Misplaced Pages

This is an old revision of this page, as edited by Dragons flight (talk | contribs) at 02:09, 15 April 2013 (→Chopping parameters from templates: comment). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

• Table of contents
• First discussion
• End of page
• New post

• WP:VPT
• WP:VP/T
• WP:TECHPUMP
• WP:PUMPTECH

Newcomers to the technical village pump are encouraged to read these guidelines prior to posting here. Questions about MediaWiki in general should be posted at the MediaWiki support desk.

view · edit Frequently asked questions (see also: Misplaced Pages:Technical FAQ) Click "" next to each point to see more details. If something looks wrong, purge the server's cache, then bypass your browser's cache. This tends to solve most issues, including improper display of images, user-preferences not loading, and old versions of pages being shown. No, we will not use JavaScript to set focus on the search box. This would interfere with usability, accessibility, keyboard navigation and standard forms. See task 3864. There is an accesskey property on it (default to accesskey="f" in English). Logged-in users can enable the "Focus the cursor in the search bar on loading the Main Page" gadget in their preferences. No, we will not add a spell-checker, or spell-checking bot. You can use a web browser such as Firefox, which has a spell checker. If you have problems making your fancy signature work, check Help:How to fix your signature. If you changed to another skin and cannot change back, use this link. Alternatively, you can press Tab until the "Save" button is highlighted, and press Enter. Using Mozilla Firefox also seems to solve the problem. If an image thumbnail is not showing, try purging its image description page. If the image is from Wikimedia Commons, you might have to purge there too. If it doesn't work, try again before doing anything else. Some ad blockers, proxies, or firewalls block URLs containing /ad/ or ending in common executable suffixes. This can cause some images or articles to not appear. For server or network status, please see Wikimedia Status. If you cannot reach Misplaced Pages services, see Reporting a connectivity issue.

Village pumps

policy

tech

proposals

idea lab

WMF

misc

• Refining the administrator elections process
• Blocks for promotional activity outside of mainspace

• view
• edit
• history
• watch
• archive
• talk
• purge

Wikidata phase 2 is coming soon

Hi :)

2 days ago the first 11 Wikipedias got the ability to include data from Wikidata in their articles. These are the Italian, Hebrew, Hungarian, Russian, Turkish, Ukrainian, Uzbek, Croatian, Bosnian, Serbian and Serbo-Croatian Wikipedias. This means they are now able to make use of the structured data that is available on Wikidata. It includes things like conservation status for a species, ISBN for a book or the top level domain of a country.

The next step is to roll out phase 2 of Wikidata on English Misplaced Pages. We are currently planning this for Wednesday next week (April 3). We are currently carefully monitoring performance on the first 11 Wikipedias to make sure everything is ok. If problems arise we might have to postpone the deployment a bit longer but I wanted to give you a heads-up that this is the current plan.

We have prepared an FAQ for this deployment and are looking forward to your testing and feedback. You can already test it on test2.

And some more details about how this will work: There are two ways to access the data:

• Use a parser function like {{#property:p169}} in the wiki text of the article on Yahoo!. This will return “Marissa Mayer” as she is the chief executive officer of the company.
• For more complicated things you can use Lua. The documentation for this is here.

We are working on expanding the parser function so you can for example use {{#property:chief executive officer}} instead of {{#property:p169}}. The complete plan for this is here.

Thank you so much for helping us bring Wikidata to all Wikipedias!

Cheers Lydia Pintscher (WMDE) (talk) 14:45, 29 March 2013 (UTC)

Hey Lydia. I was going to email you and Denny about this, but I figure here is as good a place as any to ask, since some community members might be wondering the same thing. To be clear: this is a pretty broad question, not one that I think should hold up enabling Phase II.

When it comes to implementing use of Wikidata properties in infoboxes and the current state of the syntax, what is the actual, step-by-step workflow for updating content? I see description of the syntax linked from the FAQ, but not a clear description of the steps involved in editing something should we implement properties in infoboxes or directly in articles willy-nilly. I am of course looking at this through the eyes of my own editing as an admin on enwiki, and with my editor engagement hat on. This is my understanding: Previously, if an editor wanted to update an infobox, all they'd need to do is click Edit and change the value. For new editors, the hurdle was scary infobox wikitext, but in theory you could figure out the relationship between the markup, values, and what the article looked like. Now, when you click edit, you can't actually change the content in a wiki page by clicking edit here, you have to know to go to Wikidata and update the property.

That workflow is extremely problematic from a user experience design perspective, because it adds a number of potentially annoying and complicated steps to do what once a two-step "edit-save". There seems to be zero indicator of how to do so that is obvious too, which is bound to confuse even experienced editors until everybody across the wikis learns the new system.

How are we going to solve this problem? I don't think asking all editors to head over to a completely different wiki to update infobox values is okay at all, because it forces end users to conform to the needs of the Wikidata software architecture, rather than using software to help us keep the content up to date. The main solution I am thinking is that we allow local updating of Wikidata values which the software then pulls back in to the main revision history on wikidata.org. This would be best accomplished with VisualEditor of course, so it's probably not going to happen immediately. However, I'd really like us to consider how we can reduce user confusion and introducing multiple new steps in to the workflow of updating infobox content. Steven Walling (WMF) • talk 18:29, 29 March 2013 (UTC)

My take: Ideally we should work towards an inline invocation of Wikidata editing for properties. Til then, "edit" links in infoboxes that point straight to wikidata.org may be an acceptable solution. Property invocations in the article namespace should probably be avoided completely and perhaps made impossible. On the plus side, having parameter-less templates will make wikitext less cluttered.

Cf. an example of a parameter-less template invocation.--Eloquence* 18:50, 29 March 2013 (UTC)

Inline edit links to Wikidata (ala the interwiki links implementation) is definitely a good interim solution, since it is an obvious pointer to where the content is coming from. That is a good point about templates without huge parameters, and it's actually quite likely to reduce headaches for first time editors around seeing massive amounts of infobox markup at the head of articles. Steven Walling (WMF) • talk 18:54, 29 March 2013 (UTC)

Eloquence, in your example, for property with multiple value ("It's named after ... Linus Torvalds, Unix, Linux kernel"), they are not linkable (Linus Torvalds, Unix, Linux kernel). Has anyone raised this issue before? Bennylin (talk) 14:45, 1 April 2013 (UTC)

This will be done. I've filed it as bugzilla:46788 so we don't forget. --Lydia Pintscher (WMDE) (talk) 14:29, 2 April 2013 (UTC)

Thanks :) . Bennylin (talk) 17:13, 5 April 2013 (UTC)

(ec) One quick-and-dirty solution would be a discreet link in the top of the infoboxes similar to the (V D E) links which currently exist for a lot of navigational templates; rather that take you to the template page, however, it would send you to Wikidata.

That aside, remember that this depends on how the infobox is set up; there won't be any changes to enwiki infoboxes until someone actually codes the explicit use of wikidata into them. (It's not like the interwiki links, which effectively superseded existing content). As such, we can't really write guidance until we have some functioning examples to explain how it'll work - wikidata can't dictate that behaviour centrally. Andrew Gray (talk) 19:00, 29 March 2013 (UTC)

One thing for sure: integrating Wikidata to infoboxes will change the dynamic of revision diff. For example let say Yahoo's infobox is automated as of today, and a couple years from now (2020) it's CEO has changed, but looking back at today's revision, you won't find Marissa Mayer as it's past (2013) CEO, instead you'll find 2020's CEO in 2013 text.. Automated infoboxes, while it surely has its advantages, loses it's ability to be checked at exactly what diff the change occur (Wikiblame came to mind). And, as a consequence, automated infoboxes would rely on other website (Wikidata) to be vandal-free, or at least have the critical mass to revert problematic edits which could affect hundreds of projects.. Bennylin (talk) 20:19, 29 March 2013 (UTC)

We already have this problem to some degree with templates, of course! Andrew Gray (talk) 21:35, 29 March 2013 (UTC)

With template,s the diff could still be checked on the corresponding transcluded template's history, which is located on the same site (and listed on the bottom of the edit page). Having said that, it would be great to have list of property used in an article on the bottom of edit page. Bennylin (talk) 03:33, 30 March 2013 (UTC)

Isn't there the same issue with Commons as well? Alterations to the file or even deletions could easily happen without a corresponding change to the revision history locally. They just don't happen often because files are less prone to mutability than text. Steven Walling (WMF) • talk 22:50, 31 March 2013 (UTC)

Yes, and that is recognized as enough of a problem that the {{keep local}} template exists. Regards, Orange Suede Sofa (talk) 22:57, 31 March 2013 (UTC)

No. No matter you view the page locally or in Commons, they all have their file history on the image information screen. It's working great already. See if Wikidata can do the same thing, for example, for every "data"/"property" provided by Wikidata using parserfunction, one can click a different colored link (like, black?) which ideally go to a page complete with the revision history on the page itself, akin to image namespace. Bennylin (talk) 14:42, 1 April 2013 (UTC)

Update: We encountered some small issues that we'll have to sort out. Deployment will probably not happen on Wednesday as initially planned but hopefully soon after. I'll keep you posted. --Lydia Pintscher (WMDE) (talk) 20:16, 29 March 2013 (UTC)

I would hope that the wiki page history would reflect changes to both the wiki page and any data which it fetches. Perhaps at the same time, the similar issue with template changes not being reflected in wiki page histories could be handled. There is also the related issue of source citation, the elements of which probably need to be properties of the data (covering multiple data items), which could then be pulled into the wiki page cites. —— 23:13, 29 March 2013 (UTC)

I have a request for anyone involved in updating infobox templates: do not make the templates completely parameterless, but instead allow optional parameters to override data from Wikidata. It's going to take time for the data we have in infoboxes to be added to Wikidata, so we still need to use the data in wiki pages. As with the interwiki links, there will probably be some situations where we need to override data from Wikidata. Also, since Wikidata is not so easily edited as Misplaced Pages (both in terms of usability and technical requirements), editing the page on Misplaced Pages is the only way some editors can change infobox content. (Ideally, a bot could pick up such edits and transfer them to Wikidata. The infobox should still display local overrides while waiting for the bot to transfer them to Wikidata.) – PartTimeGnome (talk | contribs) 22:32, 30 March 2013 (UTC)

As just an editor: having stuff like {{#property:p169}} is as un-wiki as it can be - (wiki as in "everyone can edit", without much trouble) It should definitely be only a short temporary use, probably betrer not to be used at all (if if it short and temporary let'«s skip such a incredibly nasty looking thing). Something like {{#property:chief executive officer}} is still hardly easy to understand to the non-programmer mind (I presume, I have a programmer mind, I think it is easy :-), but the gains may outweight that. But please don't go to "The president of {{#property:p123}} is {{#property:p456}}, since {{#property:p789}}", not even "The president of {{#property:company-short-name}} is {{#property:president}}, since {{#property:term-begin}}". - Nabla (talk) 23:18, 31 March 2013 (UTC)

Indeed. In a lot of cases it is probably not a good idea to use it in the text of the article as opposed to the infobox. --Lydia Pintscher (WMDE) (talk) 12:00, 5 April 2013 (UTC)

Update: Heya folks :) It seems things are fine again now. We have scheduled a new deployment for Monday (April 8) evening to night UTC. I'll let you know if anything changes about this plan. --Lydia Pintscher (WMDE) (talk) 11:58, 5 April 2013 (UTC)

• Actually, I am quite strongly opposed to this implementation until such time as the English Misplaced Pages community decides that it wants this, and frankly I'm doubtful that we do. I cannot understand why we will be sourcing our articles to another site not under our control. I cannot understand why updating an article will require users to go to another website. And I cannot understand why a change made by a user who has never edited English Misplaced Pages will be automatically viewable on English Misplaced Pages. This project is already heavily conflicted over the use of infoboxes in the first place. This does not help. How about deploying on projects that actually want this software before it's activated on one that hasn't even had a discussion about it? Risker (talk) 01:30, 6 April 2013 (UTC)
  • The feature will be available for us to use. Whether we use it or not is up for us to decide. --Rschen7754 01:34, 6 April 2013 (UTC)

Anyone who's spent any amount of time on English Misplaced Pages knows that statement is untrue. As soon as it is enabled, it will be used, even over the objections of other editors, because there's no rule against it. We already have edit wars about infoboxes: both their existence and the content within it (and in some memorable cases, which infobox(es) should be present on an article and in what order). We know from experience that a handful of editors can, in only a few days, take a new extension and completely negate existing policies and practices (q.v. revision-deletion, which is constantly being used to delete content against policy), and in this case we don't even have existing policy. Today, if a new editor sees an infobox that says "Spouse:Marry Smith" instead of "Spouse:Mary Smith", they can click on the edit button and find the misspelling and probably fix it - if the wikitext doesn't scare them off. With Wikidata, they won't be able to find it, they'll just see "Spouse: #property:P169" instead. What's the point of making editing easier with Visual Editor (a very worthy endeavour that recognizes the challenges involved in editing), when we then make infoboxes impenetrable and unfixable unless someone goes to another site with an even more incomprehensible editing interface? How does this attract new users, or keep the old ones?

I'll repeat what I've said in another forum: I'm supportive of offering the extension to any project that wants it. I can see it being particularly useful for smaller wikis with a very tiny community, to help them build up content. But deploying this change means that the WMF is actively affecting content on every project where it is linked, and I don't believe that the good people behind Wikidata and the WMF haven't realised that: it's the entire purpose of the extension. The moment that this extension is turned on, it will be an irreversible fait accompli. Major change should never be initiated without advance preparation: community discussion, policies and processes developed, education available, resources established; that's Change Management 101, and it's being done well with Visual Editor and is practically non-existent on this project with Wikidata. I suppose this is what is to be expected from an organization that has decided to put "tech" before "information" or "community". Risker (talk) 05:24, 6 April 2013 (UTC)

Risker once again is speaking eminently sensibly—although in a community not dominated by the mindset of "invent a technological feature, implement it widely, then decide if it has any benefit", her perspective is also entirely obvious. Why she is, so far, the only person to come close to broaching the change management and big-picture issues here is a mystery to me (or is it?).

I also oppose the implementation of Wikidata's functionality on the English Misplaced Pages until such time as the community holds a discussion to determine what benefit, if any, Wikidata functionality provides, and what good use cases there are, if any. (And can that discussion even be held with available information? See second paragraph.) (I can tell you now that there are many bad use cases. I don't see anyone making that distinction, except Risker.) Wikidata is an interesting concept, but from the perspective of any single, highly active Misplaced Pages project, the cost–benefit analysis of retrieving data from it upon first impression is so obviously weighted to disadvantages that the default position should be "do not use until we figure out where it provides benefit". I am retired from this project, but this topic interests me very much—it has the potential to further reduce the editability of Misplaced Pages—and I may continue to comment. I hope that someone more active in the community will begin discussions about goals, use cases, and caveats, before adding a single Wikidata invocation to this project.

