Misplaced Pages

Revision as of 01:03, 11 March 2006

I don't hang out on this particular wiki as often as I'd like, so may never read anything you write here. Please report problems with the software at http://bugzilla.wikipedia.org/ --Brion 17:50, 15 Sep 2004 (UTC)

You might also find me in the Misplaced Pages:IRC channels.

If you have a sensitive issue, e-mail me with details rather than being vague if possible. :D --Brion 08:46, 26 December 2005 (UTC)

Special:Ipblocklist listing truncates non-ASCII usernames

The Special:Ipblocklist listing severely truncates non-ASCII usernames (eg Russian usernames).

This can lead to severe confusion and can lead to mismatches between block and unblock listings in the Special:Log/block listing, making it appear that a non-existent username was unblocked and making it difficult or impossible to tell the actual username that was unblocked (unless you can deduce it by guessing). See discussion at

-- Curps 06:02, 26 December 2005 (UTC)

I think there's a bug filed on this since some time ago. On user blocks the name actually isn't the operable part, though, but rather the id field. --Brion 08:48, 26 December 2005 (UTC)

testing bugzilla:4411 again

Raw pages always exist, now.

Until just recently an &action=raw HTTP transaction would yield a 404 status for a non-existent page. It is now yielding 200 for all pages, whether they exist or not. This broke several of my tools, which expected this fundamental part of HTTP to be adhered to. ☺ Was this change intentional? Uncle G 07:24, 1 January 2006 (UTC)

• The 404 return was not specific to action=raw, and it was in fact rather broken. It was reverted due to massive breakage of edit pages for many users with slightly dodgey browsers or proxies or something. --Brion 22:54, 3 January 2006 (UTC)
  • Which are you saying to be broken? Status 404? Or status 200? Uncle G 00:54, 4 January 2006 (UTC)
    • Status 404. --Brion 02:46, 4 January 2006 (UTC)
      • Would it be possible to have status 404 reinstated just for &action=raw? If not, how do you propose that we can tell, using &action=raw, whether a page exists or not? Uncle G 02:40, 6 January 2006 (UTC)

kreanto volata, serĉata

Brion, ĉu vi konas iun, kiu povus helpi nin krei vikipedion en la pensilvangermana ĉe pdc.wikipedia.org? Ni estas akceptitaj approved kaj ni havas viki-on kiu estas komencita pensilvangermana viki-o, sed nun ĉio sajnas stagni. dankon. Stettlerj 00:40, 3 January 2006 (UTC)

still a problem with Recent changes "Show new changes starting from..."

Hi,

The problem I reported earlier is still occurring; the largest gap I've seen is 8 Recentchanges entries missing between the top of the previous screen and the bottom of the next screen, although usually it's just one or two or zero.

Is there any way to specify the "from" parameter as a version id rather than a timestamp? Ie, instead of:

• http://en.wikipedia.org/search/?title=Special:Recentchanges&from=20060104205710

it would be some syntax like:

• http://en.wikipedia.org/search/?title=Special:Recentchanges&fromversion=33885462

Presumably this would avoid the problem with the gaps. -- Curps 21:07, 4 January 2006 (UTC)

categories "parented" by an article

Hi - There's a discussion at Wikipedia_talk:Categorization touching on, among other things, what to do with "topic" categories (categories with the same name as an article, like Category:George W. Bush). I'd just as soon put this particular category in Category:Mass murderers, but assuming I won't be able to get a consensus behind this idea (and it really doesn't generalize very well) what would you think about the idea of providing a mechanism for a category to be "parented" into an article? The idea is that this type of category really doesn't have much to do with anything but the article of the same name, so the "category" listing might as well display only the article. This can be done with a see also link from the article, but this doesn't address what the "parent" of the category is. Assuming we might want to explicitly allow an article to be a "category parent" of a category, I don't know what the syntax might be for the category reference from the category, maybe something like ]. There might be some value in using a syntax that would allow other namespaces as well. Some folks are arguing the reverse, i.e. such an article should only be in the one "topic category" and the "topic category" should be in the categories you might expect to find the article in. Thoughts? -- Rick Block (talk) 03:47, 5 January 2006 (UTC)

Fundraising

In recognition of your generous offer to assist in fundraising efforts by recording video of yourself wearing a bronze pointy brassiere and a horney helmet singing "Give us cash," I hereby award you this horney helmet. Can't wait to see the film! -- Essjay · Talk 04:07, 6 January 2006 (UTC)

Couldn't help myself! ;-) -- Essjay · Talk 16:32, 6 January 2006 (UTC)

Regarding template default parameters and nesting

I saw you posted on bugzilla:364 regarding template default parameters. I replied there as well, but I'm not sure if you're fully aware of how much that feature has given rise to a number of conditional functions (see Category:If Templates and Category:Boolean Templates). From a practical standpoint, I don't see where, unless it is the intention to support conditionals, that any practical use of templates would be harmed by the following changes:

• Default parameters - There should be no more than 2 levels of evaluation of a parameter default. {{{A|text}}}, {{{A|{{{B}}}}}}, and {{{A|{{{B|text}}}}}} should work, but {{{A|{{{B|{{{C}}}}}}}}} and deeper should not. Additionally, parameter defaults should not accept template calls in the form of {{{A|{{template}}}}}} - that would be an avenue to skirt the restriction and there's no need for that for any practical use.
• Template nesting (Misplaced Pages:Avoid using meta-templates) - The parser should not evaluate more than one level deep on any template. For example, if a page uses Template:X, which in turn has a call for Template:Y, it works, but if Template:Y in turn calls Template:Z, then Z is not read from the database. This prevents the more outrageously complex template nesting practices, but gives a little flexibility in certain areas.

