page, I have a link on the text "See below for editing instructions." So far, all of our complete wiki novices who have edited the table have done so without needing additional help. I would be happy to e-mail a copy of my editing instructions to you in wikitext format, so you could edit them to be appropriate for your table, but Special:Emailuser/Shelton1234 says you have elected not to be able to receive e-mail. --Teratornis18:43, 4 March 2007 (UTC)
Yet another comment: it is much easier to edit a row rather than a column, because all of a row's cells group together in the wikitext. Therefore, if you expect people to edit multiple entries for a given person rather than a given time, you should rotate your table 90 degrees so each row represents a person, and each column represents a time. This probably eliminates the need to display column headings vertically, because you can use column spanning to group all the dates under a given month (or date, if you have several times for one day). For example:
April
14
...
Name
8am
9am
...
Joe
X
B
...
Tom
P
B
...
...
Then the logical way to divide the big table into little tables would be by date range, for example maybe by month, depending on how many time columns you have room to display. To make the wikitext easier to edit, you could insert some HTML comment lines that would act as pseudo column headings when someone views the edit window. --Teratornis19:24, 4 March 2007 (UTC)
Thank you Tra and Teratornis for the excellent ideas. I will definately rotate the table. Also, I have changed my user preferences to receive email. Please send me a copy of your instructions! Thanks for considering my issue. The computer geek inside me still feels like there should be a way to edit a simple list and pivot table it into a summary table, but using wiki to coordinate this group is still light years ahead of other methods! Thanks! --Shelton123418:23, 5 March 2007 (UTC)
I still can't send you e-mail; see: Special:Emailuser/Shelton1234 for the error message. Try sending yourself an e-mail message from that form. Check your user preferences. Send yourself an e-mail confirmation (there's a link in user preferences to do that). If it doesn't work, then you can send me an e-mail message, and I should be able to reply. Assuming your e-mail gets through all the SPAM filters, etc. I suppose I could just paste my sample instructions on your user talk page, and sanitize out anything proprietary.
This template added to your monobook.js adds a tab at the top of the page that when clicked adds a line to an AfD debate that says the debate has been transcluded into a specific deletion sorting. {{subst:deltab|Australia}} for example adds a line to Wikipedia:WikiProject Deletion sorting/Australia. I add debates to more than one deletion sorting pages so I tried to add the template more than once to my monobook.js file to get more than one tab. They appear to conflict. Has anyone tried to do this and can explain how it works? I asked around early last month, including on the talk page of the editor who wrote the template, but I have no responses. Can some technical minded person help? --Bduke03:27, 10 March 2007 (UTC)
I still don't get it. I understand transcluding, but is the template page the *source* page that the info comes from, or is it the resulting page that is formed from transcluding say, a template and a page? If the resulting page is the one that is including documents from another page, then why would the transcluded text even show up on the template, aka source, page? Kat, Queen of Typos05:46, 10 March 2007 (UTC)
If you look at the template page directly, you'll "see" items within "noinclude". For example, there might be category:useful templates. I put "see" in quotes because you don't see categories in the inline text, they get shown at the bottom (or top) of the screen. You don't want the article in this category, only the template.
If you look at the page which transcludes the template, you "see" items within includeonly. For example, there might be category:templated articles, so all articles which transclude this template will be in the templated articles category, but the template itself won't be.
Items in the template which are not in either includeonly or noinclude sections will appear in both places. Typically, this is just text.
So, an example template is:
<includeonly>[[Category:templated articles]]</includeonly><noinclude>[[Category:useful templates]]</noinclude>Hi, you have transcluded this template!
I'm working on a private wiki for my club. I'd like to be able to store several common forms (PDFs, DOCs and XLS files) on the wiki, and be able to set up some "packets" of the forms, where a user can download all the forms in the packet at one swipe. I'm not interested in using a ZIP file, because if one of the forms changes and the form is used in multiple packets, I don't want to have to rebuild all the packets. Oh great wikipedians, any ideas? Thanks in advance, --Shelton123421:07, 9 March 2007 (UTC)
I don't know how to do anything like that with built-in MediaWiki features. Maybe an m:Extension would be necessary. A quick glance turned up m:FileSystemListing, but that doesn't sound like what you need.
Perhaps you should consider using a structured wiki such as TWiki for this, rather than MediaWiki, because you seem to want some form-driven wiki applications. (I still have not gotten around to trying TWiki, but I promise it will be real soon now). MediaWiki might eventually feature-creep its way into a structured wiki, but wiki application features were not necessary for the initial mission of building an encyclopedia. Now that thousands of organizations are running MediaWiki, there is a demand for that sort of thing, but all these other users are incidental to the primary mission for MediaWiki, which is to run the WikiMedia Foundation's wikis. In most organizations, there are some technical specialists, and perhaps many users who work with some sort of regularly-structured or repetitive data, and are not so technically inclined. They need IT staff to design simplified form-driven applications and reports for them. Having that capability in a wiki is obviously advantageous, because then you also get the shared whiteboarding and collaborative editing of in-house documentation, features traditionally missing from business software. --Teratornis17:46, 10 March 2007 (UTC)
Does the 'random article' function really generate random articles?
Yes, I'm aware that it can't be truly 'random' because it's generated by a computer, but I've often found while using it that the random article function directs me towards articles on subjects that I have been recently reading about on the Internet. Anyone care to explain this phenomenon?--h i ss p a c er e s e a r c h16:21, 5 March 2007 (UTC)
Coincidence? I'm fairly sure Wikipedia doesn't install spyware on computers to monitor what you have been looking up - but that might explain long hours on Wikipedia. :P x42bn6Talk17:06, 5 March 2007 (UTC)
It's always possible that it's just a coincidence, but at least two or three times now I've had an article about something I'd been reading about online recently appear in the random page, such as something I was discussing on a messageboard that day.--h i ss p a c er e s e a r c h17:28, 5 March 2007 (UTC)
For a second, just for a second, I thought you believed Wikipedia pieced an article together everytime you clicked Special:Random :-) Anyways, remember the Six degrees of Wikipedia theory. Anything you are reading is related to something you will read. But yes, coincidence as far as I know. Unless you have clicked on a link in the site that took you to Wikipedia ;-) -- ReyBrujo17:39, 5 March 2007 (UTC)
I believe that's a common type of cognitive bias: one tends to notice things related to subjects one has recently learned of or had one's attention brought to, where one might previously have paid little attention to those same things. I'm having trouble finding a specific article on the effect, however, if we even have one.
