[00:40:45] ohey Gloria [00:40:53] I found your sekrit page [00:42:32] https://bugzilla.wikimedia.org/show_bug.cgi?id=15889 [00:47:27] Withoutaname: Sekrit page? [00:49:32] yes, sekrit page [00:50:54] i haz found ur sekrit page! [00:59:46] It's so secret I don't remember where it is. [01:08:08] hmm, hard to find an image that doesn't exist [01:26:07] What happens if you move a LiquidThreads page? [01:27:54] Inception. [03:29:42] fhocutt: the performance guidelines are OUT and I am flopping on the couch :) [03:33:19] Also I think it is cool that the new head of the Wikimedia Foundation wrote her blog post in English + Russian https://blog.wikimedia.org/2014/05/15/hacker-osmosis-ideas-european-hackathon-zurich/ [04:15:42] Hello #mediawiki, if around 90% of page_id's (for example, 2000-70000 while the latest page had id 75000) in my Wiki database are unoccupied (caused by `Delete default messages'), will it be advisable to shrink the id space in order to ease future maintaince? [04:24:08] Morning people! :) [04:28:47] CasperVector: No. [04:28:56] CasperVector: There are plenty of integers left. :-) [04:29:29] Though it sounds strange that such a large block would be unoccupied. [04:30:37] Gloria: Thanks... I was concerned about that because rebuildall.php always wastes quite a lot of time process these blocks :) [09:18:59] I have a question I am using the MediaWiki software version 1.22.6 and I would like to get the Module Namespace, how do I go about getting it. By module namespace like ones that Wikipedia has. [09:25:56] Never mind, figured it out./ [10:55:29] eek http://mappings.dbpedia.org/index.php/Special:Version [15:27:51] hello, is it possible to limit permission "protect" to only allow semi-protecting? [15:29:05] guest9910: Well you could give editprotected rights to more people, which would weaken what full protection does [15:29:15] guest9910: or you could try messing with $wgRestrictionLevels [15:29:25] !wg RestrictionLevels | guest9910 [15:29:25] guest9910: https://www.mediawiki.org/wiki/Manual:%24wgRestrictionLevels [15:29:45] Will mediawiki 1.21 absolutely not work with PHP 5.2.17? I'm on a Dreamhost shared server where I can't easily upgrade the PHP binary. [15:34:46] JordiGH: I think we use closures from php 5.3 in mediawiki 1.21 [15:35:13] Hm. Well, shit. [15:35:26] Looks like I also need to manually update the schema. [15:35:53] https://www.mediawiki.org/wiki/Thread:Project:Support_desk/upgrade_problems:_1.11_--_1.21 [15:36:17] JordiGH: Yeah, some things in the MessageCache will need php 5.3 [15:36:28] * JordiGH wants to link to a thousand web rants about upgrading and the sorrows of computers and life. [15:37:23] JordiGH: That bug was fixed in newer releases of the 1.21 branch, you shouldn't have to worry about it [15:37:37] (Also, how do we not have some sort of automated test for that sort of thing. *sigh*) [15:37:48] Wait, so if I upgrade from 1.18 to 1.21 following the upgrade manual, I don't need to worry about that? [15:38:21] You should not need to worry about https://www.mediawiki.org/wiki/Thread:Project:Support_desk/upgrade_problems:_1.11_--_1.21 [15:38:30] It was fixed in https://gerrit.wikimedia.org/r/#/c/76627/ [15:41:23] Oh, neat, there actually *is* a newer PHP in /usr/local. [15:41:56] So... unless the webternets are lying, all of this should be easy to do. [15:42:58] JordiGH: Keep in mind, what matters is what version of php the webserver uses. Often this is totally different from the command line version of php [15:43:26] Yeah, I gotta figure out how to do that... [15:44:23] Create a file with a .php extension, containing only " Then view it via the web [15:45:07] Or just try and install MediaWiki and see whether or not it works ;) [15:48:28] Ooh, neat, already using php 5.3 [16:04:27] Hm, do extensions do releases? [16:04:38] Or do they all just "release" from git? [16:05:18] I guess I can clone the git repos and look for tags. [16:06:56] * JordiGH sees no tags. [16:07:04] Alright, I guess there are no releases, then. [16:10:24] JordiGH: They usually have release branches [16:10:37] e.g. a branch named REL1_21 [16:10:54] Well, that's weird. Isn't that what tags are for? I see the branches now. [16:13:02] Well the branches aren't really static (in theory) authors