This page contains discussions that have been archived from Village pump (technical).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.
From my understanding of templates, their use on any page increases the time taken for that page to be assembled from the database. Navigational templates are continuously being refined, and often, people decide to shunt off repeatedly used parts of a template into a separate template. This wrapping can take on extremes where (I have seen) three templates are folded inside one another. Is this going to cause a performance problem for Wikipedia in the long term, and should it therefore be discouraged? Should a guideline be written to address this question? - Samsara (talk • contribs) 23:12, 28 November 2006 (UTC)
If nesting templates five levels deep helps editing and management of any article to be more straightfoward and easier to accomplish the goals of editors, let the system suffer! I'm mostly serious: Any structure which promotes better articles is to be commended and copied—not discouraged. One good example is periodic table. To the uninitiated, the wikilogic is bewildering. But once the inner templates are understood, it's awesome! It would be better to reengineer the system to more efficiently pre-transclude (for example) articles to make it perform better than worry about what is efficient and isn't. — EncMstr23:59, 28 November 2006 (UTC)
Please, not this again. This has been debated to death already, almost entirely by one persistent, disruptive user, who was banned twice for it.
Our lead developer has said that it doesn't cause any noticeable problems, and even if it did, we shouldn't concern ourselves with technical issues like this. Focus on content and ease of use. If the servers start to clog up, the developers will handle the problem on their end by making the code cleaner or limiting things in a technical way. By all means, use nested templates if it makes the encyclopedia better. — Omegatron00:11, 29 November 2006 (UTC)
When you're at a diff of an old edit you've made, there's a new option: "undo." What does this do? I tried it out in the sandbox but it said it couldn't undo the edit due to intermediate edit conflicts or something.
Also, the footer at the bottom of the page, in Safari 2.0.4, is far wider than it used to be. On all pages the horizontal arrow stretches farther than the text, which is out of the ordinary. --Fbv65edel / ☑t / ☛c || 22:21, 28 November 2006 (UTC)
I imagine it's a self-rollback? Though it might not be working correctly at the moment? It only seems to appear for older edits, but of course in that case, it gives the "The edit could not be undone because of conflicting intermediate edits" message. One would think it's supposed to give the undo on new edits only. --Interiot22:37, 28 November 2006 (UTC)
I suppose that's it, but I wonder how it could possibly have been run live in its broken state without someone stopping it. I am looking through svn for it, but I don't see it yet... Prodegotalk22:49, 28 November 2006 (UTC)
It only works when the software can unambiguously figure out how to do the older revert, without accidentally messing up the newer changes. I finally got it to work when User A made a change to one section, User B made a change to a different section, and then User A was able to undo their change. (but the sections have to be pre-existing...) Now everyone is going to call for MediaWiki to have better AI powers to properly discern how to fix conflicts. --Interiot23:39, 28 November 2006 (UTC)
I wrote this feature. What account you use has no bearing on whether or not the undo will succeed. undos will work if the software can figure out how to unambiguously revert a single not-on-top edit. This means that the edit to be undone needs to be the LAST unreverted edit to make changes to that particular paragraph. I'm considering making the (undo) button only appear if one is possible. — Werdnatalkcriticism07:07, 29 November 2006 (UTC)
I've tried to add a bit of documentation about this to m:Help:Reverting, and would appreciate it being improved by someone more aware of what this does (I didn't know it only worked on paragraphs, for instance). --ais523 13:58, 29 November 2006 (UTC)
As far as I understand, "undo" only undoes a single edit. If so, it should not appear on diffs covering more than one edit, as the user may get the impression that they are undoing all these changes. Tizio16:25, 29 November 2006 (UTC)
include bot should not make changes inside nowiki tags
I had on my user talk page
<nowiki>{{welcome}}</nowiki>
The WinBot Bot turned the "welcome" into "subst:welcome". That seems wrong. But I guess a work-around is for me to do more like what I have done in this posting. I.e. break up the template code into separate nowiki sections. Is that going to be the recommended course-of-action for the foreseeable future, or will the Bot be taught to understand nowiki context? I could believe recognizing that context would be very difficult. And perhaps occasionally inappropriate.--SportWagon15:25, 28 November 2006 (UTC)
The welcome template itself expands to include
<code><nowiki>{{helpme}}</nowiki></code>
I'll test with code tags, but I can't believe that makes the difference. Obviously the helpme template is not designed to be used with subst, but if later it was changed to be, would it be correct to retroactively change a whole bunch of supposedly expanded welcomes?--SportWagon15:38, 28 November 2006 (UTC)
Leave a message on the bot owners talk page. The bot shouldn't subst inside <nowiki> or <pre> land. That's an error. --Ligulem15:42, 28 November 2006 (UTC)
On my talk page, as an experiment I also had a welcome template without subst. The claim is that when WinBot came to fix that, it erroneously fixed the nowiki version too. I.e. the problem would not have occurred if only the nowiki version had been created. I'm not certain whether the problem is fixed or not, but this information qualifies the problem even further.--SportWagon13:52, 29 November 2006 (UTC)
A further experiment suggests the problem, per se, has not been fixed. However, the incorrect substitution will not happen unless the process which activates the Bot first finds a genuine "welcome" template without "subst" on the same user talk page. I.e. my experimental "bad" use of the welcome template, caused the "subst" to be added in my nowiki non-use. My suggested obfuscation for nowiki did "protect" another example, too, as one would expect. So this will, in general, be a minor problem for people who have incorrectly used templates, and the revision log does show what is done, so this is mostly just FYI now.--SportWagon17:25, 30 November 2006 (UTC)
links to wrong items with same name
I found three links to elementary schools on one page which are directed to other elementary schools in various parts of the country (rather than within the county.) How can I kill that link, but keep the text in red; as though it's a dead link?
Sorry for the laymanspeak.
If you want to see specifically what I mean, go to "Loudoun County Public Schools" and under elementary schools, click "Horizon Elementary". Note how you're now in BC, Canada.
You could change the link in the Loudoun County Public Schools article to [[Horizon Elementary School (Loudoun County, Virginia)|Horizon Elementary School]]. I would, however, just remove the link until such a time as someone writes a meaningful article about the school. (And it should never be "Horizon Elementary", but "Horizon Elementary School". User:Zoe|(talk)19:53, 28 November 2006 (UTC)
Featured articles not visible in classic skin
Hi, for people who are using classic skin, there is no hint that an article is a featured one, because the little star in the right upper edge is not visible. Is ist possible to fix that? --Nina10:14, 28 November 2006 (UTC)
No, because of the way interwikis are done in classic. In monobook, the page has a sufficiently ordered structure that some horrible CSS hacks can be used to put a star in the corner; in classic, the number of interwikis varies by page (as in monobook), but they're displayed above the page, so the star would be in the wrong position on some pages. (Correct me if I'm wrong; this answer is given from memory.) --ais523 10:19, 28 November 2006 (UTC)
Mhh, ok, so it's the interwikis. But I don't mind if it is a little star or a category that indicates the featured article. For example, in de: the FA-template includes a Category:Featured article. Maybe it is possible that this category would be visible only in classic? --Nina12:02, 28 November 2006 (UTC)
Thank you, I will try that. But I thought that it is of general interest to indicate featured articles, so I hoped that someone could fix that for every classic skin-user. --Nina13:16, 29 November 2006 (UTC)
Compare FA with no interwikis, FA with 2 interwikis, FA with 31 interwikis, FA with 59 interwikis (if I've counted correctly) using the classic skin (included in the URL) and with the CSS fix linked above; the position of the star relative to the page depends on the interwiki count. This could cause problems (especially on the last link) if the font size used by the browser for the interwiki links happens to cause the star to overlap the name of a language at the end of the line. It might be able to get the CSS to fix the interwikis too so this doesn't happen, but I'm not sure. --ais523 13:41, 29 November 2006 (UTC)
Yes, I see the problem. A category instead of the star would solve that. There is a lot of discussion on Template talk:Featured article, and I didn't read all this to find out if they have decided about a category already. --Nina13:48, 29 November 2006 (UTC)
Can't see my userpage
I am having trouble viewing my userpage. All I can see is in very small font up the top of the page:
I cannot see anything apart from that including the edit this page etc. buttons. It was perfectly visible yesterday. Any help would be greatly appreciated! Thank you, Jam0106:51, 28 November 2006 (UTC)
I can now see the page. Sorry for the inconvenience. Does anybody have any idea why it happened though? Thank you, Jam0106:53, 28 November 2006 (UTC)
Could have been an intermittent server problem: Wikipedia is a complex array of interoperating servers to dish up pages. The ball could have been fumbled in dozens of ways, which doesn't happen too often, but often enough that we all occasionally see weird stuff like that. You might try clearing your browser cache and clearing the server cache next time you see it. — EncMstr06:59, 28 November 2006 (UTC)
I'm attempting to use an external editor (any of the several that I have) for editing. I can get the data from the edit box window and into the editor, but it is input as a text file rather than *.wiki (The instructions for vim, for example, are to program vim to autodetect a file of type *.wiki so that the markup language will be used) Is there a special way to pass this this data on as *.wiki?
SFinside04:53, 28 November 2006 (UTC)
Unexpected Edit Due to Lost Connection
First, let me say that I find about 15% of the above comprehensible. I know this is a technical discussion page, but please "dumb down" any replies. My background is as an old 8-bit BBS sysop, the web was new when I was in school, I have used Windows and Mac computers at work (desktop publishing, GIS, cartography, &c.), but had little to do with maintenance, and I got my first "real" computer (Windows 98, which I am still using on a salvaged computer) a little over two years ago. I am relatively new to Wikipedia.
Here's the problem (and I'm not sure if this is the right place): I had a large talk page open for editing when I lost the connection to my (dial-up) server and was logged off. (It may have been the other way around.) I decided to call it a night, and when I checked the talk page later, a chunk of the section I was editing was deleted by an IP address associated with my server. I was not in the process of deleting the missing material, and nothing that I had added was saved.
I can understand a dropped connection saving the work in progress as it was, but why would anything have been deleted?
A problem that may be associated with this one is that sometimes the same large talk page will not load completely in my browser (Firefox), although my status bar reports "Done". Several attempts (clearing the cache in between) resolves the problem, but I have to pay international rates (the price of living in paradise) to access the internet.
What I do to get around the problem is that I usually compose any edits offline and paste them in. However, I am just learning how to use the interface, so I spend a lot of time on formatting. I am also going to archive the page in question now that I know how.
First of all, I also have a dial-up connection, and I feel your pain. About your question, it is possible because of the way Wikipedia saves edits. When you edit a page, it saves the whole contents of the page as a new revision. It is because of that that it is possible to blank a page. When you lost your connection, it saved a little of the content of the page, but not all of it. Therefore, it appears that it deleted some content. One way around this might be to use section editing. Instead of clicking the "edit this page" tab, you can find the place you want to edit on the page, then go up and find the closest headline. To the right, you can click the "edit" link, and it will only edit the content of the page within that section. That should help to prevent this happening in the future, and even save time when loading the edit page and saving the page. Hope this answers you question. Shardsofmetal 03:14, 28 November 2006 (UTC)
Got it. Now that I think about it, there's no way not to be overly technical dealing with a local telephone exchange-radiotelephone-sattelite phone-dial-up server-&c. connection. There's too many confounders. It's a wonder this doesn't happen more often. I was worried about frivolous complaints of "vandalism", but it looks like this is a known issue.
Section editing is a good tip, and I have used it (sometimes I forget). In my case, even the section I was editing had gotten rather long. Thanks for your help.
Unfortunately, I had gone to another editor's Talk page to ask for a comment on part of the section that was deleted. I was a bit disappointed that he or she did not revert the (obviously--and I also let him or her know that there was a problem I did not know how to fix) missing section before making comments.
I did a cut-and-paste revert. What's the etiquette on this? There is a discontinuity that readers should be aware of, but I don't know the appropriate way to point it out on a Talk page. If it was the article, the edit summary would seem to be sufficient. Is that enough where the deletion has disrupted the discussion on a Talk page? Should I insert [bracketed] editorial notes within someone else's comments? Should I do more? Less? Nothing other than the edit summary?
I don't want to be accused of some sort of mischief, but I could see myself making similar accusations if I (as a new user) was on the other end.
Is there any libraries for bot developing on Macs? I know a Python library out there, but if any Mac-friendly library such as Automator or AppleScript extensions, Safari helpers... available? Yao Ziyuan09:51, 27 November 2006 (UTC)
You said that I could contact you if I had something to ask or talk about. Well, I was recently on the “Christina Aguilera” Wikipedia page looking her up, and then I went onto the pages about her singles’ and then her albums.
One of the pictures, either on her main page, on her singles, or her albums’ pages, is infected with a virus. I believe it’s the “Dirrty” or “Stripped,” although it could be one of the others. My virus checker certainly picked up a virus from one of those pages.
This is extremely important. I thought that it would be a wise idea to immediately alert you about this. Viruses are no more welcome on Wikipedia than vandals.
I know you’re on a Wikibreak, but you’ve been talking on your Talk Page. You’re the first administrator I could think of, and the only one who has left a comment on my Talk Page. This is an urgent issue. I am not fooling around: my virus checker did pick up something from the Christina Aguilera sites (the main one, singles, or albums.)