In any case, the distribution of random articles given by Special:Random isn't quite uniform, but should not be biased in any systematic way. (To expand on that a bit: due to details of the underlying implementation, some articles will get returned somewhat more often than others, but the probabilities assigned to the articles should be pretty much random and independent of things like the name or age of the article.) As x42bn6 pointed out, MediaWiki certainly isn't, as far as I know, monitoring your browsing habits. —Ilmari Karonen (talk) 18:12, 5 March 2007 (UTC)
I think of all the "Isn't life strange?" observations, like "Learn a new word, and suddenly you'll see it being used repeatedly in all sorts of contexts." Well, yes, let something catch your attention for a while, get sensitized to it, and it will jump out of the background. That's how our minds work. Hunt mushrooms for meals, and you'll start noticing them in places you never saw them before. Track animals, and how quickly the ground becomes a vast pattern of hoofprints and pawprints. Read How to Lie with Statistics, or a good text on logical fallacies or misleading rhetoric, and you start seeing those tricks used everywhere. In such situations, selective attention can be a survival trait. -- Ben20:50, 5 March 2007 (UTC)
Can someone please give me a brief, understandable tutorial on how to interpret the results of using WHOIS or Reverse DNS lookup to determine whether an IP address belongs to a proxy server? Thanks.--Vbd (talk) 02:42, 16 March 2007 (UTC)
The WHOIS record will inform you of the Company & Person to whom a Public IP, or block of Public IP addresses is assigned to. It may be hard to determine the use of an IP just based on WHOIS alone. A DNS lookup on the other hand, could provide a clue regarding the use of an IP address, for example, from the command line, the entry "ping -a 64.233.183.104" tells you the FQDN of the IP address as "nf-in-f104.google.com". Sometimes this can give a clue to the use of the IP. In this case it doesn't really tell us much, other than the server may be located in the first floor computer room in room 104, for example. If it response was proxy1.google.com this may give an indication of the use. However, this is just a hostname assignment and does not explicitly determine the exact use of the IP address. You may wish to use some other tools to carry-out an more in depth analysis of the IP address belonging to the server, port scan, OS fingerprinting (use with caution). If the proxy is transparent, you may never know about its existence. § SCF - 17 March 2007
Sign up instructions
Following a recent spree of improper (per WP:U) company-usernames that have been listed in WP:RFCN, there was a short discussion in its talkpage. It seems that the information that no company names are allowed, is one of the very few restrictions of WP:U that are not explicitly listed in the create an account page. A few concerned editors came up with the proposal to add the following restriction below the "names of celebrities":
existing company names (including not-for-profit organizations) and trademarks
We think that this will seriously help in limiting the backlog of improper usernames in RFCN and AIV. Furthermore, the selection of those usernames, may actually be in good faith, since people may be creating accounts from e.g. their office. I'd be really mad if I were in their shoes and found myself listed in RFCN or AIV without a warning from the sign up page. Can we include that in the sign up page please? NikoSilver20:50, 15 March 2007 (UTC)
Hope that helps. --ais523 16:33, 15 March 2007 (UTC)
Converting rtf - or other - to Wikimedia format.
Hey,
I hope this is the right place to seek advice for a technical question. Draft texts relevant for the Wiki will often be written in another format than the Wikimedia, usually doc or rtf. I wonder, and have searched the net without any success, if there exist some kind of software that can convert such files to wikimedia format, so that we keep the formatting (bolds, italic, tables, etc). Thanks Bertilvidet11:02, 15 March 2007 (UTC)
When populating a WikiProject category by placing the project template on article talk pages, is there a way to provide a sort key so that pages are sorted alphabetically by article title rather than by talk page title? For example, category navigation is far easier if "Talk:An artice name" appears in the "A" section and "Talk:Different article name" appears in the "D" section, yet frequently all would be listed in the same huge section ("T" for "Talk"). I've noticed a bit of inconsistency in the way this happens though, as some talk pages do seem to be alphabetized by article title, while others do not. See Category:WikiProject Media, where only a handful are sorted by article title, yet the majority are lumped into the large "T" section. I've peeked backstage at various talk pages sorted differently, but could not account for the differences in the way they are sorted. -Tobogganoggintalk21:54, 14 March 2007 (UTC)
You need the {{PAGENAME}} magic word as a sort key. It removes the namespace, so for this page it produces: Village pump (technical)/Archive AK. I have updated {{WikiProject Media}}, so the contents of the category should change shortly when the job queue gets round to updating the pages. The articles that were sorted beforehand were all people, and they were sorted because {{WPBiography}} contains {{DEFAULTSORT:{{PAGENAME}}}}, which sorts all categories on a page with no sortkey. mattbr23:37, 14 March 2007 (UTC)
Some technical help on User Warning Templates appreciated
We just discovered that (woops) the uw templates were broken when used a certain way. We use the subst=subst: trick to allow the editor to remove all the logical statements used in the template. {{subst:uw-test1|Article|subst=subst:}} and {{subst:uw-test1}} work great, but {{subst:uw-test1|subst=subst:}} is completely broken (as if the else part of the if couln't work). We have this issue on all our templates. You can see the problems and play with the template at User:Lucasbfr/Sandbox.
This seems to be the ParserFunction parameter bug (bugzilla:5678) where {{{1}}} in a substituted ParserFunction sometimes refers to the first parameter of that ParserFunction rather than the first parameter of the template; see m:ParserFunctions/5678 for the known details about this bug. Can anyone figure out how to reword the template to work without triggering bug 5678? --ais523 10:53, 14 March 2007 (UTC)
(Note that in the 5678 examples grid, giving a parameter default sometimes seems to help, but it doesn't seem to have helped here.) --ais523 10:55, 14 March 2007 (UTC)
Sadly, the bug can't be avoided, from what I've tested in the past. Either use both "subst" and the article name, or nothing at all. Unless someone can find a way to get around it. GracenotesT § 16:59, 14 March 2007 (UTC)
I am writing an artical and have added a bunch of footnotes. However, although the footnote number appears where I place the footnote, the "notes" section at the bottom of my artical does not contain any notes. What am I doing wrong? (I am using the format [1]). The artical is currently on my user page user:vzmetzger. Please help. Vzmetzger19:22, 14 March 2007 (UTC)
You needed {{reflist}} at the point where you want the footnotes to go. I've added it for you. See {{reflist}} for more information and options.-gadfium19:39, 14 March 2007 (UTC)
Commons compatibility?
Working with various pictures from Ohio, I came across Image:OH_31_and_OH_38_West_View.jpg (I'd link to this if I remembered the right coding!), which was user-created PD, but which was not on Commons. Therefore, I uploaded it to Commons under that name. Now I'm wondering: what are the effects of having two different files under the same name this way? Thanks! Nyttend18:26, 14 March 2007 (UTC)
Last week, my laptop's power cord failed, so I did little editing over the weekend. Since I got a new cord yesterday, I've had trouble editing: every time that I edit a page, it tells me that the page does not exist. For example: a little while ago, I added a picture to Goshen Township, Hardin County, Ohio. As soon as I hit "Save page", instead of sending me back to that page, it sends me to http://en.wiki.x.io/w/index.php?title=Goshen_Township%2C_Hardin_County%2C_Ohio&action=submit , which is a "This website cannot be found...", entitled HTTP 400 Bad Request. Moreover, since I just took the picture yesterday, I had to upload it, and I got the same "cannot be found" page from the Commons. Has something odd happened with the servers, or is this (far more likely!) an effect of the power problems? Thanks! Nyttend16:08, 14 March 2007 (UTC)
There have been a few of those errors recently, but not as many now; so server delays could be a possibiliy. There may also be a problem with your internet connection. Acalamari18:29, 14 March 2007 (UTC)
Identifying broken links
I can't seem to find a template with which to tag broken (external) links for review. I'd swear I'd found one before through either Wikipedia:Template messages or a judicious use of Special:Prefixindex in the Template: namespace, but the closest I found in my most recent attempt is {{Failed verification}}, which isn't quite the same thing. Is there something like a {{brokenlink}} somewhere? ~ Jeff Q(talk)15:18, 14 March 2007 (UTC)
It seems like we should have a template for this, for folks who want to note a problem but don't have the time or knowledge to fix it. I'm experimenting with q:Template:Deadlink over at Wikiquote, for those who are interested. ~ Jeff Q(talk)17:46, 14 March 2007 (UTC)
Radiant: in this case, you'd want to do {{subst:drv top|[[Article name]]|result and reason}}. Your sig should be added automatically. GracenotesT § 16:57, 14 March 2007 (UTC)
Yes, but for people who work on more than one process, it would be nice if all those templates worked in the same fashion. >Radiant<17:06, 14 March 2007 (UTC)
Deletion Review is quite different from XFD in structure, so I would expect the same about the template, right? Unless I'm misinterpreting your statement. GracenotesT § 19:03, 14 March 2007 (UTC)
Tables not being sorted
For some reason the tables in WP no longer seem to obey the instruction class="wikitable sortable". Is this a known problem? JavaScript is not disabled on my system.
