This page contains discussions that have been archived from Village pump (technical).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.
Latest comment: 17 years ago2 comments2 people in discussion
I believe an article called "What Do Christians Really Believe About Evolution?" was uploaded to wikimedia on or about February 7, 2007:[1]. However, this article is no longer present, or at least that address is wrong. Did it violate copyright or some other rule? Was it uploaded improperly? Thank you.--Filll20:19, 8 August 2007 (UTC)
Is there a way to get ParserFuncs to output straight Wikitext?
Latest comment: 17 years ago5 comments3 people in discussion
Without having to go use the Meta ExpandTemplates page, is there a way to get a template to expand with {{subst:}} such that ParserFuncs output actual WikiText instead of the ParserFunc code? Here's the situation:
When putting Fair Use Rationales on hundreds of images that all read very similarly to each other, it's inconvenient to have to copy the FUR from one image, paste it to another image, then modify the things that vary between the two images (like the title of the image's content or the page it's being used on). I created a user template to take care of this, which takes several parameters to customize the output. Because the fair-use patrol bots don't accept user templates, I'm using subst with the template. But since there are ParserFuncs in the template code, those parserfuncs end up in the subst'd output.
Meta talks about ways to do "optional substitution" and composite templates, but I tried the techniques they described there, and I still ended up with the same result.
No, see, I don't want to have to go and use ExpandTemplates to create custom template code that I have to then copy/paste to each image. That's even more inconvenient than just directly copy/pasting the code from image to image. — KieferSkunk (talk) — 23:05, 1 August 2007 (UTC)
Latest comment: 17 years ago2 comments1 person in discussion
What suddenly changed on all the talk page project templates? They seem much larger, and have no border or background color anymore. Was something changed? 199.125.109.3707:26, 1 August 2007 (UTC)
The discussion centres around readability and display issues, but I don't see any significant problems with it, and certainly none that the other editor reported. I have viewed it on two PCs (with TFT and CRT) just to be sure, so I think it must be an issue that is local to his PC, not in the table itself. That raises the question of whether it may be a common enough occurrence to worry about, and I can't really answer that. AdrianM. H.18:17, 2 August 2007 (UTC)
Looked fine to me. There didn't seem to be anything funny with the code either. Unless the editor is viewing Wikipedia with a different skin and that is causing the problem (unlikely, but you never know), then it must be a local problem. - 52 Pickup18:54, 2 August 2007 (UTC)
Template help (CSS)
Latest comment: 17 years ago3 comments2 people in discussion
I'm trying to reduce the size of some university templates. I used {{Johnny Cash}} as a base and converted {{University of Oklahoma}}. I am working on converting {{University of Texas at Austin}} because it is huge but they wanted to know if I could work the image into the template and I couldn't figure out how. The template I have created in its place is located at User:Nmajdan/Test1. How can I add an image to this template? It would be easier if it was using tables, but it strictly used divs, so a CSS person is requested. Thank you.↔NMajdan•talk11:52, 2 August 2007 (UTC)
You do have a table in there, which should make adding the image easy enough. Add a colspan="2" to the single cell in the top row, and add a new cell |width="100" valign="middle"|[[Image:Hookem hand.svg]] to the end. Or is that not what you want? Anomie12:15, 2 August 2007 (UTC)
Latest comment: 17 years ago4 comments2 people in discussion
Does anyone know what the cause of the white space is in Girona between the dab notice and the beginning of the text/infobox? It doesn't seem to be in the infobox or the page source code. Thanks. —METS501 (talk)22:00, 1 August 2007 (UTC)
Latest comment: 17 years ago9 comments4 people in discussion
I'm trying to work out how to do auto-signatures in a template, so that whoever leaves a template also ends up signing it either by choice or by default. This is in an effort to get assessments signed. One method, seen at User talk:Psychless/WPBiography, involves relying on users to fill in "user" and "date" parameters correctly. It would be much more convenient to be able to enter ~~~~ directly, but I am struggling to find adequate documentation on how to do this. What I have discovered so far:
Auto-signing was in the history of {{Test5-n}}, as seen here, and used the code {{{2|~<includeonly>~~</includeonly>~}}}.
The ideal functionality would allow the template to be added by humans and bots without any auto-signing, but as soon as certain parameters are filled in (such as class, or requesting a photograph, or requesting an infobox - anything that is likely to require discussion), then the template auto-signs for the person adding that parameter. Subsequent changes to that parameter would then probably have to be manually signed. If auto-signing is not feasible, then the next best thing would be some way to have a parameter like "assessment signature=~~~~", that anyone subsequently editing the talk page would see as an expanded signature that they could then replace. Alternatively, if all this is too difficult, is there a way to get an assessment to link to the article version that was assessed, and the talk page diff that produced the last assessment? Does anyone have any ideas on this, and how to separate out the bits that should be signed from the bits that don't need signing? Carcharoth10:55, 1 August 2007 (UTC)
There is currently no magic word or other feature in Wikipedia which would return the username of the person who made an edit as text displayed on the page. There are thus essentially three possible ways of doing this;
Have the user include their name as a parameter in the function call. This isn't particularly useful as most usernames are longer than ~~~~... so it'd be easier for them to just sign normally. The Psychless/WPBiography and s/block examples you cited above work this way.
Substitute the template so that a ~<includeonly>~~</includeonly>~ or somesuch evaluates to a hard signature on the output. Drawbacks here are that the user has to remember to subst: the template (which again is more extra characters than just signing normally), ALL of the template logic will be copied onto the page, and if the user does sign normally they'll end up with two signatures. The Test5-n example you cited above works this way.
Have a bot which traces all uses of the template, checks for signatures, and adds one if none is present. So far as I know, no such bot exists, and to make one it would need to cycle through the edit history to find the edit where the template was added and then get the name of the user who made it.
It is also possible to combine the first two options to create a template which, when substituted, inserts a call to another template with all parameters passed through and an additional parameter automatically set equal to the user's signature. Something like; {{<template to display>|1={{{1|}}}|2={{{2|}}}|name=~<includeonly>~~</includeonly>~}}. Again, the drawback here is that the user has to subst: the original template. --CBD12:30, 1 August 2007 (UTC)
Subst is not really an option for WPBiography, or for any assessment templates. From what I can see, signing without substing is possible. If I write: {{s/block|~~~~}}, the signature appears inside the template even when I don't substitute it. Like this:
But the signature and timestamp has expanded when I view it after I save, and someone else could overwrite my signature and timestamp with their 'four tildes' next time they update the template. So how does that work? And can it be used in assessment templates? Carcharoth13:19, 1 August 2007 (UTC)
This has often been requested (I've personally been asked about it and looked into it about 10 times in the last 2 years). The easiest-use solution I've seen involves one subst level. For example: [[Template:Foo]] has content: {{Bar|name=<includeonly>~~~~</includeonly>}} (similar example given above), and then is easily placed with {{subst:Foo}}, which results in something like: {{Bar|name=[[User:Example|Example]] 12:00, 1 August 2007 (UTC)}}, which can then be tested with #if in the non-subst'd bar. Note that this is no different than {{Bar|~~~~}} manually (as above), but removes the ugly tildes (using subst: instead).
As for "someone else could overwrite my signature and timestamp with their 'four tildes' next time they update the template", that is true of any signature on any unprotected page. That is the nature of a wiki. Your original request is only slightly possible while allowing it to be modified by others later. --Splarka (rant) 07:18, 2 August 2007 (UTC)
Sorry, I must have been unclear. Such updating is desired. Whenever anyone changes the assessment, or reconfirms it, they should sign again. The page history then shows the assessment history. Carcharoth09:56, 2 August 2007 (UTC)
Having the user input ~~~~ as a parameter of course works, but then they might as well just sign it. Use of a 'two template' structure as described by Splarka and I above has the advantage of the user not having to remember/decide to sign. If each time the assessment template were updated the person used the top level template then their name could be automatically signed to it... you could also put the template into a 'Last updated in August 2007' type category based on the substituted date. The only problem would be if someone decided to just edit the 'second level' template directly, but then didn't update the name and/or date. That might be mitigated by having the second level template output display the input format for the current 'top level' template call. There's no foolproof solution though. --CBD10:52, 2 August 2007 (UTC)
References
[...]
Commons image not displaying in a gallery
Latest comment: 17 years ago2 comments2 people in discussion
In Kentucky State Fair, the gallery image with the caption "Freedom Hall during Fair" isn't displaying. This image was recently moved to Commons. I purged the description page in Wikipedia, as well as the page in Commons, twice to no avail. Is there something I'm missing? Stevie is the man!Talk • Work02:52, 10 August 2007 (UTC)
Latest comment: 17 years ago3 comments2 people in discussion
{{WikiProject College football}} has an extra linebreak below where it says "The article has not received a rating on the importance scale." The HTML code has an extra <p><br /></p> that should not be there but I cannot find in the code where this is coming from. Any ideas?↔NMajdan•talk15:49, 3 August 2007 (UTC)
I see. Thanks. I never understood the reason for having those HTML comment lines over line breaks. I never use those in my templates but just assumed they were they for a reason. Guess not.↔NMajdan•talk17:00, 3 August 2007 (UTC)
edit tag for lede section
Latest comment: 17 years ago10 comments6 people in discussion
I don't know if any proposed this before but a "[edit]" link on the lede section of every article would be handy. Every other section have these. At least it could be an optional element for registered users. Caveat lector13:37, 3 August 2007 (UTC)
What's wrong with the edit tab at the top? If you really want to edit only the lead section, click an edit link and change the last digit to 0. AdrianM. H.14:26, 3 August 2007 (UTC)
The "[edit]" tags are very useful for quickly editing long pages, over a slow connection. It just looks like a curious omission from the wiki. Caveat lector14:48, 3 August 2007 (UTC)
I too would prefer an edit link for the top section (not a tab). Editing the top section of long pages really is a drag, even on broadband. I'm hacking up a template to add such a link to certain pages. --Edokter (Talk) 15:48, 3 August 2007 (UTC)
That is what Alex was refering to. But my template positions itself above the top line, next to the page title, while Edit-first-section just floats to the right. --Edokter (Talk) 19:50, 3 August 2007 (UTC)
DEFAULTSORT in talk page templates - problems
Latest comment: 17 years ago4 comments2 people in discussion
The WPBiography template uses a listas parameter to define a DEFAULTSORT, and has listas=Lawrence, Aaron; but the WP:LBGT template has "DEFAULTSORT:PAGENAME". Because the WP:LBGT template appears on the talk page below the WPBiography template, the article sorts under 'A' in all its categories. If you swap the templates round, it will sort under 'L'. This is expected behaviour for DEFAULTSORT (the last occurence on a page is the one that is used), but in terms of different templates using DEFAULTSORT, this can only lead to chaos.
I think what is needed is to deprecate the use of DEFAULTSORT inside templates, and to have only a single DEFAULTSORT on a page. Otherwise you get the silly situation where a talk page with WPBiography on it is correctly sorted under the person's surname (using listas), but then the sorting gets undone if someone puts a template on the page from another project, putting it on the page after the WPBiography template, where the new template only has "DEFAULTSORT:PAGENAME", which will then sort the person under the page title (ie. in this case by their forename).
What would be simpler would be if it would be possible for talk pages to automatically DEFAULTSORT to any DEFAULTSORT value present on the talk page's article. Would this be possible? Carcharoth01:26, 3 August 2007 (UTC)
This shows that even if a template has a DEFAULTSORT tag, it can be useful to put the sortkey also in the category tag in the template. This way the sorting in this category is not affected by possible DEFAULTSORT tags further down the page in which the template is called.--Patrick07:39, 3 August 2007 (UTC)
This issue applies in every namespace. Having DEFAULTSORT in an article also apply to the talk page is a separate issue.--Patrick07:58, 3 August 2007 (UTC)
I thought of having the sortkey also in the category tags in the template. Unfortunately {{WPBiography}} is a bit bloated. Have a look at Template:WPBiography/Categories. When that list was made, the WPBiography template was populating 212 categories. Putting PAGENAME as the DEFAULTSORT value, and listas as the manual over-ride would work, and might ultimately be the best solution. Then no-one has to worry about placement any longer. This might be the best long-term solution as well, as it doesn't matter if "DEFAULTSORT:PAGENAME" is in every template on a page - the effect will be the same. Carcharoth09:32, 3 August 2007 (UTC)
Issue with small and numbered lists
Latest comment: 17 years ago8 comments6 people in discussion
Oh, and it doesn't "skip" 2, what actually happens is that a non-visible extra li element is added between the desired #1 and #2. <li style="list-style: none"></li> so you request a list of 4 elements, but mediawiki generates a list with 5 elements of which one is invisible. --TheDJ (talk • contribs) 14:47, 2 August 2007 (UTC)
This seems to be an unexpected error (well, sort of expected — shouldn't be wrapping blocks in inlines anyway) in the way the HTML cleanup and the wikicode conversion routines interact. Writing the HTML out by hand (<small><ol><li> ... </ol></small>) gives me correct numbering when the code is fixed, as does wrapping the list elements in a block-level element, such as <div>, which is normal behavior in the first place. It's possible to work around the parsing bug by adding a block element (<br /> will work fine) to the end of the first list item (e.g. # a<br />), but, well, that's a hack. This markup will only affect the items inside the list and not the auto-numbering, so you might as well move the <small> tags inside the item definitions. —TangentCube, Dialogues21:55, 2 August 2007 (UTC)
Or you could wrap the entire list in a <div style="font-size: smaller"> as with
a
b
c
d
Of course that also makes the numbers smaller, but then I'm not sure what you were trying to achieve with the formatting. older ≠ wiser11:40, 3 August 2007 (UTC)
Latest comment: 17 years ago3 comments2 people in discussion
I just made an edits to an article, then made a correction and marked it as minor. But when choosing to hide minor edits in my watchlist the article doesn't show up at all. Shouldn't the next newest non-minor edit show up in the list instead? If it doesn't then this has the potential to hide non-minor edits when minor edits are made after them. If a vandal makes an edit and someone makes a minor edit after it they could intentially or unintentially hide the vandalism from many users who choose to see the vandalism. -- Andrew | flame him | stalk him13:17, 2 August 2007 (UTC)
If I correctly understand your problem, just go to your preferences and check Watchilst → Expand watchlist to show all applicable changes and Recent Changes → Enhanced recent changes (JavaScript). This should be the default for new users because otherwise watchlist is imho unusable ∴ Alex Smotrov13:27, 2 August 2007 (UTC)
I was actually referring to when an edit is minor and hidden the next newest edit which is not hidden should show up. This shows all the edits prior to that but I suppose it's better than nothing. -- Andrew | flame him | stalk him17:53, 3 August 2007 (UTC)
Technical problems in the Latin Wikipedia
Latest comment: 17 years ago2 comments1 person in discussion
Since yesterday we are having technical problems in the Latin Wikipedia. Some/many talk pages are not displayed, I mean, they are displayed blank, there is not even a navigation on the top. See la:Vicipaedia:Taberna#Blank_talk_pages.