I don't know a thing about this sort of business, and I need to head out right now, so I am bringing this to you. Can someone follow up on this? Thanks. --Chris Griswold (☎☓) 19:08, 26 November 2006 (UTC)
I don't think there's anything to worry about, as webpages shouldn't do any harm to you, as long as your computer has all the right updates etc. Tra(Talk)19:21, 26 November 2006 (UTC)
I believe it was either the "Dirrty" picture or the "Stripped" picture. User: Tra said that they edited the pictures. However, if the virus reappears on my virus checker, I will find out which file it was. Acalamari16:24, 27 November 2006 (UTC)
Have you considered the possibility of the virus coming from something else, like a different website or a different computer program? JDtalk16:26, 27 November 2006 (UTC)
I didn't need to. My virus checker clearly pointed out that the virus came from one of the "Christina Aguilera" pages on Wikipedia. I hadn't been to any other website when the virus appeared, and since I've never been to Christina Aguilera's website before, it wasn't difficult to work out that the virus came from a picture on Wikipedia. If it appears again, I will report it again; and hopefully, I'll have some more evidence. The thing I want to know is this: how did the virus get onto Wikipedia in the first place? Acalamari17:17, 27 November 2006 (UTC)
I don’t see the relevance, but my virus checker is AVG; Program Version: 7.5.431 (at the time of this writing, that is the most advanced version, as of November 27th, 2006.) Might I ask why there is disbelief in my warning? Why are we discussing my virus checker when we should be trying to isolate the virus? I will attempt to isolate the virus myself (assuming it’s still on Wikipedia) and find out exactly where it is very soon. Acalamari22:09, 27 November 2006 (UTC)
Sorry, I forgot to sign my previous comment. Anyhow, I've corrected that. I returned to the Christina Aguilera pages (main page, albums, singles) to look for the virus. The pictures are all clean. User: Tra obviously fixed the problem when doing the "printscreen" trick with the pictures. For now, this debate should be over, unless someone has something else to say about it. Acalamari22:09, 27 November 2006 (UTC)
You should also be aware that virus checkers constantly produce false hits - claiming a virus is there when there isn't one. This doesn't mean that things aren't to be taken seriously, but it does mean that careful examination is necessary with a number of different tools. Notinasnaid09:45, 28 November 2006 (UTC)
There was a virus, and my virus checker picked it up. I doubt that this occasion would be one where my virus checker would lie, and a good virus checker shouldn't register "false hits." As I said this isn't a subject to be treated lightly, and, for now, the pictures, and other data in the Christina Aguilera pages, are all clean. Acalamari15:59, 28 November 2006 (UTC)
What kind of an image, a JPEG? A JPEG, like a GIF, a BMP, or a PNG, isn't executed and the only thing a browser does with it is display it. It can't cause any kind of harm. Meanwhile, the way virus checkers find viruses is, in many cases at least, by finding patterns of bytes that match patterns of bytes found in known viruses. Once in a while, a file that isn't a virus is going to happen to have a set of bytes that match some such pattern. In other words, a false positive is possible. --Largo Plazo15:49, 1 December 2006 (UTC)
Sadly this is not necessarily true, because your browser isn't necessarily bug-free and because downloaded data isn't protected against inadvertent execution. See for example [1]. —Blotwell18:31, 1 December 2006 (UTC)
The file was likely to be a JPEG, but I don’t see haw that is relevant. The virus is gone. I been to the Christina Aguilera sites, and they are fine. Nothing is wrong with those sites now. I don’t see what the big deal is over the virus at the moment. The virus was removed, and I said that if it came back, I would report it again with more evidence. I don’t know why you’re all disbelieving both my virus checker and I. I’ve been to many sites and not once has it picked up a virus. Someone uploaded a virus-infected picture to Wikipedia. You should be finding how the virus got onto Wikipedia instead of implying that I lied. I didn’t register with Wikipedia to tell lies. Acalamari19:53, 2 December 2006 (UTC)
Your virus scanner picked up an infected file. According to the book "Troubleshooting, Maintaining & Repairing PCs (Second Edition)" (ISBN 0-07-913732-6), Section area "Virus Symptoms and Countermeasures" part "Virus Myths" (Page #1329), and I quote (from the bottom):
A virus cannot hide inside a data file Data files (such as images) cannot spread a virus on your computer. Only executable program files (and files containing executable macros) can spread viruses. A computer virus could infect a data file, but it would be a useless effort---because a data file is not executed, only loaded, the virus would not be able to run or replicate itself.
An image is a data file. Images cannot have macros. An image is only loaded, not run. Those three statements sum up the paragraph of text above. There is nothing to worry about. But...
If it had the infection on it upon uploading, the computer should be scanned.
If it didn't have the infection on it upon uploading, all the WikiPedia servers should be scanned, in addition to a virus notice being given out instructing everyone to scan their computers for possible viruses (very unlikely, as there are no runnable applications on WikiPedia, but it is javascript-powered - javascript is like one big macro)
If you need any other information, contact me. Don't forget to help me (details on user and user talk pages).
Computer and Cricket master,
--Cricket Boy03:35, 4 December 2006 (UTC)
Your book is, quite simply, wrong. It omitted the critical proviso "if there are no major flaws in software on the target's computer". If you don't know what a buffer overflow is and how it can be exploited to spread a virus via a data file, I suggest you drop the appellation of "computer master". —Simetrical (talk • contribs) 07:30, 5 December 2006 (UTC)
Wikipedia will not load in Firefox or IE, only on AOL
Hello everyone, I was hoping you could help me out with this little problem I've been having. I am unable to access Wikipedia when I use Firefox or IE on my computer. The strange thing is, it works fine in the AOL browser. I use Firefox practically all the time so this is quite frustrating as I visit Wikipedia quite often. I have tried uninstalling all the add-ons in Firefox, but seeing as the same problem occurs in IE, I'm pretty sure that isn't the cause. I don't have the knowledge to even guess what could be wrong. All I know is that I made no major changes to either browser that I think could cause this. Thanks to anyone that can help me out!
What does "unable to access" mean? Do you see anything at all? Did it ever work previously on that computer? How about uninstalling Firefox and reinstalling? — EncMstr08:00, 26 November 2006 (UTC)
Nothing loads at all. It stays blank in Firefox and continues to show the little "Loading..." message. In IE, after about a minute of not doing anything, it takes me to a "Were you looking for..." page which just lists the link to Wikipedia.
I think I know the problem. You probably have your internet options set incorrectly. If it is Windows, try Windows Update, as it may patch IE. You may also need to patch Firefox, but I don't know how, as I don't even know what the company is called. As for me, that sounds pretty weird, as I am using Internet Explorer to edit this thread. The version # is 6.0.2900.2180. In addition, because you can only access it via AOL, your standard IP address may be having problems with WikiPedia, as AOL uses several different IPs. Of course, I cannot be certain, as I have never experienced this problem. Oh, and "unable to access" probably means just that, the browser cannot find the page on the Internet.
Signing off, as a cricket and computer expert, --Cricket Boy 01:55, 3 December 2006 (UTC)
This may be a perennial question, but I didn't see the answer anywhere. On wider screens, sometimes left-aligned and sometimes right-aligned images will bunch up and pull away from the side of the page, creating a gap with no text. I ran into a particularly bad example tonight, on a featured article, no less. Four images stacked up, consuming over half of my 1920px wide browser. I've seen case-by-case fixes that sometimes work, but they never seem conclusive (eg. even if I fix an article now, if I upgrade to a 30" Dell/Apple display, will it break again?). Is there a good way to definitively fix this issue for all possible resolutions? --Interiot09:13, 25 November 2006 (UTC)
<br style='clear:both'> might be the solution. Put it after an text and it will create a blank line that extends to both sides of the page, avoiding that class of gaps.
This will be the result:
Lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum lorem ipsum too lazy to see how lorem ipsum continues.
Lorem ipsum.
More lorem ipsum.
Anyways, that article is now formatted using mixed alignement. But this can also be useful
Or you could put the images in a table, but that might not look so good on the article; it has a lot of images on it. JDtalk11:49, 25 November 2006 (UTC)
You indicate that the first picture bad example surely you mean a good example as it precisely demonstrates your point. This is a good example of a very commmon gramatical mistake :) Dainamo14:01, 25 November 2006 (UTC)
Personally, I'd use the {{clear}} template. I've adjusted the example above to use it so you can see what I mean. 'Course, the template is just a shortcut to the br code you listed. It's just easier to type for those that don't know HTML. --Brad Beattie(talk)15:52, 25 November 2006 (UTC)
We certainly don't want "dog" and "Dog" used distinctively; that is, we don't want two different articles.
But none of these considerations apply when the first letter is not a Latin character. It is a continual annoyance in the math and science areas of WP that we cannot have articles starting with a lowercase Greek letter; in these cases, the letter in question is never uppercased; it would be quite wrong. Moreover, the cases may even be used distinctively -- for a long time I've been about to write an article Real Soon Now on W. Hugh Woodin's Ω-logic (for a description see Woodin, W. Hugh (2001). "The Continuum Hypothesis, Part II"(PDF). Notices of the AMS. 48 (7): 681–690. But there's an entirely different notion called ω-logic (see for example Weaver, George (Summer 1992). "Unifying Some Modifications of the Henkin Construction". Notre Dame Journal of Formal Logic. 33 (3): 450–460.{{cite journal}}: CS1 maint: year (link))
So in my opinion, in en.wiki at least, the rule should be modified so that articles beginning with Latin letters are uppercased but those beginning with non-Latin letters are not. There might be an exception for near-Latin letters like the Polish l-with-a-slash-through-it, or umlauted vowels (actually I'd be happy if just Greek letters were excepted from uppercasing). Obviously things would have to be different on Greek WP, but surely the software is agile enough to allow that, or easily could be. --Trovatore07:25, 25 November 2006 (UTC)
In the medium-term future, it will be possible to mark articles to display with a lowercase title. In the longer-term future, we should eventually migrate to a cleaner case-insensitive title system. (This is not a trivial problem given back-compatibility with millions of pages, so don't hold your breath on that one.) --brion20:46, 25 November 2006 (UTC)
It does not sound like either of those solutions will allow us to have one article called Ω-logic and a different one called ω-logic, which we really ought to be able to have. Can't we have case-sensitivity when the first letter is non-Latin? --Trovatore05:28, 26 November 2006 (UTC)
No. Use a disambiguation marker to make the difference clear; remember that failing to do so will also complicate searches, linking, etc. --brion01:12, 27 November 2006 (UTC)
As a software engineer myself, I really cannot believe this is an insurmountable technical problem (though I can believe you don't really want to mess with it :-) and I see no advantage, in en.wiki, to automatic uppercasing of Greek letters. Can anyone tell me why we should want to, taking English wikipedia as a standalone? If not, can anyone tell me what problems, specifically, would be caused in the interaction between English and Greek WP, by the difference in rules? --Trovatore16:41, 27 November 2006 (UTC)
Mainly, I figure, you would have to write your own case-checking functions. Regex functions with case-insensitivity switch? Nope, need to rewrite them if you want to handle Unicode and at the same time only respect some uppercasing, presumably in C(++) as an extension. Likewise for all string library functions that care about case. Doable? Certainly. Something that we probably want to do? Not really, not for the sake of a couple of articles. Your best bet is finding a slightly different name for one and having disambig links. Which you'd want anyway, since the names are so similar. —Simetrical (talk • contribs) 02:33, 29 November 2006 (UTC)
That slightly different name is easily enough done, too. Just keep it at Omega-consistent theory, for example, instead of moving it for the sole purpose of justifying the "lowercase" template that had been removed as being improper. Gene Nygaard19:19, 3 December 2006 (UTC)
Gene's version of the facts is quite incorrect. The move was not intended to justify the "lowercase" template (which was proper even where it was). The purpose of the move was to use the Greek letter rather than the Latin name for the letter.
Gene's suggestion to "keep it at Omega-consistent theory" is a non-sequitur as a response to the problem of distinguishing between Ω-logic and ω-logic, as neither of these would point to Omega-consistent theory (ω-logic is somewhat related to ω-consistency, but not closely enough to go in the same article -- in particular "consistent in ω-logic" is quite a bit stronger than "ω-consistent"). --Trovatore20:20, 3 December 2006 (UTC)
It would probably not be difficult to change, but it requires a developer to make the change, and they are somewhat busy. However, if you know how to program in the language MediaWiki is written in you could submit your code and a request to impliment it at bugzilla. The source code of the wiki software is accessible here. Prodegotalk16:53, 27 November 2006 (UTC)
Accidentally and prematurely saving a page
Twice recently I've accidentally pressed "Enter" while in the edit summary box and saved the page before I meant to. I don't think this ever happened to me before the recent (and very welcome!) changes to the summary box. Am I just becoming more butterfingered, or has some change been made along the lines of "save page" becoming a default option that it wasn't before? --Blisco10:40, 18 November 2006 (UTC)
Butterfingers. I remember having to be cautious "down there" ever since, lest to prematurely ...can I reword that? Femto11:01, 18 November 2006 (UTC)
There was a thread here some time ago on this issue. The default of web forms is that pressing enter in any field that is not a textarea causes the default button (or the selected button, if any) to be pressed. I also made such mistakes here and there; make a null edit to post your full edit summary. Tizio12:45, 18 November 2006 (UTC)
Actually, pressing enter is not the same thing as pressing the default button. Many websites (including Wikipedia) handle them both the same, but it is possible for a different action to take place depending on whether you press enter or click the button. For example, Google gives you a 'tip' message if you click the search button, but not if you press enter after searching from the homepage (compare this to this). Tra(Talk)18:07, 18 November 2006 (UTC)
I personally prefer having Save as the default button, because it allows me to just press enter once I've filled in the edit summary and minor edit fields without having to tab all the way to 'Save page'. It might be useful to allow this setting to be changed in the preferences, however. Tra(Talk)14:38, 25 November 2006 (UTC)
I use ctrl/alt/s for that. I click 'save' otherwise, and the warning preference prevents me from posting prematurely. Leaving the edit summary till last is then the most efficient way of avoiding accidental saving. qp10qp03:49, 30 November 2006 (UTC)
Shoddy SVG versions
I'm writing to complain about shoddy SVG versions of raster graphics being uploaded, then shoved into articles. Sometimes SVG versions are okay; often they're not. I'm not going to cite specific examples because this will only lead to people defending on a case-by-case basis and taking their work personally. This is a general problem.
Don't download PNG images to your local machine, run them through some sort of car washer, and upload an inferior SVG. You aren't actually achieving anything; the engine will render a PNG to most browsers anyway. You're not saving any storage, since the PNG remains -- and deleting it won't solve anything except to upset the uploader. We're not so strapped for cash that deleting 100 Kb will put us in the money.
SVG is a problem format. It has its advantages, yes. But it handles fonts poorly. It can be used well but I don't like some of the work I see. I create some graphics with a vector editor, Macromedia FreeHand, and rasterize an exported EPS in Adobe Photoshop. If you want to convert the EPS to SVG, I'll gladly make that available. But if you try to re-draw vectors from a raster image, you either need to do a great deal of unnecessary work or you mess it up.