I'd also like to note that a) Wikidata still does not support basic property values like "date" and "number". It is, frankly, ridiculous to rush to implement "Phase 2" on the client side (the client being a particular Misplaced Pages project) when the basic infrastructure to store data is not fully in place. Everyone should slow way down here. (How much of a rush can you be in to replace "birthplace: London" on a biography with "birthplace: ?) b) Wikidata is still being populated (duh--and will be, for a long time), and provides no contract, not even a loose one, that any data set for any particular range of topics, or particular range of properties, exists on Wikidata. Ironically, since the majority of data on Wikidata have been copied from Wikipedias, you could draw some pretty Dilbert-esque modelling diagrams of the information flow at work here, that, to any independent observer, would seem circular and nonsensical. Riggr Mortis (talk) 23:27, 6 April 2013 (UTC)

Risker's view is wise. I too find Risker's view to be sensible and having much merit. The community should decide if this is wanted or not. It should not be deployed by fiat, or through mostly off-English Misplaced Pages decision. I haven't actually seen a single example yet of a page with an infobox filled out by Wikidata. Why aren't examples being linked? I couldn't even find any linked from Wikidata. I'm sure there there somewhere but why are they hard to find in the first place? (If somebody could find a dozen or so examples from the 11 languages using it, that'd be great.) I also find Steven's questions about usability very important. How is this anything but 100x harder for new editors to figure out? I'm an experienced editor but since I haven't seen any examples yet I'm still confused about it. Using Wikidata for the language links made HUGE sense. It's one of the greatest things to happen around here in a long time. Intertwining Wikidata with the articles however raises many questions. They should be discussed thoroughly. What I really hate to read are myopic comments "The feature will be available for us to use. Whether we use it or not is up for us to decide.", as made above. Risker himself pointed out the problems with simple take. We have an amazingly valuable resource here. We need to treat it with care and caution, and proceed slowly and only when the risks/rewards are well understood. Jason Quinn (talk) 03:45, 7 April 2013 (UTC)

But being realistic here, the infoboxes would still have to be recoded. We don't have a functional demo of an infobox ready yet that I know of, anywhere on WMF sites. So that takes time. Secondly, a lot of our infoboxes are full-protected, or at least they're supposed to be. Full-protected pages generally require consensus before being edited - admins who don't get consensus before throwing in the Wikidata fields are taking a bit of a risk. I really don't see the need to panic. I don't think a sitewide RFC is necessarily the answer; I think this sort of decision should be made by subject area. However, if folks want a RFC about this whole thing, rather than complaining about WMDE doing this, it's probably better to spend that time drafting a RFC. --Rschen7754 04:20, 7 April 2013 (UTC)

Yes by all means if anyone wants to start an RFC about how to make use of it please do. I am even happy to link to it in the announcement with a text like "The English Misplaced Pages community asks you to not make use of this in the main articles until RFC foo is closed and it is decided how to make use of this initial deployment of Wikidata phase 2." or something similar. --Lydia Pintscher (WMDE) (talk) 07:24, 7 April 2013 (UTC)

Maybe as some starting point for someone who wants to tackle this: The Hebrew Misplaced Pages already has phase 2 and they decided on some groundrules. They're not using the parser function in the article itself but only in templates. Template changes to move to Wikidata need to be discussed on a specific page. Controversial data isn't taken from Wikidata for now. This seems to be working quite well for all I can tell. --Lydia Pintscher (WMDE) (talk) 07:49, 7 April 2013 (UTC)

Hebrew Misplaced Pages, with its small community, had that discussion well before Phase 2 was deployed, and it was a conscious decision on the part of their community to do so. I have no objection to deploying it for projects that make such a decision. However, this is being done without any discussion with the English Misplaced Pages community (believe me, this page is not representative of the community at all), without any action plans, and based on what has been pointed out below on the workflow, at a stage of development where it is not useful and potentially very harmful to this project at a time when we are trying to retain editors and attract new ones. There are projects that are happy to help you to develop this software, and I encourage you to work with them until you have something that is ready to consider on this project. Risker (talk) 03:14, 8 April 2013 (UTC)

It could be good for the {{Authority control}} data. -- WOSlinker (talk) 14:14, 6 April 2013 (UTC)

I was thinking about this today, as it happens; it's a pretty good case of a template where we can "silently" call in a variety of cross-language data without needing to embed them in the wikitext, where they're not likely to be changed on a regular basis (unlike infoboxes, which are reasonably frequently edited), and where there's a clear benefit to a central db. In many ways, the focus on infoboxes is somewhat unhelpful for considering the use of Wikidata on Misplaced Pages; it's not where the major benefits to us are likely to come from. (Consider, for example, running a sortable "list of cities in X" and being able to dynamically populate the properties of each line). Andrew Gray (talk) 20:57, 8 April 2013 (UTC)

Could this be the reason why there has been a flood of new pages in the last few days, each listing a completely unnotable village in Turkey/Serbia/Poland etc.? Is Misplaced Pages to become a gazetteer? eg Nowe Staroźreby, eg Sığırlıhacı, Çubuk, eg Polu John of Cromer in Philippines (talk) mytime= Sun 08:09, wikitime= 00:09, 7 April 2013 (UTC)

I don't think so. --Lydia Pintscher (WMDE) (talk) 07:24, 7 April 2013 (UTC)

None of those three pages was created "in the last few days". Nowe Staroźreby was created 23 October 2008‎; Sığırlıhacı, Çubuk was created 13 July 2012‎; Polu was created 1 October 2012‎. I suggest that you contact each of the three different individuals as to their motives (and of the three, Dr. Blofeld (talk · contribs) is one of the most prolific creators of new articles that we have); but you can be sure that the presence (or not) of an entry on Wikidata was not the reason for creation. --Redrose64 (talk) 14:24, 7 April 2013 (UTC)

It was a manifestation of bugzilla:46978 I think, when whole swathes of similar articles suddenly popped up on the error list, even though unchanged. Sorry

John of Cromer in Philippines (talk) mytime= Tue 14:48, wikitime= 06:48, 9 April 2013 (UTC)

Risker's concerns are spurious, technically illiterate and nothing but spreading a cloud of FUD. At this thread, you can see multiple technical people trying to painstakingly explain this to her - that the deployment does not alter anything about the syntax, that no editor need use it unless they want to, and that nothing need change unless the editor base vote with their feet. This is all electioneering to build a base for December's arbcom elections. It is spreading FUD in bad faith, and needs to be called out for what it is - David Gerard (talk) 20:45, 8 April 2013 (UTC)

That's complete nonsense, but ABF asides are not worth pursuing here so I'll just reiterate that there are two important points from above (from Steven Walling and from Risker):

• "what is the actual, step-by-step workflow for updating content?" (and see section just below).
• "As soon as it is enabled, it will be used, even over the objections of other editors, because there's no rule against it."

Risker's concern expresses the situation precisely: as soon as some technical gizmo is available, it will be forced on everyone by the small but enthusiastic crowd of editors who aim to make Misplaced Pages readable by machines. Johnuniq (talk) 23:55, 8 April 2013 (UTC)

I obviously don't fully agree with Risker, but to accuse her of bad faith in this is quite inappropriate. --Rschen7754 07:14, 9 April 2013 (UTC)

• When it comes to deciding what to implement, I don think it makes sense to go separately by subject area. Rather we should go step by step with types or items of data, starting with the most obvious. I don't know what he wikidata team is ready for first, but one obvious place is dates in infoboxes, and then geodata in infoboxes. I would be reluctant to use personal name early on, because the conventions here are different for each wiki. What I really hope for, but I understand is going to be technically difficult, is references to sources.

I don't think it true or relevant that the enWP has little to gain. We have the greatest diversity of editors, who will be greatly helped by standardized data if it is done in the proper editable fashion. We have an immense number of articles on other language areas--I would suppose surely more in the aggregate than any other single WP. And, since our articles are often copied into other WPs, particularly the smaller ones, we have a responsibility for the data they use--we have to gain by being a up-to-date and trustworthy source for this. DGG ( talk ) 17:16, 10 April 2013 (UTC)

• I love what the idea of Wikidata and the results of its work so far. We should all be happy about the reductions in duplicated labour it has brought do far and promisses to bring in the future. For this reason I want further implementation of WD to be as seemless, positive and as user friendly as possible. In this regard I think Risker has a point in that including people in a consultation and changing WD for the better in any way which is practical is a good idea, one which I'm surprised the WD team didn't think of themselves. This should be done. I really don't see a down side to it, unless there's some sort of factor of urgency present which I'm unaware of. Having said that, I feel I have to object to the idea that it would be perfectly acceptable for the english wikipedia community to say: "This doesn't benefit us, so the rest of you can go to hell." A big part of why I contribute to the English version of Misplaced Pages raughter than to the Slovene one is because it's always been clear to me that work done on EnWP would, over time, difuse into the other versions. This is a big step in the right direction and it shouldn't get messed up for the sake of speed! --U5K0'sTalk_{Make WikiLove not WikiWar} 19:28, 10 April 2013 (UTC)

Workflow

I started a page at Misplaced Pages:Wikidata/Workflow. I personally consider some of the issues mentioned there to be blockers to deployment here. --MZMcBride (talk) 16:18, 6 April 2013 (UTC)

Thanks MZ. I think focusing on the practical problems with the direct experience (for editors) is the right way to look at this issue, instead of dickering about the way deployment might or might not have been communicated about to everyone's personal satisfaction. With a list like this, which could easily translate in to bugs to be addressed, the community can make a decision not based on generalities. Steven Walling (WMF) • talk 22:08, 8 April 2013 (UTC)

Navboxes and edit tools broken

Just how soon will all the navboxes on Misplaced Pages be fixed? While we're at it, when will be get the tools at the top of the pages back? -------User:DanTD (talk) 00:22, 31 March 2013 (UTC)

They're all working fine here... - The Bushranger _{One ping only} 00:31, 31 March 2013 (UTC)

Not really. The navboxes haven't been collapsible for the past several days, and there's no edit tools above my page. I used to be able to click on an icon for my signature, or bold or italic type, or images, etcetera, and that tool bar is gone. -------User:DanTD (talk) 15:18, 31 March 2013 (UTC)

I'll bet you inadvertently disabled javascript on your browser.קיפודנחש (aka kipod) (talk) 15:48, 31 March 2013 (UTC)

What kipod said. To check, click on this link; if it says "NO", then you've disabled JavaScript accidentally. jcgoble3 (talk) 18:26, 31 March 2013 (UTC)

Nope. It said "Yes." There were other complaints about this earlier this week, but I can't find them. -------User:DanTD (talk) 22:39, 31 March 2013 (UTC)

Sounds like you have something stuck in your browser cache. —TheDJ (talk • contribs) 08:43, 2 April 2013 (UTC)

Well, whatever it is, clearing it didn't work. -------User:DanTD (talk) 01:30, 5 April 2013 (UTC)

UPDATE - All it really did was get me signed off of other webpages that I'm on, including the commons. -------User:DanTD (talk) 13:54, 5 April 2013 (UTC)

Your description certainly sounds like JavaScript isn't working, but maybe it's only here. Does it work when you are logged out? Does http://commons.wikimedia.org work? How about https://en.wikipedia.org? Which browser do you have? Can you try another? PrimeHunter (talk) 14:02, 5 April 2013 (UTC)

(edit conflict)

At this point, I might be willing to bet that you have some kind of custom JavaScript that is causing an error OR you changed one of your Special:Preferences:

• On your "Editing" tab
  • Show edit toolbar (requires JavaScript)
  • Enable enhanced editing toolbar
  • Enable dialogs for inserting links, tables and more
• On your "Gadgets" tab (In the "Editing" section)
  • CharInsert, adds a toolbar under the edit window for quickly inserting wiki markup and special characters

⇑ Those are the next things I would check, as they may have accidentally got unchecked somewhere on the way. If those all look normal, I'll try to think of some more things! Technical 13 (talk) 14:35, 5 April 2013 (UTC)

My commons page works fine, and all things that should be checked, are checked. If any changes were made in my preferences, they were made by Misplaced Pages itself, not by me. The thing is, I don't have an edit toolbar, despite the fact that it's checked in my preferences. And what's worse, is that it's really not just the navboxes and edit tools. It's everything that's normally collapsible. BTW, I just checked my JavaScript, and it claimed it was okay. -------User:DanTD (talk) 03:24, 6 April 2013 (UTC)

If you want me to keep guessing then please answer all my questions. When I ask whether something works I mean things requiring JavaScript, for example collapsible table of contents. PrimeHunter (talk) 13:31, 6 April 2013 (UTC)

Some of your questions I just don't know the answer to. I have Internet Explorer, but I'm not sure if it's 8, 9, or 10, and the PC won't give me any way of finding out(unfortunately, this is one of the hassles with Windows 8). I don't log out on this PC, so I can't tell you if it works when I'm logged out or not. -------User:DanTD (talk) 07:02, 7 April 2013 (UTC)

If you're using Windows 8, then unless you've purposely downgraded, you're probably using IE 10, which is what comes with Windows 8. To check for sure: using the desktop version of IE, press the Alt key to bring up the menu bar, choose Help → About Internet Explorer, and it should tell you what version you're using. If you're using the Metro app, on the other hand, then that is IE 10. jcgoble3 (talk) 07:12, 7 April 2013 (UTC)

I had to struggle to get the info, but it's Windows 10. -------User:DanTD (talk) 13:25, 7 April 2013 (UTC)

I think you mean IE 10. Windows 10 doesn't exist (at least not yet, outside Redmond). LeadSongDog come howl! 20:33, 10 April 2013 (UTC)

It's hard to help without more answers. The first thing to check should often be whether a problem persists when you log out and get the default interface. If wikipedia.org has been marked an unsafe website in your browser then the browser may disable JavaScript here. If you have collapsible table of contents at other Wikimedia sites but not here when you are logged out then this is a possibility. PrimeHunter (talk) 10:47, 7 April 2013 (UTC)

I logged out briefly last night, and tried cleaning my cache again, then logged back in. Misplaced Pages.org has not been marked as an unsafe, and as far as I can tell my JavaScript is okay, but everything else still doesn't work. -------User:DanTD (talk) 13:25, 7 April 2013 (UTC)

Table right floating next to TOC and after image problem

Can someone please try to get the table of issue organizations ratings I just added to the end of the intro on Political positions of Joe Biden to come after the photo on the far right instead of in between the Table of Contents and the photo, squishing the TOC? I could not see how to do this in the Table Help page. Thanks in advance. EllenCT (talk) 01:59, 6 April 2013 (UTC)

Thanks very much to User:Wbm1058 for trying, but there was no obvious solution that worked very well, so I punted by moving the photo down to replace one from the 1980s. Everyone looking at that article probably knows what he looks like anyway, and he has an icon-sized photo in the navbox. EllenCT (talk) 03:18, 6 April 2013 (UTC)

I posted another solution for you. See Help:Table#Floating images in the center, which also discusses far right behavior. Using class="infobox wikitable" or class="wikitable infobox" seems to work, although that isn't documented at Help:Table. – Wbm1058 (talk) 15:51, 6 April 2013 (UTC)

That looks much, much nicer, too. EllenCT (talk) 19:39, 6 April 2013 (UTC)

Please don't use 'infobox' unless something actually IS an infobox. Classes like these have scope and meaning, they are not tools to apply functionality. In this case, you can use the class "floatright" which has similar behavior to infobox and a right side thumb, but without the special meaning and the special styling of those elements. —TheDJ (talk • contribs) 10:14, 8 April 2013 (UTC)

What is a class? Can you disambiguate it? Given that I don't understand what a "class" is, I have no idea what you mean by the "scope" and "meaning" of classes. I have a general idea of what a (wiki)table is supposed to look like. Lacking decent documentation, what else can editors do but trial & error to see what works? Wbm1058 (talk) 04:42, 9 April 2013 (UTC)

A CSS class is a reusable group of CSS styling instructions. We have a lot of common classes used for various purposes. To use them in wikicode, you use the HTML class attribute, like this: class="nameofclass". This applies the class "nameofclass" to the scope of that block of code. You can even have multiple classes apply to the same block class="nameofclass anotherclassname".