Latest comment: 17 years ago5 comments5 people in discussion
I'm new to using the popup tool in reverting vandalism. I'd like to ask, how do you set it so that every time I revert the vandalism, it's listed as a minor edit? Thanks in advance.--Kylohk08:01, 3 August 2007 (UTC)
I can't see why all reverts are considered minor edits. I see reverting a page blanking and other vulgar vandalous edits as a major saving grace regarding the article's fortunes. Lradrama15:31, 4 August 2007 (UTC)
Since reverting vandalism doesn't actually change the content of an article, such reverts are concidered minor edits, just like any other 'housekeeping' edits. See Help:Minor edit for more info. --Edokter (Talk) 16:12, 4 August 2007 (UTC)
IMHO, Popups is correct in not marking reverts (and delinks) as minor. You might be reverting something significant, and anyone watching the page should see it. — Randall Bart(talk)22:52, 4 August 2007 (UTC)
Edit Summary
Latest comment: 17 years ago5 comments4 people in discussion
I don't think the edit summary's maximum length is big enough, the reason I say this is 3 times I had to shorten the length of the summary, the reason for that is that I like clicking on "edit this page" and fixing everything I can find, and then leaving a detailed edit summary e.g. [3]. I understand there is a need for an edit summary size limit but it could be a bit bigger. Jeffrey.Kleykamp18:48, 2 August 2007 (UTC)
I'm inclined to agree, to an extent. There are occasions when I need to add something to a reversion summary and do not wish to lose the automatic part of the summary. On the other hand, as you rightly said, there is a need for a limit. It seems to equate to exactly two lines on the watchlist on a 1024 viewport. AdrianM. H.19:36, 2 August 2007 (UTC)
That's exactly the case when in comment you say "see talk page". Although personally I would try to make my comment more compact ∴ Alex Smotrov21:29, 2 August 2007 (UTC)
It's possible to make the edit summary length allowed slightly longer using this script (by Gracenotes) in your monobook.js:
//A script by [[User:Gracenotes]] to allow longer edit summariesaddOnloadHook(function(){varsumBox;if(sumBox=document.getElementById('wpSummary'))sumBox.setAttribute("maxlength","250");});
The reason that this isn't used by default is that you have to be careful using it; the maximum edit summary length you can use reduces by two characters for every accented letter, special character or other character not in ASCII-7 (English letters and numbers are alright, but most other languages aren't). Still, the extra fifty characters do come in useful sometimes. If a really long summary is needed, write 'see talk' and put the explanation on the talk page. --ais523 16:48, 3 August 2007 (UTC)
Wouldn't huge edit summaries look rather irritating to the eye on article histories and your watchlist? I like the neat little list of usernames and their edit descriptions, but, it's what ever floats your boat I suppose... Lradrama15:29, 4 August 2007 (UTC)
Export points of interest as KML; see them on Google Maps
Latest comment: 17 years ago9 comments2 people in discussion
Pages marked with {{coord}} can be exported as KML (for use in Google Earth, for example) via Brian Suda's site, in this format:
There was big controversy over having coordinates in articles' titles at all when they were first introduced. The link now leads to a list of dozens of tools. Why should articles have a link to any single service over all the others? I can see how getting many locations plotted on a map all at once might be useful, but people have different preferences, which is why the links to geographical information services are collected on a single page where people can make the choice themselves. The proposed template would be used in all articles that have more than one instance of coordinates, and that's too much unnecessary cruft. It might be useful to have such a kml export link in the toolbox, but as far as I know it would then have to appear in all articles, even those that have nothing to do with a location. The best way to make it possible now is to have the coordinate templates give the name of the Wikipedia page to GeoHack, where that link of yours can be put alongside all the others. --Para12:21, 2 August 2007 (UTC)
"The proposed template would be used in all articles that have more than one instance of coordinates" - it would? Wow, that's an impressive crystal ball you have. Unless you're just inventing more FUD, of course. Perhaps you could list the other site(s) offering this service, so we can see what other preferences people might have? Feel free to implement your GeoHack suggestion; and remove instances of the proposed template (which is no different than, say, the IMDB templates) once you have both done so, and found a way to make it clear that a link to a single set of coordinates should be used to see all coordinates on the page displayed on a map. In the meantime, I won't be holding my breath. Andy Mabbett | Talk to Andy Mabbett13:12, 2 August 2007 (UTC)
Congratulations for having found a new type of service then, it can probably have a little section of its own in GeoTemplate. There is however no reason that it should absolutely be available at this very instant today; you can wait until the proposed change to coord is done. Otherwise people will have to clean up after you and delete the redundant links from articles. --Para15:10, 2 August 2007 (UTC)
GeoTemplate exists to help people "find maps and data about the location", which covers locations defined as both single or multiple points. You can start thinking of a good description to put in GeoTemplate to make that clear to readers. The script is just another service to add to the list, and your personal preferences have no special status to make it so prominent on Wikipedia. Now, can we please have this discussion at a single location? I suggest keeping it here. --Para15:55, 2 August 2007 (UTC)
Latest comment: 17 years ago5 comments5 people in discussion
Hi, My user name is borgx from Indonesian Wikipedia. I would like to ask the strategy used for English Wikipedia in monitoring their recent changes. As far as I'm concern it has thousands of edits an hour (exact numbers escape me). My questions are:
Did you ever miss an edit? (Yes, there are a lot of tools for monitoring recentchanges, but these also can't make sure no vandal edits unmonitored).
If yes, what if they were vandals edit? What is your solution for these?
Indonesian Wikipedia now has approx. 1,300 edits per day and the numbers will only grow in time. Although the numbers (comparing to yours) considered low, we also have a very limited number of sysops -- specially when out of 14 syspops, we only have a few active ones and during their monitoring period -- which is not a continuous-- they only check the top 50 or 100 and go about and did something else. These actions leaves gap out of the 1,300 edits a day + we have no mechanism that gives an information to public user saying " the changes had been checked" so we might be checking the same thing several time.
I understand that we have a "patrolled edit" mechanism, but this mechanism is not going well this time because new users may edit something up to 20 times in a certain time and we need to marked patrolled edit 20 times <-- this is not practical. (I have submitted bug 8697 for this but there's no action from developer for my request).
Yes, we miss edits, we do find month old vandalism, though we get most of them very fast. We use a lot of automated and semi-automated tools. We have a few bots that watch every contribution and look for blatant easily detectable vandalism. The bots revert, warn, and report vandalism.
We also have a few wonderful tools for humans to use, my favorite is WP:TWINKLE. It allows for one click reverting, warning, and reporting of vandals. As an admin the tools allow for fast deletion and removal links for images and articles, it allows for the fast posting of warnings and block messages. With Twinkle one person can do the work of 3. We have WP:AIV for reports of vandalism that can be handled with very little investigation, on this page vandals are reported and blocked on wholesale levels.
Most importantly we have a huge army of anti vandal patrolers that do a wonderful job. I am sure that all of our solutions can work on the Indonesian Wikipedia, the tools just need to be translated. We don't even use patrolled edits here. Until(1 == 2)00:23, 1 August 2007 (UTC)
I've worked on implementing this idea of collaborative patrol before, yet neither my attempts nor others' (i.e., patrol, an extension by Rob Church whose name I'm forgetting, etc.) have proved particularly successful. The primary problem is not developing a robust framework with a competently deployed server-client model--the problem is motivating users to make use of it, which has not been too successful in the past. With one of my attempts I even introduced a fairly complex and accurate "priority" system that ordered the edits displayed to reviewers based upon the likelihood of their being vandalism. This did not succeed in reducing the boredom users felt much at all. In fact it does seem to be this element of boredom vs. excitement that is the motivating force--with things like twinkle and similar scripts people feel as though they're taking part in some sort of first-person shooter, whereas with collaborative patrol tools they feel as though they're working. What we really need to do is get ourselves a video game developer to find some interesting way to integrate vandalism patrol with Counterstrike :D. In any case, if you have some firm ideas about how you would like to implement this on that wiki and could provide me with some sort of concrete specification, I'd be glad to work on implementing something for you guys. I'm just not going to promise that it will be successful, as in the past it's been quite the opposite. AmiDaniel (talk) 00:38, 1 August 2007 (UTC)
the key is not to aim at efficiency, but redundancy, and to make sure that gaps are eventually covered. Perhaps we should simply think of a calendar-like tracker, where you'd fill in the time period you covered. I use something of the sort for some specialized tracking. DGG (talk) 04:27, 1 August 2007 (UTC)
There are now so many people using Wikipedia that the recent changes is now like trying to monitor an immense swarm of insects. There are so many edits that identify as vandalism that yes, initially, some of them are missed, but never obvious ones like large-scale deletions or page blankings. Sometimes, an RC Patroller maybe unsure whether a certain edit is vandalism because he or she may not be specialised in the subject point of the article, and it may be missed that way too. However, I think that in time, most vandalism gets iradicated, because nearly all articles are assigneed to a Wikiproject or two (or even a fair few), and the article's main editors will identify it one way or another. Little stubs/articles not assigned to Wikiprojects and which are infrequently edited are the problem thus. Lradrama15:25, 4 August 2007 (UTC)
Please show wiktionary when no wikipedia entry
Latest comment: 17 years ago7 comments6 people in discussion
When someone searches on something that doesn't have an article here but does on wiktionary, it just seems like common sense that the wictionary entry would be offered or shown. There are a lot of one-paragraph articles that have been moved over there, and when you search on them you get nothing. Would this be hard to fix? 209.77.205.223:25, 31 July 2007 (UTC)
Since Wikipedia is not a dictionary, it may be confusing to people when they look for an encyclopedic article and then get a dictionary definition. They may be even more confused when they don't realize they have left Wikipedia and arrived at Wicktionary. Until(1 == 2)00:14, 1 August 2007 (UTC)
Then just offer a link to it in the missing-page message. I wish someone who knows how to work bugzilla would put it in. 209.77.205.202:02, 1 August 2007 (UTC)
An automatic transfer to another site would be very confusing to people who are logged into Wikipedia but don't have an account at Wiktionary and see an entirely different interface. Corvus cornix19:32, 2 August 2007 (UTC)
Although I agree with the above comment, maybe Wiktionary could be advertised better on this website? That may help. Currently, there's only a link on the main page down at the bottom, which not everyone will notice, especially as it's surrounded by other Wikiproject names. Just a suggestion... Lradrama15:20, 4 August 2007 (UTC)
For everyone? Because if it's only for yourself, I do know a way to do that. Just click 500 so you get that in the url like this: [4]. Leave the offset alone but for the limit just type the desired number and that'll work. :) -WarthogDemon01:38, 6 August 2007 (UTC)
Latest comment: 17 years ago3 comments2 people in discussion
When I'm at a wikipedia page, Alt-F doesn't work in IE7 (and previous IE versions too). Instead of bringing up the file menu, it takes me to the search box. Is there a way to disable this? I use alt-F pretty frequently. 24.17.110.22300:05, 6 August 2007 (UTC)
I don't know of any way to disable this, but if you type "Alt", then let go, then type "F" you should get the file menu. Obviously, someone decided to designate "Alt-F" as "find" on wikipedia. Dragons flight00:32, 6 August 2007 (UTC)
Spam Filter Filters Everything . . . Including Speedies
Latest comment: 17 years ago7 comments3 people in discussion
Just recently I was trying to put up a {{db-copyvio}} for Iranians or Persians?. Unfortunately the spam filter decided to halt my attempts to add the url into the speedy. Is there any way of getting around this short of just posting db-copyvio and putting the link in the edit summary? -WarthogDemon22:36, 5 August 2007 (UTC)
Didn't see that one coming. That's a question for the devs to answer, I think, so I recommend a trip to Bugzilla (checking first to see if it has already been raised). AdrianM. H.22:48, 5 August 2007 (UTC)
Forgot to add that I think you would be OK to leave the URL in the summary (unless that gets caught as well). Makes Twinkle impossible though, because that requires the URL to be added to a popup box. AdrianM. H.22:50, 5 August 2007 (UTC)
Registering there seems to mean using my email as a username, something I would rather avoid. Would it be possible for you to, or is there someone else here on Wikipedia I could bring this to? -WarthogDemon23:40, 5 August 2007 (UTC)
No, I don't use Bugzilla either. But plenty of editors do, and one of them might either have the answer to this or know that it has already been raised or be willing to raise it. If the e-mail address is your only concern, you could just create one for that purpose. AdrianM. H.23:57, 5 August 2007 (UTC)
Full URL and square brackets appearing in footnoted references
Latest comment: 17 years ago6 comments3 people in discussion
I am having persistent problems with references in which the URLs in refs, which ordinarily shrink to a redirect logo, are appearing in full. Examples are in the footnotes to West Gate Bridge, Consequences (album) and 10cc (oddly enough, the problem references are reference No. 2 on each occasion. Weird.) As far as I can see, the markup is correct. Can anyone spot the problem? Thanks. Grimhim04:00, 5 August 2007 (UTC)
The problem is that you put the brackets around the entire citation. THe brackets should only be arounf the url. The comments next to the url should not be inside the brackets. --Hdt83Chat04:09, 5 August 2007 (UTC)
Not sure this answers the question. The example in 10cc is a better one. Compare the markup for references 1 and 2: they're identical, but produce different results. I want them to look like ref no. 1, as illustrated in "External link" in Wikipedia:Cheatsheet, where just the summary of the URL appears.Grimhim04:17, 5 August 2007 (UTC)
Thanks, Patrick, that's what I was after. I take it you're saying a whole line space was in there somewhere between the URL and the summary? It's not an easy thing to see when the line of unbroken text is so long, and I can't see how an extra line could appear, but I'll have a play round with it if it happens to me again. Thanks again. Grimhim06:44, 5 August 2007 (UTC)
In fact the most convenient way to check for newlines in the wikitext of an external link seems to be simply to check the link in preview. If in doubt, delete with backspace the character which comes before a new line, and add a space.--Patrick00:03, 6 August 2007 (UTC)
Download speed with Mozilla Firefox
Latest comment: 17 years ago6 comments3 people in discussion
The speed at which pages load has been significantly slower for me over the last few days, about 25-50% the speed they downloaded previously. This even happens when I use the forward and back buttons. I'm using Firefox 2.0.0.6. I still get good speeds if I use Internet Explorer, but Firefox is my preferred browser.