PNG is a fine format for images; unlike GIF, it is completely free libre and there are no fears of Unisys coming to bite our balls over it. If you have nothing better to do with your time and the amazing tools on your desktop, go ahead and see if you can squeeze a few bytes losslessly out of existing PNGs. Please do not replace them with SVGs unless you can show real benefit and no harm. Thank you. John Reid °09:27, 10 November 2006 (UTC)
SVGs are, in principle, much easier to maintain and update than PNGs, and of course scale much better. At some point we'll probably have in-wiki editing of SVGs, and maybe some kind of semantic metadata so that we can (for instance) auto-standardize map colors and so on. If a good-quality SVG should be made of a PNG, it should be, just for future advantage. (Font handling is certainly a problem, I admit.)
Note that you can use PNGs or GIFs in place of autogenerated SVG thumbnails where the behavior is currently faulty, like so:
SVG is no more of a "problem" than any other vector format. And, given its wide range of capabilities (transparency, filters) and its fully open and documented status, it is definitely better than any other vector format in a lot of cases. Its font handling is neither better nor worse than that of HTML, because they both use CSS. The ability to specify a generic substitute for a missing font (such as "sans serif") via CSS is a great advantage of SVG, and for the Web it is actually a more important advantage than exact font fidelity (which you, as a graphic professional, may value more). I agree that simple automatic tracing of a bitmap into SVG is almost always a degradation, and even manual redrawing of an image may be a degradation unless certain effort is spent and relevant skills are in place, but your attitude towards SVG as a format seems unfounded. Trapolator22:37, 19 November 2006 (UTC)
Let me stress: I'm not talking about the preferred choice for original work for someone with tools that can create both formats. In some cases, SVG may be better; I can't say -- I don't have those tools. They aren't common.
I'm only upset when I see people coming along, taking an existing PNG, uploading an inferior SVG imitation, then forcing it to replace the better-quality PNG. You cannot do a straightforward conversion of a raster PNG to a true vector SVG. In some cases, the creator of the PNG may have left the community; nothing you can do about that. If you put enough time and trouble into the re-creation, you may come up with an acceptable SVG; that's fine too.
But please do not butcher PNGs into inferior SVGs when the creator is available! You -- yes, you -- can always contact me, if it is my work under consideration. I will be happy to make available vector format workfiles for any PNG that was created entirely in this fashion. You can convert to SVG to your heart's content. This is fine. I only want to express how annoying it is when I spend hours producing a beautiful graphic and somebody else comes along in 15 minutes, does a crappy conversion, then pushes my work aside. I put it here into the common bin and license under GFDL; anyone can edit, yes. But please make it better, not worse. Okay? John Reid °07:45, 15 November 2006 (UTC)
Oh, I agree completely with that. A higher-quality PNG should be used over a lower-quality SVG. But all things being equal, SVGs are superior. It would undoubtedly be appreciated if you converted the images to SVG yourself, or provided them to someone who could do that (it should be very simple to convert something like a PSD to SVG, I would imagine, although I could be wrong). —Simetrical (talk • contribs) 05:09, 16 November 2006 (UTC)
tools that can create both formats... aren't common? It's not quite true. Inkscape is pretty common, especially here on Wikipedia where thousands of images are created with it. Adobe Illustrator in the latest versions can also export valid SVG. Both of them can rasterize to PNG too, of course. (Freehand is not SVG-capable as far as I know, unfortunately, but it's rather an exception than a rule these days.) As for your request to coordinate with you any vector conversion efforts on your images, I completely sympathize, but I think your user page and/or the pages of your images may be better places to place such requests. Trapolator22:37, 19 November 2006 (UTC)
All things are rarely equal. I hardly ever do a straightforward conversion from vector to raster; I twiddle and tweak in Photoshop. Sometimes, the Photoshop side of the job is major, although the result may look as if it came right out of FreeHand. SVG stinks for rasters. I'm not sure I agree that vectors have any real practical use to the reader anyway; most browsers can't display SVG (or any other vector format) directly and MediaWiki serves a PNG anyway.
In any case, I don't have the niche tools that output SVG. Like other graphics professionals, I work in the industry-standard EPS format. We may not like that format and we can grump around all day about how it's "not free" but it is widely supported. I'd be happy to upload EPS and let somebody try to convert to SVG. It's not allowed. John Reid °21:46, 19 November 2006 (UTC)
SVG stinks for rasters how exactly? Are you saying that it is not appropriate for images that are raster by their very nature? True, but the same applies to any other vector format. Still, with its transparency and filters (e.g. blur), SVG can efficiently approximate a much wider range of imagery than, for example, EPS.
most browsers can't display SVG: you have a strange definition of "most". By the last count, only one of the major browsers didn't support SVG.
For heaven's sake. That (IE) is the browser still used by an overwhelming majority of users (and I say this as someone who avoids it whenever possible, but recognizes the obvious need to support it). (There's an Adobe plug-in but precious few people have it.) If you have an image, or any content, worth contributing to a Wikipedia article, supply it in a form that more than 20% of people will see. --Largo Plazo16:36, 29 November 2006 (UTC)
niche tools that output SVG: I don't want to start a holy war, but neither Adobe Illustrator, nor Corel Draw, nor even Inkscape can really be called "niche tools". All of them do SVG. Inkscape in particular is completely free - try it out and you may like it.
I sympathize with John, the best version of an image should be presented to the reader. The creation of crappy SVG versions of uploaded PNGs rarely benefits anyone. If the SVG is actually superior then it should be used, but this seems to rarely be the case with SVGs created after the fact from uploaded raster images. A lower quality SVG might be linked from the PNG image description, but it should not replace the PNG simply because "vector graphics are better". The overriding consideration should be which version is better from the point of view of the reader. Dragons flight04:36, 20 November 2006 (UTC)
Thanks for the suggestions. I stand corrected on the range of SVG support. Still, I personally can't do it until you buy me a new toy, Inkscape doesn't run on my box, sorry. In any case, I prefer reasonably high-resolution PNG for distribution, period. We're not running out of bytes. I don't see value in hosting/serving vector graphics. This is another case of carrying a philosophical ideal to an extreme. Anyone can edit a PNG and it's a libre format, too. John Reid °11:16, 21 November 2006 (UTC)
Well, in principle anyone can program in machine codes... yet high-level programming languages do exist :) Same with SVG. Of course there are many types of graphics that simply do not fit into (modern) vectorial formats. But many others do fit, and vectorizing them has many advantages. For example, if you want to translate a drawing, clicking on a text object and retyping it in another language using Inkscape is so much easier than wiping out a label in a bitmap editor and creating a new one instead (especially if it's not over a white background). To borrow your own example, replacing the too-thin letters by something bolder and nicer in this SVG image takes a couple of minutes; doing the same in the PNG original is infinitely more work due to the checkers.
So, if an image is of the vector nature (such as a technical drawing, a diagram, a map, etc), I think that the advantages of converting it to a vector format tend to outweigh the possible loss of some polish and gloss that the bitmap version had. Of course sometimes, SVG may be so shoddy that it's not the case, but even in these situations it's not a big deal to take the bad SVG file and improve it. In fact, perhaps the main reason for preferring SVG on Wikipedia is because it's more wiki-friendly than bitmap (although setting up the "external editor" for mediawiki is admittedly a pain, but it's not SVG's fault).
As for Inkscape, can you elaborate a bit on what exactly prevents you from running it? It works fine on Windows, OSX, and Linux, and does not require exorbitant speed or memory (unless you're editing very complex graphics). Trapolator19:55, 21 November 2006 (UTC)
Update: I went ahead and cleaned the SVG of the billion cubes image, see here. I used a bolder font, fixed stroke widths, and improved the shape of the connecting swooshes so they now look even better (IMHO) than in your PNG. I also used clones for the 4 copies of the swooshes - which is another vectorial nicety: edit one of them, all four update automatically. Altogether it took maybe 10 minutes of my time. With PNG, I wouldn't be able to do any of this without a disproportionate amount of tedious labour. Here's SVG advantages for you :)
And there's another important reason to prefer SVG for this particular image: the tiny copies of checkered cubes in the corners of big cubes are barely visible in PNG, even at full size. Yet they are important for understanding the concept. So if I'm going to use this image to explain the idea of a billion to my son, I will very much prefer to load the SVG into Inkscape where I can freely zoom in and out. Trapolator20:37, 21 November 2006 (UTC)
Adding a link to the Cheatsheet in editing-mode helpnotes?
I'd like to propose we add a link to Wikipedia:Cheatsheet, in the editing-mode layout, next to the "Cancel | Editing help (opens in new window)" links. eg:
Friends of mine who only edit very occasionally, have expressed frustration concerning finding reminders for basic wikicode easily (eg piping links); and are either daunted-by or disdainful-of the size/complexity of the Help:Editing page.
Strong support. I'd actually rather see Cheatsheet replace "Editing help," with a link on editing help to advanced help, but this'll do as well ;) --Wolf53018:58, 4 November 2006 (UTC)
How?: I'm not an admin, and I don't know how to code a suggested change for those 2 pages (I can see how it's linking "Editing help", but can't I derive how one would add an unlinked "&" and a 2nd link into it...)
It looks like we can only replace the "Editing help" link, but not add next to it. I think a greater consensus would be needed to perform that change. -- Renesis (talk) 00:24, 1 December 2006 (UTC)
Ah, software change required, I believe is meant. Bugfix submitting isn't my area - I'll stop championing this idea for now then, though anyone else is welcome to pickup and run with it. (Learning Mediawiki is on my to-do list, but not at the top ;) -Quiddity02:18, 1 December 2006 (UTC)
I actually meant we'd need greater consensus to replace the current link, instead of just adding to it. I don't know if anyone even wants that, I was just mentioning it as an option. -- Renesis (talk) 06:45, 1 December 2006 (UTC)
Well, I'd be for that, but I've already herded this idea through VPp and VPt twice, with only 6 people giving feedback, so I don't envisage greater consensus occurring (unless another route is open/suggested? I guess perhaps I could promote the proposal simultaneously via the CBB?) Hmm. Would you support that change? ;) -Quiddity20:14, 1 December 2006 (UTC)
Honestly, I don't think I ever noticed that link until I saw this discussion, so I have neither an attachment to it, nor a strong opinion of how it could be improved. The Cheatsheet link does look more like what you'd want to find when in the middle of editing a page however, so I would probably support. Another thing we could consider is placing it in the list below the "Your changes will be visible immediately text", below. -- Renesis (talk) 16:43, 2 December 2006 (UTC)
I cannot seem to locate the source of the whitespace atop the Satveer Chaudhary article. It doesn't seem to come from extra line breaks, nor from the templates. Anyone care to investigate this further? Circeus01:33, 27 November 2006 (UTC)
Removing the infobox makes it clean. Hmm. Can't see what's messy with that, though, if it displays okay elsewhere... Shimgray | talk | 01:38, 27 November 2006 (UTC)
Yes, the problem is in Template:Infobox Politician. Richard Henry Lee is another article that uses the same infobox, has the same space at the beginning, and has no apparent reason for the space. If you blank the office2, term_start2, and term_end2 parameters to the template [2], that gets rid of the space at the top. If you run it through Special:ExpandTemplates, you notice that several blank lines are created in between the </tr>..<tr> near the office2/term_start2/term_end2 arguments. If the template can be tweaked so these blank lines aren't generated, it will fix the problem. But the template is pretty complicated, it's not clear to me what changes are needed. --Interiot03:59, 27 November 2006 (UTC)
These are tags used on templates. If you put text between these tags, they show up when you view the template page, but not when the template is transcluded onto another page. Tra(Talk)19:23, 26 November 2006 (UTC)
Charon-1,205km across is just the text ((hangon))- yes, in parentheses. If you go to the edit page, however, it's a redirect- and if you finish editing it it shows a successful redirect page- but if you (for some reason) typed in charon-1,205 km across, you would *not* be redirected, just see ((hangon))... we could probably delete that redirect altogether, fix it cleanly, but wtf? Cantras17:23, 26 November 2006 (UTC)
It's working fine for me. I suggest clearing your browser's cache and trying again. It may not think the page has changed and you need to force it to download the newest version. -- JLaTondre17:29, 26 November 2006 (UTC)
Is this possible? I've noticed that various people keep recreating Category:2010 films (amongst others). It carries the somewhat confusing tense confusion "Films originally released in the year 2010." Pretty much every article that's been in it has been mostly rumour, or so scant on verifiable information that they've been deleted or repointed. Semantic arguments about including films which "were released in the future" aside, there won't be enough verifiable information on any film slated for release at that time for at least another year or so, so is it possible to prevent such categories from being recreated and repopulated any time soon? Chrischeesewhine19:04, 25 November 2006 (UTC)
Salting prevents recreation of the category's wikitext, but I believe it can't prevent repopulation of the category if someone really wanted to... --Interiot20:26, 25 November 2006 (UTC)
Hi, all. I'm wondering if it is possible to request particular page sections via a php script. For instance, how might I include the (more or less) plaintext of a page's contents box or summary without all of the of the wikipedia chrome or the rest of the article? Possibly, dare I say it, with functioning links? I imagine such a mechanism exists, due to the neet-o wikipedia lookup finctionality in the Trillian chat client, but my best efforts have turned up very lttle in the way technical details for including wikipedia content on another page. I saw somebody mention using 'action=raw' in the url, but that seems to be only part of the answer, if any answer at all.
anyway, any help or direction as to where this question might be answered effectively are much appreciated
for the record, I looking to develop what amounts to a wikipedia skin/reader with some nifty browsing developments, not just to leech wikipedia's content.
69.114.196.1318:43, 25 November 2006 (UTC)Mike Weber
-- 01:42, 25 November 2006 (UTC)
You could try getting the page with &action=render on the end, then scrape these results to get what you want. For example, for the summary you would want everything up to the first <h2> tag; for the contents you would want everything inside <table id="toc">. If you are thinking of doing this for lots of Wikipedia pages, however, you might find the database dumps useful. Tra(Talk)19:03, 25 November 2006 (UTC)
-- I am in no position to host my own wikipedia mirror, but action=render may indeed be handy (very possibly the tool I need). But is there any central repository for information about querying wikipedia? what types of parameters can be passed and the like? for instance, I see that there is a 'section' parameter that can be passed, but it only seems to work for edits. Were can I learn more about this sort of thing? (alternately an explanation of how to work the 'section' parameter or something like it would do too, in "give me a fish, but no fishing skills" sort of way). Thanks, and stay beautiful, folks.