This is the simple part. The more complex part is that class="nameofclass anotherclassname" not ONLY applies the styling. In fact, it is more a 'naming scheme' to identify blocks in the page, that has automatic inheritance of styling information with identical names. So using class="infobox" does two things. It designates the block as being an infobox AND implicitly makes it receive the styling named 'infobox'. To designate something to be an infobox, just to get similar styling is therefore incorrect. Most of our classes have designated meanings like this, there are only a few exceptions; floatleft, floatright are some of those 'purely cosmetic' classes that would be an exception.

When there is no designated or cosmetic class that suits your need, you usually should create a template that does what you want, using style="" statements that contain individual CSS styling instructions to achieve the behavior you want. If the use case is common, usually the people who work on these kinds of things on the English Misplaced Pages will eventually turn a template with style statements into a new class. If you have trouble understanding such styling instructions, then just ask for help here, that is better than reusing a class for a different purpose than what is was intended for. —TheDJ (talk • contribs) 09:52, 9 April 2013 (UTC)

By CSS class, do you mean pseudo-class? Cascading Style Sheets doesn't define class, just pseudo-class. See this. Some have been uncomfortable with the title of Misplaced Pages:Catalogue of CSS classes (see Misplaced Pages talk:Catalogue of CSS classes#Suggested move). Given that Misplaced Pages:Catalogue of CSS classes#Classes has columns for both CSS and HTML, perhaps rename to Misplaced Pages:Catalogue of classes, which catalogues both CSS pseudo-classes and HTML class attributes? I've been working on updating related documentation, which still needs work.

Is Help:Table#Classes talking about HTML class attributes, CSS pseudo-classes, or both? I'm unclear on whether this edit was an improvement. Wbm1058 (talk) 23:04, 9 April 2013 (UTC)

Pseudo-classes are really only of interest when you want certain portions of a page (such as links) to change styling (including, but not confined to, colour) when you do interactive things like click on a link - they're only encountered when you're building a style sheet (there are some in MediaWiki:Common.css in the section marked "Highlight clicked reference in blue to help navigation" - it's the three instances of :target). In fact, don't worry about what pseudo-classes are, since they don't come up in wikicode - in this, there's a six-line table; ignore the last three lines and concentrate on the first three. Therefore, in the context of wikicode like {| class="wikitable infobox" the class attribute is class="wikitable infobox" and the classes (or class names) are wikitable and infobox. --Redrose64 (talk) 07:13, 10 April 2013 (UTC)

Man down?

I cannot access the toolserver. Anyone else experiencing this? Do the hamsters need replacing? Thanks.--ukexpat (talk) 03:01, 6 April 2013 (UTC)

I first noticed it was inaccessible about 6 or 8 hours ago. I don't know why. Dragons flight (talk) 03:44, 6 April 2013 (UTC)

Me too. --Makecat 05:36, 6 April 2013 (UTC)

Seems to be back up now.--ukexpat (talk) 14:24, 6 April 2013 (UTC)

Man down?! Methinks technology is overvalued in that description. - Ac44ck (talk) 14:51, 6 April 2013 (UTC)

Wondering if it was taken down for a bit to catch up... Before it went down it had a 28 hour overhead for me, now it is 4 hours... Technical 13 (talk) 15:04, 6 April 2013 (UTC)

Down again! -(t) Josve05a (c) 11:30, 7 April 2013 (UTC)

Hopefully it might have cleared the replag when it comes back up.--Gilderien Chat|List of good deeds 12:22, 7 April 2013 (UTC)

Is toolserver broken?

I noticed that at Special:Contributions you enter a username, then at the bottom there appears a box containing a number of links. One of those links, Global contributions, doesn't seem to work. It seems the whole toolserver is unreachable, although most services listed at http://status.toolserver.org/ seem to work properly. -- Toshio Yamaguchi 12:49, 7 April 2013 (UTC)

It's been down since 04:00, 6 April 2013 (UTC). See Template:Toolserver. The information in that template is updated hourly by BryanBot. If there is no update after an hour, it is likely that the entire Toolserver is down and there have been no update since 04:00, yesterday. According to the template doc, it is possible that while the user bot (BryanBot) is up, the database (MySQL) is down. When MySQL is down, bots and web services that require MySQL will not function. Also see the post above #Man down? --Ushau97 (talk) 12:57, 7 April 2013 (UTC)

Well, I just found http://meta.wikimedia.org/Future_of_Toolserver. -- Toshio Yamaguchi 13:05, 7 April 2013 (UTC)

Well, that's sad news it's getting discontinued. Next time there is a discussion on how to use the funds, I will suggest some should be spent on maintaining the servers used by Toolserver. What do you say? --Ushau97 (talk) 13:11, 7 April 2013 (UTC)

It also means Template:Coords and related ones don't work. Since most articles about places use those templates, it's a matter of some urgency. Jim.henderson (talk) 13:25, 7 April 2013 (UTC)

Its a known issue, the toolserver roots are working on it. Werieth (talk) 13:36, 7 April 2013 (UTC)

Quick status update

The toolserver appears to be back up to full function, thanks to Nosy's hard work. For the record, Toshio is correct that the toolserver is slated to be decommissioned, but it's an orderly process – it's not going to just be turned off at some fixed date.

In the meantime, thew new home for tools, the Tool Labs, is almost ready to house every tool. Those maintainers who wish to move early can do so, provided their tools do not rely on the database replicas (which will be deployed in a few weeks); web tools and typical bots can already be moved (and, indeed, some already have). — MPelletier (WMF) 13:41, 8 April 2013 (UTC)

I think that link was supposed to be mw:Wikimedia Labs/Tool Labs. --SoledadKabocha (talk) 20:51, 10 April 2013 (UTC)

• Right now, coords are not working, though. So I have my doubts about "full function". - Jmabel | Talk 06:35, 13 April 2013 (UTC)

  The toolserver has gone down again. See Toolserver is down again below. – PartTimeGnome (talk | contribs) 14:47, 13 April 2013 (UTC)

• It comes and goes, I think. I was just doing a contribution search on an article. It was slow, but brought up 100 results. Then I clicked on "show next 100", it brought up the "404 not found" issue. — Maile (talk)

New vector image won't update

I've tried the purging steps mentioned in Misplaced Pages:Purge and haven't had any luck in getting File:Duke Energy logo.svg's 200px version to purge itself. Subsequently the wrong image is being displayed in the article Duke Energy. Any help would be great. Shep 03:39, 6 April 2013 (UTC)

I applied a very temporary fix, but someone should look into the underlying issue. jcgoble3 (talk) 05:05, 6 April 2013 (UTC)

The 200px thumbnail version is the updated one for me (I'm accessing it from Europe). Is this still an issue for you, and from which continent? (On a general note, it's welcome to enter reproducible software or server issues in the bug tracker by following the instructions How to report a bug. This is to make developers of the software aware of the issue.) For your interest, a list of known software problems with thumbnails can be found here. --AKlapper (WMF) (talk) 10:09, 6 April 2013 (UTC)

I checked from another device just to be sure and it's still an issue for me, I'm in the USA. Shep 02:58, 7 April 2013 (UTC)

This problem should be fixed now, thanks to Leslie of the Operations team (might need a purge though). If it is not, please leave a comment in bugzilla:46976. We're sorry for the inconvenience caused. --AKlapper (WMF) (talk) 02:36, 10 April 2013 (UTC)

false negatives in category:Pages with missing references list

There is a very annoying problem here, with many pages being added to the list even though they have no cite error. There was a discussion here with the final suggestion that the conversation moves here. NB user:John of Reading is not the same person as me user:John of Cromer in Philippines

From what I have seen so far, about 80% of the entries on the list are false negatives, i.e. there is nothing wrong with them, but the only way to get them off the list is by null edit. There is no way of knowing other than by inspection whether a page is correctly on the list. Inspecting and nullediting take a couple of minutes; meanwhile it is quite feasible (and often happens) that the list grows.

Having an error in a template is quite rare, usually when an editor includes <ref>s there and relies on there being a {{reflist}} on the invoking page. as soon as the template is fixed, all the using pages immediately drop out of the list.

Question is how/why do pages get on the list, and how to get the false negatives off?

As an aside, I think it would be a good move to ensure editors have reviewed, and resolved any errors, before enabling the save button.

John of Cromer in Philippines (talk) mytime= Sun 07:34, wikitime= 23:34, 6 April 2013 (UTC)

How they get into the category is beyond my level of knowledge, but getting them out of the category is simple: a mass forcelinkupdate-purge via the API (equivalent to null edits). I ran the API over them (using the API sandbox), and the category is now down to just 51 pages. Does that look better? jcgoble3 (talk) 00:22, 7 April 2013 (UTC)

It may have been, I was just out buying more internet time. Now it's back up to 144 pages.

I wish I knew what your answer meant.

John of Cromer in Philippines (talk) mytime= Sun 10:17, wikitime= 02:17, 7 April 2013 (UTC)

What I meant is that normally a standard purge of an article does not update the link tables (which is what category pages are built from), requiring a null edit. However, purging via the API has two advantages: first, you can purge many articles with one command, and secondly, the API offers a parameter, not available through this web interface, called forcelinkupdate, which causes both the article and the associated link table to be updated—the equivalent of a null edit. At any rate, I ran the category through the API again, and it's back down to 53 pages.

If you want to run it yourself, go to Special:ApiSandbox, select "purge" from the "Action" dropdown, check the box next to "forcelinkupdate" (first row of the table), select "categorymembers" from the dropdown next to "generator", then enter "Category:Pages with missing references list" next to "gcmtitle" (first row of the "Generator" table). Finally, find the "gcmlimit" row and enter a number larger than the number of pages in the category. Then click "Make request" and go fix yourself a sandwich; purging many articles at once can take a few minutes. If you get a red error message, give it another try. Once done, go back and purge the category page directly to see how many pages are left. You can't break anything with the purge action.

And in the time it took me to type that second paragraph, the category is back up to 125 pages, so I'm running it through the API again. There's gotta be a bug in here somewhere... jcgoble3 (talk) 02:46, 7 April 2013 (UTC)

Bug ticket filed: bugzilla:46978 jcgoble3 (talk) 03:00, 7 April 2013 (UTC)

Could anybody provide a hint when this problem started, approximately? Thanks! --AKlapper (WMF) (talk) 09:55, 8 April 2013 (UTC)

It was "always" a minor irritant, but then the Thursday (my time) before Easter it suddenly escalated - overnight the entries on the list jumped from less than 20 to 300 or more. (usually there are perhaps 40 true bad 'uns). This actually took more than a week to clear, as each entry had to be inspected and nulledited at a minimum. Then it suddenly, last Friday I think, jumped back from clear to in excess of 200, which is when I started this particular ball rolling. I was monitoring yesterday: it is not a steady trickle but a whole batch at once. There seemed to be three waves. Fortunately I am now able to purge the set through the sandbox. I ran this once in the(my) morning, which removed more than 50, then again later in the afternoon which removed a similar number. Finally in the evening the list jumped from around 10 to more than 50 and I ran it again. I still have the log for that, but it doesn't tell you much, doesn't have a timestamp or anything. Hope this helps. John of Cromer in Philippines (talk) mytime= Tue 06:57, wikitime= 22:57, 8 April 2013 (UTC)

Comment I noticed a single instance of a phantom entry at Category:Pages containing citation needed template with deprecated parameters for the article Environmental terrorism. (See this discussion.) Near the beginning of AprilOn March 20th, I started cleaning that category and have made about 1000 edits (roughly 10050 per day) as part of that. During this time, I ended up making quite a few minor changes to some inline templates. The biggest was that the tooltips now show the a date in the format "(April 2013)" whereas before it was "from April 2013" (see this discussion at {{fix}}, change made on April 4th). I have also made some minor changes to the wording and punctuation of the "title" parameter of a bunch of inline templates like {{citation needed}}. I don't see how any of this could could responsible for problems with missing references but the timing and number of phantom entries is coincidence enough to make me mention it. Jason Quinn (talk) 20:44, 13 April 2013 (UTC)

Deleted user pages don't give a 404 error

If an account exists then a url in its userspace never gives a HTTP 404 Not Found in my tests at http://404checker.com/404-checker. Deleted pages and pages that never existed both give 200 OK. Is this a bug or intentional? I definitely think bug. The current behaviour can cause deleted user pages to remain in search engines. See for example the Google search "This page has been deleted. The deletion and move log for the page are provided below for reference" site:en.wikipedia.org ("About 139,000 results"). Examples of red links versus 404 checks: User:Shastri09 was deleted (404 check). User:Shastri09/test never existed (404 check). If an account doesn't exist then it gives 404, for example User:Shastri09test (404 check). Deleted and never-existed pages in other tested namespaces both give 404 as they should. Examples: Lunatic labs was deleted (404 check). Lunatic test never existed (404 check). I think deleted user pages once gave a 404 but I'm not sure. PrimeHunter (talk) 12:24, 8 April 2013 (UTC)

The list from Google... How long ago were those pages deleted? I believe Google has an archive of pages cached that expires after (iirc from the report I did on Google in one of my classes last summer) three months. So, any page deleted within that time frame would still show up there. The other stuffs I'm in no position to comment on. Technical 13 (talk) 13:02, 8 April 2013 (UTC)

If Google visits a page and gets a 404 error from the server then the page should be removed right away. The search terms show that Google has visited the pages after they were deleted, but they didn't give a 404 error. Deleted pages in many other namespaces like Lunatic labs and Category:Use mdy dates from May 2010 display "A page with this title has previously been deleted". The Google search "A page with this title has previously been deleted" site:en.wikipedia.org gives no deleted pages as far as I can tell (only existing pages quoting the message). That supports my experiments with the 404 checker: It is only in userspace that deleted pages don't give a 404. PrimeHunter (talk) 13:40, 8 April 2013 (UTC)

I did some research on Google that I hope can be helpful: FAQ→Remove a page or site from Google's search results and Meta tags→Using the robots meta tag. Technical 13 (talk) 14:00, 8 April 2013 (UTC)

http://lists.wikimedia.org/pipermail/wikitech-l/2013-February/066748.html Bawolff (talk) 14:33, 8 April 2013 (UTC)