sometimes backport stuff [16:13:25] Then you use a new tag with a new revision number. [16:13:29] Oh well. [16:14:23] I think I'm at the right revision, if my gitoloogy is correct. [16:15:09] git checkout origin/REL1_21 [16:15:16] usually works pretty well [16:17:48] I always get confused between that and git co REL1_21. One creates a local branch and the other doesn't? [16:19:17] Wait, the RELEASE-NOTES-2.0 file for the Math extension at revision REL1_21 say this is not a release. [16:19:18] Hm. [16:19:25] Should I checkout an earlier branch? [16:23:05] I think you should just ignore that warning [16:23:23] unless it doesn't work [16:28:57] I don't want to run stuff that devs think is unstable. [16:31:56] Blegh, I can't figure out which version I'm supposed to get for mediawiki 1.22.6 [16:32:03] https://git.wikimedia.org/branches/mediawiki%2Fextensions%2FMath [16:32:17] I guess REL1_22 ? [16:32:25] yes [16:32:33] Ah, that would make sense for why REL1_22 had that scary warning on it. [16:32:40] Er, REL1_23 had the warning. [16:32:58] most of the time. Versioning is up to individual extension maintainers. Some do it better than others [16:34:05] Keep in mind that the math extension is used on wikipedia. So its supposed to be kept "working" at all time [16:37:14] They're not allowed to break the Math extension even in a git branch for a future release? [16:39:36] They're allowed to do so in a branch. They're not allowed to in the master branch, and it would be considered very bad form to break it in one of the REL1_XX branches [16:40:32] About once a week, the code in the master branch of all the extensions gets deployed to wikipedia et al [16:40:56] (Well all the extensions wikipedia uses. Extensions that wikipedia does not use can do whatever they want) [16:42:05] hi [16:44:00] bawolff: You've been helpful. Got a bitcoin address I could send a small tip to? [16:44:38] Sorry, haven't quite joined the bitcoin revolution yet, but thanks for the offer [16:45:10] It's just a matter of downloading e.g. electrum. It's packaged for Debian. [16:45:39] If you prefer, I can also send you paypal money, but I'm trying to stop supporting their business. [16:54:20] JordiGH: If I did that right, 1CvTSrXWqfsEVFzuFVxHgDtk77jqHfpA7A [16:55:13] Sent. Thanks. [16:55:25] Cool thanks [17:00:56] Don't mean to be an evangeliser, but I can't help myself: if you've got questions about how bitcoins work, let me know. I'm probably too enthusiastic about them for my own good. [17:03:15] I'm a fan of the general concept, and have read up on it before, I've just never experimented with it [17:14:40] Wow, purging the cache is taking a loong time. [17:14:47] I hope our host doesn't kill the MySQL process. [17:15:14] I think it's a huge table. I hope maintenance/update.php knows how to do this cleanly. [17:26:25] Hm, it's just running DELETE /* DatabaseUpdater::purgeCache */ FROM `objectcache` [17:26:28] This is very slow. [17:26:49] TRUNCATE TABLE should be faster [17:27:05] /should/would/ [17:27:25] So, interrupt the maintenance/update.php script and manually clear this table? [17:27:32] With TRUNCATE TABLE? [17:27:58] Are those C-like comments? I didn't know MySQL could do comments like that. [17:27:58] if you want it to end fast, yes :P [17:28:20] yes, those are SQL standard comments [17:28:52] MediaWiki does funky things with the comments in sql files, to replace them with various variables [17:45:41] Okay, this is taking way too long. I'm gonna try truncate table. [17:46:02] I hope update.php knows how to recover from an interrupted operation. It seems to check the schema. [17:47:39] I don't know why it doesn't use TRUNCATE TABLE in the first place. Maybe for compatibility with other RDBMS that do not support that? [17:48:44] But doesn't it know the db engine? [17:50:07] Blegh, now mysql is taking a long time to kill the bad DELETE query. [17:51:01] yes, but it was introduced in the SQL:2008 standard, so it was probably chosen to use the statement that works for all instead of adding exceptions [17:51:10] call it lazyness ;) [17:52:23] Looks like I was too optimistic: http://octave.1599824.n4.nabble.com/Wiki-is-down-for-maintenance-td4664129.html [17:54:35] JordiGH: you can use the KILL command of mysql