4:41, 25 November 2006 (UTC) Mike Weber
I'm not aware of any documentation of that type, but you might be interested in query.php which is useful for getting machine-readable lists of, say, pages in a category. Tra(Talk)22:29, 25 November 2006 (UTC)
Technical question re: categories
As you all know, on category pages, each subcategory has a plus sign next to it, with which one can view that subcategory's next layer of nested subcategories. I've heard, however, from at least one user who's said that doesn't work properly for him; instead it stalls on "Loading" and doesn't actually produce a list. It works properly for me, however. Is this just an issue of needing to upgrade his javascript, or is there something that the technical team should look into? Thanks for any help. Bearcat00:08, 25 November 2006 (UTC)
WP:UBX gives general information. For images, you could start with {{userbox|orange|yellow|[[Image:Crystal kthememgr.png|40px]]|A userbox is a small box that looks like this.}} which gives the userbox shown on the right then modify it to show the correct text and image. Tra(Talk)01:55, 25 November 2006 (UTC)
Variation in how Monobook and Classic skins align inline images
I've discovered that the Classic skin (which I have used since time immemorial) and the slightly more common Monobook skin have a difference in how they vertically align inline images. I was passing through Wikipedia:WikiProject Stargate and saw that the glyph in the title there was way out of alignment with the rest of the line, as illustrated by this partial screenshot: . I "fixed" it, but was then told by someone using Monobook that for him the fix had thrown the image out of alignment in the opposite direction: . I don't know my way around CSS very well and so haven't found the root cause of this different behavior and haven't been able to fiddle my way around it blindly. Does anyone have advice on how I might resolve this? Bryan06:37, 24 November 2006 (UTC)
(The page has since been updated with an all-image banner, for reference the two versions where this problem first showed up are [4] and [5]Bryan08:00, 24 November 2006 (UTC))
Something has happened with the links we have for coordinates. See most geographic articles, for example Milwaukee has a link to [6] where we used to find various online mapping tools to get an overview of the area, but all those pages have now been replaced with a variation of {{deletedpage}}. What happened? Sjakkalle(Check!)10:26, 23 November 2006 (UTC)
Great, just great. 1. Because I am now on DirecWay (with no other options for a server because I am in the deep woods) I cannot stay logged in *PERIOD*. 2. Before this I went to the trouble to take a photo of J. sambac in my yard, and I edited the size (bytes) for a quick load. I added it (5 months ago) to the Jasmine page. 3. Now the Wiki Police remove it, and I cannot stay logged on to put it back! It's not my fault Wikipedia has some glitch that won't recognize Direcway users, and yes I went to your help desk to no avail. Will someone PLEASE go to my website and get the pic: http://www.MagiaLuna.net/jsambac1.jpg and put it back on the darned page? The removed text under the pic read: "J.sambac in bloom along with an unopened bud. The flowers smell exactly like the tea." Please excuse my frustration. Thanks, and BTW I can't find my user page now?? And of course I logged in to post this and now I'm not logged in... Magialuna20:34, 22 November 2006 (UTC)
The user page for Magialuna is at User:Magialuna. The talk page for the ip you're using now is at User talk:69.19.14.41. To make sure you stay logged on, try making sure that you have cookies enabled in your browser, and you have enabled the option to remember you. If this fails and you still can't log in, please give the source and license of the image so that the uploader can note it on the image page. Tra(Talk)02:05, 23 November 2006 (UTC)
Been there, done that re: cookies enabled. Source? You mean my actual name? Uh uh. I took the pic, and I can put the HUGE unedited original digital photo on my website if no one believes me. License? Huh? Magialuna21:19, 22 November 2006 (UTC)
Users on satellite ISPs often find an improvement when they use the https server (which doesn't give any extra privacy, but makes satellite connections more likely to work); there's a link at the top of this page (WP:VPT), and also on my userpage (User:ais523). --ais523 09:27, 23 November 2006 (UTC)
Someone using HughesNet found instructions to bypass the Turbo Acceleration proxy: [7]. This *may* work for some of the other satellite providers. --Splarka (rant) 21:13, 23 November 2006 (UTC)
Magialuna, I think you were doubly frustrated and posted two questions in one section. Your image is OK as I understand, but it would have been fixed quicker if treated separately. As for your connnection, good luck! -- DLL .. T20:08, 24 November 2006 (UTC)
I've been around for a year and a half, was given admin chores, performed some history merges, and I still have no idea whatis going on here. I can see diffs in the history that don't appear when comparing diffs. Look at this diff] and then hit the next link. Look at the resulting diff, particularly the time gap between them. Then look at the history of the article. It skips ~30 edits made by others and attributes all their changes to an editor ~30+1 edits later. Since all of the edits do appear in the history if one compares diff by diff I guess it isn't too much of a concern, but I don't understand why it is doing this. I also have a personal concern as 25 of those 30 are mine, making a large expansion of a stub.—WAvegetarian•(talk)12:54, 24 November 2006 (UTC)
When I look at this and click through, edits through May 25-June 22 are attributed to Jumbo Snails, but I can't find him nor the person with the May 25 edit in the history either. I think it is indeed corrupted. - Mgm|(talk)13:08, 24 November 2006 (UTC)
edit conflict: User:Jumbo Snails is formerly CAPS LOCK. His first edit to the article was on the 22nd of June. This shows up if you look at the history page. Prior to my edits on the first (representing the 5th-... edits to the article) it was edited in February, March, and May. The two edits in May don't appear when you use the next diff links either. Is there someway to uncorrupt such problems, like deleting everything and manually restoring? All of the edits this article has had to it appear in the history, they just don't show up in the next diff continuum. This article has resided at around half a dozen different titles in the last year and a half. Should I take this to WP:VP/T?—WAvegetarian•(talk)13:21, 24 November 2006 (UTC)
The pages Special:Deadendpages, Special:Uncategorizedpages, Special:Uncategorizedcategories and Special:Wantedcategories are created regularly and remain unchanged in the intervals. I understand that compiling these lists dynamically would take too much time, but once they exist, it shouldn't be too demanding to monitor changes to listed pages and remove those that no longer belong. This would make these special pages more useful, because currently it can be quite challenging to find anything in them that hasn't been fixed already. --Derlay00:29, 23 November 2006 (UTC)
Every time I try to log in from my PC I get a prompt to download a file after I click log in. I do not have Use external editor checked in my user preferences and they shouldn't have an affect anyway, I think? Does anyone know what is wrong? 81.104.37.8120:29, 22 November 2006 (UTC) (usually user:Lcarsdata)
I have a couple of users on Apple PC's who have (or currently are) seeing the text in articles displayed in Hindi or some Indo-Dravidian text.
I know I've turned this error off in the past but can't seem to remember how.
They are all using Tiger, and Safari. Problem does not come up in other browsers. Also seems to effect Craig's List and some few other sites.
Any help will be appreciated! Thanks
Merel O'Rourke
(e-mail removed for privacy reasons)
Wed 22 November 2006
This seems like an error in you default browser settings. Go to the Safari menu, down to Preferences... (⌘,). Then click on the Appearance tab on top. At the bottom of the Appearance options you will see "Default Encoding." Set this to Unicode (UTF-8) and you shold be all set.—WAvegetarian•(talk)13:40, 24 November 2006 (UTC)
It is possible that the User:Curtisaallen account was registered some time ago but never utilized till today. When such accounts are layed away for nefarious purposes they are usually known as sleeper account on Wikipedia. (→Netscott)16:01, 22 November 2006 (UTC)
What would it require for the program to be changed so that users with no edits so far would be treated as if they were brand-new? Wahkeenah16:07, 22 November 2006 (UTC)
I'm not entirely certain that the change would be that helpful. Vandals who use sleeper accounts have already shown some degree of sophistication. All it'd mean is that they'd have to make a single superfluous edit before laying the account to rest for a while. GeeJo(t)⁄(c) • 16:29, 22 November 2006 (UTC)
An excellent idea. Or 50, or 250, or something that would at least require them to expend some time and effort... especially if the number of edits needed is kept secret somehow (probably not an acceptable idea, but worth fantasizing about). Wahkeenah17:21, 22 November 2006 (UTC)
There's no way to get an editcount directly within MediaWiki, as the number of edits is not recorded anywhere in the database (that means, there is no editcount table, or editcount field in the user database table). It can be derived through a separate expensive query, like it is done in the Toolserver, but it would be IMO too expensive to make the same query until a certain number of edits are made to the site, and the user gets the autoconfirmed flag. Titoxd(?!?)23:59, 23 November 2006 (UTC)
In short, it would require not just a program change, but a database change... and thus it's a rather lower likelihood that the project managers would make this kind of task a priority. :\ Wahkeenah02:23, 24 November 2006 (UTC)
A pipe format with two | signs, for words and languages with nominative singular endings
The usual pipe format is e.g. "[[scuba set|my old aqualung]]", as is well known. The trouble arises with inflections. In English, most inflected word forms have a bare-stem form, so linking from an inflected form is easy: e.g. "the [[cat]]'s tail"; "the [[patrol]]men were [[shovel]]ing seized kit into their incinerator."
The trouble comes with words where the bare stem is not a valid word. That occurs more in other languages, such as with nouns in Latin and Lithuanian and Icelandic, and with feminine and neuter words in Slavonic languages, and verbs in Romance languages, where in some inflected forms the end of the listed form drops out. That results in having to repeat the whole word at the link, e.g. "[[incineration|incinerator]]", "[[inflate|inflating]]". This is very common in some foreign Wikipedias, adding up to many megabytes of extra text. It would be useful to be able to write e.g. "[[incinerat|ing|or]]", "the star ξ [[Sagittari|us|i]]", "[[inflat|e|ing]]", instead. Anthony Appleyard11:11, 22 November 2006 (UTC)
Redirecting categories
It would be useful if this feature was put into Wikipedia. If e.g.:-
Is there any piece of Javascript or equivalent I could use to filter out any entries on Special:Newpages below a certain size threshold? I do a fair amount of searching there for potential DYK entries, and being able to avoid having to scroll over the 80% of articles smaller than 3kb would be very handy. GeeJo(t)⁄(c) • 09:53, 22 November 2006 (UTC)
Thanks very much! Certainly going to save a good deal of time, and it looks like I really underestimated the number of small articles being created. Seems to be closer to 95% on running this tool. GeeJo(t)⁄(c) • 13:51, 22 November 2006 (UTC)
Templates: Passing list of parameters as one parameter
Assume page A calls template B, which calls template C, which calls either of (X,Y,Z) depending on a parameter supplied by A. X,Y,Z have some parameters in common, which are supplied by C, but the remainer needs to be supplied by A.
So, in pseudocode:
Case 1
A calls B with params (X,1,2,3,4,5)
B passes these on to C
C calls X with params (p,q,1,2,3,4,5)
Case 2
A calls B with params (Z,alpha,beta,gamma)
B passes these on to C
C calls Z with params (p,q,alpha,beta,gamma)
Is there any way to pass the multiple values from A via B to C as one parameter, so that B does not have to know the maximum number of values that may be passed through? In other words, is there a way to place a piped parameter list into one parameter and then somehow make C "expand" the single parameter again into a list which can be fed into the X/Y/Z template call? Or am I stuck with passing them one by one?
-- TeraBlight07:25, 22 November 2006 (UTC)
The architecture is silly in this case of course, never mind that. The point is that a set of values needs to be passed along, in this case either four "booleans" or two numbers. Using (p1,p2,p3,p4) in the caller-side (/TemplateLab) and intermediary (/Month) code is quite ugly, while using (morning,afternoon,evening,night,first,last) all the way through is less ugly but quite clumsy. What I'd like is to be able to use a single param like so (pseudocode)
|timeType=1 |timeParams=[yes|yes|no|no]
|timeType=2 |timeParams=[12|16]
in the caller-side and like so
|timeParams={{{timeParams}}}
in the intermediary code.
The problem is that I would have to somehow "unwrap" this into the constituent values on the callee-side, either in /Day or in /Time. TeraBlight23:29, 23 November 2006 (UTC)
It's not possible to split up a text string using ParserFunctions. You would just have to pass all the paramaters across seperately. This shouldn't matter too much if it makes the template look ugly, because they generally aren't viewed and edited as often as articles. You may want to look at m:StringFunctions which might do the job but is not enabled on any Wikimedia wikis. Tra(Talk)23:53, 23 November 2006 (UTC)
Hmmm, yes, using direct string manipulation would kinda solve the problem, but since they're not currently available... I thought there might be some way to use a single container without turning the contents into a string (a sort of basic list or array functionality), or some way to turn a string back into wikitext so that the pipes and/or equals would work from some point on, like an inverse nowiki tag - anyway, thanks. TeraBlight00:11, 24 November 2006 (UTC)
It is impossible to parse strings in any way / convert a singular parameter into multiple ones. You may want to either:
rethink what you're doing - is there an easier way?
not be so complex in your templating, does it really need this functionality?
as suggested, pass parameters along one by one.
In the 3rd case, you should work out the maximum parameters that are likely to be passed, and code for just those. --Alfakim-- talk06:11, 24 November 2006 (UTC)
Bot Needed
I'm editing pages such as 7, 8 and 265 to solve the WP:BUNCH problem.
It's rather formulaic. I wonder if a bot/script/whatever could be written to save my time?
The problem seems to carry on right up to and beyond 1876.
See 1, 2, 3, 4, 5, 6 and 362 to see what I've done so far.
Also, there is little consistency in the disambiguation; sometimes it's {{otheruses}} and sometimes it's {{otheruses-number}} and sometimes it's both. perhaps they should all just be {{otheruses-number}} and any such page as 5 (disambiguation) should be merged with 5 (number).