Thanks for the link. It sounds like a poor decision to me, and why does it include user subpages? Google says in : "We recommend that you always return a 404 (Not found) or a 410 (Gone) response code in response to a request for a non-existing page." If we want to claim that non-existing pages exist then we should at least noindex them. We don't do that now for many of the deleted userspace pages (we do for some, I don't know the system, maybe it depends on when they were deleted). A deleted userpage with a company name may remain the first Google hit on the company. That sounds like an unreasonable "punishment" for creating a promotional userpage. PrimeHunter (talk) 15:01, 8 April 2013 (UTC)

But the page does exist (links to the user logs are valid for instance) It's an empty page, but existing. You name the exact problem. You want to 'punish' the user for doing something we don't approve of in the Google statistics. I guess adding NOINDEX to deleted user pages is a way to do that. That doesn't fix promotional username without a userpage though. —TheDJ (talk • contribs) 15:24, 8 April 2013 (UTC)

ttbomk, it really does not matter if the page exists or not - if the url structure is correct (<WP>/something in this case), WP will never return a 404. instead it returns a page that politely informs you that such an article/template/user page/whatever does not exist on this wiki, and will you be interested in creating one?. adding "noindex" magic word will not induce 404 either, but it will cause the returned page to come back with the "noindex" meta tag, which may induce google to drop this page from its DB. peace - קיפודנחש (aka kipod) (talk) 17:19, 8 April 2013 (UTC)

Kipod, Misplaced Pages does return 404 errors for non-user pages. For example, use a tool like Wireshark to monitor your network traffic, then go to a deleted page like this one. You should see a 404 status code returned to your browser. (If you don't get a 404, then some misbehaving proxy server is getting in the way.) Most web browsers only show the page content, not the status code. The status code can still be 404 even if the page content doesn't mention "404". – PartTimeGnome (talk | contribs) 22:12, 8 April 2013 (UTC)

you are correct - i wrote my previous response and then i did edit it to add "ttbomk" because i was not 100% sure. i did try to find the message status code before i wrote it, i just did not look in the right place... BTW: you do not need to download any external tool like wireshark if you use chrome, or FF with firebug - the status is available under F12-Network => Status. peace - קיפודנחש (aka kipod) (talk) 22:50, 8 April 2013 (UTC)

There are web services like http://web-sniffer.net/ that show web page headers. That service shows status "HTTP/1.0 404 Not Found" for a Misplaced Pages page that does not exist (and "HTTP/1.0 302 Moved Temporarily" for the red link above, which does show a page). Johnuniq (talk) 00:05, 9 April 2013 (UTC)

The issue I'm posting about is that the behaviour is different for non-existing user pages like http://en.wikipedia.org/User:Shastri09 (deleted) and http://en.wikipedia.org/User:Japloessl (never created). They both say "Status: HTTP/1.0 200 OK". There is no noindex tag on the pages so there is nothing stopping search engines from indexing them. Google is apparently indexing more than 100,000 of them. It seems the only way to remove the pages from Google with the current software is to create them and add __NOINDEX__ to them (or add a template which adds __NOINDEX__). Users cannot be expected to know that, and it seems illogical that you have to create a page to avoid it being indexed. And if you do create it then mirrors may copy it without including noindex, so it ends up being indexed elsewhere. PrimeHunter (talk) 01:10, 9 April 2013 (UTC)

But why is being indexed a problem ? These users exist. They should be indexed in my opinion... (And that was also the point of why the change was made). —TheDJ (talk • contribs) 08:06, 9 April 2013 (UTC)

Or do you want to add a user preference for it ? "Include my user pages in search engines" ? —TheDJ (talk • contribs) 08:09, 9 April 2013 (UTC)

As an aside, the original justification (for openid) I don't think applies anymore (given the concerns about user renames when using user pages as open id identifiers). Bawolff (talk) 11:41, 9 April 2013 (UTC)

There is an arguable case for the main user page of a registered user to return 200 OK (and be indexed) even if the page doesn't exist, since the page title also refers to the user who does exist. However, that shouldn't apply for all possible sub-pages of the user page (of which there are practically infinite possibilities). The title of a sub-page only represents a page, which doesn't exist, so should return 404 to my mind.

You might be interested to know that user pages of IP addresses also return 200 OK (and hence can be indexed by search engines), even if the address has never edited, and even if it is technically impossible for the IP address to ever edit. I really don't think User:0.0.0.0/sandbox should give a 200 OK response, but it does at present. – PartTimeGnome (talk | contribs) 20:34, 9 April 2013 (UTC)

"Added photo to page" edit summary

Lately on RC, I've noticed a lot of new editors with the edit summary of "Added photo to page" where they add in an off-topic photo or a broken photo link. For example, , , , or (topic NSFW, image off-topic). Obviously, this is a new "feature" from somewhere, but I'm not sure where. Where is it coming from and can we turn it off or at give some editor assistance on what is appropriate? -- Gogo Dodo (talk) 00:13, 9 April 2013 (UTC)

Wasn't mobile uploading just enabled? I dealt with something similar to this here and on Wiktionary (I deleted the image from Commons though, as it was just an illustration of Cyborg (comics)). EVula // talk // ☯ // 00:30, 9 April 2013 (UTC)

@EVula - yep, it's mobile uploading. —Theopolisme (talk) 01:17, 9 April 2013 (UTC)

Is there a good place to report problems? In a "providing feedback about a new feature" sort of way; all of the cited examples of mobile uploads have been crappy, but that doesn't mean all of them are bad. EVula // talk // ☯ // 05:37, 9 April 2013 (UTC)

In addition to submitting bug reports, you're welcome to come hang out/lurk/yell at us in our IRC channel, #wikimedia-mobile – that's the fastest way to get in touch. I read the Village Pump with an eye to mobile stuff, so dropping a note here works, too; I'm also happy to receive notes on my talk page :) Lastly, the mobile team has a mailing list, mobile-llists.wikimedia.org, but it's pretty quiet. Though maybe this is a good time to plug it, in order to solicit, erm, spirited conversation... Maryana (WMF) (talk) 20:54, 9 April 2013 (UTC)

So far, I haven't seen very many good ones. A lot of them have either been possible copyright violations with the wrong licensing or people snapping photos with their cellphones: William of Normandy's TV or random things around the house. -- Gogo Dodo (talk) 07:30, 9 April 2013 (UTC)

There is a good place to report problems: In general (not here as Maryana has commented on Commons that this is already being worked on), problems that sound like a potential issue in the code of the MediaWiki software or the server configuration should be send to the 'Bugzilla' bug tracker by following the instructions on How to report a bug. This is to make developers of the software aware of the issue. --AKlapper (WMF) (talk) 08:39, 9 April 2013 (UTC)

See also lot's of discussion about this on Commons. —TheDJ (talk • contribs) 07:51, 9 April 2013 (UTC)

Thanks for the discussion link. I'm glad to see that it is going to be "fixed" RSN. I'm really disappointed in how poorly this seems to have been planned and monitored. It seems to me that somebody had one of those "this is a great idea to get new editors!" moments without considering that a lot of new editors are not as smart as you think they are. -- Gogo Dodo (talk) 08:13, 9 April 2013 (UTC)

AGF run amok. :) EVula // talk // ☯ // 17:07, 9 April 2013 (UTC)

Trying to modernize our archaic multi-domain, images-on-a-separate-wiki uploading infrastructure; crazy, I know! :) Thanks so much for helping out with tagging/triaging the uploads over on Commons. I'm watching the uploader numbers now as we deploy. They should drop back down in an hour or so, once we're done. Maryana (WMF) (talk) 20:42, 9 April 2013 (UTC)

Gogo Dodo, this feature wasn't intended solely for new editors, and, in fact, some experienced Wikipedians are taking advantage of it quite happily. Put simply, it's just a nicer, more modern uploading workflow. I'd urge you to try it out if you have a camera phone and let me know what you think.

As for your comment about this being not properly planned or monitored – true, a lot more new users flooded in to try out the feature than the mobile team anticipated. I thought most people wouldn't care/understand why they should add an image, which would keep the numbers of mobile uploads low enough for the community to handle. After just a few days of full release, however, we saw that this was not the case; the usage was high and the quality was low, so the mobile team is responding to this today (right now, in fact) by making sure the feature only appears to existing Wikimedia users. This will significantly reduce the number of newbie mobile uploaders/crappy uploads while we work on improving the educational UI to ensure less noise. You could argue that a week is too slow to deploy a feature, gather data and feedback, and throttle usage/iterate on UI based on that data/feedback – but I'd say those are some very high standards for software experimentation you've got there :) Maryana (WMF) (talk) 20:31, 9 April 2013 (UTC)

How about restricting mobile image uploads to autoconfirmed users? --Redrose64 (talk) 20:51, 9 April 2013 (UTC)

Good call... on Commons, that appears to be registered for four days, which isn't a terribly high bar but should be enough to cool the heels of obvious vandals.

Let's see how the situation looks after today's deployment. If the number of low quality uploads is still too high, that could potentially be an alternative method of bringing it down. Maryana (WMF) (talk) 21:03, 9 April 2013 (UTC)

It may not have been intended solely for new users, but when the team decided to display the upload button to logged out users, you were bound to get a majority of new users clicking the button. From my own experience, if I'm not logged in (different computer, login timed out, etc.), the very first thing I do as an editor with an account is to login. I'm not going to poke around and try to edit or click a button. I suspect a lot of existing editors do the same thing.

But going back to the topic at hand, it seems a bit better, but I still see an occasional vanity photo. One other thing that I think needs to be addressed as could be considered a bug instead of a misused feature is the number of "undefined" additions such as . The form shouldn't allow an undefined file to be saved. -- Gogo Dodo (talk) 05:28, 10 April 2013 (UTC)

Hmm, that is very strange. I hadn't noticed this undefined bug; thanks for catching it. Let me know if you notice any more weirdness like that! Tracked in Phabricator
Task T49103

FYI, we're working on a pre-upload EXIF data verification system that will allow us to filter out the majority of the obvious copyvio, which should cut down on the number of newbies messing around and uploading random things they found elsewhere on the Internet. Maryana (WMF) (talk) 22:10, 10 April 2013 (UTC)

Two very unrelated questions

1. On File talk:Breakfast!.jpg, why does the tag for WikiProject Breakfast list the file as NA-class instead of File-class? AutomaticStrikeout (T • C • Sign AAPT) 00:49, 9 April 2013 (UTC)
   • It actually seems to have it listed as both. Technical 13 (talk) 00:55, 9 April 2013 (UTC)

   Well, it lists it as a file for WikiProject Food and Drink, but not for WikiProject Breakfast. AutomaticStrikeout (T • C • Sign AAPT) 01:06, 9 April 2013 (UTC)

   If it was I that was curious, I would ask them on WikiProject breakfast. :) Technical 13 (talk) 01:14, 9 April 2013 (UTC)

   It's because Template:WikiProject Breakfast has an incorrect value at "QUALITY_SCALE" and is accidentally using the default quality scale. See Template:WPBannerMeta/doc#Assessment. -- John of Reading (talk) 06:55, 9 April 2013 (UTC)

   John of Reading is correct. {{WikiProject Breakfast}} has |QUALITY_SCALE=Misplaced Pages:WikiProject Food and drink/Assessment which is invalid (see Template:WPBannerMeta#Assessment). For File-class to be recognised, it should be set to |QUALITY_SCALE=extended --Redrose64 (talk) 11:44, 9 April 2013 (UTC)

   Does this solve the problem? AutomaticStrikeout (T • C • Sign AAPT) 15:31, 9 April 2013 (UTC)

   Judging by this, yes. Please note that it may take some time before they all show up in Category:File-Class Breakfast articles, because of the job queue. --Redrose64 (talk) 17:45, 9 April 2013 (UTC)

2. As you can see at my signature history, the word "Automatic" is red (like a redlink) in two of the more recent regular versions. However, I am quite certain that it was actually blue, so why is it showing as red? It is also appearing this way on other pages. AutomaticStrikeout (T • C • Sign AAPT) 00:49, 9 April 2013 (UTC)
   • I was actually somewhat befuddled by that behavior myself when I was working on helping you rework your signature. I think it has something to do with the fact that one of the styles was invalid or syntactically incorrect so the whole style definition was being dropped, but I'm not entirely sure. Technical 13 (talk) 00:55, 9 April 2013 (UTC)

   Is there any way to fix that? AutomaticStrikeout (T • C • Sign AAPT) 01:06, 9 April 2013 (UTC)

   It shows as the color blue that I thought you wanted now, are you asking if there is anyway to go back and fix all of your previous signatures? If so, yes, but it would require bot access to go through all of the pages you've ever signed and replace the old signature with the new one without disrupting anything else. Technical 13 (talk) 01:12, 9 April 2013 (UTC)

   I'm talking about my old signatures, some of which contain red where blue should be. I'm trying to figure out why that is. It would appear that something got deleted. AutomaticStrikeout (T • C • Sign AAPT) 01:15, 9 April 2013 (UTC)

   You can write darkblue (one word) instead of dark blue. PrimeHunter (talk) 01:29, 9 April 2013 (UTC)

   Or just use hex like you do now since "#00008B" is shorter than "darkblue" anyways. :) Technical 13 (talk) 01:36, 9 April 2013 (UTC)

   I'd really prefer not to go back and change my signature a bunch of times. My opinion is that something must have been deleted in order to cause the blue in my signature to become red. Is there some way to correct this, perhaps by undeleting something? AutomaticStrikeout (T • C • Sign AAPT) 04:09, 9 April 2013 (UTC)

   I've been back through a bunch of your old posts and can find plenty of sigs that are blue and orange, also some that are orange and blue (mainly from Oct/Nov 2012), but cannot find signatures which are partly red. Please give links to examples of these. --Redrose64 (talk) 12:35, 9 April 2013 (UTC)

   • He's talking about the ones he linked here From March 11, 2013 to April 3, 2013. Technical 13 (talk) 12:56, 9 April 2013 (UTC)

   I've put my verdict at User talk:AutomaticStrikeout/signature history#Markup problems. --Redrose64 (talk) 14:13, 9 April 2013 (UTC)

   It's because <span style="text:#808080 0.2em 0.2em 0.2em"> isn't valid code in any way, shape, or form. I dunno why it suddenly doesn't show up correctly; I remember always seeing your signature as red, and it never occurred to me that it was displaying correctly at all. Maybe your browser updated, and so it fixed whatever bug that caused it to show up as you intended (when it shouldn't have)? EVula // talk // ☯ // 05:55, 9 April 2013 (UTC)

   ⇑ is exactly what I meant when I said "I think it has something to do with the fact that one of the styles was invalid or syntactically incorrect so the whole style definition was being dropped..." Technical 13 (talk) 12:03, 9 April 2013 (UTC)

   Yes, my browser was updated. I guess that explains it. Oh well, I had red in my signature the whole time without knowing it! Thanks everyone for your help. AutomaticStrikeout (T • C • Sign AAPT) 14:49, 9 April 2013 (UTC)

man

can you make a pattern kits with commons?