I am having the same problem with Firefox 2.0.0.6. I was wondering whether it was my browser or Wikipedia. So it is Firefox? Drag, since it makes many tasks onerous. It's enough to make me consider changing browsers. Regards, Mattisse12:28, 11 August 2007 (UTC)
Different platform (Firefox 2.0.0.3 on Linux), same problem. With the occasional white screen of death, of course. It's a wikipedia problem, time to give the devs a bump. MER-C13:42, 11 August 2007 (UTC)
Problem solved. It was due to the use of a template redirect, which has now been replaced with its correct title. AdrianM. H.00:53, 7 August 2007 (UTC)
(External) link colors
Latest comment: 17 years ago4 comments2 people in discussion
In a template, I have an edit link to the current article. It's a full URL so it's not a wiki link. By default, these (technically) external links have an arrow and have a different color the wiki links. I lost the arrow by using .plainlinks, but I can't seem to find a way to change the link color to behave like wiki links (instead of the always pale blue-ish external link), at least not without overriding the stylesheet. And I don't want to do that.
So, how do I create a full-url link without arrow and wiki-link colors, without overriding the current stylesheet? — Edokter • Talk • 17:05, 6 August 2007 (UTC)
That's exactly the thing I'm trying to avoid. I was hoping I could somehow invoke the regular a.link and .plainlinks class at the same time from whatever stylesheet is active; my CSS knowledge is quite lacking in that respect. — Edokter • Talk • 22:04, 6 August 2007 (UTC)
Latest comment: 17 years ago3 comments3 people in discussion
On my user page, there is a long space between the "Identity" section and the "User subpages" section. I can't find what's causing it. Is it the userboxes, transcluded at the right? There doesn't seem to be anything noticeable that's causing this. Thanks. (I have a wide screen, BTW, so if you aren't seeing a huge space, it may just be because your screen is skinnier than mine.) --Fbv65edel / ☑t / ☛c || 15:50, 6 August 2007 (UTC)
Latest comment: 17 years ago7 comments5 people in discussion
I'm unsure if I've done this correctly, see here. I hit the F5 button, but nothing comes up. I don't use FireFox or Modzilla, just Internet Exploror 6. Lord Sesshomaru21:59, 5 August 2007 (UTC)
If you want everything, you currently have it. Kind of the point of my question. What you end up with is a whole bunch of bits that you might not use and some bits (admin stuff) that you may not be able to use. Academic, though, because as our fellow editor has pointed out, IE is out.AdrianM. H.22:46, 5 August 2007 (UTC)
You may have the same problem that I encountered. If you are using Internet Explorer, TWINKLE won't work - much to my annoyance! The same thing happened with the 'Admin-like revert function' with me aswell. After over 7 months on RC Patrol, I'm still going about the process old-school style! Lradrama09:56, 6 August 2007 (UTC)
Repeated references
Latest comment: 17 years ago3 comments2 people in discussion
I notice that the same references (books, articles, etc) reoccur in many of the articles which I edit. To copy the text which defines a references, for example, in the form of a cite or ref tag, is perhaps not a big deal. But I also notice that different editors write the same reference in slightly different ways in different articles, with different styles, etc. Again, not a big problem, but it occurs to me that it could be nice to have a reference defined (author, title, etc) once and for all in one place and instead use pointers or links in the different articles to connect to the reference data. In the case of ref tags this could also mean that the edit text becomes less cluttered with the ref data and instead a short link can be used. --KYN12:56, 3 August 2007 (UTC)
As a guide, footnote layout and styling should follow WP:CITET. No need to use the templates, obviously (they create extra typing/pasting anyway and can be more difficult to view subsequently) but they provide a good style guide. That does not include any variation in the availability of certain fields, of course; sometimes you have to omit certain fields, such a publishing date. To address your suggestion, some editors keep libraries of oft-used references; perhaps you would want to do the same for references that you use a lot. AdrianM. H.14:31, 3 August 2007 (UTC)
Thanks for this input. But already from the WP:CITET guide it is clear from the two columns of examples that variation is possible. Also:
Even with a guide like WP:CITET it is up to the editor to chose if a possible subtitle should be included or not, or if the page number to a journal article is added. This may lead to one and the same reference having inconsistent appearances in several Wikipedia articles.
Minor error in one reference (e.g. ISBN number, page numbers, etc) are copy-pasted to new articles
if such errors are detected, it appears to be rather tedious to find and correct all occurrences in copy-pasted references.
Proposal for reference namespace
I'm not suggesting that these are major problems that keep me awake at night, but from a technical point of view they appear rather straightforward to fix? Why not have a Reference name space where all references (used in WP) are given a single definition and we can link to there from the articles? This also make sense since it would provide the right place to describe and discuss a reference. This can of course also be made using an ordinary Wikipedia article, but a reference also contains metadata in the form of author, title, etc, which can be managed more easily in a separate namespace. --KYN15:20, 3 August 2007 (UTC)
The number of potential problems here is enormous. For example, many references include a link to an online source such as JSTOR. This is OK, but if we were to do it in a centralized way, we would need a way of also indicating any other sources. This is particular problematic for newspaper citations,which are available from a variety of free and paid services. Equally difficult will be obtaining agreement from the several hundred thousand individualists here about the exact standard form, considering we have been arguing for months about merely whether to included the period inside or outside quotation marks. This sort of service requires both a considerable number of dedicated staff and a way of making and enforcing decisions. A top-down organisation could do it. I've said elsewhere it would be worth the effort--but its an enormous effort. 07:47, 6 August 2007 (UTC)
Horizontal
Latest comment: 17 years ago2 comments2 people in discussion
Does anyone know where I can find the code for <div class="horizontal">. I've searched most the css pages - however I only found one on list, and I don't think that's it.--danielfolsom04:04, 14 August 2007 (UTC)
Latest comment: 17 years ago4 comments3 people in discussion
I was directed here from the Help desk. Here's my question: Does anyone know how to go about fixing the links in an image's metadata? I assume there is a script that does this automatically which means an admin will most likely have to take care of it. If you look at the metadata for pictures taken by a Canon camera, the link points to Canon, not Canon (an example of a pic I uploaded yesterday is here - scroll to the bottom). Thanks. ♫ Bitch and Complain Sooner ♫02:37, 8 August 2007 (UTC)
No, we can't do anything about it. Whatever the camera says is automatically linked, and we can't add exceptions. Prodegotalk02:52, 8 August 2007 (UTC)
Some common brands of digital cameras place metadata in digital photos which identify the camera manufacturer, camera model, etc. However, they may use specific metadata values which are not compatible with Wikipedia article names.
Statement of the proposed solution:
There is no solution. We (Wikipedia) can't do anything about it. (Implication: Camera manufacturers are expected to provide Wikipedia-article_name-friendly metadata values)
This strikes me as unreasonable.
My guess is that there is a standard or a developing standard somewhere concerning the contents of the fields labeled "Camera manufacturer" and "Camera model" in Wikipedia Image pages such as this. It may be convenient to provide metadata redirects (e.g., for a "Camera manufacturer" field content of "Canon", to provide a Wikipedia page named Metadate_mfg_Canon or Metadata:mfg_Canon or Metadata:Camera manufacturer/Canon or somesuch which redirects (in this example case) to Canon (company). Then, wherever contents of the "manufacturer" field is extracted from an Image file displayed, and wikilinked wikilinked, it could be wikilinked with the appropriate indirection prefix ("Metadata:mgf_" or whatever. All that remains, then should be to manually maintain the indirection redirects according to published standards vs. Wikipedia article naming.) (that was just a top-of-the-head illustrative suggestion -- there are no doubt better ways to do this.) -- Boracay Bill04:19, 8 August 2007 (UTC)
Bug ID 9213
Latest comment: 17 years ago2 comments2 people in discussion
I have no trouble with exporting the monobook pages, though I haven't tried importing them. Where are you having trouble? Prodegotalk20:27, 7 August 2007 (UTC)
Changing "header" section on collapsible tables - can it be done?
Latest comment: 17 years ago4 comments3 people in discussion
Given that a standard collapsible table works like this:
non-hidden content
hidden content
...is there any way to make a table where the non-hidden content changes when the table is not collapsed? - 52 Pickup10:19, 6 August 2007 (UTC)
Thanks. I can see the problem there. Unfortunately, what I want to place in the "header" has a variable height. If-functions don't appear to help either. I wasn't sure if this was possible, so thanks for the confirmation. - 52 Pickup08:02, 7 August 2007 (UTC)
Lack of notification of edit conflict
Latest comment: 17 years ago5 comments5 people in discussion
I have been finding recently that if I try to add a comment to the bottom of a discussion (only seen it on WP:RD/s and WP:ANI), but someone else comments before I finish, instead of informing me of an edit conflict, my comment just gets tagged onto the end of whatever the thread looked like when I hit the "save page" button. Is this a bug, or is it supposed to happen? Someguy122107:28, 6 August 2007 (UTC)
That sometimes happens with me too. To be honest, I can't stand that blasted edit conflict thing, I much prefer the way you just mentionned! Lradrama09:58, 6 August 2007 (UTC)
I'm not sure but I think it was changed to like that on purpose. You only get an edit conflict when the system is unable to resolve the conflict. If it can, it just adds it based on who edited first or something. Most of the time, it works well IMHO Nil Einne12:14, 7 August 2007 (UTC)
Latest comment: 17 years ago3 comments2 people in discussion
I uploaded Image:Questionmark.svg, and is it just me, or is there a large white rectangle in the bottom-right corner? Just in case it is just me, I'm running Firefox 2.x and made the picture in Inkscape. Does anyone know how to fix it? I tried uploading new versions a couple of times with slightly different saved files, and this always happens. Here's the XML if that helps (I don't know if it does). I uploaded Image:Questionmarkblue.png as a bitmap alternative to it (that's what it should look like). — Bob • (talk) • 21:09, July 30, 2007 (UTC)
Hi there :) It's possible that the server doesn't have the font you used for the "?". You might want to convert the "?" to a path. -- JaeSharp07:17, 31 July 2007 (UTC)
Latest comment: 17 years ago4 comments2 people in discussion
I've just noticed at the bottom of every page I view, a "Retrieved from [http address of the page I'm viewing]" message. Is this a fault, or a new (and apparently pointless) feature? DuncanHill16:43, 14 August 2007 (UTC)
No, that has always been there--even if you haven't been aware of it :). It's text that it is hidden through styles--when spiders, such as the googlebot, or others who mirror and replicate Wikipedia's content crawl the site, they typically only fetch the HTML of pages and not the full style sheets. As such, when you see a page in google's little stubs [5] or in its cache -- or on other mirror sites -- that line is present, so that visitors can both find the original page and to make it a tid bit more difficult to commit GFDL violations when mirroring content. If you check the source of the page you're viewing now, you'll see:
near the bottom. Note that "printfooter" is a class that's set to display:none in an imported style sheet. What probably happened is that your browser failed to render the styles completely -- fairly frequent problem, especially with Internet Explorer. HTH. AmiDaniel (talk) 06:02, 15 August 2007 (UTC)
Latest comment: 17 years ago2 comments2 people in discussion
I believe an article called "What Do Christians Really Believe About Evolution?" was uploaded to wikimedia on or about February 7, 2007:[6]. However, this article is no longer present, or at least that address is wrong. Did it violate copyright or some other rule? Was it uploaded improperly? Thank you.--Filll20:19, 8 August 2007 (UTC)
Latest comment: 17 years ago1 comment1 person in discussion
I'm getting this set of logs generated each time anything is run on my wiki:
[Wed Aug 08 14:49:59 2007] [error] [client (IP)] File does not exist: C:/[System path to wiki]/wiki/common, referer: http://sesdb.kareeser.com/wiki/Main_Page
[Wed Aug 08 14:49:59 2007] [error] [client (IP)] script 'C:/[System path to wiki]/wiki/missing.php' not found or unable to stat, referer: http://sesdb.kareeser.com/wiki/Main_Page
The second line I know says that missing.php is not found, because that's what I set ErrorDocument 404 to in httpd.conf. But the only reason why missing.php is being called is because of the first line, where it seems to look for the common directory?