--User2400:01, 22 November 2006 (UTC)
I've done a couple of changes to MediaWiki:Searchnoresults, to include links to site searches on the three linked engines - what with the index being so far behind (and the mediawiki search engine being so rubbish not very good). As this is likely to affect most people at some time, I'm posting here to invite discussion (crossposting to WP:VPN, please comment on this (WP:VPT) page). Thanks -- Martinp2322:53, 21 November 2006 (UTC)
The links seem to work well. Would it be possible to also put the links on the footer of the results page, to cover the situation when results are found but they aren't the ones you want? Tra(Talk)23:02, 21 November 2006 (UTC)
I'm not sure what the system message for the footer is (if there is one). The closest I can find is MediaWiki:Searchresulttext, where I've put the text, but commented out. This is the text at the top of search results which are successful - an admin can feel free to comment that out if there's a consensus to have it there (the footer would be much better, but I can't find the message). Martinp2323:22, 21 November 2006 (UTC)
What do you mean by "count"? Do you mean the links that you sometimes see at the left of an article which list other language Wikipedias with an equivalent article? Notinasnaid21:27, 21 November 2006 (UTC)
For the gory details, see Wikipedia:Interwikimedia link. To just see how it's done, pick an article which has these links, and use edit; look at the very end of the article. (Interwiki links don't have to be at the end, but that is strongly recommended). For example at the end of Blues I see [[wa:Blouze]]. You just need the two letter code for the other Wikipedia ("wa" in this example), and the name of the article ("Blouze" in this example). Notinasnaid21:51, 21 November 2006 (UTC)
There are codes to do this, which corrospond with the first two letters in the particular wikipedia site. This, for example is en, since the webadress is en-two.iwiki.icu. Just link to that, by typing [[CODE:ARTICLE NAME]], where CODE is the two letters mentioned, and ARTICLE NAME is the name of the article you are linking to. Prodegotalk21:54, 21 November 2006 (UTC)
I get an error message that there is a typo in the image name about 99% of the time before I finally succeed. I know the general rules about naming images. Is there an extra trick? Thanks! Mattisse(talk)20:02, 21 November 2006 (UTC)
Most of the time when that happens for me it's because I used a space rather than an underscore ( _ ) between words. The underscore'll show up as a space everywhere outside of the URL, but the software likes to check that the change is fine with you before it uploads. GeeJo(t)⁄(c) • 09:45, 22 November 2006 (UTC)
Uploading image problem
I want to upload a singles cover "kaw-liga_Hank_Williams.jpg" to the article Kaw-Liga (song). I have renamed it various ways but when I try to upload it I always get the message: The file you uploaded seems to be empty. This might be due to a typo in the file name. Please check whether you really want to upload this file. Yesterday I uploaded a similar file to Dust My Broom successfully. I don't know what I am doing wrong now. I have read the direction on how to upload. Any ideas? Thanks! Mattisse(talk)16:26, 21 November 2006 (UTC)
I have changed the image name to "Kaw-liga.jpg". It is the same in the article name. I cannot figure out what is wrong. I've tried every conceivable combination. Thanks! Mattisse(talk)17:30, 21 November 2006 (UTC)
Special:Upload asks you for two names: the first is the name of the file you have (e.g., the name of your file in your hard disk); the second is the name you want to give to the image once uploaded here. The error message means that probably you made a typo in the first name, that of your local file. I would also check whether the file you have and are trying to upload is actually empty. Tizio19:49, 22 November 2006 (UTC)
Bug #471 discusses the RSS/Watchlist idea. It looks like someone has a patch too. The biggest issue is that watchlists have normally been private, and therefore need password protection. The solutions seem to be (hopefully I'm summarizing these accurately) ... 1) those who read RSS within their browser probably already have a login-cookie, so they can just use that, 2) for those who use an external RSS reader, the only option is to have the user agree to make their watchlist semi-public, and then give them a URL with a semi-secret token in it to give to their RSS reader. --Interiot15:18, 21 November 2006 (UTC)
Interesting. The cookie looks like a nice idea (although I'm uncertain as to how long the account will stay "open" enabling the stream). Nice to see all the work that's going on behind the scenes... yandman15:25, 21 November 2006 (UTC)
The cookie would be the same one that you use whenever you access Wikipedia. Eg. You only have to log in every couple months if you never log out. --Interiot21:40, 21 November 2006 (UTC)
<nocat> tag?
I've run into a number of situations over time on Wikipedia, especially those involving categories embedded within templates, for which it would have been incredibly useful to have a tag that would cause the MediaWiki software to disregard all categories within the tags.
Is there currently a way to do this?
If not, do you think that this should be a requested feature?
Well, you can use <noinclude></noinclude>. Categories inside those tags will only categorize the template, but won't make articles to be categorized under them. -- ReyBrujo04:53, 21 November 2006 (UTC)
Thanks , but that's not quite the same problem - imagine that there is a template on a page which would be cumbersome (or inappropriate) to substitute, but which contains an undesired category. The page will be categorized with that category unless nowiki or includeonly tags prevent the template from rendering - which I can't imagine a solution for. Nihiltres05:48, 21 November 2006 (UTC)
Specific instances of this general problem are almost always fixable by using noinclude or an appropriate #if. What real problem are you trying to address? -- Rick Block (talk) 14:20, 21 November 2006 (UTC)
What's currently done is, on the template, to reference the category in the form: {{{category|[[Category:Example]]}}}. The template will still continue to work normally. Then, if you want to transclude the template without the category, you reference it in the form: {{example|category=}} Tra(Talk)23:38, 21 November 2006 (UTC)
Making Images Smaller
How can you change the image size for something? I mean not forever, but for, say, a userbox. The current image is too large, and I want the userbox to be of a reasonable size.--Kookoo27501:27, 21 November 2006 (UTC)
Without having put much thought into the matter, my understanding has been that only Wikipedia articles were supposed to be indexed by search engines (enforced by robots.txt or the equivalent). But I just noticed that my talk and user pages are the fourth and fifth hits on a google search for "opabinia regalis" - [8] - far outranking our article on the actual fossil, located at Opabinia and redirected from Opabinia regalis. Even my rarely-used Commons userpage outranks our actual article. This may make some sense, as my userpage is linked to more often, but all those links should be coming from sigs, which should also be in non-indexed talk pages.
I can't think off the top of my head of any other users whose usernames are plausible search terms, but I'm sure they're out there. Google giving unusually high ranks to Wikipedia articles is not surprising (see this signpost article), but doing the same to things out of articlespace is a bad sign, and just a further indication of how strict we should be on not allowing spam or advertising to stay in userspace long enough to get indexed. Opabinia regalis05:12, 13 November 2006 (UTC)
Right - I was trying to think of users with expectedly non-unique usernames - names of things you'd reasonably search for. Seems to be that it's a bad thing if we're just adding to the google noise by indexing userpages. Opabinia regalis14:16, 13 November 2006 (UTC)
It would quite possibly be a good idea to exclude userpages from robots.txt; OTOH, I'd be against excluding Wikipedia:, Help:, Category:, Portal:, or Image:. Not sure about Template:, MediaWiki:, or the various talkspaces (I can just about imagine an instance in which people would be googling for article talkspace, but the technical namespaces should give hits on sites about MediaWiki rather than Wikipedia). --ais523 14:21, 13 November 2006 (UTC)
Yes, I agree Wikipedia: (and WP:), Help:, Portal:, and Image: should be indexed - not sure how useful it is to search Category: but I don't see why not (possibly not anything under Category:Wikipedians?) IMO everything 'internal', such as User:, Template:, and the talk spaces ought not to be indexed. To be fair, the internal Wikipedia search function is poor enough that people might actually use Google to search these areas. Opabinia regalis01:31, 14 November 2006 (UTC)
The problem with this is that if someone did want to search for discussions/user pages etc, they would be forced to use Wikipedia's search engine, and would not be able to use Google (which is better) because the pages they are looking for have been restricted by robots.txt. It might be possible to deliberately lower the pagerank of non-article space pages by perhaps putting rel=nofollow on some (not all) of the internal links pointing outside of article space. I'm not sure if this would work, though... Tra(Talk)02:04, 14 November 2006 (UTC)
Maybe the problem is with the usernames? Though as my name is also that of an article, I guess I should stop right here! :-) Carcharoth02:08, 14 November 2006 (UTC)
Opabinia, I think the answer is obvious. Google has simply decided that you are more interesting than a fossil ;-). For the record though a search on Opabinia gives the fossil page. It is only search for the more detailed Opabinia regalis that gives you first (which is after all the title of your page, but only a side item on the fossil page). I think this is probably something that the world can learn to live with. Dragons flight05:41, 14 November 2006 (UTC)
Hmm. For Carcharoth, I'm seeing our article first on the Google search, and then my user page, suitably indented from the main result. That seems reasonable. Still, just to be sure, and as a matter of principle, I'm putting a WP:HATNOTE redirecting people to the article in case they come to my user page, or talk page. Carcharoth12:16, 21 November 2006 (UTC)
When I make changes on a page, I add it to my watchlist, but sometimes I only want to watch it for a few days to make sure the page is not vandalized again. Would it be possible to create a temporary watchlisting option?
By this, I mean, you could add an item to your watchlist for 3 (or 7) days, after which it automatically would be removed. If you edit the item during the 3 (or 7) days, the item's time on your watchlist would be renewed, so that it would be on for 3 (or 7) days from that point.
One of the benefits is to prevent editor (particularly admin) burnout by automatically removing short-term watchlist items so the editor does not become overwhelmed by a tremendous watchlist that he feels responsible for. This would be especially useful in fighting short-term vandalism on article, which takes up half or more of an editor's tim on Wikipedia.--Chris Griswold (☎☓) 00:28, 9 November 2006 (UTC)
Something you could do (but it is a bit more manual) is create a page like: User:ChrisGriswold/watchlist. When you edit a page, add a link to it there, and remove any more than 7 days old. You can check recent changes of the links on a page by clicking related changes, which would be -> Special:Recentchangeslinked/User:ChrisGriswold/watchlist. Now, if we could just get the devs to add a link to that on Special:Contributions.... --Splarka (rant) 08:34, 9 November 2006 (UTC)
I kinda like that idea. Whenever I go to remove user talk pages I've left various notices on from my watchlist I have a hard time remembering who I want to keep watching and who I just gave a "no image source" notice 3 months ago. The alternative would be to not watch the pages I guess, but sometimes people do reply to these things. Also my browser freeze for about 5 minutes when I click "remove selected items from watchlist" after a manual cleanup, and I "only" have ~8500 items on there. It should not be too hard to do either. Just add a "watch for X days" field next to the "Watch this page" checkbox, if left emtpy it's added permanently while if you type in a number an "expiration date" is added to a new field in the watchlist database. Then have a script remove all "past due" items for the user whenever Special:Watchlist is loaded. --Sherool(talk)08:41, 9 November 2006 (UTC)
I find checking User Talk pages for replies is best done from my contribs rather than my watchlist. --ais523 09:03, 9 November 2006 (UTC)
I heartily agree. A "watchlist for X days" option would be utterly marvelous. Even if it used a fixed quantity, like 60 days. I wonder if it's possible? --Quiddity09:52, 9 November 2006 (UTC)
Time of removal would be enough, the system just need to know when to remove the item from the watchlist, not when it was added to it. And yes it would require a database change, but adding new fields is not a huge problem, we just need to file a request for change on MediaZilla, get a few people to vote for it and hope some dev can be bothered to implement it ;) --Sherool(talk)18:09, 9 November 2006 (UTC)
No, not that way. I don't want to have to fill in a field or hit a check box every time I submit an edit. Also, I don't know that I can say, at the time of editing, how long I'll want the page on my watchlist.
I'd rather see a text column on the watchlist page telling me how many days since I last watched it. This needs to display as well when I edit the list. Then I can easily go down the list, saying, "Hm, I haven't had anything to do with X in 6 months and I can't think of any reason to keep it watched. Bye." That's how I'd like it. An export feature would be nice, too. John Reid °04:24, 11 November 2006 (UTC)
Well obviously it would be optional. You wouldnt have to fill in or check anything, it would be an optional extra.
As mentioned above, a specific case like leaving welcomes/warnings on user talk pages is where it could come in handy. But your method is good too, and could work in conjunction with each other. -Quiddity22:18, 11 November 2006 (UTC)
Sherool summed it up perfectly just above. Bugzilla makes me nervous, else I would do it. And expect to wait a long time, enhancements are low priority and devs are too few. --Quiddity08:07, 13 November 2006 (UTC)
If the toolserver ever gets back up, I planned to implement exactly this. Basically, the recentchanges table keeps the last.. 7? days of edits. Basically, use recentchanges to get a list of all articles the target user has edited that are still in the table, and then check the recentchanges table again for all changes to those pages. I haven't gotten a chance to run the query on the toolserver to see how fast it is... I assume it's a little slower than the watchlist, but the recentchanges table is supposed to be fast... --Interiot06:33, 16 November 2006 (UTC)
Please do. I almost took a huge break from Wikipedia but realized that the major weight on me when editing was trying to get through my watchlist. This will make things so much easier. --Chris Griswold (☎☓) 09:20, 24 November 2006 (UTC)
We can only use this link suggester when the syntax error has been resolved, please can someone help? Tom13:11, 21 November 2006 (UTC)
Tiny images are too small to be read and high resolution download shows text rather than a larger image
Hello:
Some of the images on Wikipedia are too small to be read by the human eye alone. For example,the link: French monarchs family tree contains a very small image: Image:France-2ndCapet.png. This image is far too small to be read; And the high resolution download shows text rather than a larger image. How can these tiny images be made bigger and more legible for the human eye?
The page looks fine to me. What OS and browser version are you using? As for Image:France-2ndCapet.png, after clicking through to see the image, using some browsers it does load to fit the height of the browser window, resulting in illegible text - clicking on it once zooms in to 100% size (at least in Firefox 1.5.x). The latter isn't an actual usability issue, of course. Tempshill06:18, 21 November 2006 (UTC)
Saving, but not posting...