I'd like to know if this is something you would look into and perhaps address. -- Netoholic @ 18:22, 7 January 2006 (UTC)

Speard your wings!

I am hereby awarding you these wings for your assitance with the vandal issue. You have more than deserved them. --Cool Cat 21:54, 7 January 2006 (UTC)

Image:Ascorbin.gif has been listed for deletion

Reasons:

1. OB (obsolete) - The image has been replaced by a better version.
2. OR (orphan) - The image is not used on any pages in Misplaced Pages.

Gnj 20:08, 10 January 2006 (UTC)

Not mine. Any reason you're asking me about it? --Brion 20:52, 10 January 2006 (UTC)

• Very probably because you are listed as the first editor in the image page's history. Uncle G 21:18, 10 January 2006 (UTC)

Rollback button

Could you add on a Vandalism link next to rollback that rolls back the article but with its own different MediaWiki edit summary like "reverted vandalism by $2". That way, vandalism takes less time to count for blocking decisions and we can do other things more. Thanks a million for semi, not it is time for a additional rollback button :).Voice of All 16:55, 12 January 2006 (UTC)

Pleeease. This would distinquish vandalism from other random edits.Voice of All 02:43, 14 January 2006 (UTC)

Slovenian Wikisource

Hi! I would just like to ask when will the Slovenian wikisource be created. According to this page, it has enough support already. --Eleassar 13:36, 13 January 2006 (UTC)

Changing edit attribution

Brion, an editor asked me whether bureaucrats have the technical ability to change attribution for edits (as from an IP to the username). I said that only developer could do that as I would need database access. He understood you to say bureaucrats had that ability. Please clarify this. Thanks, Cecropia 16:41, 13 January 2006 (UTC)

Either Special:Renameuser can do this, or it should be fixed to do so. --Brion 18:13, 13 January 2006 (UTC)

Does that mean that you can enter an IP address in Renameuser and the name of a registered user and the attribution will be changed without wiping the history of the registered user or generating an error? If it does work, would it be a big deal to modify Special:Makesysop to give Renameuser permissions to specific current admins? -- Cecropia 22:06, 13 January 2006 (UTC)

Image:California map showing counties (source).xcf.gz

Please list the source for this image or it will be deleted in 7 days --Admrboltz (T | C) 21:18, 13 January 2006 (UTC)

It already had source information, but I've expanded the detail a bit. --Brion 01:07, 14 January 2006 (UTC)

Templates and server load

Are there any server load issues we should be aware of regarding often-used templates, categories on those templates, or images on those templates? Many editorial and/or userbox templates are frequently used and contain both category (which would be redundant with Special:Whatlinkshere) and image (which sometimes is a scaled down version of a really big image). Radiant_>|< 22:39, 15 January 2006 (UTC)

Deleted edits

I have found the ability to look at an editor's deleted edits particularly useful in evaluating RfA candidates (or potential candidates), as a high number may signify a lot of work sending bad pages for deletion. Could the system be set up so that a user could view his own deleted edits (in the same way that we can observe our own watchlists), and further perhaps to authorize others to view that user's deleted edits for limited purposes? Cheers! BDAbramson T 19:44, 16 January 2006 (UTC)

User:Brion VIBBER/Cool Cat incident report

Again thank you for that. As you well know the ban on User:MARMOT is symbolic as he uses sockpuppets and proxies like there is no tommorow. He commited the "Cool Cat incident" by using a proxy for example according to Kelly Martins report. As MARMOT have used vandal bots before he will continue to vandalise wikipedia with vandal bots like there is no tomorow. He will also abuse/stress out wikipedia servers to find more such exploits etc causing you more work. With that exploit he got at least one person (me) blocked for vandalism by spoofing my IP.

I propose you (You or board member(s) or any other employee of wikipedia/wikimedia) take this to NTL or at least we cooperate on what kind of a thing we want to send to convince NTL to take action against MARMOT.