Basically, I want to know how to stop these from being generated, as I think something's wrong...
Latest comment: 17 years ago7 comments6 people in discussion
I just made Template:Rounded that creates a box with rounded corners. It's a bit of an experiment. It uses 10 divs to give the rounded corners effect. My question is: are there any foreseeable (technical/webdesign) disadvantages of the usage of this many divs for creating only one box with rounded edges? Freestyle21:15, 6 August 2007 (UTC)
Unfortunately it looks really ugly in IE6 (SP1 6.02800.1106 on Win2000). So I guess you could use it on your user page… ∴ Alex Smotrov22:06, 6 August 2007 (UTC)
Okay thanks for the info, If I can find a computer with IE6 I might try to fix that. P.s. Don't forget that most things look really ugly in IE6 :P .Freestyle22:13, 6 August 2007 (UTC)
There, I think I fixed it The keyword here was adding "overflow: hidden" to the divs. I also added an extra line of rounding, but it may seem a tad too much? Someone please check if it still works under Firefox. — Edokter • Talk • 16:49, 7 August 2007 (UTC)
It's fine in Firefox 2. --ais523 16:53, 7 August 2007 (UTC)
Latest comment: 17 years ago5 comments2 people in discussion
Is there a way to track the use of magic words on pages? I ask because it would be an immensely useful clean-up category to be able to find all the biographical articles (at a first pass, those with {{WPBiography}} on their talk pages) that lack DEFAULTSORT. Humans using AWB could then suggest a defaultsort based on the category pipes and/or article talk page. However, identifying those pages that lack defaultsort is a problem. Possibly the only way to do it at present is to use a bot to read pages and output a list of those lacking DEFAULTSORT. It would be preferable though to have a software solution. What is needed is a category inside the magic word, or a "what links here" type thing for to track a magic word like DEFAULTSORT. Is this possible? Carcharoth13:29, 5 August 2007 (UTC)
It isn't tracked. You could use a database dump to find them quicker and more efficiently than with a bot. However, I'm not sure what the urgency is. I'd say it's best to just tackle them as they are found since no user visible behavior is impacted by leaving them, assuming the sort keys that are in place are consistent. I have a user script that helps me replace the explicit sort keys with DEFAULTSORT and I usually check it whenever I edit a biographical article, but I don't think I'd go looking for them. Mike Dillon06:39, 7 August 2007 (UTC)
You say "as they are found". I'm proposing to find them in one go, and start a clean-up project to try and ensure that all biographical articles are correctly sorted. If they are not correctly sorted, then biographical categories are a hopeless mix of articles sorted either by surname or by first name. There is little point in being able to sort articles if you cannot tell which ones are sorted without looking at individual articles. It is a bit like the categories Category:Year of birth missing and other clean up categories like that. This would simply be Category:Default sort key missing. If a category is too obtrusive, a "what links here" function would help. Another reason for wanting to track DEFAULTSORT is the point raised at this discussion, namely that DEFAULTSORT is being extensively used in talk page templates. This is silly because only the last occurence of DEFAULTSORT on a page will work. I suspect the vast majority of DEFAULTSORTs in talk page template are "DEFAULTSORT:PAGENAME", which is why the "only the lowest DEFAULTSORT works" behaviour is rarely spotted. And when a template tries to use a different DEFAULTSORT, it rarely works unless it is the lowest or only template on the talk page. But there seem to be strong arguments for deprecating the use of DEFAULTSORT in templates altogether (templates 'hide' the defaultsort, making it very difficult to tell what is going on when one template's DEFAULTSORT is over-riding the others), or limiting it to just DEFAULTSORT:PAGENAME. As I say in that discussion I linked to, it would be nice to be able to do a "what links here" for the DEFAULTSORT magic word, and restrict the results to template namespace. Failing that, the options seem to be a bot search, a database dump search, or a search like this. Now do you see why I'm saying that tracking magic words would be a very useful feature? Who do you think would be able to actually answer the question of whether this is possible? Carcharoth10:32, 7 August 2007 (UTC)
I guess I was assuming that most biography articles are sorted right, with or without default sort, since that's the case with most biography articles I've seen. In addition to what I said above about editing default sorts if I'm already editing a bio article, the other case that prompts me to do it is seeing something out-of-order in a category (it just doesn't seem to happen often). Given what you're talking about, I guess it would be useful to find the intersection of {{WPBiography}} articles and those that don't use DEFAULTSORT to check that they are actually being sorted correctly (since there isn't much chance of someone adding DEFAULTSORT in a biography article for a reason other than sorting correctly).
As for whether it's possible to do, yes, it is possible, but it would require changing MediaWiki to do it. Finding uses of magic words (or parser hooks for that matter) would be a lot like tracking uses of categories. The change would require some parser changes to keep track of magic word use and a new database table to track it. Whether or not this is something that will be implemented is another issue, but it doesn't seem like it would be that hard. Mike Dillon04:09, 8 August 2007 (UTC)
Thanks. You've answered my main question about whether it is at all possible. Taking other questions to your talk page. Carcharoth15:19, 8 August 2007 (UTC)
Microformat markup in article space
Latest comment: 17 years ago8 comments2 people in discussion
I have recently become worried of article modifications by Wikipedia:WikiProject Microformats, which intends to introduce microformat html markup in wikitext as such, instead of implementing it inside the usual data entry templates. This unnecessarily obfuscates wikitext with keywords that are meaningless to most editors, making articles harder to edit. Examples of the damage can be seen here and as single purpose inline microformat templates (such as used on [7] and [8]). What could be done to keep the microformat features but not show the special purpose markup to article editors?
I have tried combining the semantic annotation and the microformat functionality of those special use templates into the standard templates to keep articles editable by proposing a change to add a name parameter to the templates, but a single editor keeps insisting on a minor aesthetic detail as a reason to keep using his own templates. The proposed change would make the obscure
{{hcard-bday|Firstname Lastname|birth date and age|1955|06|08}}
to
Firstname Lastname, born {{birth date and age|1955|06|08|name=Firstname Lastname}}
Having failed to obtain support for your campaign elsewhere, you appear to be forum shopping. Your claim of "[intention] to introduce microformat html markup in wikitext as such" is bogus. Your attempts at modifying these templates are, as explained previously, broken (and the issue is not a "minor aesthetic detail"). Andy Mabbett | Talk to Andy Mabbett14:17, 2 August 2007 (UTC)
Requesting more opinions to a discussion where a single contributor manages to make it seem as if there's no consensus is not forum shopping at all, it's lack of participation, which apparently needs to be requested. People, please contribute, the editor's well known repetitive argumentation has repeatedly stopped admins from closing editprotected requests because on a quick look it may seem that he still has something new to say. The topics are on the long side, my apologies for feeding the troll. --Para14:30, 2 August 2007 (UTC)
There has been some more opinions to one of the microformat merge proposals, but so far nobody has commented on the issue of including raw special case HTML in articles directly. In its current form it seems to really be getting out of hand with edit obstructions such as here or here. How can the situation be improved to keep articles editable? Is there perhaps some semantic extension somewhere between the standard MediaWiki and a full blown Semantic MediaWiki installation? --Para11:04, 8 August 2007 (UTC)
The edits you describe as "edit obstructions" are, of course, nothing of the kind - the articles are still fully editable. I also note that you've previously opposed the kind of templates which would obviate the need for the in-line-class mark-up used in the examples you give. Andy Mabbett | Talk to Andy Mabbett13:27, 8 August 2007 (UTC)
Technically oriented people may still be able to edit articles with microformat html mixed in with the text, but it's total gibberish for most others. The whole point of wikitext is to make it easy for people to participate without having to touch the raw html, and these microformat edits are moving away from that principle.
What microformats seem to need are containers for fragments of data, and a wrapper to tie them all together. In most cases this requires the data to be clustered close together as overlapping wrappers aren't allowed. I am not expecting to see templates to wrap pieces of information together and as a side-effect create a microformat wrapper (like {{PoI}} for example does), as such meta-templates would make no sense logically, wouldn't be easily editable either as the syntax differs between meta-templates, and would take away all the editor freedom to format and order the content. I'm thinking more of something like an extension that 1) allows information to be tagged to tell what it is, and 2) reacts to a special keyword with some attribute syntax to explain how additional markup should be generated. The wikitext rendering module could then generate the microformat html and other similar additional formats that can be included in the standard html, based on the semantic markup. That way people wouldn't have to look at keywords of some obscure output format, but general semantic ones readable for anyone. SMW does something like this, but I feel it's bit of an overkill with all its query features. --Para15:03, 8 August 2007 (UTC)
Latest comment: 17 years ago3 comments2 people in discussion
I want to comment on a bug - see here, but I don't want to create a bugzilla account. I found Wikipedia:Bug reports, but that said to create a bugzilla account! Is there anywhere I can post something for people to move over to the relevant pages? In this case, I want to point out that the feature being voted on could disrupt the linkage between pages that redirects help to maintain. My point against a feature that allows users to suppress redirect creation can be summed up as follows: "unless the software checks for links to the page and alerts people "THERE ARE 500 INCOMING LINKS THAT NEED TO BE REDIRECTED, SUGGEST YOU DO NOT SUPPRESS THE CREATION OF A REDIRECT", then inexperienced, tired, careless, users and admins will use the feature when it shouldn't be used". If someone could post this over there, I'd be most grateful. Carcharoth16:41, 9 August 2007 (UTC)
If you'd like, drop the bug #, description of what you want to add, and the severity on my talkpage, and I'll add the notes for you. ~Kylu (u|t) 00:22, 10 August 2007 (UTC)
Latest comment: 17 years ago11 comments7 people in discussion
The image Image:Waterhouse-sleep and his half-brother death-1874.jpg doesn't seem to display when scaled to 310px. This hasn't been a problem before though, the image has been displaying fine at 310px for some while now. Here is a revision where it was scaled at 310px and here is a revision where I changed it to 309px and it displays fine. This same problem appears to happen on allpages it is scaled to 310x at, but only seems to happen with this image. --Android Mouse06:01, 9 August 2007 (UTC)
Looks fine here. Most likely a problem with your local cache or the WP's caching server you are using. Clear your browser cache, then try viewing the revision with the 310px version by adding &action=purge to the URL. (PS, fixed first link.) — Edokter • Talk • 11:14, 9 August 2007 (UTC)
This problem happens a lot with SVG images, too. Some images simply refuse to display at certain sizes. - 52 Pickup
Now it's displaying fine for me aswell, probably something wrong with the server's cache as I've disabled caching entirely in my browser. --Android Mouse17:47, 9 August 2007 (UTC)
== "string") return el;
if (typeof el == "undefined") { return el };
if (el.innerText) return el.innerText; // Not needed but it is faster
var str = "";
var cs = el.childNodes;
var l = cs.length;
for (var i = 0; i < l; i++) {
switch (cs[i].nodeType) {
case 1: //ELEMENT_NODE
str += ts_getInnerText(cs[i]);
break;
case 3: //TEXT_NODE
str += cs[i].nodeValue;
break;
}
}
return str;
}
</syntaxhighlight>
The problems is in the line "if (el.innerText) return el.innerText;" innerText is an IE extension. on FF it is not used because it is undefined. On Safari the innerText call IS present but unfortunately, it is not the same as the IE version. On Safari it does not return "display:none" text. Therefore the whole sortkey is unusable on Safari.
There are 2 ways to deal with this issue:
remove the entire line. I'm not sure if it's actually that much faster on any platform, and the FF way seems to work just fine.
add "if (el.textContent) return el.textContent;" right before it. This basically is the FF variant of innerText and Safari supports it as well. Some very old browsers might not have it, but they can use our current FF for-loop. I tested textContent on Safari 2/3, FF 2.0 and Opera 9.
As some might know, there have been some issues with sortable wikitables left on Safari. I have found the cause of some of my problems. Some tables use a sortkey trick. Take the following tablecell: <td><span style="display:none">02005-06-02</span> <a href="/wiki/June_2"
title="June 2">June 2</a>, <a href="/wiki/2005" title="2005">2005</a></td>
The idea that the span is hidden from the reader, but is used in sorting. This was not working in Safari (2 and 3). The cause lies in wikibits.js in function ts_getInnerText(). This function retrieves the textual contents stripped from the html tags.
functionts_getInnerText(el){if(typeofel
Transclusion
Latest comment: 17 years ago5 comments2 people in discussion
Latest comment: 17 years ago8 comments3 people in discussion
Hi everyone. I got tired of handling word wrapping in long link lists, for instance in navigation boxes. So I have made a new template similar to {{nowrap}}, but that only prevents word wraps in links. That is, it allows wraps between links and in normal text. Thus one single {{nowraplinks}} tag can surround the whole link list instead of having one {{nowrap}} tag around each link or using lots of .
I would like if some expert template editors (preferably with some CSS experience) take a loot at it to see that I didn't miss anything. I think this template might be used a lot once we deploy it.
Oh, I probably should mention that this template will need the addition of one line of CSS code to common.css. So I think this needs to be thoroughly checked and discussed before we deploy it. (And I am not an admin so I can not add that line of code myself.)
Edokter: Yes, I tried that too and Wikipedia just filtered the tags to text output. I even tried some nowiki tags etc to try and get it to stick, to no avail.
BenB4: And yes, that "nowraplinks" class can be very useful. If/when it is in the common.css it can be used together with pretty much any HTML tag or any box to prevent links to word wrap. For instance like this:
<div class="nowraplinks"> Lots of text and links </div>
Or like this:
<span class="nowraplinks"> Lots of text and links </span>
That span code is really what the {{nowraplinks}} tag is packaging in an easy to use way.