It must be something to do with my WP Preferences, because the problem disappears when I log out. I don't remember changing anything in the Prefs, however: any suggestions? I've tried bypassing my cache, but that didn't help.
Have you made changes to your monobook.js page recently (I see the last was Feb 25)? Try undoing the changes or force reloading of pages per the instructions at the top of User:Ndsg/monobook.js. --MECU≈talk14:36, 13 March 2007 (UTC)
Hey, I remember that at one time, articles like iPod had the first letter capitalized due to technical restrictions. I must not have been here the day that changed. When and how was it made possible for articles to begin with lowercase letters? It's certainly an improvement. Thanks. Paul Haymon05:24, 14 March 2007 (UTC)
The "technical restrictions" title fix was first discussed on this village pump page and then implemented on November 24, 2006. It uses JavaScript to hide the redirect and display the proper title. The code is viewable here. --MZMcBride05:54, 14 March 2007 (UTC)
Near the beginning, there is a long stream of caps words. I don't want to just delete it myself, I want an admin to see it so they can punish the user.— Preceding unsigned comment added by Glutted (talk • contribs)
Hi. I seem to have enabled some features such as having pop-up boxes (well, more like mouseovers really) when I highlight links and a feature in editing that's hard to describe ... when I select text, the text appears above the editing box. Anyway, I can't recall how I turned these features on, and as I'm finding the latter feature very annoying (as it interrupts my selection), I'd like to turn it off. Of course, I can't figure out how. Can anyone assist? Many thanks. --CPAScott03:41, 14 March 2007 (UTC)
Question about the capitalization of search terms.
Greetings.
I'm currently in the process of compiling and posting material for the University of Arkansas - Fort Smith (I'm the University editor). I've discovered that if I type "University of Arkansas - Fort Smith," I'm taken directly to the page. If I don't use capital letters, however, I get a page of completely unrelated results.
Please sign your comments on talk pages using: ~~~~. The reason is: Because capitalization matters. You can create redirects to the article. Also, since you claim to be the University editor, please see the policy about neutral point of view and conflict of interest. Thanks.
It is because if there is not page as you type it, it goes to our search engine which has a lot to be desired. HighInBC(Need help? Ask me)23:47, 13 March 2007 (UTC)
How to print "How to Edit a Page"
Printed versions of this section cut off most of the example boxes on the right of the page. It would be much easier to paster Wiki markup if one could consult the examples from a printed copy, but I have yet to find a way of producing one.
Ozmirage20:34, 13 March 2007 (UTC)
This is a long-standing bug of Internet Explorer. Use just about any other browser to view and print and it will look fine. —EncMstr21:33, 13 March 2007 (UTC)
Internet explorer is blue. To the left is 1994, to the right 2006. Well since IE is still the most popular browser that is a problem that should probably be fixed on our end, if possible. The printable version looks fine to me in IE though. Prodegotalk21:45, 13 March 2007 (UTC)
I just did a Print Preview of Wikipedia:How to edit a page in IE7 with the default settings and it worked fine. Perhaps you need to upgrade to IE7. We don't need to expend resources to make a fix for IE<7 if the vast majority of users experiencing this problem can just upgrade to IE7. Better yet, users could just use Firefox, which uses the standards-compliant Gecko layout engine to make sure pages are displayed correctly. —Remember the dot(talk)22:17, 13 March 2007 (UTC)
Lack of Quality assurance -- SVG Image scaling problem
I increasingly come across images, perhaps especially maps, which do not fill in. The one thing they all seem to have in common is they are SVG format whatever that is. Suggest switching to good old png, which is evidently more modern. On the following I can but see the top 15-20% of the image, and the rest is blank. This seems to be a typical manifestation of the problem, and is happening on three different computers, and three browsers: IE6, IE7, and Firefox 2.x. It's really annoying to have to click and enlarge an image to see what it's general gist is trying to communicated. Please fix this soon. I love browsing Wikipedia's History and geography articles, but this steals much of the pleasure.
SVG is a vector-graphics format that scales a lot better than PNG does! I'm not seeing the problem, and the .svg images are sent to your image in .png format anyway (right-click on them if you want to check this), so the most likely possibility is that the image server's thumbnailer has got temporarily confused and some duff pictures are stuck in your cache. Try bypassing your cache and report back if the problem persists past that. --ais523 17:22, 12 March 2007 (UTC)
We are trying to diagnose the problem; of course readers are not supposed to have to do it. Whether or doing that fixes it for that image can help us find out where the problem is. --cesarb20:03, 12 March 2007 (UTC)
I see the problem right now (on Firefox 2 for Mac). I've never loaded the page before, so it doesn't seem to be a caching problem. I agree that SVGs are superior to PNGs for maps, but there's something broken about the way WP is rendering SVGs as PNGs. rspeer / ɹəədsɹ 17:47, 12 March 2007 (UTC)
Even if you never loaded the page before, the Wikimedia servers have a shared caching layer, which could have cached a broken image even after the problem cleared (or sometimes the problem happens in the caching layer itself). Doing a forced reload (Shift-Reload or similar) sends a header which tells the squid servers to fetch the data directly from the origin server, instead of returning a cached response. --cesarb20:03, 12 March 2007 (UTC)
It has to do with Imagemagick's limited svg support, it does go faulty when scaling some svg features. I think this could be avoided if the mediawiki software simply used image magick to render the svg at it's default scale as a PNG, then scale that PNG. This is an issue for the developers. HighInBC(Need help? Ask me)18:13, 12 March 2007 (UTC)
No, Wikipedia doesn't use Imagemagick to render svg images, it uses rsvg. The problem is probably not with it; the servers have been acting up lately. --cesarb19:56, 12 March 2007 (UTC)
I am not experiencing any problems with viewing SVGs. Can you be more specific about what the problem is?
And yes, the SVG format is much superior to PNG in a number of ways. However, Microsoft chose to use their own proprietary vector graphics format, VML, instead of the now-industry-standard SVG format. Thus, Internet Explorer has little-used VML support and no built-in SVG support.
While we're on the topic, why doesn't MediaWiki just send out the SVG (instead of converting it to PNG) if the user-agent is Firefox? —Remember the dot (t) 19:42, 12 March 2007 (UTC)
For security reasons, IIRC. To be able to send the svg directly to the browser, it would need to first sanitize it (like is done with HTML tags). --cesarb19:56, 12 March 2007 (UTC)
I'm sorry to start a new SVG topic, but this one seems to be different. The problem is that the thumbnails at 1 − 2 + 3 − 4 + · · ·sometimes do not display at all. I have never witnessed them to display partially, as described above; what happens is that the thumbnail box has the appropriate dimensions, but there is nothing inside: no HTML broken image indicator, no alternate text, just whiteness.
I created the images in Illustrator and uploaded them to Commons this month. It may be significant that I did the same thing on 0.999..., and it went on the Main Page in October 2006, but I don't recall anyone reporting similar problems then.
I've tried all the tricks at Wikipedia:Purge, both here and at Commons, and I've tried uploading tweaked versions of the images. Doing these steps enough times and in the right order finally made the thumbnails show up consistently for me, but it wasn't straightforward or easy, and it apparently didn't fix the problem for other people.