I am not thinking of a legal action although I would not object it but that would perhaps be too extreme at this point. --Cool Cat 11:16, 18 January 2006 (UTC)

Deleted Edits

Could the Undelete feature be restored to non-admins with the edit summary removed? I don't think there's any harm with that. According to the email linked from Kate's editcounter, it only said that the edit summary had "private, embarrassing, vandalistic, libelous, etc stuff". Thanks. -- King of Hearts | (talk) 00:57, 20 January 2006 (UTC)

:-)

Thanks for the note at AUM. Dragons flight 03:28, 21 January 2006 (UTC)

Same here, thanks for that. —Locke Cole • t • c 03:57, 21 January 2006 (UTC)

Editing statements

I'm sorry for editing you statement and leaving your signature, but I was of the understanding that you had overstated your opinion. Also, the strongest words fit better on a talk page than a policy page. Now, I'll stay away for the time being. --Eddi (Talk) 04:41, 21 January 2006 (UTC)

Images/templates and server load

Thank you for clarifying that. Radiant_>|< 11:45, 21 January 2006 (UTC)

Misplaced Pages:Avoid using meta-templates note

Thank you for the clarifications :) --Cool Cat 13:20, 21 January 2006 (UTC)

Thank you for the clarifications!  :-) I thought the policy made little sense. Too bad it's already been "enforced" everywhere, subst'ing templates and permanently removing links and dependencies, cluttering up the wiki markup with unnecessary HTML (remember that wiki is supposed to be editable by anyone) and using kludgy hacks to perform functions that could be done much more elegantly with the "forbidden" template functionality.

You said "You should avoid metatemplates if they're ugly, hard to use, or fragile. That's just common sense; don't worry about "server load" for them."

This will be the next problem. People will create their own (mis)interpretations of what is "ugly, hard to use, or fragile", and start the same trouble again. Some people really don't like templates, and they'll use any excuse they can get. As you can see, your changes to WP:AUM are already being revert-warred.

Some similar comments are over here on Jamesday's talk page: User_talk:Jamesday#Templates_and_server_load.

Note that this isn't just about "meta-templates" or templates within templates. The same "policy" has also been enforced against regular templates, conditional templates, templates that generate HTML, and so on. (Netoholic has called conditional templates "the Antichrist", for instance, and called for their complete deletion to enforce WP:AUM against the "disobedient masses" who didn't believe in his interpretation of Jamesday's words.)

This is a big issue. Please don't make vague statements and then vanish from sight like Jamesday did. Then it just becomes different users' interpretations of "divine revelation", and the most disruptive users generally win. We need you to stick around and give us clear explanations of what is bad and what is not, and stop all the misinterpretations and BS before they start; before they turn into more time-wasting talk page wars.

Any rewrite of the page will basically need to be done by you, since everyone else will just fight about what "you really meant to say". — Omegatron 19:26, 21 January 2006 (UTC)

I found Jamesday's original comments about server load, and the only benchmarks I have ever seen regarding this issue: Template_talk:Sisterproject#Technical_impact_of_templates_like_this — Omegatron 19:02, 22 January 2006 (UTC)

I would like to add my thanks. It is also notable that the triumvirate of Snowspinner, Netoholic and Sam Korn are still insisting that they must be correct and that you must have missed a bit. SK is getting a little strident and his anti-template prejudices are showing. It remains to be seen what their reaction to your kind offer to consider limited built-in functionality will be. HTH HAND —Phil | Talk 14:47, 23 January 2006 (UTC)

Thanks!

Hi Brion, thanks for tacking the time to make your cogent, concise, and timely answers to my questions at Misplaced Pages:Village pump (technical)#Questions re server load from images in userspace. Very enlightening and useful. Herostratus 14:24, 21 January 2006 (UTC)

Username Change

Hello I was wanting to follow up the short conversation we had at Apachecon about the posibility of a username change for myself. The problem lies in that I would like to use the name A on Misplaced Pages but someone with no intent to contribute beat me to the punch. I put in a request on the change user page, but I've been told that a dev would have to help me with this name change rather then a bcrat. A friend recommended I ask you here for some help. Thanks in advance for any help you can be. I can be reached via my talk page or via the email I have associated with my username. —A 06:19, 23 January 2006 (UTC)

I'm pretty sure a bureaucrat can do that, if there's consensus to rename the other account. --Brion 07:34, 23 January 2006 (UTC)

Conditionals

Oppose

Please don't add conditionals to MediaWiki. I have made an open challenge to people on the AUM talk page to put forward any template that really needs conditional syntax. I am absolutely convinced that any practical template does not need it. The CSS hiddenStructure method I advocate wasn't meant to be an alternative to conditionals. It is meant only as a "would be nice" feature for hiding rows in an Infobox that otherwise would be blank. It works just fine for that purpose. Conditionals also aren't needed because changing how the template is used/setup or by splitting the template into multiple similar ones to avoid "optional" requirements are both good solutions.

