[04:55:22] does $wgRCFeeds have an equivalent of $wgRC2UDPPrefix [05:01:37] petan : Thank you for assigning the task to me [05:01:56] Can you please guide me on how to carry on about it [05:02:13] I am talking about this : https://www.google-melange.com/gci/task/view/google/gci2014/5899133453860864 [07:35:35] petan : hi , I found that there are quite a few changes in the menu toolbar of Huggle.Do I have to document them as well ? for instance , the browser menu doesn't exist in the new huggle. [08:54:41] hey could someone please tell me how to find bug reports on the phabricator? [09:27:36] PK1: Search on https://phabricator.wikimedia.org ? [15:25:56] I'm having a bit of difficulty adding an API to an extension I'm working on [15:26:04] When I load api.php it just gives me this error [15:26:07] Fatal error: Allowed memory size of 134217728 bytes exhausted (tried to allocate 36 bytes) in A:\xampp\apps\mediawiki\htdocs\extensions\OreDict\OreDict.api.php on line 85 [15:30:06] ~ 128MB, that should be more than sufficient for MediaWiki [15:30:57] That line is in the getAllowedParams method [15:31:05] which simply returns an array of parameters [15:31:07] WalrusPony: check your code, maybe you have a recursive call somewhere that keeps loading objects in memory until it reaches the memory limit [15:31:50] the point where this exception is raised usually isn't the problem [15:32:48] Well, I certainly don't see any suspicious loops in the code https://github.com/Telshin/OreDict/blob/api/OreDict.api.php [15:34:31] It must be triggering a loop in the api documentation somehow, I just have absolutely no idea how [15:35:02] me neither :S [15:36:21] (╯°□°）╯︵ ┻━┻ [15:37:03] New plan, start ripping out code until it works [15:39:20] Vulpix: I copied the example from the API:Extension page, and it causes the same error [15:40:08] Wait, I had getDescription twice [15:40:29] ah, now it works, now to start re-adding the code [15:40:34] until it breaks [15:40:47] :) [15:47:21] ah ha, I figured it out [15:47:26] it was the needsToken thing [15:48:35] Okay, what about needing a token would cause an infinite loop? [15:53:27] Is there a way to list the most active users except Extension:Contribution Scores? [15:54:30] SQL query? :P [15:55:31] I thought of an unknown Special Page! :) [15:55:32] On most wikis, Special:ActiveUsers is enough :) [15:55:51] I need the total ;) [15:56:09] ? [15:56:31] The MOST active user from the beginning of the wiki. [15:56:41] hi, can someone explain me how to load a class, i added in /includes/password/ in mediawiki 1.24 [15:56:47] A list with the top contributors and the num of contributions [15:56:51] I added a corresponding $wgPasswordConfig in Localsettings, but i get an Clas not found error [15:57:23] And Extension:Contribution Scores happens to be super expensive [15:57:54] Could the API help? [15:58:19] Usually, inactive users are not considered MOST active users [15:58:31] So you're just looking for a list of all users by editcount? [15:58:59] yep. I'd only need the top 100 or 500 users. [15:59:10] I could hack Special:ActiveUsers tho. [15:59:21] Unlikely to be a good idea [16:00:22] Right, it's also used for stats etc. [16:08:11] likio: Add it to includes/AutoLoader.php [16:08:25] Subfader: I think https://www.mediawiki.org/wiki/Extension:UserDailyContribs might be your best bet [16:08:39] Uh, wait likio.. It looks like it changed in master [16:09:34] Subfader: because I don't see other extensions pulling user_editcount/getEditCount which are the only viable options to check all users [16:09:42] likio: autoload.php in the root of mediawiki, if you're using master [16:09:43] https://github.com/wikimedia/mediawiki/blob/master/autoload.php [16:10:03] Okay, if needsToken returns false everything works, if it returns a string, everything works [16:10:09] Except [16:10:09] / This file is generated by maintenance/generateLocalAutoload.php, do not adjust manually [16:10:11] Thanks! I might hack that instead :) [16:10:20] but if it returns either false or a string depending upon a condition, everything breaks [16:10:31] Am I not allowed to conditionally need a token? [16:10:37] You can [16:10:40] Depending on the condition [16:11:21] Has anyone seen superm401 recently? He is a mentor of my GCI task and it is due in 9 hours. [16:11:29] Reedy: Well when I try it just causes an