Thanks BenB4, I didn't know that. I have read up on it now and it seems you are right about the {{editprotected}} procedure.
Edokter: Oops? I explain that in the talk page of the {{nowraplinks}} template, but I should have mentioned it here too. Sorry for being unclear. So, any one else reading this, go to that talk page for explanations and for some test examples.
Latest comment: 17 years ago2 comments2 people in discussion
In Kentucky State Fair, the gallery image with the caption "Freedom Hall during Fair" isn't displaying. This image was recently moved to Commons. I purged the description page in Wikipedia, as well as the page in Commons, twice to no avail. Is there something I'm missing? Stevie is the man!Talk • Work02:52, 10 August 2007 (UTC)
Latest comment: 17 years ago11 comments8 people in discussion
In the article Great Upheaval, under the the governors letter, just above the heading After the expulsion there is a link in the form of a footnote that redirects back to the same page. The problem is that I cannot find that note in the editing window to remove it. I cannot figure out where it is coming from. I would appreciate it if someone would clear up this mystery for me. Thanks! Regards, Mattisse22:59, 8 August 2007 (UTC)
(EC) Go and take a look again, I added a "notes" section that displays the text of the reference ("ref" tag) that you will see if you edit that specific subsection of the article. If this doesn't quite make sense feel free to ask again or try the Wikipedia help desk for more detailed info. Wikipedia:Help desk. dr.ef.tymac23:17, 8 August 2007 (UTC)
Thanks. I see it in the Footnotes but I still cannot find it in the text in the edit window. I understand about {{reflist}} and <references/>. Oh well. Thanks anyway. Regards, Mattisse00:14, 9 August 2007 (UTC)
Wow! Have you not worked that out in over a years worth of editing on Wikipedia? Sorry if I sound rude, I'm only curious and can't word it any other way! ;-) Ah well, one learns something new every day! Lradrama14:42, 9 August 2007 (UTC)
Lradrama, are you addressing Matisse? and do you mean that the answer provided by Smotrov is trivial? It occurs to me that you "can't word it any other way" because you aren't actually saying anything. And Carcharoth, I think a typo somewhere prevents me from understanding your reply too. My policy is to employ complete, coherent sentences when flaming :-) Pete St.John17:29, 9 August 2007 (UTC)
My dear Pete St. John, I was merely indicating surprise at the question that was from a very experienced editor. I just didn't want to sound rude you see. As for the answer given by Smotrov, it is a very fine answer indeed, and exactly what I would have written. Lradrama10:33, 10 August 2007 (UTC)
Bot archival setup
Latest comment: 17 years ago5 comments2 people in discussion
The New Contributors' Help page desperately needs a better archival system, but I'm not 100% sure which bot to use and how to do it. Normally, when something needs doing, I just go ahead and fix it, but I would rather get this right first time. MiszaBot II looks like the best choice, but the instructions are a little too vague, particularly with regard to selecting the correct archive page (unless that is done automatically?). Can anyone advise someone who has never used a bot before? AdrianM. H.17:00, 8 August 2007 (UTC)
Before choosing a bot, it's best to see what type of archiving system needs to be automated. At the moment, the page is archived by month, but this could be changed. Some possible options are to archive every time the page gets to x sections/bytes long, or to have a weekly archive. Another idea might be to follow what happens in the Village Pump and remove completely any sections that are older than two weeks. If a bot doesn't contain any options in the interface to do what is required, it's possible to ask the operator to set it to do those functions. Tra(Talk)18:30, 8 August 2007 (UTC)
I think a bi-weekly archival would be suitable. Is the existing archive page sequence suitable, and does the process need any regular human intervention? AdrianM. H.19:10, 8 August 2007 (UTC)
I've had a look, and I don't think the bot has an option to do bi-weekly; it can only do monthly. However, you might be able to sort something out by contacting the bot's operator. Tra(Talk)21:29, 9 August 2007 (UTC)
Latest comment: 17 years ago2 comments2 people in discussion
Semantically, each page on Wikipedia should have an address element. If this were marked up with class="author" an hCardmicroformat, the hAtom microformat could be added to relevant pages, such as Portal:Current_events (everything else required is tested and possible). It seems to me that the page footer:
Wikipedia® is a registered trademark of the Wikimedia Foundation, Inc., a US-registered 501(c)(3) tax-deductible nonprofit charity.
or part of it should be in that element, with "Wikipedia" as the hCard's class="fn org" property.
(I have tried adding an address element, manually, to a sandbox page, but it's not recognised by MediaWiki.)
I don't think you've made clear exactly what problem this change would help solve. What exactly are the benefits of this change, or the disadvantages of the current situation? Or, if you will, how do you answer the question of "Why?", in response to your first sentence above? -- John Broughton(♫♫)20:41, 10 August 2007 (UTC)
Templates and "What links here"
Latest comment: 17 years ago3 comments3 people in discussion
When a page is linked from a common template, "What links here" will show a lot of pages that actually all include the same template. Would it be possible to group all those pages under the template name? The same goes for "File links" on image pages. --Apoc240000:34, 12 August 2007 (UTC)
That would perhaps be useful, but already you can select a namespace (e.g. Template:) in "What links here", and, if desired, then apply "What links here" to a particular template.--Patrick00:40, 12 August 2007 (UTC)
Latest comment: 17 years ago5 comments4 people in discussion
Greetings! In Song Dynasty, I reduced the images to default size, per FAR rules. Now there is an image that overlaps the text on the right side, and no matter where I move it the same thing happens. The article is now in FA review. It is under the heading "Southern Song". Would very much appreciate any help and tips about preventing this sort of thing. It rarely happens but when it does......Thanks! Sincerely, Mattisse02:08, 11 August 2007 (UTC)
Yen sign looks OK to me with IE 7.0.5730.11, but it has overlaps with some window geometries with Firefox 2.0.0.6. Welcome to the wonderful world of browser & browser-version CSS compatibility weirdness. -- Boracay Bill11:38, 11 August 2007 (UTC)
Latest comment: 17 years ago4 comments4 people in discussion
Hello, there is that annoying yellow messages bar up there. Problem is the messages are about a couple months old (I didnt do it) and it stays there after reading the page. I tried clearing the cache , deleting cookies but its still there. How do I get rid of it? Does anybody else have this? —The preceding unsigned comment was added by 71.112.246.219 (talk • contribs) 21:50, 10 August 2007.
The only way I know of to ensure that the new messages bar goes away once it's appeared for an anon (thanks to Bug 9213), is to get someone to delete your user_talk page. Failing that, you can wait a few hours/days/weeks and it might go away on its own. However, you seem to have no messages on your user_talk page, so why you'd even see the new messages bar at all is a bit of a mystery. Presumably you're having this problem on a different IP than the one you used to post this question?--VectorPotentialTalk21:58, 10 August 2007 (UTC)
Yep, it was on a diffrent computer, anyway thinks for the quick response and I hope you get this fixed soon. I've managed to ignore that bar. —The preceding unsigned comment was added by 71.112.246.219 (talk • contribs) 22:14, 10 August 2007.
I would recommend getting an accoutn as the above point states, as some IP addresses are operated by more than just one person, whether they know so or not, and messages may be recieved for what other people have done, and not necessarily you. Lradrama10:32, 11 August 2007 (UTC)
Template middot {{·}}
Latest comment: 17 years ago3 comments2 people in discussion
Expert template coders needed to double check a high risk template.
Template middot {{·}} is the dot used for most dotted lists in most navigation boxes here on Wikipedia. So it is used in extremely many articles.
Some day ago on that template's talk page I suggested an improvement of the code for the template, but I have received no responses at all. So before we add the code to the article I would like some experienced template coders to take a look at the code so I didn't miss anything. Because if I have it will effect a big part of the articles...
I that assume no one had noticed it because you did not use {{editprotected}}. You don't have to, but if you don't, there's not that much of a chance someone will happen to find your suggestion (depending on the template). I've added it; hope this helps. I'll take a look at your comments. GracenotesT§23:41, 12 August 2007 (UTC)
He, I guess I am to careful but I thought some double checking before asking an admin to "immediately edit" the template. Three minutes after I asked here you added the {{editprotected}}, and another three minutes later the admin TKD had responded and copied my code suggestion to the template. Things are moving fast around here! :))
Thanks for the confidence in my code. I checked some articles that use it and ran some more provocation tests. It seems to work perfectly. Now all I have to do is to update the documentation.
Latest comment: 17 years ago6 comments3 people in discussion
The speed at which pages load has been significantly slower for me over the last few days, about 25-50% the speed they downloaded previously. This even happens when I use the forward and back buttons. I'm using Firefox 2.0.0.6. I still get good speeds if I use Internet Explorer, but Firefox is my preferred browser.
I am having the same problem with Firefox 2.0.0.6. I was wondering whether it was my browser or Wikipedia. So it is Firefox? Drag, since it makes many tasks onerous. It's enough to make me consider changing browsers. Regards, Mattisse12:28, 11 August 2007 (UTC)
Different platform (Firefox 2.0.0.3 on Linux), same problem. With the occasional white screen of death, of course. It's a wikipedia problem, time to give the devs a bump. MER-C13:42, 11 August 2007 (UTC)
Latest comment: 17 years ago5 comments4 people in discussion
I swear something strange is going on here. Although a bot is counting the number of good articles on English Wikipedia, it appears to be undercounting. As of the time of this post, the number of GAs according to a bot count is 2,748. But if you go to Category:Wikipedia good articles and ignore the subcategories and first 7 listed articles which are all WP pages and count the good articles from there (by clicking next 200) you should find you have 2,760 good articles. The bizarre thing is if you count backwards from the end using previous 200, you arrive at a smaller number, the same number as the bot count. Now I looked into why there is a discrepancy...the reason is when you count backwards, an article that is between the first article of one page and the last article of the previous page disappears from the listing giving you the apparent smaller number.
This isn't an isolated incident. There are supposedly 1,531 featured articles when again by checking Category:Wikipedia featured articles there seem to be 1,537. But count backwards through the category and again the count is 1,531. What's I can show the larger figure is the correct figure in both cases by simply listing them all in Excel.
There was a similar issue regarding featured pictures sometime ago, see WT:FP#Discrepancy in the featured picture count. It's probably due to lazy closers not updating the count or people deciding their article is now "featured(tm)". As for fixing it, you'll have to do it the slow way. MER-C12:39, 13 August 2007 (UTC)
The problem here isn't people forgetting. The problem here is the bot is miscounting. A bot updates the count every day but the bot is counting incorrectly. Centy – reply• contribs – 14:22, 13 August 2007 (UTC)
There is no problem here. The bot counts the number of articles on the list page in both instances. There are exactly 1531 articles listed on WP:FA, and because the FA category there corresponds to the FA list, the category also contains 1531 talkspace articles. Almost the same algorithm is used to count the articles listed on WP:GA. According to WP:GA, {{GA number}} is the count of articles listed at WP:GA. It has nothing to do with CAT:GA. There are 2748 articles listed on WP:GA. Because the GA process is not so tightly controlled, there happen to be 2757 talkspace in the GA category. If you're not getting these numbers, perhaps you are not excluding pages from the categories that are not main talkspace? Gimmetrow02:59, 14 August 2007 (UTC)
The category traversal does seem to show different contents in the forward vs. the reverse direction. It might help if you could paste them all into Excel and let us know the differences (at least one of the "missing" FAs when traversing backwards is a userspace article). This looks like a bug of some sort in the category traversal code to me. As Gimmetrow says, the official count of FAs is from WP:FA, not Category:Wikipedia featured articles, but whatever else is in the category should be there whether you traverse it forwards or backwards. -- Rick Block (talk) 04:21, 14 August 2007 (UTC)
reCAPTCHA
Latest comment: 17 years ago16 comments11 people in discussion
Currently we use a CAPTCHA on the sign up page. [10] Has anyone considered replacing it with reCAPTCHA to help digitize books? (Mediazilla appears to be down right now.) — Omegatron00:31, 4 August 2007 (UTC)
The article says that "reCAPTCHA is a free service, except for users who would require a prohibitive amount of bandwidth." Since Wikipedia has a lot of signups each day, they may want to charge money. Tra(Talk)01:02, 4 August 2007 (UTC)
We also use captchas whenever new accounts and IPs try to add a new external link to an article, or whenever a login fails. The load could add up to quite a bit. Dragons flight01:05, 4 August 2007 (UTC)
That's ~7 / minute, we have a couple new account creations per minute on enwiki alone, so I'd be pretty sure that collectively there are more than 10000 captchas served per day. How many more I'm not sure. Dragons flight01:17, 4 August 2007 (UTC)
I've had a look at the log, and I estimate there to be approximately 7300 account creations in the last 24 hours. If we assume that only 10% of people who get to the account creation form actually sign up, that's 73000 requests to provide a captcha each day. Then you need to add in the number of people who get it wrong and need another captcha, plus the number of IPs adding links and the number of incorrect passwords. I think they probably will want to charge some money. Even if it was affordable, the foundation might not want to do it since they tend to want to use free software (bear in mind that these numbers are just estimates, and may be hopelessly out). Tra(Talk)01:28, 4 August 2007 (UTC)
For us it would make more sense if to use our own re-captcha for helping out with our own OCRing efforts. Sadly our efforts are not organized enough to do that, yet. --Gmaxwell14:20, 5 August 2007 (UTC)
Would it make sense to help them in the interim to see how well this works? maybe if we get them a lot of digitization they'd put our own OCR stuff in as part of what gets done? ++Lar: t/c14:23, 5 August 2007 (UTC)
The fact that they refuse to open-source the software is kind of a showstopper for us for being involved with it. --brion01:30, 6 August 2007 (UTC)
I'd suggest they be asked. Given whom we are, they might be very willing to accommodate the load. But I would think the Foundation would want to be asked in any case. DGG (talk) 07:40, 6 August 2007 (UTC)
Brilliant idea! Literally LOL! But if the meaning of the text is unknown, how is that a verification? (E.g. gobbledigook -> biteme. Sounds good. Welcome, bot.) Saintrain17:57, 10 August 2007 (UTC)
Saintrain: Follow the links! I first wondered how it worked too. But this is Wikipedia and if you look above you see that Omegatron did link to the article reCAPTCHA and it explains it. Pretty nifty idea actually. Although it is strange that they charge money from organisations (server owners) that help them digitize their texts... --David Göthberg18:37, 10 August 2007 (UTC)
It also seems that the reCAPTCHA's audio version for the visually impaired is not well-tailored. I found it extremely difficult to pass. This might not be a good solution for that reason even if they were willing to allow its use freely. (I suspect the lack of open source code is to make it more difficult for crackers to beat). -- Myke Cuthbert(talk)04:06, 14 August 2007 (UTC)
Help posting image
Latest comment: 17 years ago3 comments2 people in discussion
Latest comment: 17 years ago8 comments4 people in discussion
Can I ask that articles tagged for speedy deletion be left for at least 24 hours (or longer! - maybe 7 days!) before acting on these tags? I say this since after several months someone tagged an article of mine for speedy deletion. This can be done by someone in the US or elsewhere at a time when I may not log in for several hours, by which time the article has disappeared. This particular case said the article appeared to be blatant advertising - it was not. However I was in no position to argue this or to change the article to improve it. By the time I got there, it had gone!