(Oh, and I've edited that article to use PNG versions of the illustrations; the last version to include the SVGs is here. Melchoir18:40, 13 March 2007 (UTC)
How to make columns
I wrote the Wikipedia article Jerome's De Viris Illustribus. In Libri I it is a list of 134 authors. I would like to make a "Table" spreadsheet so that it consists of 4 columns wide (with no showing lines between them). This way then it will be 4 columns wide and 34 rows deep (perhaps easier to navigate). I would then like to STILL have it that each name can be edited for the future. In other words, to be able to change or edit the name through the normal way of "Editing" modifications for the future. --Dougtalk13:36, 13 March 2007 (UTC)
The following template is the "start" template that is then used with several other templates: {{col-begin}}. There are detailed instructions on how to use these templates. Let me know if you have any other questions.↔NMajdan•talk13:42, 13 March 2007 (UTC)
Just happened to notice that the image for the tireless contributor barnstar seems to be acting up, and fixing it is way outside of my scope. If someone can take a look at it, or figure out who should....the link is TC Barnstar. Risker21:34, 12 March 2007 (UTC)
Hitting the "stop" button on your browser before a page has loaded also makes the Tireless Contributor Barnstar play up (actually, it just stops). Acalamari18:14, 13 March 2007 (UTC)
Granting specific IPs or new accounts ability to edit semi-protected pages?
I'm wondering if it's possible, maybe with some modification of MediaWiki or something, to make it so that admins can grant certain IP addresses or new accounts, which have demonstrated good editing and no abuse of Wikipedia, ability to edit semi-protected pages. 69.168.140.188, for example, has made numerous excellent contributions to various articles such as World Trade Center which has at times needed to be semi-protected. 172.132.67.217 also expressed interest in making good edits to the article, which I would like to allow, without having to endure extensive vandalism originating from other IPs. I'm not sure if this has been considered at all? --Aude (talk) 21:54, 11 March 2007 (UTC)
It is certainly possible with new users, as "autoconfirm" is just a flag, it would just need an interfaced programed in, and of course consensus for such a new feature. Giving it to IPs may be prevented by the layout of the code, and it must be remembered that IPs change owners. Not sure if there is enough demand for it though. HighInBC(Need help? Ask me)21:58, 11 March 2007 (UTC)
While this might be an option to consider adding into the MediaWiki software, it would be a bad idea to grant certain IPs special permissions on public projects like Wikipedia. The problem is that an IP may at any time be reassigned to another person. For example, an IP that for years was used by a good-faith editor could be reassigned to a vandal without warning. It's far better to just create an account and have your own personal history of contributions. —Remember the dot (t) 22:03, 11 March 2007 (UTC)
Since special priviledges should not be given to IPs and the limit for priviledges for registered users is only 4 days, I don't see this as being a problem. What would it achieve besides opening more possibilities for abuse by people who should not have the priviledges? If that can be answered, then this may have some value - otherwise, I can't see it as a good idea. Nihiltres22:16, 11 March 2007 (UTC)
If User talk:69.168.140.188 changes owners and starts vandalizing, then we would remove the flag. It's been the same user since they started editing last April. User_talk:68.39.174.238 is another good, long-time contributor. It would be nice if they created accounts, but for whatever reason, these people are content with editing from IPs. I think the instances of "flagging" IPs or new users would be rare. --Aude (talk) 22:24, 11 March 2007 (UTC)
Semiprotecting pages is a regretful option, as it is anti-wiki and shuts out good IP editors. Unfortunately, some articles are vandalized so much beyond what we tolerate, so are semi-protected. If there is anything we can do to lessen the impact of semi-protection on known, good editors, it would be helpful. --Aude (talk) 22:27, 11 March 2007 (UTC)
Is there a reason why these anons can't set up an account? I really don't see a reason for this. If you want to edit semi-protected pages, set up an account and wait 4 days or request unprotection. Mr.Z-mantalk¢Review!22:30, 11 March 2007 (UTC)
They can just create an account, being anonymous means that we cannot tell if you are on the same person from hour to hour, that is a technical limitation, creating a user avoids those limitations. HighInBC(Need help? Ask me)22:33, 11 March 2007 (UTC)
I don't know why some anons. don't want to create accounts. Indeed would be easy, but for whatever reason some good users edit from static IPs which don't change from hour to hour. --Aude (talk) 22:37, 11 March 2007 (UTC)
Anyway, back to the original question, if anyone is still interested. It is not possible with the current code to give one IP an autoconfirmed (or any kind) flag. It is no IPs or all IPs. Prodegotalk19:02, 12 March 2007 (UTC)
Yes; and if all IPs and new accounts could edit semi-protected pages, semi-protection would become nothing more than a decoration. Also, some IPs are shared; and if an IP that a good contributor and a vandal shared was given the ability to edit semi-protected pages, the good contributor would end up getting blocked as the vandal takes advantage of the situation and vandalizes. Acalamari18:19, 13 March 2007 (UTC)
An icon, such that, when clicked, a new browser window opens
There's an article i'm contributing to, a biorgraphy of a living person. I could not find a free picture of the subject of this article. I'd like, instead, to put in the article an icon of a picture, such that, when clicked, a new browser window would open to a webpage displaying a photograph of this person. How do i accomplish this, please? Itayb19:48, 10 March 2007 (UTC)
And clicking on it just goes to the image page for that camera picture. The link below it jumps to the part of the page where there is in turn an external link to the picture, which is rather convoluted; why not just put the external link there in the first place? Anyway, I don't think any of those links try to open new browser windows (I wouldn't know, since I've got the Mozilla SeaMonkey browser configured to ignore such requests anyway), which is good since opening new browser windows is really annoying; I know how to open a window or tab when I want to. *Dan T.*21:09, 10 March 2007 (UTC)
Yes, i'd deliberated with myself whether i should put the external link there in the first place. I finally decided to separate the icon from the link for two reasons: (1) i wanted there to be a detailed description of the reference, so the link can be more easily fixed in case it ever gets broken; putting the whole reference text as the underscript of that small icon seemed disproportionate, (2) I wanted all the links to be concentrated in a single, easily identified place, rather than starting spreading them all over the place. Itayb22:37, 10 March 2007 (UTC)
Please don't include external links to pictures. If they're free, upload them to commons. If they're not free and can be uploaded under Wikipedia fair use, upload them here. Otherwise, links to such pictures are a bad idea. Plus, are we favoring certain websites over others regarding which picture to link to? Inevitably. GracenotesT § 18:10, 11 March 2007 (UTC)
This really isn't a good idea. If readers want a picture, they can Google for one easily. Linking to a non-free image can be confusing to readers (they might click the picture of the camera instead of the link), and it also discourages the creation and uploading of free images. —Remember the dot (t) 20:35, 11 March 2007 (UTC)
Gracesnotes and Remember the dot, i appreciate your comments; they force me to think more carefully about my editing, and about interesting issues that are relevant to Wikipedia. All the same, i do not see this matter eye to eye with you.
The article under discussion, Azmi Bishara, featured a picture of its subject a couple of days ago. The picture was uploaded to Wikipedia under fair use license. The picture was deleted from Wikipedia (and from the article) a couple of days ago, by the administrator Angr, for being potentially replaceable by an equivalent free picture.([6], [7]) I searched for a free replacement, but my endeavors bore no fruit. It was then that i added an external link to the page displaying the deleted picture.
Gracenotes, you write: "links to such pictures are a bad idea". Could you please explain why you think so? You write "are we favoring certain websites over others regarding which picture to link to? Inevitably." But aren't we still favouring certain wesites over others when we directly embed fair-use pictures in articles? I believe that choosing certain websites and other sources over others is an important part of the work of most Wikipedians. The website i've linked to is not just any website, it is the official site of the Israeli parliament, of which the subject of the article is currently a member.