Hello
I would like to say i am very upset that wikipedia does not have a "save, but do not post" button. If, in fact they do have such a feature it needs to be more prominant. I had a very upsetting moment when i was wokring on a new article and I walked to get a drink. When I came back, wikipedia had been closed and all of my hard work lost. It could be placed right next to the save changes and post button. The button would save the article to your contributions. please correct me if I am wrong but I do not believe there is such a way.--Gwakamoley18:07, 20 November 2006 (UTC)Gwakamoley
All changes to wiki pages are immediately live, by design, so you can't save a draft and not affect the article while you're still editing. If you'd like to make a draft copy to work things out, you might try using a subpage either in articlespace or your userspace (like a personal Wikipedia:Sandbox). Copy all the text into User:Gwakamoley/Draft sandbox or Talk:Article name/Temp, for example, work out your edits there, then select all again and copy it back to the article, being careful to re-include any edits made in the meantime. Does that help? -- nae'blis18:16, 20 November 2006 (UTC)
Yes, but I still think that such a feature should still be added... unless it is impossible (if thats what your sayin).--Gwakamoley18:20, 20 November 2006 (UTC)
Pretty sure there isn't a way to do that. If you're doing a major overhaul of a project, I'd suggesting doing your writing in a program on your computer; you can save your progress as much as you want. Other than that, I'd suggest using a subpage of your userspace so that you can work on it anywhere. EVula// talk // ☯ //20:36, 20 November 2006 (UTC)
Some more good reasons : WP is not your own archiving device ; any good text processor uses proofreading functions that are recommended before posting ; you are allowed to read again once (preview button) things that you would not like to post raw, &c. -- DLL .. T21:37, 20 November 2006 (UTC)
I've gotten hosed, too. The solution is to select-all, and hit ctrl-C to copy to the clipboard. And then paste it into Notepad or Word (or whatever) if you're extra paranoid. This has saved me a couple of times. It sounds like a low bang-for-buck as a Wikipedia feature, and I would think a 'save without posting' feature sounds confusing in any case. Tempshill06:21, 21 November 2006 (UTC)
When I write a lot, I too copy the text to the clipboard before posting it. If something goes wrong, I can then try again. Sometimes clicking 'back' on the browser helps, but if the browser dies, or the computer dies, then there is not a lot that can be done. For major rewrites, working in an external program with 'save' functions is best. But, as others have said, remember to re-add edits made since you rewrote the text, and credit them in the edit summary (say something like 're-adding edits first added in version xyz'). Carcharoth11:43, 21 November 2006 (UTC)
Horizontal scrollbar
Since this morning I've been seeing a horizontal scrollbar in Wikipedia article pages (including user and talk pages, but not the watchlist) in Firefox 2.0. The horizontal scrollbar does not appear in IE. Does anyone know if the CSS or ouputted HTML has changed in MediaWiki. -- Jeff300023:48, 19 November 2006 (UTC)
Maybe this depends if the text contains "pre" tags ; the line is not automatically cut at the end of a window line, and firefox then offers a scrollbar. -- DLL .. T21:39, 20 November 2006 (UTC)
Hi, on the above article section, there's a link that points to a postscript file on someone's web page, but when the article is viewed (using Internet Explorer), it does not show the usual graphic to indicate that the article is a URL rather than a wikipedia entry. --Rebroad10:42, 19 November 2006 (UTC)
In IE7, I notice these external links show correctly when they are on one line, but if they wrap to another line then the arrow doesn't show. I think this could be due to aa problem with how the css is rendered. Tra(Talk)14:48, 19 November 2006 (UTC)
I've been converting a list I'm working on into table form, located at User:VegaDark/Sandbox and I've been inline sourcing every entry. There are so many football players that use the same source that I have gotten past bz on the reference count at the bottom of the page and it is showing some sort of error. Only a developer can add the ability for ca-cz and beyond (by the time I am done, I may need through the d's or e's). Any chance that can happen? Thanks, VegaDark01:49, 19 November 2006 (UTC)
I just realized this could be considered a bug/feature request and should be posted on bugzilla, but I don't have an account there...anyone who does feel like submitting it? Thanks. VegaDark02:10, 19 November 2006 (UTC)
Thanks, that worked. It now goes through dz. However, now it stretches the page instead of word wrapping. Any way for it to word wrap without having the developers change the programming? VegaDark07:58, 19 November 2006 (UTC)
Done, let me know if it works (they're all on one line on your sandbox page for me). But seriously, maybe consider an alternative referencing scheme if you need up to dz for a single source. Opabinia regalis02:51, 20 November 2006 (UTC)
It worked after I purged the page, thanks. I originally had a different referencing scheme but when I put the page through peer review it was recommended I inline cite every entry, or it wouldn't get to featured status. Feel free to take a look to see if you can come up with a better idea. VegaDark04:11, 20 November 2006 (UTC)
I know nothing about featured lists specifically, but I'd say the inline citation creep lately has gotten way out of hand if we're getting up to dz in the footnotes. People have, in the past, evaded this problem by including individual notes with page/section/etc numbers and the author's name, and listing the full citation for the work in a separate section (I understand this is common in WP:MILHIST-supported articles). A simple option would be to use the old {{ref}}/{{note}} system, which is clunkier in some ways but doesn't produce individual notes for each time a work is cited, or just include a linked asterisk that goes to this particular oft-cited work in place of a note. Don't know if any of those would suit your specific purpose, though. Opabinia regalis05:40, 20 November 2006 (UTC)
Mmmmm. Hmmm, I see that now, but I'm inclined to agree there should be another solution then. Sorry for misreading the discussion. -- nae'blis20:48, 20 November 2006 (UTC)
The tool generated some information which was considered private. As a result, the statistics are unavailable until this information is purged, and the tool is corrected so that this does not happen again.
Inferring from a thread on wikitech-l, the tool's developer, Erik Zachte, and our CTO, Brion Vibber, are communicating to this effect, so the statistics might be restored in the near future. 164.11.204.5620:41, 20 November 2006 (UTC)
Changes in auto-summary feature?
I noticed that there is now a (replacing page with 'new page content') auto-summary when a vandal blanks a page (example); however, the one I really liked - the auto-summary that was provided when you make a redirect - is gone. Can the developers bring this one back? :-( Kimchi.sg13:30, 17 November 2006 (UTC)
I figured it out. Apparently the new redirect autosummary requires there be no space between the "#REDIRECT" and the [[. Kimchi.sg05:35, 21 November 2006 (UTC)
The same code is used to check whether to generate an autosummary and to check whether something is a redirect. Could you link to an edit where the summary wasn't generated but the resulting redirect worked? —Simetrical (talk • contribs) 06:08, 21 November 2006 (UTC)
Where are Wikipedia article authors located?
I posted this query on one of the help pages, and somebody suggested I try the Village Pump, too. I'd love to find information on something I know is technically feasible, I'm just not sure we actually do it. I'm interested in what countries' residents give back to the Wikipedia.
For instance, where do the people who contribute to en-two.iwiki.icu live? Are they primarily in countries where English is the first language? Or are there thousands in places where English is only a second language, but widely spoken, whether that means the Netherlands or India? I'm not interested in whether the contributors are really native English speakers or not (well, I am, but we can't really know that about each one without asking them.) I'm simply interested in where they are, because we can track that through their IP addresses. And of course I'm not interested in who they are, either, this is not an invasion of any individual's privacy. Yes, articles originating in India (according their IP address) could have been written by, say, a Briton temporarily (or even permanently) residing there, but if there are a few hundred thousand articles from non-English speaking countries, I think it would be safe to assume that they weren't all written by expats!
Where do the people live who contribute to es.wikipedia.org? Spain, South America, Central America, the U.S.?
Where do the people who contribute to zh-two.iwiki.icu live, or ko.wikipedia.org or ... are these people in their respective diasporas, socalled "overseas" Chinese, or Taiwanese, or people living in the People's Republic?
Wouldn't this be easy to track? And might it not give insight into why some Wikipedias grow so much faster than others?
You could try to find out by checking nationality userboxes on those wikis that use them; you could also check 'Recent changes' for anons and check the IPs to see which country they came from. --ais523 09:40, 17 November 2006 (UTC)
There have been some cursory statistics done on this; see meta:Edits by project and country of origin. No pretty graphs or anything, but the data is there. Perhaps most interestingly, for Chinese: HK: 28.6%, TW: 25.9% ... CN: 6.1% (though bear in mind that we'd have been blocked in mainland China for much of that time). Shimgray | talk | 20:54, 20 November 2006 (UTC)
Trouble with SVG image
Caption
I'm having trouble with the following SVG image. The file loads fine in IE and FireFox. Is there something I can do to fix it? -SharkD09:41, 16 November 2006 (UTC)
It's a known bug in the version of libcroco (used by librsvg) that is used on Wikipedia. It's fixed in upstream in libcroco, but an old version is still used on the Wikimedia's servers. See the bug report for a workaround… –Gustavb16:26, 16 November 2006 (UTC)
Workaround: convert the text to paths (in Inkscape under Path->Object to Path). Don't use line endings, these are problematic as well. Set the line ending to "none" and use separately drawn triangles or arrowheads, positioned appropriately.
BTW, as the image is supposed to show a sphere, I think a perspective similar to this picture, but rotated maybe another 30° to the right might be better. Lupo12:05, 21 November 2006 (UTC)
Name technical restrictions workaround
Someone pointed out above that the fr:IPod article displays its title as "iPod", rather than "IPod" as we do here. This looks much nicer than making the user read "due to technical restrictions, blah blah...". The french wikipedia accomplishes this by adding a site-wide javascript function that checks for the presence of their {{lowercase}} template, removes the lowercase notice, and changes the title to be the lowercase version. (search for "RealTitleBanner") If the user doesn't have javascript enabled, it just falls back to displaying the "due to technical restrictions..." blurb.
Gerbrant has managed to get this working on enwiki as well. You can make this work now by adding {{subst:User:Gerbrant/realTitle.js}} to your monobook.js.
Two things: first, as an enhancement, it may as well fix the HTML title as well as the displayed one. Two, I strongly suggest that to avoid confusion, you be careful with something like C Sharp where the desired title won't work in wikilinks. It would be neat if this were committed sitewide, but for initial case only. Being unable to display lowercase is a silly flaw in the software, but it's a bad idea to make the name unpasteable. Maybe okay with an explanatory message, but not much point then, is there? You could still fix the title for the HTML title and for print, though, even where it's no good for pasting. —Simetrical (talk • contribs) 05:22, 16 November 2006 (UTC)
No idea how much of that will break horribly, and no idea how to get different output for print, and certainly no idea how to get correct display for anything other than direct viewing of the article (e.g. edit pages). —Simetrical (talk • contribs) 05:36, 16 November 2006 (UTC)
Okay, the title works okay, see User:Interiot/js/RealTitle.js (I based it off of the french code, since it has a bit of error-checking and has been in use for a while). Per your suggestion, I made a new div id, "RealTitleBannerUnpastable", for when the title given is intended to not be pastable (in this case, I only change the document title, not the H1 as you said). (is it pastable, or pasteable?) However, it's not always straightforward to predict whether it will be pastable. Most articles tagged with {{underscore}} substitute a space in the name where an underscore would be, however a few do not (Shift-JIS) or can not (NSAKEY, though "_NSAKEY" is mostly pastable). Even with the more straightforward {{lowercase}}, articles can give a completely different title as the first argument... this is typical for (disambiguated) titles such as Occam (programming language). I'm not sure what to do about some of those.
Fixing the title for print would probably be done with something related to the CSS @media print thing, but it's not obvious right this second how to make the mechanics work. --Interiot07:37, 16 November 2006 (UTC)
One thing that could be done to make sure the name is pastable is for the javascript to actually check whether one name maps to the other (eg. if you take the RealTitle, uppercase the first letter, change all underscores to spaces, if you get the current page name, then it's okay to display the RealTitle instead). Besides fixing small mundane issues, that would also prevent more extreme attempts to have fun^W^W confuse people. --Interiot17:09, 16 November 2006 (UTC)
That's what I did in my JavaScript above, or at least that's what it's supposed to do. rtb.substr(0,1).toLowerCase() == doctitle.substr(0,1).toLowerCase() should only be true if they're the same except for their initial letter's case. Except looking at it again, that's totally stupid and wrong, and should be rtb.substr(0,1).toLowerCase() + rtb.substr(1) == doctitle.substr(0,1).toLowerCase() + doctitle.substr(1), I think. Other transformations could also be made, like underscores/spaces (although I guess only ones that are neither the first nor last character?).
Also, maybe something could be worked out for the messy cases that preserves pasteability (no, I don't know whether that e belongs there). Like, what if you did something funky with transparency and z-index? Put the real title on top but invisible, the display title underneath and visible. The problem with that would be if the real title is longer than the display title (C Sharp), people would be prone to select less of the title than they should. Also, it would be distinctly unexpected for the user. I'd stick with canonically equivalent names only for now. —Simetrical (talk • contribs) 00:22, 17 November 2006 (UTC)
Simetrical's right - titles should remain pasteable. Any problems with usability veto things looking nice, as annoying as that may be. My personal suggestion, although it risks breaking format, would be to use a double title - note what the subject is with one and where it is listed with another. For articles like C sharp, we'll have trouble with intuitive links anyway - people knowledgable about the subject but not about Wikipedia may still write [[C#]] and hope to get to the right page. Nihiltres00:44, 17 November 2006 (UTC)
Would it be possible to implement the fix for every article, and then add a mouseover or select event to the title text so that the pasteable text will either appear and/or be copied respectively? --DavidHOzAu03:03, 17 November 2006 (UTC)
Sneaky, but I'm not sure I'm in favor. Some people will be inclined to type out a link to the title rather than copying it, and there's nothing we can do to stop that. —Simetrical (talk • contribs) 01:11, 19 November 2006 (UTC)
Okay, I implemented the "check in real-time if it's pasteable, and only replace the <h1> if it is" code (see User:Interiot/js/RealTitle.js). At the same time, that solved another problem I had, that usually a page's namespace usually isn't mentioned in the "due to technical restrictions" blurb (but sometimes it is). So I also automatically detect whether the namespace is there, and add it if it's not (again, to make sure the name is pasteable). At this point, I see no obvious downsides to the script (other than possibly its complexity), so I'd like to try to add it to monospace.js soon unless there are objections. (the minor downsides are that I don't have printing working, and a few weird cases like lac operon don't work, but I think for 97% of its uses, it does the best thing possible) --Interiot04:42, 21 November 2006 (UTC)
I'm also having this problem - both IE and Firefox seem unwilling to load from upload.wikimedia.org. I've tried changing block settings but the fact that both browsers are unhappy seems very odd. Any ideas? Greycap23:09, 20 November 2006 (UTC)
Wikipedia is spelled wrong
The new spell checker is great; but did anyone notice how wikipedia, Wikipedia, WikipediA, no matter where you put capital letter is underlined with a red line? I think thats kind of weird and even though it doesn't matter that much, I still think it should be corrected. Pseudoanonymous21:50, 13 November 2006 (UTC)
The new spell checker is part of your browser, not Wikipedia ;) All sorts of things are missing from its built-in dictionary (taken from openoffice), just right-click on them and "add to dictionary". --Quiddity22:02, 13 November 2006 (UTC)
On this pic for example I see a kind of pixelized, crimsoned rings on the background. However nothing like this appeared when I viewed the image on laptop. Is it an image or monitor defect? --Brandспойт15:32, 15 December 2006 (UTC)
The picture looks fine to me; the colour-depth on your screen settings is probably set too low. --ais523 15:34, 15 December 2006 (UTC)
My user subpages
Is there some way of making my top, while-logged-in bar, display a link to my two subpages alongside my main user page? So instead of:
Bobo192 my talk my preferences my watchlist my contributions log out
I actually want it to read:
Bobo192 Sandbox (linking to User:Bobo192/Sandbox) MAP (linking to User:Bobo192/MAP) my talk my preferences my watchlist my contributions log out
Is there some uncomplicated way to add these pages to the top of my pages, by editing either my Javascript or CSS page, and is all I need to do after that to clear my cache and refresh the page? I'm not experienced in either medium but have always wanted to do this for quick access. Ideally the bar should still be on the right-hand side at the top. Thank you. Bobo.07:56, 15 December 2006 (UTC)
Thank you Splarka. That's exactly what I wanted. I attempted the other version but it didn't appear to do as I wanted, so I've put it away and will look to see what it does later. Bobo.18:11, 15 December 2006 (UTC)
Adding transcluded page edits to watchlist?