I believe a lot of people agree with this in theory (I sure do), but it is very hard to make a 24-hour or 7-day queue of articles out of a category, which is how speedy deletion candidates are located by admins. Can we use a category pipe to sort on the subst of the date? ←BenB422:33, 13 August 2007 (UTC)
This is a common request, but the fact of the matter is most articles simply don't need to stay around for very long. If you feel that an article was improperly deleted, I recommend you take it up with the deleting administrator; if you explain that it was an in-progress article, they should be willing to undelete it so you can improve it. EVula// talk // ☯ //22:35, 13 August 2007 (UTC)
I don't think new users, who are proportionately more offten bit by speedy deletion of their in-progress articles, know how or who to ask. Biron says that you can't count on undeletion. There is a chance that a deleted article can be reclaimed at any time. ←BenB422:40, 13 August 2007 (UTC)
Very probable. Should we perhaps tweak the page that says "this article doesn't exist" to include more specific instructions about how to find this information? I'm strongly opposed to tweaking the speedy deletion process since I see this sort of "false positive" as being an extremely rare thing, and stretching it out to something like 24 hours or 7 days dilutes the purpose behind it (and begins to look like prodding). EVula// talk // ☯ //22:47, 13 August 2007 (UTC)
(edit conflict) Speedy deletion is, as the name says, intended to be speedy, and is (or at least should be) only applied to a limited list of clear and uncontroversial cases. If you feel that your article has been inappropriately speedy deleted, you may ask the admin who deleted it to reconsider the issue or, if necessary, submit it for deletion review.
In the situation you describe, it may be more effective to ask the admin to just send you a copy of the deleted article without actually undeleting it, so that you may edit it into something that no longer seems like a blatant advertisement before resubmitting it. Alternatively, in some cases you might find it easier to just write a better article from scratch.
Do keep in mind, however, the if the recreated article still looks like blatant advertisement, it may and probably will get deleted again, and possibly protected against recreation if you keep recreating it over and over. Thus, it's best to make sure you've adequately addressed the problems the article had before resubmitting it. One possible approach is to first create it as a subpage of your user page, and ask for comments from other editors (possibly including the admin who deleted it previously) before moving it into the main article space. —Ilmari Karonen (talk) 22:42, 13 August 2007 (UTC)
This would be easy if the administrator who deleted the article showed more than an IP address - which was true in this case - with no existing talk page. Look at User_talk:Soarhead77 for this example! Soarhead7722:48, 13 August 2007 (UTC)
BBC Radio Ripper is a commercial program which will take Real Player stream audio and related formats used by the BBC for its Listen Again and Listen Live functionality and convert this to mp3 format. it is built by the company Alionsoft.
Latest comment: 17 years ago3 comments2 people in discussion
Earlier today, I went to add articles about Led Zeppelin (the bands article, article on there songs, etc) to my watchlist. I had no problem doing this, but to add all the articles to my watchlist, i had to go every article and add it. I want to watch similar articles on bands like The Who and The Rolling Stones, and was wondering if there was an easier way to add all the articles. Sasha Callahan04:54, 13 August 2007 (UTC)
You can use "Edit raw watchlist" (at the top of your watchlist page), and paste in a large list of pages. Might be faster for you. --Splarka (rant) 07:55, 13 August 2007 (UTC)
Latest comment: 17 years ago7 comments5 people in discussion
I made a numbered list with four items on a talk page (old revision), and the items are labeled as 1, 3, 4, and 5, skipping the number 2! Is this some sort of bug, or am I doing something wrong? Anyways, I've already submitted a report on Bugzilla. --Ixfd6403:03, 13 August 2007 (UTC)
I don't know why there is this extra list item with style="list-style: none", but it seems due to the span tag before the list.--Patrick04:02, 13 August 2007 (UTC)
And oddly enough, the numbering is working just fine in RfAs (and in a test I just did in my userspace). Hrm. EVula// talk // ☯ //04:16, 13 August 2007 (UTC)
It's because of the unclosed <span class="plainlinks"> in Ixfd64's comment above. If you close with </span> then things return to normal:
a
b
c
That is certainly a bug though. Here is the HTML generated above:
<p><span class="plainlinks">That's not it, look:</span></p>
<ol>
<li><span class="plainlinks">a</span></li>
<li style="list-style-type: none; list-style-image: none; list-style-position: outside;"></li>
<li><span class="plainlinks">b</span></li>
<li><span class="plainlinks">c</span></li>
</ol>
<p><span class="plainlinks">MediaWiki is out of 2s. WTF? ←<a href="/wiki/User_talk:BenB4"
Latest comment: 17 years ago8 comments6 people in discussion
Is it me or has there been a significant drop in userpage vandalism since the latter half of July? I've been looking at userpages which have been vandalised a real lot (mine being one of them) and there seems to be a significant decrease. I'm not complaining! Just thought I'd point it out. Have vandals taken a change of heart or has some sort of protection been inserted into the system? Lradrama18:32, 11 August 2007 (UTC)
Probably due to Bug 9213. This prevents anonymous IPs from getting that "New Messages" bar when you place a message in their talk page. SInce they are not getting warnings, they don't have user pages to vandalize. Maybe this is a good thing?? --Hdt83Chat19:31, 11 August 2007 (UTC)
Bug 9213 is not anything new, as far as I'm aware. The bug report merely presents the problem in the form of, well, a bug report. GracenotesT§23:26, 11 August 2007 (UTC)
I think it's the bots. The anti-vandal bots are getting better tuned, and the recognize the unique character of those likely to repeat vandalize (all caps, penis, etc.) and revert within a minute. Vandals new and old alike see that kind of thing and get discouraged. ←BenB406:05, 12 August 2007 (UTC)
Ah well, it's a good thing! Just confused about one thing though - if vandals don't recieve the message alert, do they know if they've got a message then? If so is there point in sending them anymore? Lradrama18:10, 12 August 2007 (UTC)
Bug 9213 is intermittent, so it's always possible to hope that the message will come up. (Sometimes everything works fine; sometimes the new-messages bar doesn't come up at all, and sometimes it doesn't go away again after the IP has read the message. It's not entirely clear exactly what's going on, though; intermittent bugs are often the hardest to pin down. There have also been changes to various interface messages to try to get IPs to check their talk page intermittently, and some more experienced IP editors have probably caught onto the bug by now and will use their own workarounds.) ais52314:49, 13 August 2007 (UTC)
Search box moved on home page
Latest comment: 17 years ago3 comments2 people in discussion
hey,
I just wanted to log my opinion on the new location for the search box on the home page. It used to be pretty far up on the page, quick and easy to find. Now you have to scroll *way* down to find it. I much preferred the old location. I use the search feature a lot, as I would imagine most users do - that's pretty much the point of wikipedia, isn't it?
I would like to see it moved back up, or even to the top of the page where it really should have been all along, IMO.
Just putting that out there in case anyone cares what users think. I looked at the bugzilla thing to put it in, but it looks like way too many steps to get into.
hi Edokter, I usually use en.wikipedia.org. And it seems to me it was always at least above the "picture of the day," if not slightly higher. Now it is several screens down (5 screens for me on firefox), when really it should be right at the top next to the welcome message on all the pages. And yes, I do see the link at the top, but that just jumps downs rather than typing in the box there; plus that is only on the welcome page, not every article page too. Just MHO, thanks for your response. —The preceding unsigned comment was added by 76.226.106.239 (talk • contribs) 02:01, 12 August 2007.
It's back!!! You're right, it is on the left, very high up as I remembered. For some strange reason the entire left frame had disappeared from my firefox, for at least several days, but now it is back. Maybe one of the automatic upgrades firefox always seems to be doing was the culprit and/or the fix. Weird.
Thanks for your help!
The first is legal, but not recognized for some reason, not sure about the other 2, but at least the second is probably ok, and I imagine the third probably is. Prodegotalk06:42, 12 August 2007 (UTC)
The first is a CSS3 property, which is not recognized by Firefox yet; its Firefox equivalent (-moz-column-count) is also used, and has almost the same behaviour as the CSS3 stylesheet. The second is probably the WebKit equivalent of the first, and is also vendor-specific like -moz-column-count (it has a valid vendor-specific prefix), so it's actually valid (no matter what the W3C validator says). The third one, as the comment on the stylesheet says, is yet another CSS3 property not recognized yet by Firefox. So, all three are not real problems with the stylesheet, and there is nothing to fix. --cesarb00:44, 13 August 2007 (UTC)
Yeah, I modified the CSS page to add "column-count" because it gives the behavior to WebKits browsers like Safari. --MZMcBride23:51, 13 August 2007 (UTC)
Latest comment: 17 years ago2 comments1 person in discussion
I am a victim of a shared IP block that expired two days ago. However whenever I edit any page on wikipedia I get a "The connection was reset" message. I have tried the secure website and other wikipedias, no luck. I have no problem editing talk pages, though. What's wrong? --Skyfiler02:51, 14 August 2007 (UTC)
Latest comment: 17 years ago4 comments4 people in discussion
My userbar is jumping around the top of the screen! Can someone tell me how to fix it? Here's my Monobook.js. I recently installed Twinkle and then deleted the text. I don't know if popups have to do with it, and I use IE and sometimes Firefox to edit. Thanks a whole lot.--Kkrouni/Ккроунл/ΚκρΩυνι22:52, 11 August 2007 (UTC)
It's an intermittent problem on IE6; it happens sometimes even with no scripts installed. I don't know what causes it or how to fix it, though. ais52314:45, 13 August 2007 (UTC)
Latest comment: 17 years ago6 comments3 people in discussion
Just noticed this link in my watchlist "Wikipedia:Obesity is for the hateful racist that is [username] who is a [poopiepants (well you get the idea)]". Not surprisingly, the link is red. How did that happen? Saintrain17:58, 14 August 2007 (UTC)
A vandal moved a page that you had watchlisted to that name, someone reversed the move, and deleted the vandalized location. The vandalized location stayed in your watchlist, as moving pages automatically adds both the old and new locations to a user's watchlist. EVula// talk // ☯ //18:03, 14 August 2007 (UTC)
Thanks: Did I understand correctly that the original page is (probably) still in my watchlist? (Since it's overlarge (and not backed up), it's hard to tell if/what's missing.) If not, can I track the changes via the deleted article link? Saintrain19:39, 14 August 2007 (UTC)
Latest comment: 17 years ago2 comments2 people in discussion
Does anyone know where I can find the code for <div class="horizontal">. I've searched most the css pages - however I only found one on list, and I don't think that's it.--danielfolsom04:04, 14 August 2007 (UTC)
A quick glance suggests that you're missing the latitudinal coordinates (hence {{{2}}} being displayed). EVula// talk // ☯ //04:08, 13 August 2007 (UTC)
Actually I figured it out. The problem was I added spaces where there should not have been any spaces. After closing the extra spaces and placing a vertical line where it need to be, it was fine. -WarthogDemon18:38, 14 August 2007 (UTC)
Getting a list of the most commonly done searches?