Remember the dot, you write: "If readers want a picture, they can Google for one easily". But can't readers easily Google for much of the content of that article, as well as the content of many other article in Wikipedia as well? Would you support a policy of putting into Wikipedia only such content that cannot be easily found by performing a search in a popular search engine? In my opinion, the crucial criteria for content inclusion are: (1) Is it relevant to the article? (2) Is the content new, rather than duplicating already existing information? (3) Is the information attributable to a reliable source? (4) Is it free for use? I think the picture of Mr. Bishara satisfies the first three criteria, don't you? If you do, why is a link to the page where it originates inadequate, seeing that the picture itself cannot be used directly?
You write: "Linking to a non-free image can be confusing to readers (they might click the picture of the camera instead of the link)". I agree with you. That's why i started this topic in the first place: i wanted to use an "icon of a picture, such that, when clicked, a new browser window would open to a webpage displaying a photograph of this person". It doesn't have to open a new window. What i have in mind is something similar the icon of a star on featured articles, which, when clicked, links to the list of featured articles. I'd appreciate it, if you could help me figure out how to do this. In the meantime, i think the solution i've come up with: a picture of a camera with a subscript that points to the external link, is a reasonable substitute.
You write: "it also discourages the creation and uploading of free images". I'm afraid i don't see why this should be so. Could you please explain?
The default skin is okay, but I'd like something better. All the other skins are awful and hard to look at. Any way a suggestion can be put in for some new skins or is there any way other skins can be installed? Aaron Bowen13:22, 12 March 2007 (UTC)
m:Help:User style and m:Gallery of user styles should tell you more than you ever needed to know. You can write your own skins using CSS and JavaScript; if you don't feel comfortable doing this, but know the kind of thing you want, suggest it here and it's possible someone might make it for you. Note that many user scripts are designed to work assuming the Monobook skin, so most attempts to modify skins start from there. Hope that helps! --ais523 14:25, 12 March 2007 (UTC)
Go to a random article's history page. Select two revisions which are more than 1 revision apart. Show changes. Now, I expect the 'undo' link to undo the diff that I'm looking at. Instead, it undoes the most recent revision, which has little or nothing to do with the diff that I'm looking at. I accidentally vandalized a page due to this nonsensical behavior. Surely this must be a known bug? Xerxesnine23:40, 11 March 2007 (UTC)
Seems to be working fine now, how long ago was this? This was a problem with page histories a few hours ago, could this have been the cause?--VectorPotentialTalk00:14, 12 March 2007 (UTC)
It has been broken for at least two months (my entire wikipedia career), up to and including the present moment. I confirmed the problem before making my previous post here, as well as just now. Did you follow my instructions exactly? The two selected revisions have to be non-adjacent (more than 1 revision apart). Xerxesnine03:16, 12 March 2007 (UTC)
As far as I know, the undo feature has only ever been able to undo changes one revision deep. If you go to the trouble of viewing the history and through it a multi-revision diff, you can just revert changes manually by clicking the "edit" button for the old version and saving with no changes. It isn't really a problem. Nihiltres05:17, 12 March 2007 (UTC)
Undoing a diff is not the same as a revert. A revert negates all revisions after the selected one, whereas a diff-undo negates only a chunk of revisions (if it worked correctly). There should be no distinction drawn between "multi-revision" diffs and "1-revision" diffs. A diff is a diff is a diff. Clicking "undo" should undo the diff which you are viewing, period. Since that doesn't happen, that's a bug. Xerxesnine06:01, 12 March 2007 (UTC)
The Undo feature is not an outright reversion tool. Iff the edit that you want to undo is not unaffected by any more recent edits, you can undo the edit. You can undo adjacent edits, really old edits, brand new edits... anything, as long as there are no interfering edits between the revision you want to undo and the current revision. Titoxd(?!? - cool stuff)05:22, 12 March 2007 (UTC)
I'm not looking for a reversion tool; see above. The rest of your comment seems to just be explaining what revisions and diffs are. Be assured that I understand the concepts. Perhaps you are suggesting that there may be conflicts in undoing a particular diff? Well, of course there might be, as there might be with undoing an adjacent diff. I don't see how that is relevant. Xerxesnine06:01, 12 March 2007 (UTC)
Still occurring. I seem to be getting the page itself out of the cache, so it looks current. However, it doesn't match with the history or edit page, which show an older version of the file. —Doug Belltalk11:09, 12 March 2007 (UTC)
Hmmm. Now I just had a page where the history/edit was up-to-date, but the page displayed from the cache was old. That one is harder to explain. —Doug Belltalk11:36, 12 March 2007 (UTC)
blue links
Is it possible to have the wikipedia links as underlined? I found this...this and image. The red links work fine as do the visited links, but the blue links still show up as blue with no underline. Can someone help out? W3stfa11/Talk to me06:27, 12 March 2007 (UTC)
Null edits don't seem to be necessary. Once the template is edited, the pages categorized by the template will be updated eventually, typically in 5 minutes to an hour. I've seen it take as long as a day, but it's always happened. Gimmetrow05:34, 12 March 2007 (UTC)
I suspect that this line: "document.getElementsByName("wpSummary")[0].value = "{{Egypt-stub}} => {{Ancient-Egypt-stub}} as per [[WP:SFD]]" is to blame, it's the only thing I can think of--VectorPotentialTalk19:45, 11 March 2007 (UTC)
Edit histories temporarily scrambled?
Causing massive edit conflicts and accidental blankings, page histories disappearing and reappearing? Is there some sort of database problem?--VectorPotentialTalk18:49, 11 March 2007 (UTC)
Something's definitely wrong in that department - returning to the page that I linked to two sections above, I can't even scroll through the page history in order anymore, as it occasionally skips two or three revisions. --TMFLet's Go Mets - Stats18:57, 11 March 2007 (UTC)
I think they better lock the database down before page histories become hopelessly scrambled, there's massive amounts of unintentional blanking going on--VectorPotentialTalk18:59, 11 March 2007 (UTC)
[9]Looks like it's happening everywhere, this is going to be a messy cleanup once the problem is solved--VectorPotentialTalk19:13, 11 March 2007 (UTC)
Agreed ... I was about to block someone for deleting evidence on an arbitration page until I realized what the problem was. I guess we'll all have to take a couple of hours and go out and enjoy the fresh air while they work on it. :) Newyorkbrad19:22, 11 March 2007 (UTC)
This keeps popping up a lot at the moment when there are no new messages. I've flushed the cache on two browsers each on Windows and Macs and quit and restarted the browsers. It still keeps popping up. Any clues to help me please? Candy18:04, 11 March 2007 (UTC)
Try clicking the "last change" to get to the diff. page, that clears the problem for me when it happens. Matthew18:06, 11 March 2007 (UTC)
I am also seeing this. It keeps recycling the last message. In one case, I even brought up the history and the last change was not listed. I've been notified about half a dozen times for the last message I got—each time I view the page to clear the message, and maybe 10 minutes later I get notified again. —Doug Belltalk18:45, 11 March 2007 (UTC)
I don't think so...I'm not an IP. Also, the message bar goes away for a while, but then the flag seems to get reset without a new message. I think the issues listed above and below this one may be related though. —Doug Belltalk18:50, 11 March 2007 (UTC)
If I had to guess, I'd say somewhere in the cluster is at least one computer that got confused by the daylight savings time switch in the US and is now not understanding timestamps correctly for the purposes of handling caching of pages and histories and events like determining whether a message is "new". Dragons flight18:58, 11 March 2007 (UTC)
It has to do with the same problem that is causing edit conflicts not to be detected, see the thread directly above this one. HighInBC(Need help? Ask me)19:18, 11 March 2007 (UTC)
This is getting strange