--2.6.124.215 (d) 8 avril 2013 à 18:43 (CEST) — Preceding unsigned comment added by 2.6.124.215 (talk) 12:16, 9 April 2013 (UTC)

when is the soocer/football pub?--2.6.124.215 (talk) 12:29, 9 April 2013 (UTC)

You can use {{football kit}} to make those. If you need help, try asking on the talk page of whichever team that is. Cheers, — Bility (talk) 16:05, 9 April 2013 (UTC)

SVG image thumbnail caching broken?

If I go to Bird anatomy#Respiratory system, I see on the right hand side a grey schematic of the bird lungs, File:BirdRespiration.svg (shown to the right). If I click on that thumbnail, I get a different, updated schematic. The update happened already in June 2012, but the cached thumbnail still isn't updated. Do other people see the same problem; is it a known bug; is there a workaround? Thanks, AxelBoldt (talk) 16:52, 9 April 2013 (UTC)

I'm seeing the same problem. I have never looked at that file before, so I know it isn't a local caching issue. Technical 13 (talk) 17:15, 9 April 2013 (UTC)

it's an image caching issue on the imageservers. if it does not go away, there is a workaround: show the image in a slightly different size, which will entice the imageservers to create a new copy (even though the image is .svg. the imageservers create an actual .png copy in the correct size which is what's actually gets sent to the readers. i believe the reason is poor support for svg on some browsers, most notably older versions of ie).

to demostrate, i added a second image of same file, one pixel wider (221 instead of 220). hopefully, when i hit "save", you'll see the new image under the old one. peace - קיפודנחש (aka kipod) (talk) 19:39, 9 April 2013 (UTC)

Kipod, I'm using the most recent version of Firefox, yet I still see the difference. Perhaps there needs to be a tag or category created on commons to tell the image server to update the .png it creates when a new file is uploaded? Who would be the contact person for that, if you know? Technical 13 (talk) 19:43, 9 April 2013 (UTC)

You seem to have misunderstood the explanation. Since some browsers cannot handle SVG files directly, everybody is served PNG thumbnails instead of the original SVG files, so the version of your browser has no bearing on the issue.—Emil J. 20:23, 9 April 2013 (UTC)

This is a bug. The image thumbnail seems not to be purgeable. A possible cause is permission problems on the thumbnail server, or an image being stuck in the squid cache. Usually forcing thumbnail generation with thumb.php and then triggering another purge should fix the latter though, and I can't seem to get that working. This requires filing a bugzilla ticket and then attention of a system administrator I think. —TheDJ (talk • contribs) 20:06, 9 April 2013 (UTC)

bugzilla:46976 may be relevant.—Emil J. 20:23, 9 April 2013 (UTC)

Could you please explicitly state whether you have tried to purge already? See https://en.wikipedia.org/Wikipedia:Purge . As I constantly get the same (wrong) result for the 220px thumbnail version on https://upload.wikimedia.org/wikipedia/commons/thumb/f/f2/BirdRespiration.svg/220px-BirdRespiration.svg.png this seems to be a different problem than https://bugzilla.wikimedia.org/show_bug.cgi?id=46976 which is about "sometimes an old version of the thumbnail is delivered, sometimes the new version of the thumbnail is delivered". --AKlapper (WMF) (talk) 17:09, 10 April 2013 (UTC)

This is a different issue than bugzilla:46976. I've reported it as bugzilla:47087. --AKlapper (WMF) (talk) 17:43, 10 April 2013 (UTC)

Scripts etc not loading

Is there a problem with user scripts not loading at the moment, was OK a few minutes ago. Also in edit mode the edit toolbar is missing and the selectable items below the edit box are not selectable but all are all displayed. Keith D (talk) 22:17, 9 April 2013 (UTC)

Someone has broken the system! Jared Preston (talk) 22:23, 9 April 2013 (UTC)

What browser versions are you using? Some old browsers have recently been put on a JavaScript blacklist (Template:Bug, gerrit:57693, gerrit:55446). I knew people were going to complain... PleaseStand (talk) 22:40, 9 April 2013 (UTC)

function isCompatible( ua ) {
	if ( ua === undefined ) {
		ua = navigator.userAgent;
	}
	// MediaWiki JS or jQuery is known to have issues with:
	return !(
		// Internet Explorer < 6
		( ua.indexOf( 'MSIE' ) !== -1 && parseFloat( ua.split( 'MSIE' ) ) < 6 ) ||
		// Firefox < 4
		( ua.indexOf( 'Firefox/' ) !== -1 && parseFloat( ua.split( 'Firefox/' ) ) < 4 ) ||
		// BlackBerry < 6
		ua.match( /BlackBerry*\/\./ ) ||
		// Open WebOS < 1.5
		ua.match( /webOS\/1\./ ) ||
		// Anything PlayStation based.
		ua.match( /PlayStation/i ) ||
		// Any Symbian based browsers
		ua.match( /SymbianOS|Series60/ ) ||
		// Any NetFront based browser
		ua.match( /NetFront/ ) ||
		// Opera Mini < 7
		ua.match( /Opera Mini\/\./ )
	);
}

Firefox 3.6.28, because I can't use a more-updated version on my computer. Jared Preston (talk) 22:45, 9 April 2013 (UTC)

Only FF < 2 would be blocked, and that block does not appear to be active. --  Gadget850 (Ed) 22:48, 9 April 2013 (UTC)

In the "extended content" above, it shows any Firefox browser "< 4", which would include Version 3. Since that includes me and I'm experiencing problems like the author of this thread, it would seem that there's a problem... Jared Preston (talk) 22:52, 9 April 2013 (UTC)

gerrit:57693 indeed says in the commit message that Firefox < 4 was blacklisted, as does the actual code change. jcgoble3 (talk) 23:00, 9 April 2013 (UTC)

I am using Firefox version 3.6.28. Keith D (talk) 23:03, 9 April 2013 (UTC)

So can a fix be found for people who don't or can't use the newest, flashiest browsers? Jared Preston (talk) 23:06, 9 April 2013 (UTC)

If you are meant to be blocking only FF < 2 then the code is in error and should be changed. Keith D (talk) 23:10, 9 April 2013 (UTC)

There is a fix: Stop using an end-of-lifed browser that's no longer supported and get a modern one. And if your end-of-lifed operating system prevents you from using a newer browser, then buy an upgrade to a newer OS or even a whole new computer if that's what it takes. Developers cannot, repeat cannot, be expected to account for every browser and operating system that ever existed. Instead, they can, should, and do code for what's still supported. At some point the technology has to move on, and if you insist on using outdated browsers and/or OSes, you can and should expect things to break. jcgoble3 (talk) 00:44, 10 April 2013 (UTC)

It's harsh, but it's true; developers can only support outdated software for so long before it becomes counter-productive. Improvements and features become harder and harder to implement over time the longer legacy software is supported. Firefox 3.6 has apparently been out for more than three years, and had its support officially dropped almost a year ago. I realize "get a new browser/computer" isn't the answer that you'd prefer, but I don't think it's fair to expect the devs to continue supporting a browser that has been dropped by its own parent company. EVula // talk // ☯ // 02:36, 10 April 2013 (UTC)

Why is it apparently forcing us ffox3.6 folks to be completely turned off rather than just not supported? Everything was working fine yesterday, did every javascript bit suddenly get updated to do things my browser can't? "You're on your own, things might start to behave badly because we're adding new-browser-only stuff" is a lot more friendly than "screw off, don't even try, no bits for you." DMacks (talk) 03:53, 10 April 2013 (UTC)

It's not always that simple. I remember a while ago there was a problem with Misplaced Pages tending to lock up for readers due to Javascript not running properly on an old version of IE. If something impacts readers badly, then often the right response is to turn it off rather than telling visitors to just live with it. Anyway, the person who blocked FF less than 4 was Timo Tijhof ttijhofwikimedia.org, you could ask him why. Dragons flight (talk) 04:38, 10 April 2013 (UTC)

The article "History of Firefox" quite rightly states that support for Firefox 3.6 expired not even 365 days ago. I don't know much about IE, but there are also a number of wp skins which are ancient and still supported, or at least still offered. At 22:00 UTC everything was still fine, so now to be "blacklisted" all of a sudden makes me feel like a criminal. I thought this was supposed to be "Misplaced Pages, the free encyclopedia that anyone can edit" – telling someone to update their OS or buy a new computer isn't very helpful. I'm really sorry, but despite having a job I simply cannot afford a new computer. And I can't for the life of me work out how one MediaWiki user can decide which browsers are to be supported, and which should not be. Well done for pissing someone off that has helped out here for almost eight years! Jared Preston (talk) 05:58, 10 April 2013 (UTC)

For what it is worth, I doubt very much that it was "one Mediawiki user". The change is traceable to a particular developer, but in all likelihood there was a significant discussion about it before hand as well as a specific reason for the change (i.e. some specific problem with the old browsers). However, those discussions likely happened within the WMF so we are unlikely to know about them unless we ask. The quickest way to get to the underlying reasons for the change would probably be to ask the developer who made it. Dragons flight (talk) 06:31, 10 April 2013 (UTC)

Firstly, let me address the "Misplaced Pages, the 💕 that anyone can edit" argument. In all honesty, this is getting old and inappropriate. The encyclopedia is no more or less free or editable by the presence of the extra javascript features. Obviously things are prettier and easier to edit with the edit toolbar and gadgets than without, but it is by no means an access restriction for even a blocked user is still able to read the encyclopedia pages.

Now, as for Firefox 3.x and below, jQuery has not supported Firefox 3.6 for quite a few releases. They currently only actively support Firefox 18 and above (though consider that number with the knowledge that Firefox now has auto-updating built-in, like Google Chrome, which means that only users of before that version are stuck. There are practically 0 users of any Firefox version between 15 and 18). Also note that this was not Wikimedia's decision to make, it was jQuery's and since we depend on it (unless we start to fork and patch it) we can't control this. jQuery could break in Firefox 3 at any time as nobody is actively looking after it during jQuery development as far as I know (see jquery/testswarm for example).

Having said that, there is a difference between a browser not actively supported (e.g. tested for before deployment and reported bugs are prioritised) and browsers that are known to have fatal errors in script execution - in which case a blacklist in ResourceLoader is appropriate.

When updating our blacklist I jumped the gun in this and should've excluded Firefox below 3 not up until 3.x (< 4). Firefox 2 is known to be incompatible with our latest scripts, Firefox 3.0 - 3.5 is a bit of a wild card. Firefox 3.6 however is an LTS (Long Term Support) release by Mozilla that (last I checked) is compatible enough to not be a problem for most users most of the time. So I will lower our blacklist for now to tolerate Firefox 3.6.

However LTS is not ELS (Ever Lasting Support), Firefox has maintained 3.6 for many years, but that stopped in April 2012. I can't stress enough the importance of upgrading to a browser that the creator of that browser supports (in this case, a version of Firefox that Mozilla supports). When Mozilla finds out about security issues and releases updates, they do not create and release patches for users of old versions such as Firefox 3.6. As such you make yourself vulnerable to security leaks when browsing the web (not Misplaced Pages in particular).

Anyhow, as said, I've lowered ResourceLoader startup check from Firefox 4 to Firefox 3 (gerrit:58465). Krinkle (talk) 07:53, 10 April 2013 (UTC)

Well I knew I didn't have a problem using Misplaced Pages until late last night. And of course, it's not just Misplaced Pages, but without javascript support, you're pretty much excluded from editing many features of Wikidata too. I appreciate that using an outdated browser leaves yourself open to security issues, and fortunately I haven't had any since April 2012 (having an anti-virus program helps a bit, I guess) and, like, I said, I just can't update from here. It isn't great being stuck in a black hole, and didn't think it made such a big difference to Misplaced Pages which browser I use to edit in... Thank you, Timo, for lowering the bar for me, and other users experiencing the same problem. I don't know whether or not you got my e-mail from an hour or so ago. I don't know what "jQuery" is (maybe I should?) but having javascript open again makes a whole lot of editing here a lot, lot easier. So, as until a time that javascript breaks uncontrollably, thank you very, very much! Jared Preston (talk) 08:14, 10 April 2013 (UTC)

Thank you for explaining the situation! In the future, these sorts of user-impactful changes might be worthy of a sitenotice, or technical note in Signpost. DMacks (talk) 11:52, 10 April 2013 (UTC)

English Misplaced Pages servers going down around 02:30 UTC every day?

For the last few days, I have noticed around 02:30 UTC, the English Misplaced Pages completely shuts down, and I receive a message that states "This wiki is having technical problems." However, the message disappears after reloading the site about a minute afterwards. I have been using this Misplaced Pages regularly for the last few months, and have never ran across this message regularly (or at all) until last week. Just wondering ... what is the cause of this, and is any action being done to resolve this intermittent issue? Steel1943 (talk) 02:40, 10 April 2013 (UTC)

It's not the normal site error notice either. Ryan Vesey 02:55, 10 April 2013 (UTC)

I think this is Template:Bug. If you look at the Server Admin Log, you will see that LocalisationUpdate runs at about that time every day. Database server overload could certainly result in such error messages as database connection errors ("Sorry! This site is experiencing technical difficulties. Try waiting a few minutes and reloading." as opposed to the text shown at WP:WFEM). PleaseStand (talk) 03:42, 10 April 2013 (UTC)

I saw the same message last night from my mobile device on the desktop site. It fixed itself in less than 2 minutes though. Technical 13 (talk) 17:14, 10 April 2013 (UTC)

I've been seeing that error in the last week or so too. I didn't take notice if it only occurs at a specific time or not. What I do know if that I basically would never see this problem unless Misplaced Pages had a major outage so it's something relatively new. Jason Quinn (talk) 21:28, 10 April 2013 (UTC)

Sudden drop in pageviews

The article List of best-selling books is (or was) one of our most-viewed articles, with about 5000 views per day. E.g. June 2011, January 2012, September 2012. Suddenly, late in March 2013, it dropped to about 1,000 a day, or less than 1/4th of the previous number. The article had no huge changes at that time, and other articles don't seem to show the same drop. Anyone has any clue as to what might have caused this? No idea if it is a technical problem or not, posting it here seemed the most appropriate. Fram (talk) 08:35, 10 April 2013 (UTC)

There is not any clue in Google searh either! --Tito Dutta (contact) 08:42, 10 April 2013 (UTC)

Indeed, I can't find a single reason for this drop (it seems unlikely that this was consistently wrong for months and months, and now suddenly has been corrected, but it is a possibility of course). Fram (talk) 09:52, 10 April 2013 (UTC)

I wonder if it's either (a) something to do with Wikidata, or (b) that something else was redirecting to the page and no longer is? Black Kite (talk) 10:20, 10 April 2013 (UTC)

Spring fever? Everyone deciding to go outside to enjoy the beautiful weather inside of looking for a book to read? I wouldn't be concerned with a change in such a stat for one month. If it persists, then I would suspect something is going on and hunt for it. Template:2c from a broke bloke... Technical 13 (talk) 10:42, 10 April 2013 (UTC)

Perhaps someone deleted a link from a highly visited page, like Book. --NaBUru38 (talk) 16:03, 11 April 2013 (UTC)