The problem is, the minute you add conditionals to MediaWiki, these same clever folks and dozens more will forget about the idea that not everyone on Misplaced Pages is a computer geek. We have historians, linguists, etc., and they are the ones that need to be able to easily make, change, and use templates to fit their needs. What will happen is that we'll be put on a one-way course towards increasingly complex structures and lose focus of what we're doing. Template space will become a growing data repository where statistical information is pulled. As funny as it is, I've already seen entire articles wrapped in templates, and I've seen templates being used just as a data array (checkout the templatelinks on Misplaced Pages:WikiProject Flag Template/Testall). That is not something I find appealing, nor would you, I'm sure. -- Netoholic @ 15:32, 23 January 2006 (UTC)

Let me provide evidence. The introduction of just one conditional template feature (that being default parameter values) has lead to in incredible growth in template complexity. It seemed innocent at the time, I'm sure, but has directly given rise to both the "ugly" CSS method and things like Template:Qif. One cannot begin to imagine what purposes more functions would introduce. -- Netoholic @ 20:53, 23 January 2006 (UTC)

Support

Please add conditionals. Not everyone on Misplaced Pages is a computer geek. The article source code is getting more and more complex over time, and is only accessible to those not intimidated by code. We need the enhanced template functionality to make article editing easier for non-technical types and avoid systemic bias.

The template forking and limited functionality advocated by Netoholic is not an acceptable or desirable solution. — Omegatron 15:44, 23 January 2006 (UTC)

Agree with Omegatron. If implemented correctly, conditionals in templates would likely be easier to use and understand than the current meta-templates. —Locke Cole • t • c 15:52, 23 January 2006 (UTC)

Indeed. --Adrian Buehlmann 16:36, 23 January 2006 (UTC)

Me too. The more robust the template options are the less complex the article and template code become. --CBD _☎ 17:12, 23 January 2006 (UTC)

Netoholic argues that we should be making it easier for people to edit Misplaced Pages. So we should make it simple if they want to do something complicated: we should provide them with a template which will take their raw data as parameters, with friendly names, and lay it out in an agreed format with the minimum of fuss. So what happens if they are missing one aspect of that raw data?

• tell them to push off until they find the last piece of the puzzle? (oh, but they've got to cite their sources, so they can't actually write their article until they've tracked down the exact editon of the book they want to reference)
• tell them to pick from one of 17 possible variants of the required template which fits the subset of parameters which they are able to specify right now? (oh, and that's without providing them with a nice pick-list or anything, they have to scurry through umpteen talk-pages searching for it…)
• give them an external link to a separate tool which will accept their parameters and pump out the required wiki-text? (oh, but if they change their mind, or find an additional nugget of information, they've got to cut & paste again from scratch)
• construct a template which will accept a subset of the full panoply of parameters with the minimum of fuss, pointing out if they have missed something vital, allowing them to add further pieces of the puzzle as and when they arrive? (oh, but…)

I know which of these scenarios corresponds to what I would think of as the "wiki way". And I know that I would like a strictly-limited set of conditional constructs built in to Mediawiki. Yes, please. Pretty please with a bow on top and a Voldemort-size cauldron of best-quality chocolate. HTH HAND —Phil | Talk 17:06, 23 January 2006 (UTC)

Phil's comments are generally about citation templates like Template:Book reference. My opinion is that we shouldn't use templates for something like this, which only propogates formatting (not content). Simple wikitext is probably best here, but if something more is desired, it should be done along the lines of meta:Cite, not templates. -- Netoholic @ 17:15, 23 January 2006 (UTC)

I thought that part of the rationale for the template system in the first place was that it enabled formatting to be standardised over a large number of loosely-related articles, with a mechanism for updating the standard formatting from a centralised location: are you suggesting that we repudiate the template system altogether? Since the only alternate solutions being offered are the opportunity to run to a bot-operator, cap-in-hand, requesting a re-format sweep, or to use an external tool the availability of which is not guaranteed and the functionality of which is not fixed, I would rather stick with the template system, which is at least subject to the same merciless editing as everything else in Misplaced Pages. —Phil | Talk 12:14, 24 January 2006 (UTC)

• I feel conditionals in MediaWiki software are the most elegant and the easiest to maintain solution, to the recocurring problem of optional parameters in info boxes and citation templates. The various pipe tricks are one of the most confusing syntaxes I've come across, very esoteric and hard to grasp for most users. (although there is something strangly intreaguing about how much functionality can be obtained from such a simple syntax). CSS hack are hacks. A good conditional syntax would enhance the value of the encylopedia and allow as yet unknow improvements to be implemented by users with little programming experience. --Salix alba (talk) 21:11, 10 March 2006 (UTC)

Categories -- How big is too big?

At Misplaced Pages talk:Categorization there is a good deal of support for making bigger categories. For instance, having Category:Film directors being fully populated with all the articles in the subcategories by nationality. The small subcategories would remain. The larger categories might have several thousand member articles. Is there a technical reason to limit categories? Do large categories slow up the servers? Should we worry about this at all, or make these decisions totally based on what makes for the best category structure? Radiant seems to think that developers don't want the categories to have more than a few hundred articles. I have not heard anything about this recently, and am wondering if this concern is just left over from the time before we had Category TOC's. Thanks. -- Samuel Wantman 01:34, 24 January 2006 (UTC)

Check out Commons for some examples of reeeeallllyyyy big categories. They kinda suck to navigate, but *shrug*. --Brion 01:59, 24 January 2006 (UTC)