I added comments to a few threads on WP:VPR but something very strange happened. Only my last comment still exists and all of the comments I made on the two threads at the end, along with the other people's comments, and the last thread have dissappeared entirely. Also, the history keeps changing. Sometimes it shows my edits, sometimes it doesn't. On one diff, [10] where I only edited one section, it shows that I deleted the section, part of the section above it, and part of another section somehow. What the hell is going on? Mr.Z-mantalk¢Review!18:01, 11 March 2007 (UTC)
It is happening to everyone, use the preview button to check if your post will wonk the page first for now. I hope they get this fixed up soon, a hint to the problem is that the history page does not show to the most recent version always(currently). HighInBC(Need help? Ask me)19:17, 11 March 2007 (UTC)
Daylight Savings Time
Due to the idiocy of Congress, Daylight Savings Time started today, 3 weeks early - but the Wikipedia clock didn't adjust for it. Is it going to adjust automatically on April 1? Don't want a double loss of hour. Is there a plan or do we just have to do the offset from UTC manually? Tvoz| talk07:53, 11 March 2007 (UTC)
I don't understand. What clock are you talking about? The timestamps are in UTC, to which Daylight Saving Time does not apply. (By the way, I don't like DST either, but that seems to be beside the point here.) --Trovatore07:58, 11 March 2007 (UTC)
Right, UTC is the official time on talk pages but watchlist and history and diffs show local time, based on the offset. I don't recall manually changing the offset when Standard time began, but I could be remembering wrong - I thought the offset was reset automatically. If not automatic, that's ok but people might need to be told that they have to change the offset now if they want it to reflect DST. And what I'm asking is if the offset is programmed to automatically change first week of April, in which case it will have to be manually overridden again. Hope that's clearer. (And right, it is beside the point that I don't like Daylight Savings, but I don't think we have to be so NPOV here - this is hardly an article...) Tvoz| talk08:18, 11 March 2007 (UTC)
Ok, I changed mine manually (grumbling the whole time), and I guesss I'll keep an eye out in April - but it's something that perhaps should be publicizd. As I say, I don't remember manually changing the offset in October, but I could be wrong about that. Tvoz| talk08:59, 11 March 2007 (UTC)
Wikipedia doesn't distinguish between "time zones" per se, it just offsets the time from UTC. It has no understanding of whether it's CST or EST, and doesn't worry about daylight savings. Harryboyles13:49, 11 March 2007 (UTC)
Hello, once I log into my account, I am unable to see the side menus (gray box with table of contents) of any wikipedia page, including my that of my userpage. However, once I log off, I am able to see them again. This has only began to occur during the past week, and I don't want to edit wikipedia using only my IP address, so what can I do to see the menus again? Count de Chagny05:04, 11 March 2007 (UTC)
I went to have a look at the Wikipedia Statistics "Charts and Table" section, and I found that it has not been updated since October. Who updates it and how - I would be happy to assist, but I feel that someone should certainly be doing it. David Spart15:03, 11 March 2007 (UTC)
Userboxen Gone Wild! ($14.99 on DVD)
My userboxes have gone a little crazy, showing up at crazy angles, and, in one case, apparently creating a fourth column for themselves out of boredom, malice, or a combination thereof. Is there any way these heathen monstrosities can be corralled into doing my bidding? Jouster (whisper) 11:38, 11 March 2007 (UTC)
That seemed to corral most, though the Kazoo is showing as on a roadtrip in its own column. Just my useragent? Jouster (whisper)
Editintro bug?
I'm pretty sure this is a bug (as opposed to a feature) -- compare the content below the edit summary box on these two pages: [11][12]. Is there a reason for the seemingly random <p> tags in the first link? --- RockMFR19:40, 10 March 2007 (UTC)
That is weird. It happens on the main create page thing on wikinews too, which I knw worked before. Bawolff08:31, 11 March 2007 (UTC)
It can go either way, depending on how many you put in, and the level of incertitude you have about the text in question. So if you put several {{fact}} tags, or if the sentence seems unlikely, consider it a regular "major" edit. − Twas Now ( talk • contribs • e-mail ) 06:11, 5 March 2007 (UTC)
I don't follow Twas Now's reasoning, I'm afraid. Adding {{fact}} is surely never a minor edit; it involves adding something substantial which those editing the article should be aware of. --Mel Etitis (Talk) 10:52, 11 March 2007 (UTC)
Werdnabot has now been blocked for several days following a malfunction. Werdna does not appear to be around to deal with the problem. It may be worth considering switching the archiving to MiszaBot II (for project talk pages and noticeboards) and MiszaBot III (for user talk pages), especially if talkpages are getting very full. To have either of those Bots handle archiving, make a request at User talk:Misza13, including the following information:
Page to be archived =
Current archive =
Age threads should reach before being archived =
Max size of archive before new archive is started =
I am using DynamicPageList on nl.wikibooks to show recipees which use a certain ingredient on the page of this ingredient. DynamicPageList makes a list out of the content of a category. If you use this categories for a certain recipee, you will have a big category-list under the recipee. The categories used for the DynamicPageList all start with KB-xxxxx. Is there a possibility with CSS to not display these categories on the recipee-article? I have found out that you can turn of categories all together with
#catlinks {display:none;}
So my question: is it possible to only stop displaying categories starting with "KB-". And if so, what is the code you should use. Thanks for an answer. I expect CSS-specialists will read this part of the village pump and might be able to help me. Londenp23:34, 6 March 2007 (UTC)
I guess what you'd want is something like:
.recipeebox li a[title^="Category:KB-"] {display:none;}
Thanks for guessing what I mean. I understand it is not easy. But I think I mean something else.
This is the page with the DynamicPageList: nl:b:Kookboek/Tomaat#Recepten_met_tomaat and this is a recipee where certain KB-categories are shown: nl:b:Kookboek/Pasta_met_budgetbolognese. You see how many KB- categories are displayed? The only category I would like to keep shown are: Pastarecept, Budgetrecept and Italiaanse keuken. So I think I would need a selector in combination with #catlinks. It might be quite easy, but I don't know that much about CSS and we don't have many contributors on Wikibooks-nl with this knowledge. That is why I ask here. Thanks for all the help! If it works I will put it in the Common.css. Londenp21:42, 7 March 2007 (UTC)
Note that it looks rather ugly, makes a bunch of vertical pipes. Categories aren't meant to be hidden in this way. --Splarka (rant) 08:29, 9 March 2007 (UTC)
I'm going to assume you're using Firefox and that you right-clicked and hit "Block images from...". If so, you can go into your preferences (Tools->Options on Windows or Firefox->Preferences on Mac OS X), click the "Content" tab, click the "Exceptions" button next to the "Load images automatically" option, and click "Remove site" when Wikipedia is selected in the list. —bbatsell¿?✍19:45, 16 March 2007 (UTC)
Article Traffic
I'm curious if i can find out how many visits a single wikipedia article page gets in a month. Is it possible? can i find out somehow how many times people are looking at an article?--24.184.167.10018:38, 9 March 2007 (UTC)
Doing that with any reasonable amount of accuracy would place far too much load on Wikimedia's servers; the feature exists, but has been deliberately disabled. However, there's some links on Special:Statistics which estimate the read levels for some of the most popular articles (less popular articles don't register at all on it). --ais523 18:44, 9 March 2007 (UTC)
If newer accounts (including IPs) make many edits in a row quickly, they have to fill in a captcha. You created an account about one or two days ago, so you're not autoconfirmed -- a status automatically given to people that have accounts for 4 days or more, or more accurately, 345600 seconds (as the server variable is in seconds). Sorry for the inconvenience! GracenotesT § 16:15, 16 March 2007 (UTC)
Must be, because it happened twice again , I added a link to Santa Paula, California and it asked for a captcha when I reported an ip vandal it did it again. I think I can see the rationale for the links but not reporting a vandal. It is a little frustrating.--Paloma Walker21:43, 16 March 2007 (UTC)
Well, template {{IPvandal}} does incorporate multiple external links, and yes, it is frustrating, and probably unnecessary, the good news is that this is yet another special privilege granted to users with the autoconfirmed flag, which means it goes away five days after you register your account--VectorPotentialTalk23:37, 16 March 2007 (UTC)
How/When will an update of the interwiki map occur?