I don't know whether this is the right place to propose a feature for MediaWiki... but since this is the largest use of the software (and the only place I have an account) I'll try to get a discussion started here. I know that when a page is edited, the MediaWiki software makes a null edit to all pages that transclude that page. My question is... would it be possible (or "feasible" -- I'm a developer so I know how annoying the phrase "is it possible" is) to be able to watch those null edits as well as regular edits, so essentially I could be alerted when anything on a certain page is edited? For instance, if I was watching Wikipedia:Requests for adminship, I would see a notice that a change was made when any of the subpages were edited. I personally would find this very useful... anyone know if it could ever get added to the software? -- Renesis (talk) 07:43, 15 December 2006 (UTC)
Here's one place that enhancement requests can be discussed, but if you want the opinion of developers it's probably best to open an enhancement request on Bugzilla. --ais523 12:05, 15 December 2006 (UTC)
recent changes
I got a quaestion. In this news article, they said that they have fixed the recent changes so you could see how many characters were added and remove to remove vandalism. But I can't see this feature, is this because I'm using the wrong computer, or is it because this feature hasn't been uploaded yet in the recent changes.--PrestonH05:26, 15 December 2006 (UTC)
The changes haven't been uploaded yet to the Wikimedia servers. The Signpost article you linked says that the changes were introduced in revisions 18237 and 18238, but at the time of writing, special:version was at r18226. It doesn't usually take this long for changes to be copied from the Subversion Repository to the Wikimedia servers though. Graham8710:55, 15 December 2006 (UTC)
Still at r18226. I recall reading on wikitech-l that the revisions weren't being uploaded onto the servers as quickly at the moment, but I can't remember why. --ais523 12:03, 15 December 2006 (UTC)
Several recent code changes involve database layout changes. While these should be relatively minor, we still need to arrange the servers carefully while making the update (there will be brief downtime while switching databases). Will get done today or tomorrow, with load near weekly low. --brion20:34, 15 December 2006 (UTC)
YouThe editor before you forgot a slash in the closing tag on the last ref in the article. Fixed. --*Spark* 04:01, 15 December 2006 (UTC)
I just noticed the change as that article was listed before this page in my watchlist. I supposed I should have noticed it. Sorry to bother you. Will (Talk - contribs) 05:44, 15 December 2006 (UTC)
Searching problem
For the last few days, whenever I search for any page on Wikipedia that doesn't exist, the "There was a problem with your search" text comes up every time, with the Google and Yahoo links. –The Great Llamamoo?02:50, 15 December 2006 (UTC)
Over at WP:AN3RR, violation reports are very time-consuming to fill out. Going back and forth from page history and noticeboard, copying all the diff urls as well as timestamps (that's a minimum of five diffs and five timestamps, back and forth makes that a minimum of twenty page switches). It would be more than awesome to have some kind of automated system that helps the user fill out a report. Anybody feel my pain? — coelacantalk — 00:41, 15 December 2006 (UTC)
I thought I had seen an automated tool for doing a 3RR report, but I do no recall where. I think it is purposely made tedious to discourage frivolous reports. --*Spark* 22:06, 15 December 2006 (UTC)
Issue with skin variable
More JS issues that I've been having. As you can see if you view the source of any Wikipedia page, there is a Javascript near the top that defines several variables, including skin="monobook"; or skin="standard"; for classic. I have tried to take advantage of this in several scripts:
function myFunction() {
//Function code
}
if(skin=="standard") {
addOnloadHook(myFunction());
}
Just to jump on the critique bandwagon: that js only loads for the standard skin, so checking if it is the standard skin is kinda silly. --Splarka (rant) 08:22, 15 December 2006 (UTC)
To clarify: my bad JS above (the unclosed brackets and badly formed comments) are not problems with my original code. My original code works, as can be attested to both by some thorough parsing on my part and by Firefox's JS console. Also, if "that js" refers to the script that assigns var skin="skin name", then it doesn't just load for the standard skin, it loads for any skin. Check the source of a page with monobook. Karl Dickmantalk04:39, 16 December 2006 (UTC)
What about User:Userscripts/Editcount link/source.js. This script is not skin-specific, and is meant to be copied to other user's scripts with document.write. Yet the addition of a test for the skin caused the script to break. Why?
copywrited image tag
I am trying to find the correct tag for some images. They are copywrited by my local newspaper company, but I have permission from the media coordinator to use them. What image tag should I use so they are not deleteted?
Cabman
In a template that I am placing on a talk page I assigning a Catagory. Right now it is set up with {{PAGENAME}}. Unfourtunatly this is setting the talk page into the Castagory where I would like the parent of the talk page in the Catagory to show. Does anyone know a method of doing this? Markco123:29, 13 December 2006 (UTC)
To do what you're trying to do, you would need to add a template or category link to the parent page. If the template is designed to only be added to talk pages, then adding the talk page to the category is normally considered an acceptable alternative. Tra(Talk) 00:07, 14 December 2006
(UTC)
Actually I did some research and it appears that {{BASEPAGENAME}} returns the Parent of the talk page. Works just placing straight on the page but I am not sure why this not working in my template.Markco117:37, 14 December 2006 (UTC)
I suspect you're doing something like [[Category:My category|{{BASEPAGENAME}}]] ? When you place a category link, the only parameter you have is the sort label which affects how it shows up on the category page, but not the link itself. --*Spark* 17:51, 14 December 2006 (UTC)
Yes you are correct that was what I was attempting and all I am doing is sorting. shoot no way around this I guess Catagory just does not allow me to do what I would like to do. I will research some more maybe I can hack something together. I just do not want to clutter the Parent page with my template but I do want to catagorize it via the template - there has got to be a way. Markco118:01, 14 December 2006 (UTC)
I suspect the potential for vandalism by allowing pages to be categorized from a different page - you'd never see the edit if you're watching the page, and it could be tough to track down - would prohibit this functionality from ever being implemented. --*Spark* 18:12, 14 December 2006 (UTC)
Searching Wikipedia
I love the site but is there dicussion of a better search engine? I can find information easier using Google to search for information in Wikipedia than searching in Wikipedia. For instance, I did a search for GDDI in Wikipedia and it returned with no information about GDDI in Wikipedia but if I do a Google search, then it refers me to the page in Wikipedia that I can find it. So this is a great concept but if you can't search it, then it is no serving it's purpose.
Google is pretty much the best search engine for most sites. Why would you want to use a local one, when you can just add "site:..." to any Google search? Notinasnaid00:23, 14 December 2006 (UTC)
I just came upon a similar problem and not sure if this needs a bug report or if I am missing something. I did a search for "Goertzel" and Wikipedia returned 0 search results, however, there is clearly an article titled "Goertzel algorithm". Why did a search for a word contained in an article title not return a result for that article? --MattWatt01:58, 14 December 2006 (UTC)
Currently it returns many results, of which the article you mention is the second listed. Can you confirm whether you encountered a transient error? If so, the system should have shown a message indicating that the search servers could not be contacted. --brion10:44, 14 December 2006 (UTC)
Ok so maybe I've just lost my brain temporarily. Whenever I use the search box on the left hand side of the page I get no results. It says "No page with that title exists.", but then it gives me the choice to search for pages with that prefix and it returns many results. I had just assumed that the search would return all articles that included that search keyword. I didn't realize it was looking for an exact match. I find it very strange that the search feature would operate this way. --MattWatt22:20, 14 December 2006 (UTC)
You might notice the search form on the side has a bold Go button, which is the default action if you press enter in the box. To search instead, click Search. --Splarka (rant) 08:17, 15 December 2006 (UTC)
New Bot Idea
I have no idea how to create a bot, but I think a bot for removing red links would be a good idea.
No removing the word (e.g. Sir Colt Watson would become merely Sir Colt Watson, the [[]] would be reomved. Any thoughts? Do we have one already?
disabling auto-refreshing with Internet Explorer 7?
Before upgrading to IE7, I could usually enter changes, visit another page, and come back without losing the changes. However, IE7 seems to auto-refresh whenever I use the "back" button, and any unsaved changes would be lost. Does anyone know how to disable the auto-refreshing? Thanks. --Ixfd6401:19, 14 December 2006 (UTC)
Click Show preview immediately before clicking on a hyperlink and your changes will still be available when you click the back button to go to the Show preview screen. However, if you then make any further changes in the Show preview screen, you will need to click Show preview again before following hyperlinks. Tra(Talk)01:38, 14 December 2006 (UTC)
The setting you probably want to change is "Check for newer versions of stored pages" in the "Temporary Internet Files" area of settings. [11]. --Splarka (rant) 09:36, 14 December 2006 (UTC)
Blocking pages
Hello,
I am a new user to editing. I just edited my school's page, but noticed that it was getting vandalized a bunch, mostly by people from my school while school was in session. Is there a way I can just block unrregistered users from my school from editing that page? — Preceding unsigned comment added by Cabman (talk • contribs)
There is nothing you can do to prevent their editing, but an administrator could either semi-protect the page, and prevent all unregistered users from editing that page, or block the schools IP, and prevent unregistered users from editing anything. However, in this case neither seems to be necessary, just revert any vandalism you see. Happy editing! Prodegotalk00:23, 14 December 2006 (UTC)
Our robot overlords
I'm a fairly experienced editor, but I've never had the occasion to use any automated robots. Right now I have a task which I think would be perfect for a robot to a one-time job. However, I have no idea how to use or create them (I'm a programmer by trade, so I'm fairly technically adept). I need to update all the links from Accolade (company) to Accolade (company). Anyone know how to get (or develop) a wiki-robot to do this? TIA! — Frecklefoot | Talk16:23, 13 December 2006 (UTC)
That sounds like the sort of job that's normally done with WP:AWB; you could also try the requested bots page. --ais523 16:40, 13 December 2006 (UTC)
The WMF uses rsvg, which this is presumably a bug in. This probably needs to be fixed upstream, since we aren't exactly going to be forking. I suggest you upload a version with text converted to paths for use over the current version, which could thus still be accessed by anyone looking to modify your work and could be reverted to when support improves. —Simetrical (talk • contribs) 17:22, 13 December 2006 (UTC)
Transclusion edits
If you transclude a page through a transcluded template, is there any way to make the little [edit] boxes next to the headings in the transcluded page not edit the template instead, or, at least, to suppress them? For an example of what I'm doing, check out the transcluded Peer review on Talk:Charles Darwin. Adam Cuerdentalk04:22, 13 December 2006 (UTC)
I've removed it from there, since it also breaks edit functionality, but since it mostly works on other pages (except the auto-set header....
I suggest removing it once everyone's had a chance to look.
Only way I can think of is to use <noinclude> and related tags in the article so the header is replaced with large bold text when transcluded and displayed normally otherwise. Anyone viewing the page directly shouldn't notice a difference from how it looks now. --*Spark* 15:25, 13 December 2006 (UTC)
I've made the change, see box above and the associated page. --*Spark* 15:29, 13 December 2006 (UTC)
IPA confusion
I'm using Opera 9.02 and the IPA transcriptions are giving me considerable trouble since ɪ and ː appear the same to me. Which font should I change to make them display differently? eszetttalk02:25, 12 December 2006 (UTC)
Until the moment, in every typical font I saw they're different:
Wikipedia default edit toolbar not displaying button images
Finally found the cause of this longstanding bug. It's a Firefox upgrade bug, one most likely to be seen by users who upgraded from Firefox 1.0 to 1.5 or later. In 1.0, there was a configuration checkbox to turn off image changing from Javascript, which was an early way to block ads. In 1.5, the checkbox was removed. But the 1.0 to 1.5 upgrader didn't reset the related configuration parameter. So now the user is stuck.
Then, a few months back, the Wikimedia developers got fancy and started building the toolbar entirely in Javascript. Images inserted that way were thus blocked.
If you have this problem, here's the fix.
Open about:config in Firefox. Just type that in the regular URL window.
Find dom.disable_image_src_set.
Right click on it, and select "reset" to reset it to the default value ("false").