http://dev.mysql.com/doc/refman/5.0/en/kill.html [17:55:00] Yeah, already killed it, but now it's stuck at "freeing items". [17:55:07] Can I killdashnine it? [17:55:17] I guess even if I could, I wouldn't want to do that. [17:58:50] yeah, it's frustrating when things go clusterfuck like this... [17:59:03] seriously, is *that hard* to stop the current deletion? [18:00:13] I guess that mysql is doing the right thing in ensuring the database is coherent, so if you stop a deletion and it started deleting things, it's probably restoring the current deletions to leave it to the original state [18:00:39] Oh, you thing it's being all ACIDy? [18:00:44] think [18:01:40] yes [18:01:53] In this case, I guess I don't want that. [18:02:03] DELETE EVERYTHING NO BACKUPS RAAWR [18:02:45] that's what TRUNCATE TABLE does [18:10:38] There, finally done. [18:17:27] Can someone help me out real quick? [18:17:34] I'm trying to add a namespace to my private wiki [18:17:44] I'm not sure I really understand the docunmentation [18:18:32] trixy: which part don't you understand [18:18:39] !wg ExtraNamespaces [18:18:39] https://www.mediawiki.org/wiki/Manual:%24wgExtraNamespaces [18:18:40] Okay, looks like it's backup. Pageloads are slow as balls, but I guess that's gonna happen until the cache rebuilds. [18:18:45] back up [18:19:30] So I defined a custom namespace with $wgExtraNamespaces [18:19:39] And that goes in LocalSettings.php? [18:19:45] yep, at the bottom [18:20:23] And then I went to my wiki and added a page in the namespace like so -> urlpath/Namspacename:NewArticle [18:20:40] Added some test text [18:20:40] Then went to urlpath/Character [18:20:44] Shouldn't it list articles in that namespace? [18:20:58] :S [18:21:29] Does anyone want to try to break my new wiki setup? http://wiki.octave.org [18:22:07] trixy: Try looking at Special:Allpages - at the top there should be a drop down list, does it include the new namespace? [18:22:16] yo [18:23:04] JordiGH: you may want to change MediaWiki:Mainpage, to point to your actual main page [18:24:16] bawolff: It includes the new namespace and when I select that namespace it shows me the test article I created [18:24:24] Or a link to it rather, durr [18:24:49] trixy: Cool, that means you succesfully set up the namespace [18:24:56] Vulpix: Just do a redirect? [18:25:40] JordiGH: The text of the page MediaWiki:Mainpage on your wiki controls what the wiki thinks is the main page [18:25:58] http://wiki.octave.org/MediaWiki:Mainpage [18:26:32] bawolff: Gotcha. Is urlpath/NewNamespace supposed to show me the list of articles in it's namespace though? [18:26:38] Or is that not the default behavior? [18:26:45] no, that's not the default behaviour [18:26:57] bawolff: Ah, okay. That must be what I was misunderstanding [18:27:01] Thanks :) [18:27:07] You could create that page with the text {{Special:AllPages/NewNamespace:}} and then it will [18:27:20] I'll try that! [18:27:44] Yay, math works: http://wiki.octave.org/User:JordiGH [18:27:47] I think that used to be broken. [18:28:20] JordiGH: you might want to fix your shorturls to exclude http://wiki.octave.org/robots.txt [18:29:16] Hm, does mediawiki have an annotate function? I want to know who was the idiot who changed a particular line. :-) [18:29:57] alas no, we don't have an equivalent to git blame [18:30:40] bawolff: That's what I needed, thanks! Everything is in order now. [18:30:44] I appreciate your help. [18:30:50] no problem [18:36:38] Aha, I found my idiot. By manually bisecting on the diffs. :-/ [18:37:19] * JordiGH revs up his flame thrower. [18:39:41] JordiGH: you're everywhere :D [18:40:04] Hm? What do you mean? [18:40:16] ive seen you in a lot of places [18:40:44] Ah, really? Probably an unhealthy obsession of mine with the internet. [18:40:54] lol [18:40:56] lol [18:42:01] atleast you admit i [18:42:03] t [18:42:22] Don't we just run into each other in various IRC channels? [18:42:35] If so, I blame you for being in them. [18:43:10] lol [18:43:19] sure my fault [18:45:49] you're playing ith the wiki? [19:11:56] Does anybody know the name of the "class" for dropdown menus regarding input? Such as "Select a state", which is then followed by a dropdown menu full of options. [19:13:02] I've been developing an extension that adds a "Birthdate" function, which allows the user to enter a birthdate upon signing up (Which can be used to determine what the user can [19:13:05] Elvana: select boxes? comboboxes? [19:13:14] (assuming you mean the but browser support for that might not be great [19:31:44] I'm trying to define this all this within the executiong UserCreate class - Mistake? [19:31:59] *execution [19:32:19] So-so sorry, I'm a bit out of it. [19:37:04] does abusefilter work with php regex rules or its own regex [19:37:55] Ah, forget my previous question. I'm just going to go the old fashion way and do it all in HTML. [19:38:54] Though I do have another question, just looking for opinions: Would it be batter to store the date of birth as an integer, or as an actual date? To be honest I'm not so sure how the "DATE" type in SQL works. [19:39:00] *BETTER [19:42:32] anyone work with the abusefilter before? [19:43:36] I have. [19:43:45] What are you looking for exactly? [19:44:09] just wondering if it excepts php regex [19:45:12] I believe it uses pcre regex syntax like commonly used in php [19:45:27] Withoutaname: it uses PHP regexp [19:45:41] thanks [19:45:56] there's a link to PHP regexp reference on the manual about AF rules [19:46:39] how about /

[^\[]*http/i [19:47:19] Elvana: MediaWiki normally stores dates as a VARCHAR(14) in TS_MW (YYYYMMDDHHMMSS) format (e.g. see wfTimestamp function in GlobalFunctions.php as well as $dbw->timestamp() [19:55:05] bawolff, do you know where i can test the abusefilter rules [19:55:50] https://test.wikipedia.org/wiki/Main_Page I suppose (provided you don't do anything too drastic and stop everyone from editing) [19:56:41] im only going to log [19:57:12] wait, how do I use testwiki [19:57:41] as admin you also have [[Special:Abusefilter/test]] [19:59:34] se4598: does that show deleted edits [19:59:59] don't know, I'm not an admin :P [20:00:14] Withoutaname: What's your on wiki username? [20:08:04] does abusefilter use delimiters or just quotation marks [20:10:13] Withoutaname: quotation marks [23:15:01] how does {{CASCADINGSOURCES}} work [23:24:35] magic [23:27:44] Withoutaname, read the source and find out? [23:28:22] To be honest, I didn't even know that was actually a thing [23:28:32] Krenair: yeah I read it, but I want to know if it was able to take a parameter [23:28:52] not a whole lot of things are cascadeprotected, so it's not used enough for me to see it in action [23:29:37] or maybe it's only used on a cascadeprotected page [23:29:49] in which case i can't find out since i don't have the perms [23:30:11] Withoutaname: Everything in includes/parser/CoreParserFunctions.php takes at least 1 parameter [23:30:28] The 0 parameter magic words are defined in includes/parser/Parser.php [23:31:33] In this case, Cascadingsources can take either 0 or 1 parameters. If it takes a parameter then that parameter is the title to work on, otherwise it defaults to {{PAGENAME}} [23:32:56] ok thnx [23:33:07] ill make a note on Manual:Magic_words [23:33:40] technically 0 parameter magic words are called "variables" and 1 or more parameters are called parser functions [23:34:02] yeah that's what they told me [23:34:19] most variables are used as paarser funcs anyway so the distinction was hardly useful [23:34:39] and then doubleunderscores are behavior switches [23:37:37] bawolff: I just tried {{CASCADINGSOURCES:Wikipedia:Main Page/Tomorrowā€ˇ}} on enwiki [23:37:42] from a user prespective, the distinction is kind of silly at this point. From an implementation prespective, its rather different (Which is also silly, but that's the way it is) [23:37:56] it didn't output anything, can you please check [23:39:02] try {{CASCADINGSOURCES:template:Did you know}} [23:39:13] and wow that's a silly output format. At the very least it should be a list [23:39:48] if its going to be a | separated string of data, then it should probably be only accessible to lua, as there's no real use for that [23:40:24] we could just change it to \n right? [23:41:26] I guess that's a little better [23:41:29] return implode( $names, '|' ); [23:41:34] yeah [23:42:21] hm, should we prepend "*" to the $names array [23:44:10] personally I'd like that a lot better [23:44:23] although I'm a bit late to the party to make criticism of it