Latest comment: 17 years ago15 comments5 people in discussion
There would be many queries done on wikipedia which do not redirect to any page, and give a search result. If we had a list of the most commonly used searches which does not redirect to any articles, we can have list of pages where creating a redirect would be useful. Does something like this exist on wikipedia, or is sometrhing like this planned? Shabda12:51, 12 August 2007 (UTC)
Such a list would be possible to make, but it seems like it would result in too much server load. Roughly (very roughly) 7 searches are made every 2 seconds, and recording all that might be too expensive. Wikipedia used to have a list of Special:Wantedpages, but apparently it was abused, or something. Most redirects are created based on editors' initiative to do so, and this system works. GracenotesT§20:00, 12 August 2007 (UTC)
Yep, I know. (I remember when it worked, actually.) I was trying to mention that special page in the context of a possible substitute, although it's not active, so scratch that. GracenotesT§20:26, 12 August 2007 (UTC)
Ah, that would work perfectly, we just need a random sampling, not an exact break-up. And n can be set up in a way to achieve perfect balance between server load and accuracy of the sample. Shabda05:10, 13 August 2007 (UTC)
It is a random sample. It works by running a piece of javascript on each page load which generates a random number from 0 to 6000, and if that random number is equal to 42, the toolserver is called to record the article view. This set-up could then be modified to record the text searched for, rather than the article name. Tra(Talk)13:25, 13 August 2007 (UTC)
What is the wikipedia development model? If such a change is accepted, what is process to get it coded and hosted on the tool server? Shabda14:13, 13 August 2007 (UTC)
I tried leaving a comment for him, but with the buttons being in German, I cant submit my comment. Shabda16:17, 13 August 2007 (UTC)
To leave a comment, click the '+' tab, then enter the subject in the Betraff field, and the comment in the large text box below. Then, click Vorschau zeigen to get a preview then Seite speichern to save the page. Tra(Talk)17:03, 13 August 2007 (UTC)
Latest comment: 17 years ago6 comments4 people in discussion
Why is the current job queue length so long (about 2.7 million)? Aside from yesterday (when it was at about 1.5 million), I had never seen it go above 500 thousand and usually it's below 100 thousand. Of course, I don't check that often, but still ... — Black Falcon(Talk)20:45, 18 August 2007 (UTC)
It fluctuates. Likely culprit is that someone made a series of changes to some widely used templates or categories. You also have to keep in mind that the job queue is not just for this wiki but for several (I don't believe all) Wikimedia wikis. In general, you shouldn't be considered by the length of the job queue unless you notice a lag occurring somewhere (populating a category, etc.). AmiDaniel (talk) 00:05, 19 August 2007 (UTC)
The job queue count shown on this wiki a very bad estimate of the length of this wiki's job queue only. Each wiki has an independent job queue, it's the job queue processing script that is shared between groups of wikis. 86.148.49.2818:31, 21 August 2007 (UTC)
A lag (32 hours so far) in emptying a template-populated category is why I actually checked the queue. Of about 100 pages, only about 40 have been removed so far. It's nothing urgent, but I just wanted to know if perhaps something big was going on that I was unaware of. :) Thanks for your response, Black Falcon(Talk)00:13, 19 August 2007 (UTC)
They are probably too specialized to go in the mostly reader oriented toolbox. However, there is no reason you couldn't add them for yourself through javascript. Prodegotalk05:24, 16 August 2007 (UTC)
JavaScript issues
Latest comment: 17 years ago4 comments3 people in discussion
I've recently been trying to clean up and simplify my user JavaScript (User:Pyrospirit/monobook.js), and apparently something I did broke everything; however, nothing I did should have done that. The problem occurred after this revision, where I replaced document.write functions with the importScript function, and also changed a parameter for popups and rewrote a few comments for clarity. The next revision after that was simply moving some code to user subpages (User:Pyrospirit/editintro.js, User:Pyrospirit/syntaxhighlighter.js), and the next 2 revisions were undoing and redoing in an attempt to fix the scripts being broken (so they cancel each other out). Could someone experienced with JavaScript explain why none of my scripts work and how to fix it? Thanks, PyrospiritShiny!16:20, 10 August 2007 (UTC)
One error in the JS file is enough to disable all scripts. Use browser console (Firefox2: Tools → Error Console, Opera: Tools → Advanced → Error console) and track all the errors one by one. You could start with adding // in front of the category at the end and fixing addOnLoadHook with capital L (Javascript is case-sensitive) ∴ Alex Smotrov16:50, 10 August 2007 (UTC)
I don't think there's a number of templates limit, but there is a limit to the total size which may be transcluded onto a page. That page is at that limit. Look at the page source, about 90% down:
Pre-expand include size: 2047955 bytes
Post-expand include size: 210070 bytes
Template argument size: 126505 bytes
Maximum: 2048000 bytes
I'm not sure if that means it is just under the limit, or if that means that the limit was applied, but either way it is probably trouble. It looks like {{Cite news}} isn't being used correctly. Perhaps fixing that will help? No, it is giving an HTML comment which says WARNING: template omitted, pre-expand include size too large, so it couldn't finish substituting them. —EncMstr22:10, 14 August 2007 (UTC)
Where do you find the oage source? Is there anyway round the problem? looking at that i don't think so, i will have to revert it then. Thanks Woodym55522:33, 14 August 2007 (UTC)
It's a function of your browser. With FireFox, Mozilla, Netscape and family, press Ctrl-U. Reverting might easily fix it, but it looks like the right thing is being done in the article, so it's something the developers should know about too. As wiki users, we're not supposed to care about efficiency or limits. —EncMstr22:42, 14 August 2007 (UTC)
I use a rather obscure Yahoo browser that doesn't have that function. I will try and alert the developers about this, see what they have to say. Thanks for your help Woodym55523:27, 14 August 2007 (UTC)
There is detailed information on this issue at Wikipedia:Template limits. If the page shows a size just under the limit, that usually means the limit was reached, because the software never allow going over the limit. — Carl (CBM · talk) 01:32, 15 August 2007 (UTC)
I have started removing unneeded datelinks and spaces and it is helping, but i fear it won't remove enough to make all the templates work. Thanks for the suggestion. Woodym55512:05, 15 August 2007 (UTC)
Retrieved from bizarreness
Latest comment: 17 years ago4 comments2 people in discussion
I've just noticed at the bottom of every page I view, a "Retrieved from [http address of the page I'm viewing]" message. Is this a fault, or a new (and apparently pointless) feature? DuncanHill16:43, 14 August 2007 (UTC)
No, that has always been there--even if you haven't been aware of it :). It's text that it is hidden through styles--when spiders, such as the googlebot, or others who mirror and replicate Wikipedia's content crawl the site, they typically only fetch the HTML of pages and not the full style sheets. As such, when you see a page in google's little stubs [12] or in its cache -- or on other mirror sites -- that line is present, so that visitors can both find the original page and to make it a tid bit more difficult to commit GFDL violations when mirroring content. If you check the source of the page you're viewing now, you'll see:
near the bottom. Note that "printfooter" is a class that's set to display:none in an imported style sheet. What probably happened is that your browser failed to render the styles completely -- fairly frequent problem, especially with Internet Explorer. HTH. AmiDaniel (talk) 06:02, 15 August 2007 (UTC)
Latest comment: 17 years ago5 comments4 people in discussion
I've noticed that some policies/guidelines address unnecessary use of system resources; WP:SIG on why images are not allowed in signatures; they are an unnecessary drain on server resources, and could cause server slowdown. WP:SUB cites a possible concern about server performance too.
Does the creation of a redirect page like Mantle (disambiguation) use resources or affect performance negatively? (Cumulatively of course, I assume any lag/drain would be proportional to the amount of pages.) Experience with other software tells me a tiny fraction is used in the form of disk space, like creating an empty folder, but creating millions of empty folders could start to affect performance on a very (very) minor level. Anynobody07:30, 14 August 2007 (UTC)
Well, images use more resources than transclusion. They both require extra database fetches but each image comes in as a whole new HTTP request. Redirects just take an extra fetch; not enough to worry about, although we recommend fixing double redirects when possible.
That could be true if we consider bandwidth to be our most valuable resource (it's surely the most expensive), but generally article size doesn't add any performance load to the servers. Watchlists and widely-used templates, however, are likely the most damaging to server performance, as they require many very expensive sql queries to be performed. As such, I'd say that probably the biggest "waste" of server resources is the use of unsubsted and unnecessary templates, such as you see with {{citation needed}} and userboxes, etc. Nonetheless, nothing the servers can't handle :) AmiDaniel (talk) 22:02, 15 August 2007 (UTC)
I'm sorry if I gave the impression I equated images or subst with (disambiguation) as far as performance effect but it sounds like you are confirming that they do indeed have the potential to cause a very VERY minor reduction in performance and not enough to mandate concern.
Is there a page to show server resources and what is using them? If not, how difficult would it be to create? Anynobody23:32, 14 August 2007 (UTC)
Latest comment: 17 years ago3 comments3 people in discussion
I'm in Firefox on a Windows machine using the default css (Classic, I guess), and the logo is looking rather messy - there's a lot of white space between the logo and the background, but I can't seem to figure out which MediaWiki page the main css is on - could someone more clueful please look into it? Thanks, Mak(talk)
Seriously... I've cleared my browser cache a number of times, it definitely looks worse that it used to... has someone replaced an image version? What's the deal? It looks like someone did a messy job of defining transparent areas or something. Mak(talk)
Actually, it does... Wikipedia displays PNG logo in IE6 using a CSS hack utilizing the DXImageTransform.Microsoft.AlphaImageLoader filter, enabling IE6 to render PNGs with proper transparency. I don't know for how long they've been doing this, (I wish they did it with all PNGs though) but I never had any problems with the logo. And the hack does not affect Firefox, so it should display PNGs like it always does. Do you have a screenshot? — Edokter • Talk • 22:56, 16 August 2007 (UTC)
The default skin should be Monobook unless you've chosen something different. The CSS for placing the logo is in [13]; look for the style "p-logo". But this doesn't sound like a PNG rendering problem as such, and it would be weird if Firefox had suddenly decided to forget how to properly render CSS. Perhaps you inadvertently changed your skin to Classic? Check out your preferences page, and set it to Monobook if it's anything else. TCC(talk)(contribs)23:45, 16 August 2007 (UTC)
Longstanding image/text space problem
Latest comment: 17 years ago5 comments3 people in discussion
There has been a serious issue in how text is formatted for some time. I assumed it would be fixed eventually but nothing has happened and it seems to be getting worse.
The problem is that if an in image is placed above some text and it runs into an image below it that was not placed directly below the first image tag in the markup, a void will be left in the text. In other words, if there is an image tag, then intervening text that will not necessarily fill the space until another image tag placed below the text, the text after the next image tag will appear in line with that next image. This behavior is wrong to me. Images should be formatted separately from the text and if there is not enough room to properly place an image it should just be stacked below the images above it.
This depends somewhat on display aspect ratio and resolution so it often goes unnoticed by editors with for example 1280x1024 displays. However on high resolution and widescreen displays it can become a severe formatting problem which leaves the article looking disjointed and very messy.
The problem behavior happens like this:
==First section==
[[image:soandso.jpg|right|thumb]]
Some text that may not format out to be as tall as that image.
==Second section==
[[image:another.jpg|right|thumb]]
This text will appear after an empty space (with any section heading left way above), at the position the another.jpg ends up.
Whereas this works properly:
==First section==
[[image:soandso.jpg|right|thumb]]
[[image:another.jpg|right|thumb]]
Text associated with soandso.jpg
==Second section==
Text associated with another.jpg. However it will not necessarily appear next to another.jpg when possible (except by coincidence).
It's clear to me that the proper behavior would be for an image to appear where placed in the text if possible, otherwise for the text to flow smoothly and images just to stack up one after another. Are there any justifications for the behavior that leaves the huge blank spaces? Are there any guidelines or style guides to point editors to to help them understand and avoid the problem? I'm spending a lot of time trying to fix this problem, especially in articles prone to picture-spam where the editors may be more concerned about getting their picture in than worrying about the formatting problems it causes. At any rate, working around the problem aside, this is only going to keep getting worse. To make the markup device-independent it should be formatted to never leave spaces in the text due to placement of an image. -- fourdee ᛇᚹᛟ 01:58, 16 August 2007 (UTC)
The problem is that this is an issue with CSS, not Wikipedia or the MediaWiki software. When you've got two things floated to the same side and they're hitting each other, the second is just going to get bumped around. You can force the top section to have its own break by using {{-}}, and you can follow the instructions at WP:BUNCH to avoid the "[ edit ]" links from bunching up when you use the second example. EVula// talk // ☯ //13:25, 16 August 2007 (UTC)
Well I'm not closely familiar with the details but isn't CSS just used by the skin and not really part of the core wiki software itself at all? I find it hard to believe there is not some way to make this work right. This is not my expertise but I am frustrated enough by this to look at the perl and css code and see what might be possible. If this is caused purely by CSS itself (for example by its lack of conditionals) and there is no way to work around it, I think that'd be good reason to dump CSS from the default skin. I'd rank the current behavior as "totally broken". -- fourdee ᛇᚹᛟ 15:08, 16 August 2007 (UTC)
I would consider the behavior you describe to be broken as well, but fortunately it is just an Internet Explorer bug and not an issue with the existing CSS. Your "proper behavior" is exactly what occurs in Firefox, and reportedly in Opera and Safari as well. As noted, a potential workaround is to use {{clear}} or {{-}} just before the second header, which will move the gap to before the header instead of between the header and the text. Anomie16:14, 16 August 2007 (UTC)
Ok cool it does work in firefox. However not working with IE seems to me to be a failure on the part of the skin since it's the most common browser. Is there no way to adjust the monobook skin's code whether the perl or css or html to work around this problem in IE? It seems likely to me that Microsoft will retain this behavior since webpages will have been designed to account for this. I'd like to see this fixed very much if there's some possible way. Perhaps I should take the discussion to mediawiki. -- fourdee ᛇᚹᛟ 16:37, 16 August 2007 (UTC)
Articles not appearing in Category
Latest comment: 17 years ago4 comments2 people in discussion
There appears to be a problem with articles not appearing in a Category. For example article Talk:Sandal Castle is in the Category Category:Yorkshire articles with comments which is inserted by template {{WikiProject Yorkshire}}. The category can be seen on the article itself but when you go to look at the category itself the article is not listed. I thought that there may be some sort of time delay but the comment file has existed since 23 July 2007 so it should have appeared in the category since then. Any ideas as to why this is occuring as it causes the comments files not to be picked up by the bot to create the assessment summary pages.
Fixed. It would appear that the category addition somehow never made its way into the job queue--likely a service that was down or having difficulties at the time the template was added. A simple null edit fixed it. If you find any more, let us know. AmiDaniel (talk) 22:07, 15 August 2007 (UTC)
Thinking about it, my guess is, the problem is that the comment files are created at some point after the change is made to the talk page. The template adds the Category when rendering the page but there is no process for putting the article in the Category list unless the talk page is actually changed. If this is the case then it will affect all projects that use the comment facility. Keith D10:25, 16 August 2007 (UTC)
Custom settings not working on public computer - Restore monobook?