out of memory [16:11:35] !seen superm401 [16:11:35] Did you mean @seen superm401? [16:11:36] superm401 (~matthew@wikipedia/Superm401) was last seen quitting from #wikimedia-tech 11 hours, 22 minutes ago stating (Client Quit). [16:12:28] Hmm, I guess I have to keep looking for him, he has not responded to the task in 27 hours :( [16:12:47] marcinlawnik: What do you need him for specifically? [16:12:50] Tried emailing him? [16:13:55] Reedy, I need a gerrit patch reviewed (Not necessairly him) and him or some other mentor if possible to mark my task as done. [16:14:17] He should have gotten notifications all over the place both from GCI and gerrit [16:14:26] gah, this makes absolutely no sense [16:14:58] If anyone could spare a few minutes of their time: https://gerrit.wikimedia.org/r/#/c/177269/ [16:18:04] Nemo_bis: It uses an own table which updates on ArticleSaveComplete. Thanks anyway! [16:18:37] Subfader: for everything? there is a line: 'totalEdits', $user->getEditCount() == null ? 0 : $user->getEditCount() ); [16:19:17] It might be easiest to alter Contribution Scores so that it uses user_editcount if asked for alltime stats [16:19:48] What do I need to do to make an API conditionally need a token? [16:20:01] So is doesn't calculate through ALL? Good idea! [16:20:02] marcinlawnik: i'll have a look :) [16:20:07] Or will I have to implement two APIs, one that needs a token and one that doesn't? [16:21:19] WalrusPony: you really shouldn't need to [16:21:29] You should just be able to put a conditional in the needsToken() [16:21:41] Reedy: But when I do that it gives me an out of memory error [16:22:01] return $this->isEditAction() ? 'csrf' : false; [16:22:06] that causes an out of memory [16:22:53] ah ha [16:23:15] I figured it out, isEditAction has to check the parameters to see if it is an edit action [16:23:39] but checking the parameters, inside the ApiBase implementation, checks whether it needs a token [16:23:44] which turns into a recursive loop [16:23:45] sneaky [16:39:12] /hotkeys [16:40:35] Reedy: thanks! [16:55:05] Reedy, added you as reviewer [16:55:35] marcinlawnik: I already had a look, but it's not something I'm really familiar with unfortunately :( [16:56:06] I sent superm401 a memoserv message, so he should get an email from it too [16:56:25] Reedy, Thank you very much ;) [16:56:57] Anyone here knows what happens if the task expires because mentor didn't check it? [16:57:39] qgil: ^^ About? [16:59:01] marcinlawnik: You'd hope the student wouldn't be penalised considering they have done the work already :) [17:00:19] Reedy, If the student does not complete the work they automatically get another 24h before it is returned back to the pool. Had ta\hat 2-3 years ago [17:00:51] Reedy, I just don't know what happens when the mentor is the one behind the schedule :( [17:01:20] Presumably the same or better [17:01:56] they actually got rid of the automatic extension this year. [17:02:10] Then the FAQ needs to be updated :( [17:02:17] (or so i think.) [17:02:41] http://www.google-melange.com/gci/document/show/gci_program/google/gci2014/help_page#values [18:06:56] Nemo_bis: does https://www.google-melange.com/gci/task/view/google/gci2014/5878174080565248 have a good description? I have ~10 more tasks like that [18:11:15] legoktm: why is the first link unclickable [18:11:26] o.O [18:12:02] Nemo_bis: fixed [18:13:15] "Examples should be sentences that could logically end with a colon" [18:13:19] * Nemo_bis getting confused [18:13:25] legoktm, Are you a code-in mentor? [18:14:05] marcinlawnik: yup [18:14:24] legoktm, Can you edit other mentor's tasks? [18:14:28] legoktm: I don't like "There is a proposal to use " [18:14:31] marcinlawnik: no; I can [18:15:07] hmm [18:15:08] Nemo_bis, do you know what happens if the mentor does not review the task and the time runs out? [18:15:32] marcinlawnik: yes, you tell me and someone gets slapped [18:15:47] Nemo_bis, https://www.google-melange.com/gci/task/view/google/gci2014/5793664760545280 [18:15:57] marcinlawnik: one minute [18:16:01] >24h ;) [18:16:38] legoktm: the description is ok in itself, but the acceptance criteria must be clear [18:17:19] legoktm: because 1) we don't want GCI students stuck in bikeshedding wars, 2) in case a patch is not merged within a reasonable time it must be clear enough even to an org admin whether the student