So we just need to worry about the ability to easily navigate, and not worry about server loads? -- Samuel Wantman 02:34, 24 January 2006 (UTC)

And, if there's no load worry, can you think of a TOC solution other than a big, ugly 2-dimensional TOC (see User talk:Rick Block#Huge categories)? -- Rick Block (talk) 05:36, 24 January 2006 (UTC)

There's another thing you could help clarify. On Commons there was this big vote about whether to use categories for everything or regular pages for everything, and it just kind of floundered. Jamesday made some comments that categories were absolutely evil and caused lots of server load. Are there are any plans to modify the way categories are computed/cached so we could use them (maybe some kind of meta-data category tagging thing)? Are there any plans to make a combined category/article like Duesentrieb's proposal? — Omegatron 05:45, 24 January 2006 (UTC)

Those comments were from last March. I'm wondering if anything was rewritten since then. That discussion was also comparing a page view with a category view. I did not get a sense from that discussion that there would be any difference between viewing a large category or a small one. If someone is browsing through a large category instead of several subcategories it might actually take less resources because the person would not be going back to the parent category between each subcategory load. -- Samuel Wantman 07:04, 24 January 2006 (UTC)

Short answer: large categories are a pain in the ass to work with; a decent user interface for them that won't be hard on the server is hard to work out. The current limitations on display are quick ugly hacks to keep it from being a server burden, but there's been little further work on it since. --Brion 08:34, 24 January 2006 (UTC)

Perhaps I'm looking for the long answer. Let me be a little more specific. With the current interface is there much difference in the server load if categories are 200 articles, 2000 articles or 5000 articles? If we want to re-populate some categories with ~4000 articles; are you saying this would detrimental to the performance of Misplaced Pages and we should not do it? -- Samuel Wantman 10:41, 24 January 2006 (UTC)

Is it more of a problem if there are a large number of images in the category? Would it maybe be sensible to use a different, lower, limit on the number of images displayed at one shot? HTH HAND —Phil | Talk 12:09, 24 January 2006 (UTC)

With the current interface navigating large categories suck. Improving the interface for large categories without making it harder on the servers is not a problem that has been solved yet. --Brion 21:12, 24 January 2006 (UTC)

Sorry to bug you about this. When you say "Suck", does that mean the categories are not fun to navigate, or does "Suck" mean that categories over a few hundred in size are not good for the servers? Is there CURRENTLY a reason OTHER THAN THE INTERFACE to limit the size of the categories? If so, what is the size of a category that is too big? -- Samuel Wantman 21:48, 24 January 2006 (UTC)

I don't know how many times I can say this. RIGHT NOW THERE IS NO ISSUE WITH THE SERVERS DUE TO LARGE CATEGORIES BECAUSE WE CHOPPED THE INTERFACE INTO A CRAPPY LIMIT OF A FEW PAGES AT A TIME, WHICH MAKES IT HORRIBLE FOR HUMANS TO TRY TO USE THEM. --Brion 22:46, 24 January 2006 (UTC)

So, at a technical level, the database query used by an Apache to retrieve the nth clump of 200 entries doesn't result in the database retrieving all the entries off the disk and then transmitting only 200 entries back to the Apache, but rather the DB query limits the internal result set to 200 (right?). This means the load on the DB is restricted to queries involving 200 entries at a time, but this limit shows through to the UI. I think Samuel is just looking for a warm fuzzy that there aren't any DB issues related to the sheer number of entries in a category. BTW - can you think of a way to do a 2-dimensional TOC other than a big ugly 26x26 table (like a cascading pop-up menu)? Thanks. -- Rick Block (talk) 01:47, 25 January 2006 (UTC)

The paging is done based on points in the sort index. This has known problems with breaking if you have a lot of similarly-sorted pages, but is cheap on the servers. --Brion 03:00, 25 January 2006 (UTC)

Thanks for your help Brion. Sorry if it seemed like I was bugging you. Without your definitive answer I'm sure there would be arguments saying that categories MUST be broken into smaller subcategories because otherwise it puts a burden on the servers. It is now clear from your answers that this isn't the case. Thanks for your help. -- Samuel Wantman 02:01, 25 January 2006 (UTC)

Great. :) --Brion 03:00, 25 January 2006 (UTC)

Thanks

Thanks for your help with the search problem...I was beginning to feel really dumb not being able to do a simple search for a page properly! Any idea when the search index will be brought up to date? Is it a server load problem? bcatt 06:47, 31 January 2006 (UTC)

The current search is updated from the data dumps; data dumps weren't being made for a while while I was fixing the system. When the current run is finished, search gets updated. --Brion 20:08, 31 January 2006 (UTC)

MW Dev Map

I saw you listed your location as Santa Ana. What are you doing here, of all places? -- NGerda 06:58, 31 January 2006 (UTC)

I... live here? --Brion 20:05, 31 January 2006 (UTC)