It's a possibility, so far I haven't found the obvious culprit though. Fram (talk) 09:44, 13 April 2013 (UTC)

The same thing happened, at about the same time, to the Parabola article, with the daily number of reported views dropping from about 2500 to about 600 on weekdays (both fewer on weekends). Other related articles such as Hyperbola and Catenary have not been affected. I think it's highly unlikely that this drop is real, especially since the same has happened elsewhere. There's a bug somewhere. DOwenWilliams (talk) 20:15, 14 April 2013 (UTC)

Show

There appears to be a glitch at this category where several entries such as James Wilson are missing. Can someone fix that please. Pass a Method talk 13:09, 10 April 2013 (UTC)

• I can see Wilson's name there! --Tito Dutta (contact) 14:57, 10 April 2013 (UTC)

Forward slash

Hello. I have a problem with the symbol /. For example Talk:C/2006 P1. If i create a subpage it would be Talk:C/2006 P1/Archive. But if i used it in a template or someone else, its regonzise that this is a subpage of Talk:C. What can i do? Xaris333 (talk) 14:47, 10 April 2013 (UTC)

/Archive would be a sub-page of /2006 P1 which is in itself a sub-page of Talk:C. What template are you trying to use it in? Technical 13 (talk) 15:05, 10 April 2013 (UTC)

C/2006 P1 and C are different articles. Xaris333 (talk) 16:09, 10 April 2013 (UTC)

Yes, but whereas the software treats a slash in an article name as yet another character, it treats it as a subpage separator everywhere else. Thus Talk:C/2006 P1 is indeed a subpage of Talk:C, despite that humans think of it as the talk page of C/2006 P1. Notice the little “< Talk:C” below the heading of Talk:C/2006 P1, which reminds to this fact.—Emil J. 16:24, 10 April 2013 (UTC)

I think archive templates may have problems. {{Talk archive}} includes the code {{#titleparts:{{FULLPAGENAME}}|1}} if its used on Talk:C/2006 P1/Archive the result would be Talk:C rather than Talk:C/2006 P1. I probably suggest do archive links by hand, although it might be posible to fix the template with {{#titleparts:{{FULLPAGENAME}}|-1}}, but this might break some other pages. See parser function doc--Salix (talk): 16:43, 10 April 2013 (UTC)

{{#titleparts:{{FULLPAGENAME}}|-1}} would most definitely fix the problem... To avoid possible issues with other pages that have been using this for any duration of time, you might go {{#titleparts:{{FULLPAGENAME}}|{{#ifexpr:{{REVISIONDATE}}<{{subst:REVISIONDATE}}|1|-1}}}} and change the documentation to reflect the new way it is working. Alternatively, you could have someone with AWB access pull up a list of the pages that transclude the template and fix the usage for them. This would probably be preferred but less likely to happen. Technical 13 (talk) 17:08, 10 April 2013 (UTC)

The page C/2006 P1 is not a subpage of C. But the page Talk:C/2006 P1 is a subpage of Talk:C. This is wrong. How can i change that. Why C/2006 P1 is not a subpage and Talk:C/2006 P1 is; Where is the different? Xaris333 (talk) 16:51, 10 April 2013 (UTC)

The difference is that C/2006 P1 is in the main namespace, and Talk:C/2006 P1 is in the corresponding Talk namespace. You cannot change that, this is how the site is set up. See WP:SUBPAGE#Articles do not have sub-pages (main namespace) for more details.—Emil J. 17:03, 10 April 2013 (UTC)

in reality, in mw system "subpage" is pure fiction. it's a convention we use for convenience, and some features of the system pretend it's real, e.g. all the "titleparts" tools, and some protections/rights (e.g., a page in "User:" namespace whose name ends with ".js" can only be modified by users with "editinterface" rights, and the user that this page looks to be a subpage of her user page). however, as i said, this is pure fiction. you can create a page called No such page/No such subpage/This doesn't exist also/Faux subpage, without having to create any of the pages that nominally "contain" it, and all will be well (from the system's page storage POV, the forward slash is merely one more legal character in a page name. we actually have an article called ], which, if you think of forward slash as "subpage separator" really should mean "a page with empty name which is a subpage of a page with empty name which is a subpage of "the" page with empty name". of course, this is not so - it's just a page with the name "//" (becaus it confuses the system, in order to link to it you'll need to write ], which may or may not be a small legal program in Brainfuck). peace - קיפודנחש (aka kipod) (talk) 19:11, 10 April 2013 (UTC)

I do agree that subpages should be enabled on BOTH an article namespaces and its associated talk namespace, or neither of them. Given that the main namespace is not configured to support subpages, the main Talk namespace should not be configured to support it too (because for some articles in the main namespace it's impossible to link to their associated Talk page, as it will reach a Talk page associated to another article !)

In fact the situation would be safer if it was the reverse : article namespaces with support of subpages, but Talk pages without, allowing to group discussions related to a page and all their subpages (this would not prohibit the creation of archived Talk "subpages", but navigating between these false subpages and the main Talk page would be complicated a bit, and would require the cration and use of navigation templates).

The consequence of adding support for subpages is the possibility of creating relative links to a subpage using "/" as the first character in the wiki link, or the initial characters "../" to create a link based from the parent page.

Enabling subpages adds new restrictions on valid titles of pages, where they may contain slashes (such as "../" or "//" within any namespaces, which will be interpreted hierarchically instead of litterally): you cannot change this setting before testing the content of the complete set of pages in that namespace to see if they start with a "/" or with "../" or if they contain "//" or "/../", because these pages will no longer be reachable.

So these rare pages will have to be renamed, and the redirection link deleted, and all pages linking to them will need to be edited to not use the redirection link but the new name in the target parameter of links.

Note that some pages are creating links using URLs instead of wikilinks (sometimes via template helpers), and they are more difficult to detect because these sources pages using these URLs are not completely indexed ; and this won't be detected if source pages are from other Wikimedia wikis, or from other external sites). To solve this last issue, we would sometimes have to insert in the top of some pages a disambiguation notice, providing the new name of the page whose old name would now conflict with the title of another unrelated article (lots of pages may be affected if ever the database, before the change of setting, contained some article titles starting by "/" or "../" or containing "//" or "/../").

The "titleparts" parserfunction should also be able to test the namespace given to see if that namespace is configured to support subpages or not. If not, it should not split the given title into multiple parts. Unfortunately, this builtin function is also used in various templates to split random parameters into multiple parts easily, allowing to pass a list of items in a single template parameter: if namespaces are tested by the builtin "titleparts" parserfunction, these templates will stop working. The only solution would be to add an additional (optional) parameter to the "titleparts" parserfunction specifying that namespaces present in its 1st parameter should be tested.

For now, the best you can do is to write an utility template that will perform the namespace test itself before using "titleparts"... But this will still return incoherent results between the main articles namespace and the main Talk namespace, the way they are currently configured on English Misplaced Pages ! verdy_p (talk) 08:40, 11 April 2013 (UTC)

For the most part subpages are very useful for the Talk: namespace. A large number of articles have archives for talk pages (94464 transclude {{Talk archive}}) and treating these as subpages makes a lot of sense. Yes there are a few problems when the corresponding article has a / in it but this only a tiny portion of articles and there are workarounds. --Salix (talk): 10:11, 11 April 2013 (UTC)

The problem i had is with a template in Greek Misplaced Pages. el:Πρότυπο:Κριτική λημμάτων. I had to change the parameter ] than can recognize the talk page of an article has name like el:C/2006 P1. Can anyone help? Xaris333 (talk) 18:23, 13 April 2013 (UTC)

archivebot.py

Hello, somebody uses this script? At me it is impossible to archive pages with its help.

   C:\Python26\pywikipedia>archivebot.py Archive-bot/archive
   Fetching template transclusions...
   Getting references to Template:Archive-bot/archive via API...
   Processing ru:Обсуждение участника:Парис "Анима" надаль
   4 Threads found on ru:Обсуждение участника:Парис "Анима" надаль
   Looking for: Template:Archive-bot/archive in [[:ru:Обсуждение участника:Парис "Анима"
   надаль]]
   Processing 4 threads
   There are only 0 Threads. Skipping
   Processing ru:Участник:Парис "Анима" надаль/Черновик
   14 Threads found on ru:Участник:Парис "Анима" надаль/Черновик
   Looking for: Template:Archive-bot/archive in [[:ru:Участник:Парис "Анима" надаль/Черно
   вик]]
   Processing 14 threads
   There are only 0 Threads. Skipping
   C:\Python26\pywikipedia>

Repeatedly I repeated start of the bot, the same result turns out. Nobody knows, what it is necessary to make in order that the bot ceased to pass pages? Thanks.--Парис "Анима" надаль (talk) 10:31, 11 April 2013 (UTC)

I tweaked your talk page see if it works now. Werieth (talk) 12:19, 11 April 2013 (UTC)

Yes, I saw, thank you very much, but, unfortunately, the bot all the same doesn't carry out archiving. :-( --Парис "Анима" надаль (talk) 12:58, 11 April 2013 (UTC)

Probably, someone uses this script. May I look at the pages of discussions archived by this bot? Perhaps, I will find, in what my mistake. Thanks.--Парис "Анима" надаль (talk) 14:35, 11 April 2013 (UTC)

Looks like the bot wasnt programmed for ru.wiki's timestamp format. Werieth (talk) 15:53, 11 April 2013 (UTC)

I believe this is the script used by MiszaBots I, II, and III. jcgoble3 (talk) 23:11, 11 April 2013 (UTC)

I agree with Werieth. You should use a regex to search for timestamp in Russian (in English, it's in the form "23:11, 11 April 2013 (UTC)"), otherwise it can't find the date of each thread. Here is my code that used on the Chinese Misplaced Pages.--Makecat 02:25, 12 April 2013 (UTC)

Makecat, would you mind submitting a patch to our tracker so we can push it upstream for others to use?

Hi, most Pywikipediabot developers don't watch this page, you're better off filing a bug in our tracker. Legoktm (talk) 02:39, 12 April 2013 (UTC)

Sourceforge is quite slow here due to some well-known reasons. I found that I can't add thread to tracker and it seems that Sourceforge doesn't allow using proxies (log out just after logging in). I hope you will fix it as my code was published. --Makecat 03:21, 12 April 2013 (UTC)

Browser table rendering differences

I am building a table with 18 equal width columns (plus one filler column). The text in the column should wrap if it is too wide. One browser does it nicely (Firefox on WinXP), but Safari on WinXP mixes up the sequence: some columns are minimised, then others are left too wide.

See {{Periodic table/sandbox}}. I used these css techniques: {{shy}} (soft hyphen) in the text, and set column width through ! style="min-width:5.35%; max-width:5.36%;" |. Suggestions anyone? -DePiep (talk) 15:26, 11 April 2013 (UTC)

I don't have any other browsers except for FF installed on this laptop, yet I suspect the edit I just made "should" have fixed the problem. I believe the problem was caused by undefined cells in the table. Technical 13 (talk) 15:57, 11 April 2013 (UTC)

This did not work (though the changes are to stay). Printscreen current Safari effect (same as previous) shown here. -DePiep (talk) 20:26, 11 April 2013 (UTC)

table column width is always problematic issue. you may want to try adding to the table style itself, "table-layout: fixed;" (picked this one from stack overflow - i do not know what other side-effects this may have, but i think it will make it do what you want). peace - קיפודנחש (aka kipod) (talk) 16:27, 11 April 2013 (UTC)

Will try this. You mean add the code to wikitable top row, like {| style="table-layout: fixed;" ? (or go ahead in the sandbox). -DePiep (talk) 20:26, 11 April 2013 (UTC)

exactly. i tried it in your sandbox (didn't hit "save"...), and it definitely makes a difference, though i'm not sure this is what you really wanted. peace - קיפודנחש (aka kipod) (talk) 22:36, 11 April 2013 (UTC)

 Done. Works. Thanks. Barnstar. -DePiep (talk) 10:35, 12 April 2013 (UTC)

because of how hairy this whole issue is, i strongly recommend testing with several browsers (at least the big 3, safari and opera gets bonus points, and if you have access to some handheld it's a medal), and trying several screen-widths with each - table should display correctly with width of 800px or less, before declaring victory. peace - קיפודנחש (aka kipod) (talk) 13:32, 12 April 2013 (UTC)

Checked some 100+ browser × OS screenshots. Per WP:ACCESS#Resolution, I took 1024px screens to be decisive. Did not see any prohibiting failings. As for the 800px screens: tried two at home, but no issues. Same for the few mobile versions I checked. I consider this acceptable. Details and tweakings continue at WT:ELEM. -DePiep (talk) 19:13, 14 April 2013 (UTC)

Timeline question

I just created User:Werieth/sandbox, is there a way to highlight every other row (white/gray alternating) so that it is easier to follow the timeline? Werieth (talk) 16:36, 11 April 2013 (UTC)

I couldn't find anything of the sort listed on MW:Extension:EasyTimeline/syntax which is where I expect it would be. Technical 13 (talk) 16:49, 11 April 2013 (UTC)

with timeline extension, almost everything is "possible", and almost nothing is easy or simple.

it is not clear what do you mean by "highlight". the simplest thing would be to simply assign a different color to the odd vs. even bars. if by "highlight" you mean using different background, i think that even though "background" per se is not supported as far as i could see, it should be possible to place different bars one on top of the other, in effect creating background "manually" (by making the "background bars" assume the full width of the row). for a very complex example, see mw:Template:Wikimedia Growth.