has been reasonably diligent [18:18:27] legoktm: hopefully some tweaks to the docs are enough? moving some secondary considerations into an own section, or something [18:19:40] "Submit a patch to gerrit that properly converts the existing documentation to the new format, and marks methods that are no longer needed as deprecated." so they just have to convert what already exists, not necessarily improve it [18:19:46] yeah, /me does [18:19:55] and I commit to reviewing and merging these :) [18:20:03] legoktm: that helps ;) [18:20:24] marcinlawnik: so, Andre already added reviewers for you, but for the future https://www.mediawiki.org/wiki/Gerrit/Code_review/Getting_reviews is useful to keep at hand :) [18:20:55] Nemo_bis, yes, I know. I'm more concerned about the melange task [18:21:55] If the patch is merged, closing the task is almost automatic [18:22:34] @seen superm401 [18:22:34] marcinlawnik: Last time I saw superm401 they were quitting the network with reason: Client Quit N/A at 12/6/2014 4:48:38 AM (13h33m56s ago) [18:23:29] I can probably review the patch. [18:23:42] https://gerrit.wikimedia.org/r/#/c/177269/ [18:23:50] It's simple css + svg icons [18:23:54] in Echo [18:24:07] How many reviews do I need? [18:29:25] marcinlawnik: ok, I closed the task [18:29:36] Nemo_bis, thank you [18:29:38] The description wasn't clear enough IMHO, not your fault [18:34:06] marcinlawnik: let me know if you need help finding the next task :) [18:34:19] Anything you can offer? [18:39:20] Nemo_bis: updated https://www.mediawiki.org/w/index.php?title=API%2FArchitecture_work%2Fi18n&diff=1305339&oldid=1305337 and added the acceptance criteria to the task description [18:45:34] Wasn't there a user pref to select "ns to be searched by default"? Can't find it on mw org [18:55:13] marcinlawnik: I mentor three kind of tasks, you can find them in open tasks [18:55:41] marcinlawnik: if there isn't anything of your interest, you can tell us what would be interesting for you :) even better if you find a bug filed in phabricator [18:58:47] legoktm: ok, approved :) how many can you manage at once? [18:59:21] I have no idea. [18:59:58] 2 maybe? I also have 2 open MM tasks that have been claimed, but haven't heard anything from the students about it :/ [19:00:03] marcinlawnik: for instance, https://www.google-melange.com/gci/task/view/google/gci2014/5004830263214080 is rather straightforward [19:01:16] legoktm: yeah, I'm often wondering what to do with those silent tasks :) I wish we had IRC nicks or usernames ;) [19:01:36] Neo_bis : I was wondering if wikimedia has any sort of python tasks except of that of the pywikibot [19:02:21] AnonGuy: that's most of it [19:02:44] We could find some python stuff for you though, if you're interested [19:02:55] legoktm: let me know when you add another task of those then [19:03:41] I was just going to create tasks for all of them now, and then they can be approved when the other ones are completed? [19:03:51] yeah , I mean I know a bit of coding in python and was wondering if I could contribuite that ways ! [19:04:21] pywikibot seems to be a bit tough for a beginner [19:04:26] legoktm: ok; how many are they? [19:04:50] AnonGuy: some things in pwb are easy enough even for me! ;D [19:05:27] Let's see if they have easier bugs [19:05:51] umm okay, I'll be happy to do whatever I can ! [19:05:51] 9 so far, I could probably find more [19:06:58] AnonGuy: like, anything interesting here? https://phabricator.wikimedia.org/search/query/uQdX1qGf9UVc/#R [19:07:34] legoktm: I ask because if needed we can mass-create them with CSV import [19:07:39] :oo [19:07:48] that would be much better than me manually creating them :P [19:08:32] see email [19:08:37] Nemo_bis : I'll look into it and get back to you in 10 mins [19:08:44] :) [19:09:50] ok, I'll create a csv after food [19:10:01] much better than using melange :D [19:18:23] Hi! Is it possible to find non-WMF MediaWiki related job? I am unable to find such job for a long time. Seems to be very rare job. [19:19:10] I worked with MediaWiki for years, yet in last years switched to various different PHP engines. [19:20:41] Nemo_bis : I have no clue what any of those mean without being able to see any of the code they are talking about [19:20:47] how do I see the code ? [19:21:04] AnonGuy: should all be in the same pywikibot/core or pywikibot/compat