WP:AUM

I'd like to know whether you think this version appropriately describes the reasons why meta-templates should be avoided, while more fairly describing the server impact. I respect that there is no evidence of actual impact, but in theory, extra database calls that can be avoided, should be. The bulk of it now concentrates on the "ugliness and fragility", as well as complexity from the perspective of non-geeky template editors. Let me know your thoughts. -- Netoholic @ 07:13, 31 January 2006 (UTC)

Actually, here is the version I'd like you to review. I added a section called "Template links". Would like to have your feedback. -- Netoholic @ 00:38, 1 February 2006 (UTC)

Lilypond

Hello, Brion. Is there a plan to enable the Lilypond extension on English Misplaced Pages in the near future? Do you need help testing it first (for instance on Wikisophia)? Regards, RobertG ♬ talk 15:27, 31 January 2006 (UTC)

No immediate plans. I'm still recommending that extensions based on shell-outs be reworked into a daemonizable form so we can isolate these large third-party programs. --Brion 02:02, 1 February 2006 (UTC)

IE7 / Misplaced Pages logo

You don't need Vista for IE 7, it is already out

~ Cheers —This user has left wikipedia 20:46 2006-02-01

It's not out, but they finally released a public beta for XP. (I've been on the closed beta program since July.) --Brion 21:27, 1 February 2006 (UTC)

Oh I see. Does the closed beta identify itself differently than the preview?

~Cheers —This user has left wikipedia 22:02 2006-02-01

Probably a different build number or something, but I haven't checked. --Brion 23:43, 1 February 2006 (UTC)

Java script

I am trying to use godmode light to get an additional "vandalism" rollback button. However, it only appears on the contributions pages, not the diff pages. I am a sysop, and that is interfereing with the script so that only the sysop rollback button shows on diff pages. What excatly in my monobook.js code is causing this? How can I fix it? Thanks.Voice of All 04:35, 2 February 2006 (UTC)

Got it, nevermind. Although I have to have it above the Current Version link. But still good though.Voice of All 05:55, 2 February 2006 (UTC)

The future of qif

The constant pressure for that missing functionality in MediaWiki lead to the invention of template:qif, which despite its ugliness is a very ingenious trick that uses the current MediaWiki template system in ways that its inventors possibly never have dreamt of. Some wikipedians sheer got sick in the second they first looked at the wiki code of it, other more pragmatic souls read that kind of wiki source like a heading in a newspaper.

Driven by a very desparate wikipedian and a group of admins that – shocked by the ugliness of qif – slapped a policy tag on WP:AUM, we have taken great lengths to look into ways how to remove qif without breaking a lot of templates. We tried hard, but we failed.

Despite its ugliness, template:qif is the closest mockup for a conditional that we have now. However ugly this might be, it was great to see how much more could be done if we had such a darned thing in MediaWiki. Thanks to this, it spread very fast into the actual wikisource of templates and thousands of articles on the en wiki depend indirectly on it today. It has thus been protected to prevent DOS attacks.

As it is now clear that we will have conditionals in MediaWiki sooner or later, is it really needed that we go and destroy each and every template that uses qif today (as mandated by the not so popular WP:AUM)? Can't we not simply continue to use that qif as a stop-gap measure?

The two currently known alternatives either break accessibility (Misplaced Pages:hiddenStructure) or just move the ugliness currently contained in qif into the higher level templates ("Weeble" code).

Brion, it would be very helpful if you could answer this question here in public as there is an increasing unrest among the editors around that WP:AUM.

Can we forget that WP:AUM hell for now?

Thank you for your careful consideration. --Adrian Buehlmann 23:55, 2 February 2006 (UTC)

• rolleyes*

I don't care if you use it for now. It's ugly. Use it if you must for the moment, but it'll probably break when we redo templates and add proper conditionals. --Brion 02:52, 3 February 2006 (UTC)

If you intend to break m:Help:Parameter default it would be wise to issue a deprecate notice in advance. qif is not the only thing that depends on this. Same holds true if you intend to limit inclusion levels on templates. Deprecation notices are generally a good idea when removing existing functionality. If it boils down to removal of conditional template inclusion I can say that we are able to go without that as we have seen that we can do nearly everything we need with Misplaced Pages:hiddenStructure, which does not provide conditional inclusion for obvious reasons. The bad thing with Misplaced Pages:hiddenStructure is that it depends on the hiding capability of CSS which breaks on non CSS browsers and older screen readers. The html from this just looks ugly. But it has the plus that the servers are not used for this kind of hiding. Well, I thought that it would be not such a big deal to implement a simple hiding function in MediaWiki. The toughest part is sure the syntax. If needed we could even use the Misplaced Pages:hiddenStructure and strip out the hidden part server side. But as everything on the code side, this is up to you. We take whatever you give us :-). --Adrian Buehlmann 07:47, 3 February 2006 (UTC)

Note that 'hiddenStructure' also relies on m:Help:Parameter default, so anything which 'breaks' qif is likely to break it for most users (it is already broken for some) as well. Parameter defaults are also just a useful feature in their own right. They have been extended to allow conditional logic in general, but the original purpose of allowing an unset parameter to default to a given value is used in many templates and would require special conditional logic to replicate each instance if the defaults were removed. --CBD _☎ 12:12, 3 February 2006 (UTC)