As a result of deleting superfluous fonts I am now getting Wikipedia pages in an ugly upper case type. What font should be used to display the headings and articles? Laurie
There's no default font for Wikipedia. Your web browser should try to display it in a sans-serif font by default, so installing any one of those (including Arial, the most common), should fix your problem. —Mets501 (talk)04:10, 13 December 2006 (UTC)
{{Computer Magazines}} is broken and breaking layout on pages it's on because it's transcluding {{sprotected2}} from somewhere, even though it's not directly included. I looked at both {{tnavbar}} and {{navbar-header}} and can't seem to figure out how it's getting into the other template. Anyone know where this is coming from? Hmm, it's been fixed now. found the culprit [12]. Night Gyr (talk/Oy) 23:16, 12 December 2006 (UTC)
Large reference sections
From Wikipedia talk:Citing sources. Some articles have understandably large reference sections. Now I'm all for this but some have observed that it makes the articles a bit unwieldy and bottom-heavy, even with two-row formatting and tiny text. I'd rather not crop references to correct this (as has been suggested elsewhere). Would it be possible for us to impliment a "hide" function for references in the same manner as the contents sections can be shown/hidden? Sockatume21:29, 12 December 2006 (UTC)
Subtle form of vandalism which I can't get rid of.
It's come to my attention that IE6 does something bizarre when previewing an edit. Try editing a page, click preview, and scroll down about to where the edit summary should usually go. The exact nature of the problem is well over my head, but it's clear that something isn't parsing quite right. Ideas? Luna Santin10:03, 12 December 2006 (UTC)
I noticed this when copyediting an article which included {{PMSA}}: you can see the effect by going here and clicking the "Show preview" button. Firefox 1.5.0.8 apparently has no such problem, so it's probably one of those CSS things which makes dealing with IE6 such fun. —Phil | Talk10:11, 12 December 2006 (UTC)
I had the same problem at Physics (edit) and my user talk (edit). On further investigation, it doesn't happen everywhere. Not sure what the link is -- that template, {{PMSA}}, may be involved, but it's not the sole cause.
After some further investigation between the two of us, we seem to have tracked it down to wikitables which include "align=center" -- case in point:
I notice that sometimes I see an item in my watchlist, then when I do a page refresh it disappears. Why is that? It makes me think that I can't rely on my watchlist to keep me informed of changes. Thanks! Tanaats05:09, 11 December 2006 (UTC)
If the article was deleted, moved, or protected, then it will not be listed on your watchlist, at least until another edit occurs to the page. It's a commonly-reported issue. Titoxd(?!?)05:11, 11 December 2006 (UTC)
I kind of doubt that the article was deleted, moved, or protected. Recently I hit "diff" on an item, went over to the article Talk page and responded to something, then came back to the Watchlist to find that the item was now missing. Is that explained by the problem you described? Tanaats05:16, 11 December 2006 (UTC)
Check your preferences. There's a setting in the Watchlist tab that causes your edits to be hidden from your watchlist, and it sounds like it could be the issue in this case. Titoxd(?!?)07:55, 11 December 2006 (UTC)
Yes, after you edit this page, this page (and subsequently, my edit) should disappear from your watchlist... did you check your preferences? Titoxd(?!?)02:04, 13 December 2006 (UTC)
Ah, that begins to explain things. What should I check in my preferences? On the Watchlist tab I have "number of days"=3, and all the checkboxes off. Tanaats04:06, 13 December 2006 (UTC)
Go to the history, find the most recent unvandalized version, click on its date: "Revision as of X time, Y date", "edit" that page without changing anything, type "rvv" or something similar for your edit summary, and save. The article will be reverted to the previous version that you selected. Alternatively, diffs now have edit links for reverting single-level vandalism. You can click "edit" in the diff on the version before the vandalism and similarly submit the old version without changing it. If vandalism has been buried by legitimate edits, there's an experimental "undo" function in old diffs - try it. Hope that helps, Nihiltres14:08, 12 December 2006 (UTC)
The semi-protectection on my userpage undid itself... I know this because my userpage was vandalized by an IP recently. (See the most recent history of my userpage.) What gives? Is this normal? There is no log entry of it being removed. Grandmasterka03:13, 12 December 2006 (UTC)
Wouldn't it be great to be able to add a page to your watchlist for a specfied period of time, say, 2 days? You could watch talk pages on which you have posted, or articles where you have reverted vandalism, without ending up with a 971 (and counting...) page watchlist. Solutions? yandman18:50, 11 December 2006 (UTC)
Yes. It should be possible to have an alternative view (either in mediawiki, or on the toolserver if it ever comes back up) that finds all the pages you've edited in the last n days, and shows you the recent changes on those. It's more indiscriminate than what you're describing, but it would be useful nonetheless and would be relatively easy to implement (it's just an SQL join of the recentchanges table with itself). --Interiot19:58, 11 December 2006 (UTC)
Fixing up columns to fit in next to an information box
Try it this way instead: rather than spatchcocking the infobox into the columns, which will totally confuse anybody wanting to edit the article after you, kill off that third column and restrict the width of the columns to allow the text to rise up next to the infobox. It still drops down if you reduce the size of your browser window far enough, but this should come as no surprise if you're mucking about like that: furthermore the content of the article should now be rather easier to edit. HTH HAND —Phil | Talk10:22, 12 December 2006 (UTC)
Diff and hist links
This is a pretty minor issue in the scheme of things, but is there a particular reason why watchlists and related changes pages list the links in the order (diff) (hist), but lists of contributions have them in the order (hist) (diff)? Amazingly, I never noticed this until today. Opabinia regalis05:18, 11 December 2006 (UTC)
No, I don't think there is. It's probably just the way someone originally coded them. Should be trivial to fix, but I'm not sure if it's worth even the minor effort. —Ilmari Karonen (talk) 16:09, 11 December 2006 (UTC)
I am entering this under your heading because it has to do with this diff/history link: when reviewing the changes under "Diff", there is a line number listed which locates the change in reference to the line number. How can I locate the line number in an artice? I couldn't find
the answer in any of the help articles. Thanks!Richiar17:20, 11 December 2006 (UTC)
Think Unix. E.g., vi (the vi editor) allows to visualize line numbers, just enter :set nu. So you just have to copy the text (editing version) and try it. If you're stuck with Windows, maybe vim (vi improved ?) could do it. -- DLL .. T18:10, 11 December 2006 (UTC)
Citation & link within article
Could the <ref name=""></ref> be somehow modified to be used both a link within the body paragraph & be listed in the </references> list?
In the Takeshima Island article, there were two links on aerial & water level views of a geographical feature of the island. I want the two links to show up both in the article as examples & also as references. (Wikimachine20:58, 10 December 2006 (UTC))
I don't think there should be too much of a problem with a link only appearing in one place in the article. If it's absolutly essential, however, you could follow a normal inline link with a <ref> tag, e.g. A search engine [http://www.google.com]<ref>[http://www.google.com Google.com]</ref>. Tra(Talk)21:52, 10 December 2006 (UTC)
Though 1) that's pretty non-standard behavior, so readers have a good chance of being confused, and 2) you really really shouldn't use an auto-numbered external link for this, you should link a word instead. Otherwise you'd end up with something like "[1][14]" (eg. Wikipedia:Footnotes notes that it's a bad idea to include auto-numbered external links alongside auto-numbered footnote links because it can confuse readers). --Interiot19:52, 11 December 2006 (UTC)
rss feeds
wondering whether i can read my watchlist via rss. i know that rc is available. asked at the helpdesk, was directed here. gracias. ... aa:talk07:56, 10 December 2006 (UTC)
Not at the moment, but it might be implemented in future. The main problem is that the watchlist needs to be kept confidential. One way of doing this would be to check for the cookie on your system that says you're logged on but this wouldn't work for web-based feed readers. The other option is to include a long number in the URL of the feed that only you know but this is quite insecure. What you could do is add the RSS feeds for all of the pages you are interested in to your feed reader. Tra(Talk)14:33, 10 December 2006 (UTC)
Hello! It's nice to learn that, where was it first announced ? I'm using the sage Firefox extension. I'll try it and explain how to if it works. -- DLL .. T18:17, 11 December 2006 (UTC)
Hi. I created DiaoyuDao (edit | talk | history | protect | delete | links | watch | logs | views) as a redirect to Uotsuri Jima (1st edit). Another new editor blanked the page (2nd edit). I rolled it back. The page now redirects to Uotsuri Jima again, but my last rollback is not in the edit history'. Not sure what exactly happened. I am just curious, since the redirect I created was a typo anyway (should be Diaoyu Dao), it could be deleted. I just wanted to know beforehand what happened. Many thanks -- Chris 73 | Talk10:39, 10 December 2006 (UTC)
From reading your contributions, it seems you reverted Diaoyu Dao and Uotsuri Jima, but not DiaoyuDao. It's possible you may have thought you were reverting the redirect when you were in fact reverting the page it redirected to. However, you did (effectively) restore Diaoyu dao to its original content after Yeahsoo had moved it to DiaoyuDao (creating a double redirect).
All these redirects, page moves and cut-and-pasting make the histories of these articles really hard to follow. There might be some call for cut-and-paste move repair here, though it would have to be done with extreme care to avoid messing up the histories even more. Incidentally, I note that there is some valid-looking content about Chinese claims in the history of Diaoyu Dao that isn't currently found in Uotsuri Jima; you might want to merge that in. —Ilmari Karonen (talk) 11:00, 10 December 2006 (UTC)
The claims (Chinese and other) are all detailed in Senkaku Islands. The new user merely choose to show only the Chinese claims and ignore the others. As for the edit, i definitely clicked on "revert" on this page. The page got reverted, but the revert never shows up in the history. -- Chris 73 | Talk17:26, 10 December 2006 (UTC)
Apologies if this is in the wrong section. The "otherpeople" template says: "for other persons named...". This is grammatically incorrect, the plural of "person" is "people". So the template should read: "for other people named..."
--Salsa man22:11, 9 December 2006 (UTC)
I just noticed this site which appears to be some sort of mirror of our content, but in a commercial venue. I'm curious as to whether or not this is appropriate usage? I recall a similar situation (possibly with this same site), that was considered "not acceptable", but I'm unfamiliar with the technicalities. If this is not the right place to mention this, could someone more experienced give me a pointer to the correct page? Thanks. Doc Tropics21:11, 9 December 2006 (UTC)
A Graphic Lab have Started on Wikipedia-en. You can help by reading its Main page, and helping to its improvement.
The Graphic Lab need some active users and graphists to start and improve it, raise graphic request,and make images improvement.
To request graphic improvement, please see the newly open Graphic Lab/Images to improve (copied from Deutsch and Français).
Please, talk about this to other users who can be interesting by graphism, requesting images improvement or creation, and people interesting by photographs. Yug(talk) 10:53, 9 December 2006 (UTC)
Thumb and other various images
Hi umm...does anybody see a giant thumb where the mediawiki logo should be? Basically this only appears when I use firefox and it is normal of Internet explorer. Earlier there was a pirate flag and a PI sign in place of the logo...but right now there is a giant thumb. What is going on. — Seadog01:33, 9 December 2006 (UTC)
Watching 'What Links Here'
Is there any way to watch a page's 'what links here', ie so that when a new link is made to that page it shows up in my watchlist? I ask because there's a couple of disambig pages I like to keep an eye on, dabbing when necessary, and it would be save me having to remember to periodically go to the page to check (they're pages which periodically get a lot of new wrong links)DuncanHill16:43, 8 December 2006 (UTC)
I don't think so; however, it could probably be done with a user script. --ais523 17:05, 8 December 2006 (UTC)
I've made a javascript tool that creates a 'What links here' watchlist (at User:Tra#What links here watchlist). The catch is it only works properly if there are no more than 50 backlinks to all of the pages in your watchlist in total so it should only be used for things like disambiguation pages that shouldn't have backlinks. Tra(Talk)17:30, 8 December 2006 (UTC)
Thank you Tra! That looks like exactly what I need :) Just hope I've done it properly, not hugely confident with User Script thingies (ie I know nothing at all about them).DuncanHill17:47, 8 December 2006 (UTC)
In "my preferences", I have the edit box as 25 rows of 80 chars. I use the classic skin on MSIE6 on Win2000 and I find that recently (several weeks) the edit box is no longer 80 * 25 but dynamically sized. Apart from the fact that I'd much prefer it if it didn't dynamically resize, the resize is wrong and the right edge of the edit box is off the right of the window pane. If I resize the window the box fits in the pane until I start typing again when it immediately becomes too big again. The scroll bar is already fully right, so access to this area of the box is impossible. What can I do to fix it? -- SGBailey09:20, 8 December 2006 (UTC)
Does anybody know why two paragraphs of source text are sometimes rendered as one paragraph on the page, and how to fix it? This is happening at Banksia epica. The Description section is rendering as one long paragraph, even though the wiki source is clearly written in two paragraphs, with a blank line between them. One blank line between the paragraphs is ignored; two blank lines is correctly rendered as too much whitespace. Hesperian04:43, 8 December 2006 (UTC)
The problem was ''B. epica''<div/>'s. The <div/> was intended to stop the triple apostrophe from rendering as bold, but that kind of self-closing style isn't recognized by many browsers and consequently not by the parser, so it interpreted it as opening a div and that led to all kinds of weirdness. If it had worked, it would end up looking like B. epica's with a line break, anyway (except that strangely something is stripping that altogether, feh).
There were some more instances of <div/> in the article, plus at least one case with no delimiter at all. I fixed them to use <span/>, which is only one character longer but less likely to cause problems (being a text-level element, not a block-level one like div). Of course, Simetrical's <nowiki> trick works just as well. —Ilmari Karonen (talk) 12:38, 10 December 2006 (UTC)
glyph in the title there was way out of alignment with the rest of the line, as illustrated by this partial screenshot: 
. I "fixed" it, but was then told by someone using Monobook that for him the fix had thrown the image out of alignment in the opposite direction: 
. I don't know my way around CSS very well and so haven't found the root cause of this different behavior and haven't been able to fiddle my way around it blindly. Does anyone have advice on how I might resolve this? Bryan 06:37, 24 November 2006 (UTC)
/Archive_AC#c-Brandmeister-2006-12-15T15:32:00.000Z-Image:Characteristics_of_AD.jpg){kind=link}
![[11]](http://www.centralaz.edu/Images/Distance_Learning/Online_Learning/IE7-Settings2.gif){kind=link}