Hey guys, I was wondering when the latest version of the Interwiki map will be rolled out to the Wikipedia, Wikibooks, etc. databases, or even if there's someone here who might be able to speedy this process along. It's been a few months (I believe) since we changed the address of our wiki from .net to .org within the interwiki map, and we've been hoping that the change would soon be reflected here so we won't have to worry about the various server maintenance kinks on our end that seemingly always manage to break the redirect when we're not looking. I don't mind waiting, but it'd be great to know when we can expect this update to take place if it's possible to find out that information. Thanks! echelontalk06:24, 18 March 2007 (UTC)
Idea for alleviating vandalism
I have an odd idea for alleviating some of the pressure of vandalism on the quality of wikipedia. Would it be possible, once an anon IP were blocked for vandalism, to put a cookie on the anon's computer which disallowed further editing from that computer without signing up for an account? Obviously the more computer literate and determined vandals would figure it out and delete the cookie, but I wonder if it wouldn't make a difference? --Bmk03:04, 18 March 2007 (UTC)
Well, in all likelihood people would not be happy with an eternal cookie. However, the real problem is that it would require the user to attempt to edit after being blocked. Since only a few do (which triggers the autoblocker anyway), it is not all that effective. And soon it would be all over the internet on to delete your cookies. I would say it is not worth it. Prodegotalk03:08, 18 March 2007 (UTC)
Fair points - I suppose it would only help for repeat offenders, which, if you are correct, are few. --Bmk03:25, 18 March 2007 (UTC)
Stop scripts from running in IE
Hi, Does anybody know how to stop scripts in my monobook.js from running in Internet Explorer. I am using Firefox and when I use IE7, the scripts don't work and sometimes error message popup saying there is a script error. So is there a way to make scripts only work on Firefox but disables when IE is running on Wikipedia? -- Hdt83 | Talk/Chat23:47, 17 March 2007 (UTC)
You can put a check for the browser around the inclusion of the scripts that don't work in IE:
if (navigator.userAgent.match(/MSIE/)) {
importScript('Example.js');
}
I got this message in one article after posting a reference (scroll down until you see a red text). What does it mean and how do we I get rid of, without removing the reference? The colour of the error message was red, but when I tried to post it here, it turns black and puts a reference number after it. It also blocks any text written beneath it. --Thus Spake Anittas22:23, 17 March 2007 (UTC)
Okay, well; I don't know in what way I am using the source wrong. Can someone point it out to me? This is not the first time I'm posting sources. --Thus Spake Anittas22:57, 17 March 2007 (UTC)
I have created the template {{sort}} according to current recommended practice at Help:Sorting, and used it at List of U.S. states by population. What does everyone think? I would like to see this used in all tables where the column should be sorted by something other than the visible text (persons' names, numbers of varying magnitude, etc). The template should probably be protected in that case, I already set it up with the /doc system in anticipation of this. --Random83203:55, 17 March 2007 (UTC)
One disadvantage springs to mind: consider a table like in Wikipedia:WikiProject Oregon/Admin where there are more than 2500 rows. If the {{sort}} template appeared on every line once, well there would be a lot of template expansion. Would it be that bad? I don't know. Judging by the new statistics in periodic table (Pre-expand include size: 102013 bytes, Post-expand include size: 91248 bytes, Template argument size: 25690 bytes, Maximum: 2048000 bytes) it looks like previous problems with template expansion size have been addressed, and it shouldn't bother the servers any more than not having/using the template. I really like the idea of separating the mechanism for sort order from the apparent text, as the tricks to do it are ugly, not to mention tedious. Sort isn't a bad name either, though I toyed with variations involving key and index. Great idea. —EncMstr04:27, 17 March 2007 (UTC)
Are there any monster tables like that in article space? It seems to me a table that large may be awkward anyway - the one you linked takes ridiculously long to run sort on. --Random83206:05, 17 March 2007 (UTC)
Not that I know of. That table is an experiment to see how useful it is for managing the wikiproject. The sort speed is dependent on the speed of your computer, so maybe you need to feed the squirrels in it more often. :-) —EncMstr06:43, 17 March 2007 (UTC)
Kudos on a good new tool in the toolbox, but it needs categorized... special effects templates of some kind (per Mike Peel or David Kernow's reorganization of template categories), and your /doc page wasn't quite standard, which I just fixed up so it's close enough for the moment--but substing that new {{template doc}} template really adds to the pre-expand size, so the pretty stuff needs put in the doc page. That at least now gets categorized as a template documentation page. // FrankB20:15, 17 March 2007 (UTC)
The point of the template is to make tables that use it more easily editable (less extra markup) and to allow it to be changed when/if table sorting provides for a better way to specify the sort key. Substing it would be against the purpose of the template. WP:PERF is relevant here I think. --Random83221:04, 17 March 2007 (UTC)
Purge squids on article lock
I suggest that when an article is locked, all squids purge their cache of the article. A vandal recently edited Sinbad to say that he was deceased, and then started sending out e-mails. As the e-mails spread, more and more people visited the article. It was locked and the vandalism reverted, but the people who clicked on the links in their inboxes still saw the text saying he was deceased, both on the disambiguation page and in the article itself, due to the caches retaining the vandalized versions. To those who do not know, if you are logged in everything (or at least, almost everything) is generated dynamically for you, but non-logged in users see cached content. A more general solution to this problem is to purge all caches of an article every time that article is reverted to a previous state. This will be a good step towards stopping readers of our encyclopedia from seeing any vandalism, as we tend to revert it quickly. Cheers, --Alterego20:04, 16 March 2007 (UTC)
An article is purged from squids when locked -- that has always been the case.
People following links to old versions, however, will see the old version because the link is to the old version. --brion15:28, 17 March 2007 (UTC)
It wasn't a link to an old version. I personally saw the old content, and when I logged in it switched to the new content. --128.138.114.15121:14, 17 March 2007 (UTC)
You could require that when the parameter is used, it places the article in a certain category. Something like: {{#if:{{{2|}}}|{{{2}}}<includeonly>[[Category:Articles that use parameter 2]]</includeonly>}} Cheers. --MZMcBride00:38, 18 March 2007 (UTC)
That's an interesting idea, but is it allowed? This would be a category, albeit temporary, that should not be on the articles. --NE200:46, 18 March 2007 (UTC)
Here's what you do. Copy the codes the MZMcBride gave you to the template, create the category page, wait for it to fill up (this may take anywhere from a couple of minutes to a day), do what you need to do, remove the code from the template, and put {{db-author}} on the category. I'm doing the same with Category:Articles with incorrect school infobox format, and it has helped tremendously :) GracenotesT § 02:21, 18 March 2007 (UTC)
Might be easier and less intrusive to use Whatlinkshere... {{#if:{{{2|}}}|{{{2}}}<includeonly><span style="display:none;">[[User:NE2/temp]] <!-- temporary to find template parameter usage --></span></includeonly>}} and then watch Special:Whatlinkshere/User:NE2/temp for population (still relies on the job queue like Categories). --Splarka (rant) 07:24, 18 March 2007 (UTC)
Wouldn't it be better to use {{#if:{{{2|}}}|{{{2}}}<includeonly>{{User:NE2/temp}}<!-- temporary to find template parameter usage --></includeonly>}}, and make User:NE2/temp a blank page? It's the trick used by Werdnabot, and does not need the equivalent of hiddenStructure. --cesarb15:22, 18 March 2007 (UTC)
My user page has gone very strange today. Not a problem, as I can fix it and improve the layout, but I a) wondered why, b) wondered if the cause may have caused other problems. RichFarmbrough, 14:19 16 March2007 (GMT).