(side comment): at one point i was considering adding line-graph capabilities to Module:Chart, by studying "timeline" syntax and using it (the idea was to use the module to translate sane and more-or-less unified way to pass parameters, and translate it to the byzantine syntax used by timeline), but i am not sure i want to do it any more - timeline syntax may be too much for me. peace - קיפודנחש (aka kipod) (talk) 17:22, 11 April 2013 (UTC)

I was looking for something like where I could color the rows and make reading it easier. Werieth (talk) 17:36, 11 April 2013 (UTC)

Problem with wgRedirectedFrom

I have some JavaScript code at User:Dudemanfellabra/NRHPstats/sandbox.js that does one thing if you've been redirected to a given page and another thing if you went straight there. The only way I know to check if the user has been redirected is

 if (wgRedirectedFrom != null) {
    // do stuff if you were redirected
 }
 else {
    // do stuff if you weren't redirected
 }

For some reason, though, the wgRedirectedFrom variable seems not to be functioning as expected. If I'm on a redirected page, it works (i.e. it isn't null), but if I'm not on a redirected page, I get an error from that check saying wgRedirectedFrom isn't defined. I thought it was always defined but null if you weren't redirected. If this is not the case, does anyone know how to check if you've been redirected to where you are now?--Dudemanfellabra (talk) 18:01, 11 April 2013 (UTC)

Sounds like wgRedirectedFrom doesn't exist if you weren't redirected... Try:

 if (wgRedirectedFrom) {
    // do stuff if you were redirected
 }
 else {
    // do stuff if you weren't redirected
 }

Technical 13 (talk) 18:08, 11 April 2013 (UTC)

Tried that earlier (and again just now, just in case); didn't work. Still said it wasn't defined, even though I was literally checking to see if it was defined. I have no idea what's going on.--Dudemanfellabra (talk) 18:11, 11 April 2013 (UTC)

This page says "wgRedirectedFrom | String | When redirected contains the title of the page we were redirected from. null when not redirected. Uses the same format as wgPageName." So it should be defined... Right?--Dudemanfellabra (talk) 18:18, 11 April 2013 (UTC)

When you try to use a variable name like that, Javascript tries to dereference it to get at its value, which will fail if the variable is undefined. Try:

if(typeof wgRedirectedFrom == "undefined")
{
   //handling of undefined redirect
}
else
{
   //handling of defined redirect
}

That should let you test to see if it's defined without actually looking at what's in it, thus avoiding the undefined error. Writ Keeper ⚇♔ 18:27, 11 April 2013 (UTC)

typeof worked. The script now does what it's supposed to do. That doesn't change the fact that the manual over at mediawiki is incorrect. That manual says it's null when it's actually undefined.--Dudemanfellabra (talk) 18:37, 11 April 2013 (UTC)

Yeah, that happens. WP:SOFIXIT. ;) Writ Keeper ⚇♔ 18:39, 11 April 2013 (UTC)

I didn't know I could edit mediawiki.. thought that required more credentials. I actually logged in for the first time there just now and was about to update the page, but I see you did. Since I've never edited there, though, I can't confirm the pending change. Oh well.--Dudemanfellabra (talk) 18:45, 11 April 2013 (UTC)

I guess it requires more credentials than either of us have to really edit it. :) Writ Keeper ⚇♔ 18:48, 11 April 2013 (UTC)

normally, you do not need to use "typeof". however, more and more scripts nowadays define "use strict", which may protest when encountering constructs like "if (wgRedirectedFrom)" i thing the standard way to access globals is "if (window.wgRedirectedFrom)", which should still work, even in strict mode, but in wikipedia code, we are supposed to use "if (mw.config.get('wgRedirectedFrom'))" instead. peace - קיפודנחש (aka kipod) (talk) 19:31, 11 April 2013 (UTC)

According to MW:ResourceLoader/Default_modules#mediaWiki.config the proper way to do it is if ( mw.config.exists( 'wgRedirectedFrom' ) ) { /* do stuff */ } Technical 13 (talk) 18:09, 13 April 2013 (UTC)

Remove page from categories

I'm sure there is a way to do this, does anybody know how to make it so that User:Jay8g/Duplicate templates is not included in the maintenance categories from the tags?--Jac16888 20:53, 11 April 2013 (UTC)

Well, if nothing else you could just substitute the templates with cats attached and manually remove the cats afterward.--Fyre2387 23:08, 11 April 2013 (UTC)

Set |category= to "no" on the translation and Wikiquote templates, and to the empty string on the policy templates. jcgoble3 (talk) 23:21, 11 April 2013 (UTC)

It seems to have been removed from the now anyway, I suspect due to the edits User:WOSlinker has been making to various templates (because of this thread?). Thanks for the suggestions though--Jac16888 19:01, 12 April 2013 (UTC)

How do I categorize a template?

I made Category:Conflict of interest templates but I can't figure out how to put templates in it. When I go to edit {{uw-coi}} the only categories I see are ones that are for the pages the template would be put on, even though that template itself is already in Category:Standardised user warning templates and Category:User warning templates.--Atlantima (talk) 14:27, 12 April 2013 (UTC)

Most templates are categorised by putting the cat on the template's /doc page, enclosed in <includeonly>...</includeonly>. Where the template has no /doc page, the cat goes on the template page but enclosed in <noinclude>...</noinclude>. --Redrose64 (talk) 14:37, 12 April 2013 (UTC)

And the existing categories on that template are coming from the inclusion of {{Single notice}}. the wub "?!" 14:40, 12 April 2013 (UTC)

Ah, yes, the noinclude. Forgot about that. Thanks--Atlantima (talk) 14:45, 12 April 2013 (UTC)

API help question

Hello, I originally asked my question here, but realize it may be awhile to get a response there. So, I'm going to ask here too in hopes of getting a speedier answer. I have found that "api.php?action=query&list=backlinks&blnamespace=0&bltitle=" + wgPageName; will tell me if there are any articles that link to the page, but how can I get the full count (if greater than 500)? I don't care "what" articles link there, just how many. I'm trying to write myself a little JavaScript which in part will expand the "What links here" link in my p-tb to include an count of pages that link to a specific NS (or two or three) and a total count of pages that link to that article. I intend to use such a script to save me some time when I'm tagging articles for improvement on enwiki to quickly know if the article is an "orphan" or not. Thank you. Technical 13 (talk) 17:04, 12 April 2013 (UTC)

You cant, that information is not stored in the database, the only way to do that is to retrieve the lists in chunks of 500 until you dont have any more. Werieth (talk) 17:06, 12 April 2013 (UTC)

That doesn't make sense. It has to be in the database somewhere. If it isn't, how does Special:MostLinkedPages have the counts readily available. If I do have to do it like you say, some of the pages on that list would require 9,000+ API requests to populate the list. Technical 13 (talk) 17:24, 12 April 2013 (UTC)

Special:MostLinkedPages is updated infrequently and probably takes quite a bit of time to run. The results are then cached until the next update. Anomie⚔ 18:10, 12 April 2013 (UTC)

• Honestly, I'm not real good with ajax or json. Would someone take a look at my script and tell me why it's not even trying to run? User:Technical_13/Scripts/Enhanced "What links here".js‎‎ I'm not looking for anyone to re-write it for me, just point to things that are not right and where I might find the documentation to fix it... If it is easier to tell me "this" should be "that", I have no problem with that, just please link some documentation so I can learn why and get it right myself next time, if possible. Thanks Technical 13 (talk) 19:38, 12 April 2013 (UTC)
  • On a quick glance, elseif isn't valid; JavaScript uses two words (else if) instead. :) jcgoble3 (talk) 19:51, 12 April 2013 (UTC)
    • I knew that one... It was a typo in the first section that got copy pasted to the second... Technical 13 (talk) 19:58, 12 April 2013 (UTC)
  • Well, you're attaching a load event to an element—is the element already loaded? — Bility (talk) 21:11, 12 April 2013 (UTC)
    • I'll have to get back to you on this one... I'm thinking of changing the .load() to a .mouseover() or something. Technical 13 (talk) 00:48, 13 April 2013 (UTC)
  • Aha! Plugged your script into my common.js, and my error console complained about a parsing error, namely an invalid return in the first if statement. return can only be used inside a function, to terminate the function, but your code isn't wrapped in a function. jcgoble3 (talk) 21:38, 12 April 2013 (UTC)
    • I've taken down the script and am "almost" entirely re-doing it. I'm working on it off-line. I'll keep that in mind though. Technical 13 (talk) 00:48, 13 April 2013 (UTC)
• Okay, I've reworked my script, and put it back up. It "should" be building a 2-dimensional array of which each inner array should be comprised of and each inner array is stored in a sequential outer array. Right now, it is set to just gather data and pop out an alert for each of the 24-26 namespaces. If someone familiar would be so kind as to take a look at it and tell me if I've got anything badly borked, I would be appreciative. Technical 13 (talk) 09:44, 13 April 2013 (UTC)
  • Still throwing a SyntaxError: return must be in a function. You have two options here, as I see it: either wrap the script in a function and then call the function at the bottom of the script, or enclose the whole thing in an if-then-else statement. The latter would involve using the namespace/link presence check as the condition, only the alert in the first part of the if block, and the rest of the script in the else block. jcgoble3 (talk) 16:37, 13 April 2013 (UTC)
    • I forgot about that return up there. I'm importScript instead of using the resourceloader... Should I be using the resourceloader? I've commented out the return and replaced it with a break for now. Technical 13 (talk)
      • ResourceLoader is beyond my knowledge. As for break instead of return, that throws yet another SyntaxError: break can only occur inside a switch statement or loop. If you want that condition to cause the rest of the script to not run, you'll have to use one of the two methods I mentioned above. jcgoble3 (talk) 17:55, 13 April 2013 (UTC)

Toolserver is down again

I cannot use the lovely WP:REFLINKS. -- Alan Liefting (talk - contribs) 20:27, 12 April 2013 (UTC)

Or the Edit Counts tool, or Wikistalk, or... Can someone update me on what the deal is with Toolserver? I've heard that we're going to lose it entirely, is that the case? Beyond My Ken (talk) 21:41, 12 April 2013 (UTC)

The toolserver, which is run by the German chapter of the Wikimedia Foundation (WMDE), will be shut down either late this year or early next year. In its place, the WMF is creating a new platform (Tool Labs) that is intended to have essentially the same role as the toolserver, but will have a somewhat different configuration. Existing tools will need to be converted to run on the Tool Labs platform; however, the intent is to continue to support all or most of the current population of widely used tools. Because Tool Labs will be supported / managed by the WMF directly, one of the hopes is that it will be more stable over the long-term than the existing toolserver. However, it is very early in the process (e.g. Tool Labs does not yet have database replication), so it will probably be many months before we know how well the migration will work in practice. Dragons flight (talk) 21:56, 12 April 2013 (UTC)

I went to Misplaced Pages:Toolserver to see what the current situation was but there is absolutely nothing about it there. Can you update that page? Cheers. -- Alan Liefting (talk - contribs) 22:25, 12 April 2013 (UTC)

Have now put a hatnote over there. -- Alan Liefting (talk - contribs) 22:55, 12 April 2013 (UTC)

It is flaky. Now I cannot use Dabsolver. -- Alan Liefting (talk - contribs) 02:36, 13 April 2013 (UTC)

Or the ever-dependable Geohack. --Slgrandson (How's my egg-throwing coleslaw?) 03:37, 13 April 2013 (UTC)

It was working fine for me just a few hours ago, but now, the whole Toolserver website is down (the main page is now only a blank page, while every other page now returns a 404 error). I'm completely aware of the financial problems that the Toolserver has had for quite some time, but with several of our projects being dependent on Toolserver (like the one mentioned at the section directly above this), and Tool Labs still under development, this could become a major problem. What now? Narutolovehinata5 03:27, 13 April 2013 (UTC)

Dabsolver, Reflinks, Commons helper ( eg{) are all flaky for me. A real shame that an important tool is now unreliable. -- Alan Liefting (talk - contribs) 10:11, 13 April 2013 (UTC)

404 error at WP:UTRS

Hey, I'm trying to get my account exempt from IP hardblocks. After over an hour of trying to find out what I need to do, I posted a justification at the toolserver link provided at WP:UTRS, only to get a 404 error (The requested URL /~unblock/p/index.php was not found on this server.) when I submitted the form. This is seriously annoying. Sławomir Biały (talk) 02:42, 13 April 2013 (UTC)

The WP:UTRS interface runs on the toolserver and the toolserver is having problems again. See #Toolserver is down again above. It's still a bit buggy at the moment. These types of problens are usually fixed within a few hours, so checking back later is probably the best advice. 64.40.54.100 (talk) 05:08, 13 April 2013 (UTC)

Display problem with sub/superscripts

Sorry for asking a question for the Chinese Misplaced Pages. I posted a request on {{su}} about display problems of related sub/superscript templates on zh.wiki. When viewed in-line, a big space is added before the sub/superscripts, like this: . It is displayed fine when put in a numbered list or table. An incomplete list of problematic templates are in the first post on this site. This problem occurs on Chrome 26.0 on Mountain Lion and Chrome app on iPad, but is not seen on other browsers. The same templates here on en.wiki don't have this issue, and a simple copy and paste onto the zh.wiki templates result in only the big space before the subscript being removed, but the superscript is still bad. Please help us solve the problem! Yinweichen (talk) 02:10, 13 April 2013 (UTC)

We can's solve the problem; it's a Webkit bug. You may want to post a bug report at Chromium. — Edokter (talk) — 09:48, 13 April 2013 (UTC)

On second thought, it seems you are using old code on zh. Try our code again and add margin-left: 0; to see if that fixes the problem. — Edokter (talk) — 09:56, 13 April 2013 (UTC)

No it doesn't help. The subscript returned to the correct position, but the superscript remains to be detached. Yinweichen (talk) 17:15, 13 April 2013 (UTC)

I think the key is to figure out why Chrome renders it differently than Safari, and why the same code works in en.wiki but not zh.wiki. Yinweichen (talk) 17:30, 13 April 2013 (UTC)

Here's the reason: it has nothing to do with the code. I turned off the option to indent the first line of every paragraph, and the problem is solved. Yinweichen (talk) 17:57, 13 April 2013 (UTC)

Good to know! — Edokter (talk) — 20:16, 13 April 2013 (UTC)

Articles randomly showing up in categories

If you look at this cat for unassessed articles, Talk:Robert Andrew Loughnan and Talk:Rangi Mawhete suddenly showed up a few days ago. However, both are assessed and neither talk page has been edited in months (and the unassessed cat is not on those talk pages). Ditto with this other unassessed cat has Talk:Jones (left fielder) which is also assessed. This has been happening a bit the last week or so. Often I can fix it by moving the talk page header below the project banners, but none of the aforementioned articles has the talk page header. Was there a software upgrade recently that caused this problem? Aboutmovies (talk) 07:20, 13 April 2013 (UTC)

Take a look up this page at false negatives in category:Pages with missing references list and bugzilla:46978. I expect the api procedure described would work for you too. John of Cromer in Philippines (talk) mytime= Sat 17:19, wikitime= 09:19, 13 April 2013 (UTC)

I saw possibly related problems here, with Harrison Brown being reported as re-assessed in every bot run, even though the article talk page wasn't touched. My first suspicion was that WP 1.0 bot keeps it's own database to track assessment changes, which could be affected by the recent toolserver problems. But I don't think that can explain your observations. — HHHIPPO 09:47, 13 April 2013 (UTC)

Update: the problem I mentioned first started with this bot-run on 27 October 2012, one month after the last talk page edit. (The similar behavior for Ernst Abbe was due to a duplicated WikiProject tag, but I don't see such problems on Talk:Harrison Brown.) — HHHIPPO 10:05, 13 April 2013 (UTC)

Another update: looking at this thread, my problem seems to be indeed a toolserver issue and unrelated to yours. — HHHIPPO 10:28, 13 April 2013 (UTC)

Templates

I have this template that's for Work In Progress articles here: , and i'm trying to make it so that when you don't give it a proper date, it should give an error. How do i do this? RocketMaster (talk) 10:26, 13 April 2013 (UTC)

What do you mean "proper date"? What do you want the date to do? The way it currently is doesn't discriminate. Technical 13 (talk) 10:41, 13 April 2013 (UTC)

Surely it already does this: {{WIP|qwertyuiop}} should give a big red "Error: Invalid time". If you want a custom message, use {{#iferror:{{#time: m-d-Y|{{lc:{{{1|{{{date|yesterday}}}}}}}}}}|<span class="error">'''Error message here'''</span>}}. — This, that and the other (talk) 12:14, 13 April 2013 (UTC)

tt-to, His template isn't on enwiki... Technical 13 (talk) 12:18, 13 April 2013 (UTC)