If hiddenStructure breaks, that's OK. All it means is that some hidden infobox rows become visible - and that should be no big deal. With {Qif}, due to the intricate conditional relationships, it's hard to imagine how many ways it could break things. -- Netoholic @ 20:26, 3 February 2006 (UTC)

Sorry for marking the Nevada map for deletion

Good thing you reverted it. I just saw something really weird as the only contribution of a newbie and forgot to check the history. Hope you can find the image again. Dr Debug (Talk) 05:22, 3 February 2006 (UTC)

That's ok... it was the only reason I noticed that someone had deleted the actual image four months ago. :) --Brion 05:31, 3 February 2006 (UTC)

can autoblocking be made optional?

Hello,

Would it be possible to modify the <form> in Special:Blockip to add a checkbox that would control whether or not the underlying IP gets blocked when a registered username is blocked? I would even suggest that the default should be unchecked (autoblocking turned off). Vandals have figured out how to use autoblocking for the purpose of denial-of-service attacks. For instance, they create accounts with names like "Block Me" which I suspect are for no other purpose (but if you ignore them, then they go ahead and use that account for vandalism, eg ). -- Curps 03:39, 5 February 2006 (UTC)

There are several possibilities for this stuff which, hopefully, we'll get hacked out and decided on soonish. --Brion 04:59, 5 February 2006 (UTC)

Image:California map showing Alameda County.png listed for deletion

Image:Nevada map showing Carson City.png listed for deletion

dbenbenn | talk 15:08, 12 February 2006 (UTC)

Weird login on it.wiki

Hi Brion, as I told you on IRC, we have a strange case on it.wiki with it:Utente:Lucas. The new user log says he registered in January 2006, but he appears to have contributions in 2003. Thanks, --Cruccone 22:02, 12 February 2006 (UTC)

You = God

That is all. Rory096 21:42, 13 February 2006 (UTC)

Thumb up

I hereby give you thumb up for your work today! →AzaToth 21:43, 13 February 2006 (UTC)

YOU ROCK!

U —WAvegetarian•_{CONTRIBUTIONS}• •21:50, 13 February 2006 (UTC)

Yay! It's fixed!

Have yourself a beer, you deserve it! the wub "?!" 22:16, 13 February 2006 (UTC)

Thank you

I don't happen to have any better images handy, so I hereby award you this pair of Republican congressmen, in appreciation of your hard work today.

Sorry.

Thanks again. --James S. 04:17, 14 February 2006 (UTC)

Thanks

A very personal thank you for restoring funtionality quickly. Id give you an image if I had one to spare. But thanks for making wikipedia a great place to be! You are truly a superior human being. Dan, the CowMan 03:26, 15 February 2006 (UTC)

Note

I have proposed a new policy at Misplaced Pages:Semi-block, and request input as to the feasibility of including such a feature in MediaWiki, should the community accept it. Thanks! —BorgHunter (talk) 20:00, 15 February 2006 (UTC)

Meta:Interwiki map

Can you add the text

|-
| PMID || http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=$1

to Meta:Interwiki map? Superm401 - Talk 03:06, 16 February 2006 (UTC)

Template substitution

Would you lend an opinion to Misplaced Pages talk:Template substitution?

The page lists a bunch of templates which should always be substituted, and refers to vandalism and server load as the rationale.

I think substituting to prevent vandalism is silly. Transcluding templates is far better. Template vandalism can be reverted just as easily as article vandalism, and will be done a lot quicker due to the higher visibility. If the templates are substituted while vandalized, it is much more difficult to clean up.

As for server load, you've said we shouldn't worry about templates and "server load" at a policy level. — Omegatron 17:04, 17 February 2006 (UTC)

Image:Downtown Los Angeles at twilight.jpg listed for deletion

This image is an orphan -Nv8200p talk 17:48, 17 February 2006 (UTC)

Urgent CheckUser request

Please see WP:RCU regarding User:Bowlhover. Thanks. --Nlu (talk) 08:05, 21 February 2006 (UTC)

Patrolled edits request

Brion, over at the simple English Wiktionary, we've decided to try having patrolled edits for sysops. 100% of the administrators (a whopping four) and the sole bureaucrat (pretty much the only active users) have decided to give it a try: see wiktionary:simple:Wiktionary:Administrators#Patrolled_edits?. I've read Help:Patrolled edit, and it seems like it requires a developer to enable this for sysop-patrols (because I can't seem to find LocalSettings.php...) . Would you mind taking care of this? Many thanks! Flcelloguy (A note?) 20:40, 21 February 2006 (UTC)

WP:RCU

I am leaving this message to all 10 people at Special/checkuser list. Therefore forgive me for its being impersonal. Please take a look at Misplaced Pages talk:Requests for CheckUser#cleanup needed. Your response and/or actions there would be very much appreciated. Thanks! --Irpen 23:25, 24 February 2006 (UTC)

Message posted on Wikitech list

Yesterday I tried to post on the Wikitech list, using gmane. At 10h47 UTC I received the following message : You are now authorized to post to the gmane.science.linguistics.wikipedia.technical newsgroup. The original message you sent to the newsgroup will be posted within ten minutes. but obviously it has not been posted.