Latest comment: 17 years ago3 comments2 people in discussion
I am accessing Wikipedia from a public computer and after I log in certain links like the ones to my userpage, watchlist, edit page, etc. are not being displayed. I have a custom monobook.js file so I guess it doesn't work in this computer. Other skins do work but I get lost with them. How can I restore the monobook file to the original, or get a copy of a standard one? Thanks in advance for the help. --ChaChaFut20:30, 17 August 2007 (UTC)
You are right, I confused the .js file with the css file. I blanked the .css file and that did it. Thank you very much for the quick response and accurate solution, Alex. --ChaChaFut20:49, 17 August 2007 (UTC)
The font of Wikipedia articles
Latest comment: 17 years ago2 comments2 people in discussion
Latest comment: 17 years ago6 comments3 people in discussion
Several weeks ago now, all of the known 'inline templates' (those little '[citation needed]' and '[original research?]' type things) were revised to use Template:Fix-inline for consistency. As a 'side-benefit' of this, these templates all are now set to class 'Inline-Template' by default or class 'Template-Fact' for two which were already using that. These classes can now be used to change the appearance of the templates, interact with them via bots or javascript, et cetera.
Likely of most immediate note, users who just want to read a page and don't need to see these 'editor comments' can add the following to User:<Your name>/monobook.css (or <Skin name>.css if not using Monobook) to suppress all such template notices;
For some that might be handy. I've heard that some people would prefer it if the many banners used for articles (wikify, cleanup, etc) could be "hidden" in the same way. - 52 Pickup14:16, 17 August 2007 (UTC)
Heh, see Template:Fix-box... still an early work in progress, but intended to consolidate and allow such options for banners in the same way that Template:Fix-inline now does for the inline templates. That's my next big template project. --CBD14:45, 17 August 2007 (UTC)
Should it not be (eventually) the other way round? The default could be that such editor notes are hidden, and editors can (and should) turn the notices on? This might, however, impact the process whereby readers are drawn to start editing. Looks like an "advanced readers" feature. Such advanced readers know that all content in Wikipedia should be taken with a pinch of salt, not just the tagged stuff. Carcharoth16:19, 17 August 2007 (UTC)
One place where I can see this coming in handy is when printing an article. If you want to make a hard copy of an article, you don't care if it hasn't been wikified, etc. - 52 Pickup19:33, 17 August 2007 (UTC)
Talk Page Archiving
Latest comment: 17 years ago2 comments2 people in discussion
I'd like my page to be archived by Werdnabot or one of the MiszaBots. Could someone come and do this for me, or do I need to do this my self? If you can do it for me, could you?
This is probably caused by the fact that Wikipedia now has built-in syntax highlighting which is automatically enable on user .js and .css pages. This probably makes the extension more or less obsolete, but if anyone does want to maintain, it should presumably be changed to either properly handle or avoid altogether code that has already been highlighted by the built-in highlighter. —Ilmari Karonen (talk) 20:59, 18 August 2007 (UTC)
Latest comment: 17 years ago4 comments2 people in discussion
How can you change the Talk: and the User: pages' titles to something else (in this case Collogue: and Uiser: respectively), I can't find it in Special:Allmessages at all. I am an administrator on the Wikipedia in question. sco:User:OchAyeTheNoo 14/08 10.47 pm
To change the names of the namespaces themselves, you need shell and mysql access. Consider filing a bug to request that a db admin change them for you. You'll need to show consensus from the members of the wiki's community for the change, however, or else it will not be done. AmiDaniel (talk) 21:41, 15 August 2007 (UTC)
However, the Wikipedia copy of Help:Section has a red link {{peisl}}, because Wikipedia doesn't have a Template:Peisl. I could create it, of course, but then it wouldn't automatically update when the one at Meta is changed. Is there a way to transclude a page from another wiki? - Fayenatic london(talk)18:57, 10 August 2007 (UTC)
You could just copy the code from the meta template and make a local template. Just link to the original in your edit summary when you do so to ensure that you're not violating the GFDL. EVula// talk // ☯ //19:48, 10 August 2007 (UTC)
...heh, so I guess I should read a bit more closely before responding, shouldn't I? :) Sorry 'bout that. EVula// talk // ☯ //06:21, 11 August 2007 (UTC)
At present time there is no support for transwiki includes. This may be something that will be approached in the future, but for now, I would recommend simply copying (or, better, importing--which at present is not enabled on enwiki) the template from meta. AmiDaniel (talk) 10:45, 11 August 2007 (UTC)
Latest comment: 17 years ago5 comments2 people in discussion
Resolved
It used to be that [[Image:I-95.svg|25x20px]] returned a 20px by 20px image, since that was the largest that fit into the specified rectangle. But now it gives 25px by 25px: Is this related to the above update? If so, and it's not a bug, what's the new format to specify height? --NE214:34, 25 August 2007 (UTC)
If you want a more distinctive signature, simply put the wikicode into your preferences and check "Raw signature". Templated signatures are a bad idea for the same reason that Until(1 ==2) mentioned. EVula// talk // ☯ //13:56, 16 August 2007 (UTC)
Thanks, I kind of thought about the template idea last night after I asked another user to personalize. I never thought about the technical stuff like overloading servers though. Sasha Callahan14:40, 16 August 2007 (UTC)
Also, there's an argument that "allowing users to have more unique signatures" is not such a good idea. (I realize it's a fuddy-duddy and creativity-stifling argument, but anyway.) You'll notice that people manage to create some pretty remarkably unique signatures even with the limited means available. But these are an arms race of sorts: in an ongoing attempt to outdo each other and become even more unique, users create increasingly distracting, wasteful-of-space signatures that don't really help the project much. Better to spend your time improving content than playing wikimarkup and HTML games with your signature. Better to distinguish your talk page posts by their wit and logic rather than by gaudy trappings. —Steve Summit (talk) 00:30, 18 August 2007 (UTC)
An extremely logical argument, and one that I happen to agree with 98%. However, mildly distinctive signatures are useful (which is why I've got the exact same signature across all my Wikimedia accounts); for example, there are several users whose signatures I recognize instantly (such as WJBscribe's) and will pretty much assume that their post is useful. However, that still reinforces your "make your posts stand out by logic rather than by signature" argument. :) EVula// talk // ☯ //16:27, 19 August 2007 (UTC)
Well, there's "unique enough to be instantly recognizable", and then there's "so unique that it takes up ten lines of esoteric Wiki code". No thinking person would object to the first; the disadvantages of the second are rarely apparent to those who use them. TCC(talk)(contribs)01:43, 20 August 2007 (UTC)
A problem with images
Latest comment: 17 years ago2 comments2 people in discussion
Latest comment: 17 years ago4 comments3 people in discussion
I am having a strange problem when I log in. When people log in; you see a header with you username, your talk, contributions etc. However, the technical problem I am having is that when I put my mouse arrow towards this header. For some reason it moves leftwards to the side of the screen, and it also does this to the "Log in/Create account" link when I go to log in. Such a problem has also happened on other wiki's with me, and I am wondering if such a problem is happening with other people, and if so it can be fixed. The sunder king13:53, 19 August 2007 (UTC)
What browser do you use? If this is happening across all wikis, it sounds like it could be an issue with your browser's rendering of CSS (which, if it's any variety of Internet Explorer, is a distinct possibility). EVula// talk // ☯ //16:20, 19 August 2007 (UTC)
Yes, cross-browser compatibility checking is a lovely way to spend a rainy afternoon. If you search the web for this behaviour, you will see a lot of threads and articles that talk about it and it is not specific to IE. It usually moves only a small(ish) degree and is often attributed to an anomalous handling of a padding declaration/line height/or similar somewhere. AdrianM. H.16:30, 19 August 2007 (UTC)
I use a yahoo browser, and sometimes internet explorer. However the incident has happened across various computers and networks. The sunder king16:38, 19 August 2007 (UTC)
Broken W3C compliance
Latest comment: 17 years ago2 comments2 people in discussion
I would just like to point out that Wikipedia currently does not validate against W3C
Latest comment: 17 years ago4 comments4 people in discussion
Someone put this formula into one of my contributions, converting acres to square kilometres: 400 acres (1.6 km2)
Where can I find other conversion formulas? Specifically, I want to be able to convert cubid metres to barrels (1 cubic metre = 6.29 barrels) (both ways) and cubic metres to cubic feet (1 cubic metre = 35.49 cubic feet).
(Strangely, the space between the last parenthesis and the equal sign seems to be necessary to get Google to recognize the search string as a unit conversion request.) Unfortunately, in keeping with Google's remarkably poor documentation (which stunningly contrasts with the general quality of Google's services), the Google calculator document does not list all the available units, nor does it describe the exact syntax necessary to make unit conversion requests work. Thus you may have to resort to some trial and error to fill the inexcusable gaps in Google's documentation. It's too bad Google doesn't put up a documentation wiki, so the millions of Google users can show Google how to write proper documents. --Teratornis14:35, 18 August 2007 (UTC)
I would like to suggest that this additional text would come after the current line:
Here sysops can make changes to the system messages specifically for the english Wikipedia. General changes that could benefit other wiki sites can be submitted to BetaWiki.
I see many changes coming in that are not in any way specific to the english Wikipedia, and these would be much more fitting for BetaWiki, that will later be easily transferable to the svg file. --Steinninn19:08, 12 August 2007 (UTC)
I've disabled this editprotected request. I'm not sure what policy is surrounding the addition of links to outside non-Wikimedia wikis. It would be better to have a discussion somewhere, possible WP:VP/T. Cheers. --MZMcBride02:18, 14 August 2007 (UTC)
End of copied discussion.
Is there any official policy about links to non-Wikimedia wikis? BetaWiki could deffenetlly be very help full for Wikimedia, at least it has been in the past. --Steinninn22:03, 17 August 2007 (UTC)
I don't see a single reason not to include external links into MediaWiki: namespace if they're relevant and useful. MaxSem20:14, 18 August 2007 (UTC)
I agree. Anyway, for the time being at least, BetaWiki most definitely is an important part of the current MediaWiki internationalization development workflow, and hiding it from those interested in participating in that process seem disingenious. In any case, it's worth noting that the page the proposed link would be displayed on, Special:Allmessages, isn't exactly high-visibility, being useful mostly only to those interested in customizing the MediaWiki system messages — i.e. sysops (here and on other Wikis), MediaWiki developers, and those interested in becoming either. —Ilmari Karonen (talk) 20:53, 18 August 2007 (UTC)
I agree completely with Ilmari, BetaWioki8 is an integral part of MediaWiki development. Furthermore, there tons of sites that we have interwiki prefixes for that aren't Wikimedia projects, see m:Interwiki map. Cbrown1023talk01:04, 19 August 2007 (UTC)
Thanks for putting it up. And yes, Allmessages isn't a high visibility place, but where is one? I have been translating system messages for about a year and I didn't know about Metawiki, where can we post a link to advertise the site? --Steinninn11:30, 19 August 2007 (UTC)
Anyone heard of this? Windows XP cannot ALT-TAB or switch tasks properly
Latest comment: 17 years ago3 comments2 people in discussion
I upgraded to Windows XP about six months ago and there's a feature called ALT-TAB.
For reference, I have "group similar taskbar buttons" unchecked because I dislike that function.
Before Windows XP, if I have a window open then I click in the start menu to another menu, then I hit ALT-TAB, it'll go to the previous window. But in Windows XP, ALT-TAB goes to a random window and not my last window. I basically have to click back and forth between then in the start menu 10 times before ALT-TAB between them works like in previous windows versions.
In additon, I may also have two notepad windows lined up as I'm using them both together--sy NotepadWindowA in the left corner and NotepadWindowB in the right corner--and well when I even click back and forth between them in the start menu (not the window, but the start menu), Windows XP refuses to keep both Notepad windows in the foreground and instead randomizes all window-on-top orders when I click between them in the start menu. I basically have to click back and forth between then in the start menu at least 10 times before clicking back and forth between them in the start menu works like in previous windows versions.
I've been trying to find information on google and it gives no information. Is there even a name for these Windows XP bugs/features so I can look for it? Juanita Hodges01:03, 22 August 2007 (UTC)
You should post this at the computing/IT section of the Ref Desk. This page is for technical queries specific to WP editing. AdrianM. H.01:20, 22 August 2007 (UTC)
Latest comment: 17 years ago2 comments2 people in discussion
Regarding the table in my sandbox, I need help in figuring out how to right align the table with article text to the left. If anyone is able to help out, feel free ot edit my sandbox and please insert any old blah, blah, blah text (as in an article) to the left of the table, and right align the table in my sandbox. Thanks, SandyGeorgia (Talk) 18:52, 21 August 2007 (UTC)
Thank you. —The preceding unsigned comment was added by Fijitodd (talk • contribs) 14:49, August 21, 2007 (UTC).
I don't know; maybe someone with a copy of a database dump will run the search and turn up here to give the results, but in the meantime you can use Special:Linksearch to check how many links there are to a particular website. --ais523 16:56, 21 August 2007 (UTC)
you get —~~~~ instead of --~~~~?
The keyword here was adding "overflow: hidden" to the divs. I also added an extra line of rounding, but it may seem a tad too much? Someone please check if it still works under Firefox. — Edokter • Talk • 16:49, 7 August 2007 (UTC)
![[2]](https://upload.wikimedia.org/wikipedia/commons/thumb/4/4a/FreedomHallStateFair.jpg/120px-FreedomHallStateFair.jpg){kind=link}
Resolved
Is this related to the above update? If so, and it's not a bug, what's the new format to specify height? --NE2 14:34, 25 August 2007 (UTC)