Also, I've visited his template and made some modifications for him that since there has been no further question here, I'm assuming took care of it. Technical 13 (talk) 16:06, 13 April 2013 (UTC)

Main Page added to watch list

It has been recommended (via my talk page) that I post here: Is it just me but main page is being continually added to my watch list with no input from me. This occurs on iPad, smartphone and laptop. Is it a common fault? Asked this on talk page but that may not have been the best place for such a question. I think it is the smartphone I remove from my watchlist and see it returned after I have accessed wikipedia via my phone. Advice appriciated. Phone HTC one S android version 4.1.1 / browser chrome and one provided with phone - webkit/534.30. Ipad is usual, up to dat,e and laptop is windows 7 via firefox. Edmund Patrick – confer 11:10, 13 April 2013 (UTC)

Categories need a purge

Usage of Template:Convert was categorised. The cats were deleted. We now have redlinked cats in over 100,000 articles. See Northwest Lineman College for example (contains Category:Pages using Template Convert and Category:Pages using Template Convert/adj=on). An edit will clear out the redlinked cats. There is a way of clearing this problem by changing a flag or something. Another fine mess. -- Alan Liefting (talk - contribs) 12:03, 13 April 2013 (UTC)

Just give it some time, the Job Queue will take care of it. Werieth (talk) 12:04, 13 April 2013 (UTC)

How long will it take? -- Alan Liefting (talk - contribs) 12:10, 13 April 2013 (UTC)

likely a half a day. maybe a full day. That is exactly the reason why many templates (with high usage) are fully protectd mabdul 12:16, 13 April 2013 (UTC)

Is it is a function that can be done by a bot? That could speed it up? -- Alan Liefting (talk - contribs) 12:27, 13 April 2013 (UTC)

Bad idea, we have the job queue for a reason, it keeps server stress to a minimum. Werieth (talk) 12:37, 13 April 2013 (UTC)

Mediawiki, again

Hello from Latvian Misplaced Pages! I have a problem of finding the Mediawiki interface message for the message after succ. move for updating Wikidata page. Can somebody help? --Edgars2007 13:57, 13 April 2013 (UTC)

Are you talking about MediaWiki:Movepage-moved? Or something else? —Theopolisme (talk) 13:59, 13 April 2013 (UTC)

No, that is easy founded :D I can't find message, which appears below this message that user should update the Wikidata page (if there is such). --Edgars2007 14:03, 13 April 2013 (UTC)

Add ?uselang=qqx to the end of the URL (or &uselang=qqx if the URL already contains any parameters). That will show you the location of all the MediaWiki messages used on the page. It's an obscure but very handy feature. the wub "?!" 15:08, 14 April 2013 (UTC)

That wouldn't work here because it's a submit url with content depending on the action which lead there. The message is MediaWiki:Wikibase-after-page-move. PrimeHunter (talk) 23:41, 14 April 2013 (UTC)

My talk page posts deleting a more or less simultaneous post from another editor

At this edit my post deleted a more or less simultaneous post by another editor – were both replying to a third editor's post. There was no edit conflict warning. This has happened to me three times now, all on this same page and, as far as I know, only when the other's post is to the same talk page section. Anyone know what's going on?

—Trappist the monk (talk) 15:54, 13 April 2013 (UTC)

The problem you are reporting sounds like a potential issue in the code of the MediaWiki software or the server configuration. If the problem is reproducible, it would be nice if somebody who has this issue could send the software bug to the 'Bugzilla' bug tracker by following the instructions How to report a bug. This is to make developers of the software aware of the issue. If you have done so, please paste the number of the bug report (or the link) here, so others can also inform themselves about the bug's status. Thanks in advance! --AKlapper (WMF) (talk) 16:19, 14 April 2013 (UTC)

Hasn't happened since the post I referenced, though I've been extra careful to make sure that other posts haven't occurred while I've been writing a new post. I'll go back to just hacking away and posting and see if I can get it to happen again.

—Trappist the monk (talk) 17:15, 14 April 2013 (UTC)

Template help

I want to make a template that shows one message when given a 1, and another message when given a 0, and an error when i give it no parameters. How do i do this? Let's say, a site status template. I want it to say "This site is {{{status}}}", where status is either 0 or 1. And an error when given no parameters. How do i do this? RocketMaster (talk) 15:57, 13 April 2013 (UTC)

"This site is {{#switch:{{{status|}}}|0= Dead|1= Alive|#default= Error!}} Technical 13 (talk) 16:05, 13 April 2013 (UTC)

Updated Misplaced Pages favicon

Hi. I didn't see this mentioned in many places around here, but the Misplaced Pages favicon is being updated this month (I think it's slowly rolling out to the various Wikipedias now). Further details are available here: m:Favicon#Misplaced Pages. --MZMcBride (talk) 19:10, 13 April 2013 (UTC)

Note that the mobile favicon, for when a wikipage is added to your homescreen on an iOS device, has been updated as well-previously it had no icon. —Theopolisme (talk) 19:39, 13 April 2013 (UTC)

Clicking an image redlink brings me to the File Upload Wizard

Royal Flying Doctor Service of Australia#The service today, the second image down. I remember seeing a place that could tell me if the file was deleted and why, but I don't see that. MeekSaffron (talk) 19:56, 13 April 2013 (UTC)

It would be the deletion log although since it's an image, you need to try the one on commons as well. Anyway, the problem wasn't a deleted image, but that somebody had altered a hyphen-minus to an en-dash, thus breaking the image link, which has now been fixed. --Redrose64 (talk) 20:48, 13 April 2013 (UTC)

Given that the red link is almost useless, what's the minimum number of keystrokes to obtain the deletion log for a deleted image? I made a suggestion at WT:POPUPS last December and got nowhere. -- John of Reading (talk) 20:54, 13 April 2013 (UTC)

If you use the mouse to copy the redlinked filename, then in the left margin go for Special pages, then Logs; click in the "Target (title or User:username for user):" box and paste, then hit Go I tally a total number of keystrokes: 0.

I've found the edit which broke the file link. Quite surprised at who did that, really. --Redrose64 (talk) 22:14, 13 April 2013 (UTC)

Stuff like that will happen if you use the search-and-replace function of the toolbar to replace all hyphens. — Edokter (talk) — 22:27, 13 April 2013 (UTC)

wikEd strikes again... I'm glad I'm not the only one that does that (I try to catch and fix right away when I do). I've been meaning to go to the wikEd creators and ask for some fixes, like it should know not to add spaces after periods in URL strings and not capitalize parameter names inside templates... Just two more examples of things to look out for. Technical 13 (talk) 22:37, 13 April 2013 (UTC)

Oops! That was a bad search and replace (using an offline text editor), performed late at night when I probably shouldn't have been editing at all (I'd been up the night before with a gastro bug). Thanks for the fix, Edokter! Graham87 11:56, 14 April 2013 (UTC)

Or rather, that was the night I had just returned home from the holiday where I probably caught the bug and by the time I'd made the offending edit ... I don't even want to talk about it. Graham87 14:53, 14 April 2013 (UTC)

Favicon

Since when did our favicon change? I just noticed it having rounded corners and being more squashed.--Gilderien Chat|List of good deeds 20:23, 13 April 2013 (UTC)

Look up. —Theopolisme (talk) 20:35, 13 April 2013 (UTC)

Oh thanks don't know why I didn't see that I did look for it :/ --Gilderien Chat|List of good deeds 20:37, 13 April 2013 (UTC)

IP tool links

What is going on with the links in the info box at the bottom of IP talk pages and edit histories? There is no longer a geolocate link, and the traceroute link goes to the DNSstuff site where I am requested to register and sign up for "30+ powerful tools and alerts" (which annoyingly I can't back out of because the previous page just redirects me back). SpinningSpark 23:43, 13 April 2013 (UTC)

You describe {{Anontools/ipv6}} which you get for IPv6 addresses. For IPv4 you get {{Anontools/ipv4}}. If you know better tools for IPv6 then you can make a suggestion at Template talk:Anontools. PrimeHunter (talk) 00:35, 14 April 2013 (UTC)

Ah, thanks, I hadn't realised the tools were different. SpinningSpark 09:21, 14 April 2013 (UTC)

Red link convert categories

Hello, I'm not completely sure if this is the best place to list this problem but I've seen a few problems happening lately. I'm not sure if these are supposed to be hidden categories but they keep showing up in articles that use the convert template. Here are some examples, .-- Astros4477 (Talk) 03:21, 14 April 2013 (UTC)

See #Categories need a purge it is a known issue that should resolve itself. Werieth (talk) 03:22, 14 April 2013 (UTC)

The three examples given and the two categories that I know of have now cleared. -- Alan Liefting (talk - contribs) 07:34, 14 April 2013 (UTC)

It's still showing up for the first two links I listed.-- Astros4477 (Talk) 13:58, 14 April 2013 (UTC)

It disappeared when I purged the articles. PrimeHunter (talk) 14:18, 14 April 2013 (UTC)

Redirects not working in the Module namespace

See e.g. Module:Cite. Any ideas on how to fix this? It Is Me Here 12:02, 14 April 2013 (UTC)

Redirects are not implemented for the Module: namespace, because these are server-side scripts. There is no fix (nor can I imagine a use-case for redirecting a module). — Edokter (talk) — 12:10, 14 April 2013 (UTC)

Redirects don't work in the MediaWiki namespace as well, but you can transclude one MediaWiki to another MediaWiki. --  Gadget850 (Ed) 13:51, 14 April 2013 (UTC)

With that said, would someone delete Module:Cite, please? —Theopolisme (talk) 13:54, 14 April 2013 (UTC)

done —TheDJ (talk • contribs) 15:19, 14 April 2013 (UTC)

Image displayed on black background

File:SOAS Crest.jpg Why do the arms of SOAS display on a black background? It makes it virtually impossible to see. Other colleges of London University appear normal. Justlettersandnumbers (talk) 20:56, 14 April 2013 (UTC)

It doesn't look black to me. You don't happen to have "Use a black background with green text on the Monobook skin" checked on Special:Preferences#mw-prefsection-gadgets and are using the Monobook skin, are you? Technical 13 (talk) 21:27, 14 April 2013 (UTC)

It looks black for me as well, but solely when used with |thumb -- when I view it on the description page, everything looks fine. —Theopolisme (talk) 22:09, 14 April 2013 (UTC)

What about the thumb lower down on that same page? I too see the full-size image in its proper colours. I've asked on commons if I can upload a screenshot of this (non-free) image for discussion purposes; relieved to hear that it isn't just me that's affected. Justlettersandnumbers (talk) 22:30, 14 April 2013 (UTC)

Do any other images do this with |thumb set? What browser are you guys using? I can't make the background black for the image in Firefox 19.0.2 Technical 13 (talk) 23:06, 14 April 2013 (UTC)

Me: Safari on a Mac (Mozilla/5.0 (Macintosh; Intel Mac OS X 10_6_8) AppleWebKit/534.57.2 (KHTML, like Gecko) Version/5.1.7 Safari/534.57.2). —Theopolisme (talk) 23:08, 14 April 2013 (UTC)

Yep, it's a Safari thing (I have the same browser and OS version as Theo); loading this page in Firefox makes the image display correctly. Off to Bugzilla ... Justlettersandnumbers (talk) 23:35, 14 April 2013 (UTC)

Chopping parameters from templates

{{Marriage}} just survived TFD via a no-consensus close; the delete voters were upset that it was heavily bloated with unnecessary components, while most of the keep voters (including me) wanted to keep it but thought we should remove the unnecessary components. I'd like to start removing some of the unnecessary components, but I fear damaging pages where those components are used — is there any way to figure out (automatically) which components are used on which pages? The template is transcluded on approximately 2,500 pages, so I'm not about to check all of them as long as there's any other way to do it. Nyttend (talk) 00:42, 15 April 2013 (UTC)

There are, depending on what kind of information you need, you can also add tracking categories to find what pages use which params. Werieth (talk) 00:44, 15 April 2013 (UTC)

I've used tracking categories before as Werieth suggests, and I want to let you know if you haven't done it before it is something that has to go through the job cue and "may" take a couple of days to show accurate results. If you are patient enough, it is the easiest way to do it in my opinion. Technical 13 (talk) 01:17, 15 April 2013 (UTC)

Please look at this revision of the Sandbox, plus its code — I've filled out all of the parameters, but nothing appears except the name of the person who's gotten married, the date of the marriage, the fact that it ended because of divorce, and the date of the divorce. Nothing about the location or the name of the spouse appears; is this because it's not designed to appear, or because I did something wrong? The main reason I came here is that I didn't want to remove text that this template was putting on pages, but if things like the location of the wedding ceremony aren't displayed in articles, we might as well begin discussions immediately about removing the parameters, without worrying about what pages use those parameters. Nyttend (talk) 01:31, 15 April 2013 (UTC)

okay, I'll set up the tracking categories and a control board page table to morrow. I'm in bed now and can't do it from my BlackBerry. (buttons too small and I can't see screen because of some odd bug that makes the line I am typing one line off the bottom of the screen.) :( Technical 13 (talk) 01:56, 15 April 2013 (UTC)

In you sandbox, the following large block appears in the HTML:

<span style="display:none">«start:<span style="display:none"> (<span class="dtstart">2000-01-02</span>)</span>–end+1:<span style="display:none"> (<span class="dtend">2014-01-04</span>)</span>»<span>"<span>Marriage: Mrs. Booth to John Wilkes Booth</span>"</span> Location:<span><span><span class="street-address">11</span>, <span class="street-address">22</span>, <span class="street-address">33</span>, <span class="locality">place1, </span></span><span style="display:none"><span class="plainlinks nourlexpansion"><a rel="nofollow" class="external text" href="//toolserver.org/~geohack/geohack.php?pagename=Misplaced Pages:Sandbox&params=5.000_N_5.000_E_scale:1000&title=January+2%2C+2000%3A+Marriage%3A+Mrs.+Booth+to+John+Wilkes+Booth"><span class="geo-nondefault"><span class="geo-dms" title="Maps, aerial photos, and other data for this location"><span class="latitude">5°00′00″N</span> <span class="longitude">5°00′00″E</span></span></span><span class="geo-multi-punct"> / </span><span class="geo-default"><span class="vcard"><span class="geo-dec" title="Maps, aerial photos, and other data for this location">5.000°N 5.000°E</span><span style="display:none"> / <span class="geo">5.000; 5.000</span></span><span style="display:none"> (<span class="fn org">January 2, 2000: Marriage: Mrs. Booth to John Wilkes Booth</span>)</span></span></span></a></span></span></span> <span style="display:none">(linkback:<span class="uid url">//en.wikipedia.org/Wikipedia:Sandbox</span>)</span></span></span>

I assume someone wants / expects it to be there even though it is intentionally not visible to the user. As to why it exists? I don't know. I've never followed this template / issue. Dragons flight (talk) 02:09, 15 April 2013 (UTC)

• Misplaced Pages noticeboards
• Pages automatically checked for accidental language links