I did a preview with {{babel}} removed and it looked pretty good. The babel template hasn't had a significant change since January 27, nor have any of the {{user en}}, etc. templates. You have a stray </div> after babel, but that doesn't seem to affect the goofiness. Dunno. —EncMstr07:03, 17 March 2007 (UTC)
I've reverted the changes made on March 15 to the template and your page now appears fine. Then I reverted back with a change to see if that would work (it didn't). Drini removed links to non-existant categories from many of the babel templates and that seems to be the catalyst for the change. It appears to have affected everyone's pages. I've left a note on Drini's talk page. Harryboyles12:28, 18 March 2007 (UTC)
Im trying to copy and past a .gif file into a powerpoint and its not working. Ive also tried saving the file and inserting it. Please advise.--68.42.172.10303:27, 19 March 2007 (UTC)
What version of Powerpoint are you using? A bit of Google searching implies that there is not consistent handling of animated GIF files in different versions of Powerpoint. If you are using Powerpoint 2003, this page may be helpful. Mike Dillon04:46, 19 March 2007 (UTC)
Pink box on "archived version" pages displays on "current version" too
The pink box displayed when you are viewing an older version of a page also appears when you are viewing the current version (if you pick the last version from the history page), such as this [13].
This really isn't appropriate to display and may be confusing to some people when they are in fact looking at the current version. What can be done with monobook.css, common.js or one of the other pages (or elsewhere) to make this warning not display on the current version of pages? --Aude (talk) 21:13, 18 March 2007 (UTC)
Thanks for the pointer. Good to know I'm not the only one noticing this and having concerns about the pink box. --Aude (talk) 21:22, 18 March 2007 (UTC)
I noticed this while I was browsing talk:Usenet - when there are two identical headings on a page, clicking any of them in the contents list at the top will cause the browser to jump to the first instance of the heading (on the Usenet page there are two comments entitled "Google Groups"). I'm assuming this happens in all cases, but I haven't been able to test it... maybe I will try it out on my user page.
I also don't know if this is a Wikipedia specific issue or whether it applies to all MediaWiki sites - I'm guessing it's more likely to be the latter, in which case there's not much anybody can do about it except tell them. --Tacetus 18:56, 18 March 2007 (UTC)
I tried it out at user:Tacetus/test and I can't figure out what on earth is going on... It seems to be random - I dunno. Meh. --Tacetus 19:03, 18 March 2007 (UTC)
Oh. I added some more headings and I think it works after all. Weird.... --Tacetus
But that one on the talk:Usenet page is definitely broken... maybe only on talk pages? Let me just go check that... --Tacetus 19:06, 18 March 2007 (UTC)
I think I know what's wrong. Normally if the headers are the same, wiki software should direct you to Header_2. The thing is there is only one Google groups in that discussion. The other is Google Groups. Wikisoftware doesn't therefore recognise them as the same and it doesn't link you to _2. If you try it in the sandbox and change them both to Google groups or Google Groups or whatever, it should be fine (it is still in Wikipedia talk:Sandbox at this time in fact). In the case of Talk:Usenet, my browser, Firefox 2.0 also doesn't recognise them as the same so it works fine. You browser I presume isn't properly case sensitive so does think they're the same and has a problem. The only think I can recommend is a newer or different browser. BTW, this sort of discussion is probably more appropriate for perhaps Wikipedia:Village pump (technical)Nil Einne19:13, 18 March 2007 (UTC)
(edit conflict x2) Please report this WP:VPT instead of here in the future. The problem is that if 2 or more headings are the same except for the capitalization ("test" and "Test", for example), the TOC with take you to the first one no matter which one you click. If the headers are exactly identical, it will take you to the right header. Test it at User:Llama man/Sandbox. I'm using IE 7 and this problem is present. –Llama mantalkcontribs19:16, 18 March 2007 (UTC)
Forgot to say - I use IE7 as well, maybe that's where the problem is. --Tacetus
I just noticed that the multicolumn markers ({{MultiCol}}, {{ColBreak}} and {{EndMultiCol}} don't seem to be working anymore. Is that an error or have they been replaced by another method? Robotman197419:19, 18 March 2007 (UTC)
Unfortunately, the template hardcoded the image size at 300px. I suggest creating an optional size parameter, defaulting to 300 to keep compatibility. -- ReyBrujo17:59, 18 March 2007 (UTC)
I added a "image size" parameter to the template, with the default staying at 300px. I also set the image size on the article to 200px. Cheers. --MZMcBride23:11, 18 March 2007 (UTC)
Images appearing higher than they should be and covering text/section titles.
For some reason images when loaded in my browser are hovering higher on the page than they should be. An example of what I mean. I looked at the links at the top of this page but found nothing that covered what I'm talking about. Can anyone help me fix this? It's really annoying. Veesicle (Talk) (Contribs) 17:32, 18 March 2007 (UTC)
Hi Veesicle. You're using my script, right? This image problem is related to the script. I haven't diagnosed this problem yet, but if anyone reading this knows how to correct such a problem, related to programmatically inserting elements into the DOM, please do let us know. The image position is not always too high, and will correct itself (in Firefox) when one interacts with the page, usually. –Outriggr§21:13, 18 March 2007 (UTC)
Ohhh, I see :) I'm such a dumbass, it didn't occur to me that it was related to the script. I guess I'll just put up with it, the script is far too useful to me for it to be worth getting rid of for the sake of things being neat. Thanks anyway. I'll just wait for a bugfix. Veesicle (Talk) (Contribs) 00:03, 19 March 2007 (UTC)
Preventing a page from being created
I've just deleted a page with an offensive page name. Is there any way of stopping that page being created? ( it's been recreated a numbrer of times already) Theresa Knott | Taste the Korn09:59, 18 March 2007 (UTC)
I just got the strangest result whenever I typed in a response to a post on my talk page. After I typed the response which was "Yes, he was indefblocked.", the finished edit was "me" redirecting my talk page to my userpage despite me not typing that at all. [14] Can anyone explain what happened. — Moe23:30, 19 March 2007 (UTC)
Sandbot is constantly resetting a template change that needs a thorough test before updating the actual templates. An email last week (plus latest link diff) to user:AllyUnion has brought no clarity (answer) on why things disappear while work is in progress. The logs show a manual request has been placed, which flys in the face of the {{Inuse}} template, and the short time interval involved. Is there any way to block or shut this thing down for a while? OR AS I'M CURRENTLY SUSPECING... Is it wiping templates when the request is for the editing trial pages--a sort of shotgun effect, as it were? It really seems to not be running on a schedule, but kicks you in the gut when you're least expecting it. // FrankB21:02, 19 March 2007 (UTC)
{{X5}} is a sandbox, which is for short-lived experiments. If someone is repeatedly disturbing your editing there (the resets seem to be manual; someone is following the link given in Sandbot's edit message), perhaps you would prefer to experiment in your own userspace? E.g. at User:Fabartus/Sandbox? -- Eugène van der Pijll22:09, 19 March 2007 (UTC)
The point is some things need to be in template space, which is why X0-X9 had been established. These aren't experiments, but consolidations of seven templates down to one or two, and long overdue. I don't think anyone is directly asking for that page (or last week, pages) to be wiped, but that it's a side-effect. The use is short term... this is happening sometimes within 20 minutes of the last change... a ridiculously short time. // FrankB22:33, 19 March 2007 (UTC)