Here is approximately what I wrote :

Teofilo talk 11:36, 4 March 2006 (UTC)

If you're not subscribed to the list your post will have been automatically discarded.

1) Hypothetically possible, but annoying.

2) Well, I for one didn't get any wikipedia-related email for a full day which rather cut into my ability to work. :) Spamcop is one of several blacklisting services out there, and many ISPs and mailers use them and others in their automated spam catchers. --Brion 12:08, 4 March 2006 (UTC)

Templates and variables

It's possible to create a template with the same name as a variable, and the invoking syntax is the same (eg, the "CURRENT" variables). Just today a vandal tried this. I trust the hardcoded variable takes precedence over any template with the same name? If not, it's a problem. -- Curps 07:52, 5 March 2006 (UTC)

Current time and local time

Brion, There is {{CURRENTTIME}} which displays 18:56 UTC, but is there anything that can be used to display a user's local time? Better than that, is there a way to find out a user's time zone offset? I've been wondering about templates that would handle differences in British/American English based on the time zone of the user. Is this at all possible? Are there any plans to deal with this in future software? I'm hoping for a user set flag in preferences that could be referenced by a simple template, for example {{Eng|categorise|categorize}} would choose the two alternate words or phrases as parameters based on the flag. One template could be used to fix every article and put an end to the never ending spelling battles and arguments. Thanks. -- Samuel Wantman 08:23, 5 March 2006 (UTC)

Material that goes into page text has to be the same for all users, so no you can't show the user's timezone selection in text. --Brion 22:40, 5 March 2006 (UTC)

Templates in signatures

• I was pretty sure that templates as signatures were certifiably banned and Bad, but now that I look, I can't find any reasoning behind this except server load.
• On the other hand, you've explicitly said that we shouldn't worry about server load incurred by templates on a policy level.

So are user signatures a special case? Do they cause a significant load problem? How do we know which things are special cases or not, so that we don't have to ask you directly about each thing? — Omegatron 06:42, 10 March 2006 (UTC)

Signatures aren't quite the same as regular content; instead of being carefully placed by humans editing on articles they're dumped everywhere you make a comment on a talk page. We've had many many problems with people putting broken markup and such into their signatures, then complaining when a bug in the software is fixed which exposes their error and damages the display of thousands of pages.

Keeping non-trivial markup out of signatures keeps things simple and non-breaky.

Actual real content which is relevant to Wikimedia's mission will take whatever resources it takes.

Pretty signatures aren't part of our mission and there is zero commitment to them. --Brion 19:31, 10 March 2006 (UTC)

This isn't about keeping non-trivial markup out of signatures, though; I'm already in complete agreement with you on that. (See how complex mine is?) :-) This is about whether transcluding them with templates is harmful and why it is banned.

Actually, your comments are sort of relevant, though. If someone used a "markup trick" to make their sig do something, and it later damaged thousands of pages when the software was changed, wouldn't it be better if all of those broken signature instances were from a template? Then you could fix all of those pages in one edit. Currently, signature templates are banned (I don't know why), so if this occurred, all the broken pages would need to be fixed by hand. I would think that using templates for signatures would actually be better for this reason.

Also, talk page source code would be much cleaner and easier to navigate through. Complex signatures would look like this:

{{User:Wikipedian/sig}}

instead of this:

[[User:Wikipedian|<span style="font-size:150%;">W</span><span style="font-size:140%;">i</span>

<span style="font-size:130%;">K</span><span style="font-size:120%;">i</span>]] <span style="font-size:110%;">P</span><span style="font-size:100%;">dee</span> <span style="font-size:90%;">u</span><span style="font-size:80%;">N</span>

<span style="font-size:70%;">n</span>^{<span style="font-size:90%">]</span>} (Ick. Copied from a real sig. Username changed to protect the... guilty.)

The only downside I see would be server load, but you have said this isn't really an issue, so I'm confused as to why they're prohibited. — Omegatron 00:31, 11 March 2006 (UTC)

As I said above, they can be disruptive. Everything bad that's been said about templates is true times a hundred for signatures, as 1) THEY SERVER NO USEFUL PURPOSE and 2) THEY MULTIPLY INCREDIBLY WITHOUT GIVING ANY USEFULNESS BACK. This means that THERE IS NO JUSTIFICATION TO USE THEM AT ALL and therefore ANY LOAD PROBLEMS, CONSISTENCY PROBLEMS, CLEANUP PROBLEMS, VANDALISM PROBLEMS, ETC ARE MUCH MUCH MUCH WORSE FOR SIGS THAN FOR ACTUAL REAL CONTENT WHICH WE ACTUALLY WANT. --Brion 01:03, 11 March 2006 (UTC)