repos [19:21:05] QuestPC: there's always Wikia, and various people running MW sites that need maintenance and such [19:21:42] QuestPC: there seem to be many "consultants" on odesk, for instance, but I have no idea what's the total size of the market and what the trend is [19:22:03] QuestPC: quite often, your best chance is convincing someone that MediaWiki will make them save money [19:22:38] Someone who wouldn't otherwise use MediaWiki, that is. And someone with a budget, of course. :) [19:22:50] MatmaRex: I worked at local uni with MediaWiki farm long time ago (1.14-1.15) however the sites switched to different platforms. Also worked with two different local MediaWiki encyclopedias (psychology and sports). But it was not well paid and both are not required in full time developer anymore. [19:23:44] QuestPC: have you tried giving a look at http://wikiapiary.com/ ? [19:24:20] There are many site admins who woould like to pay a bit for help, but they think none is available [19:25:11] Nemo_bis : To be honest , I am only a beginner in python (like a beginner beginner) and have no clue about all that. So I think I'll stick whatever I can do. I am actually enjoying makin the documentation. If there is any task I think I can do , I'll surely claim it ! [19:25:25] Nemo_bis: I checked that site for my old published extensions some time ago, maybe I'll check it again. [19:25:39] AnonGuy: we have several non-pywikibot documentation tasks as well ;) [19:25:57] yeah sure ! I'll look into whatever I can do ! [19:26:07] :) [19:26:18] Nemo_bis, Which extensions are maintained and really need the wfmsg change, mentioned in the task? [19:27:00] marcinlawnik: most of our code is nominally unmaintained [19:27:16] If something is really worthless and abandoned, we remove the code altogether [19:27:46] I can try suggesting something though [19:28:23] Nemo_bis, yes? [19:29:12] marcinlawnik: ConfirmAccount is an important one [19:29:45] Nemo_bis, should a separate Phabricator thread be created? [19:29:54] Nemo_bis : suggestions for me ? [19:31:32] marcinlawnik: no, use the general one [19:31:45] marcinlawnik: PageNotice is also in use and alive [19:32:24] AnonGuy: well, I'm always going to suggest something I mentor :p https://www.google-melange.com/gci/task/view/google/gci2014/5015535200763904 [19:33:27] AnonGuy: otherwise, in python this should be easy enough, there's a link to the code in question https://phabricator.wikimedia.org/T75170#774143 [19:39:37] Finally [19:39:58] What. [19:40:35] figuring out the shit i've been aware of for years now [19:41:42] still would help if i had someone to just show me, but either way [19:43:54] Nemo_bis , I might seem annoying , but I didn't even get what the problem was. Can you please explain what problem is there in that code? I didn't get what's wrong with the namespace. [19:44:21] I know this will be frustrating for you , so I don't mind if you don't explain it to me [19:44:45] As I told you , I just started with coding and python was my first language [19:47:31] AnonGuy: do you know what a namespace is? [19:48:06] yeah [19:48:23] I know about globals and local variable [19:48:42] Good. That's not what we're talking about :p [19:48:58] Nemo_bis, care to assign? [19:48:59] https://www.google-melange.com/gci/task/view/google/gci2014/5004830263214080 [19:49:00] umm , then ? [19:49:19] https://www.mediawiki.org/wiki/Manual:Namespace [19:50:01] ohh , I was talking about the namespace in languages [19:50:05] my bad [19:50:09] marcinlawnik: done [19:50:26] AnonGuy: so you should find a reference to that namespace, either by name or by number, and you should remove/update it [19:52:03] I'll look into this stuff after I'm done with my current task [19:52:08] Thanks for all the help [19:52:14] Appreciate it ! [19:52:18] :) [19:52:44] uitzoeken die kak Ek het al bewus van vir jare [19:53:41] AnonGuy: ok, when you're ready to claim the task let me know and I'll create it [19:53:52] yeah , sure ! [19:54:06] I'll be happy to learn something to do something new [19:54:16] SO I'll go through the wiki [19:54:18] manual [19:54:24] and then decide [20:00:30] I kind of miss regex titlematch as API parameter (and nottitlematch). Is it worth requesting this in bugzilla? [20:01:02] ** or the new bugzilla system [20:10:32] AntiSpamMeta is too spammy, soon gotta be banned ;) [20:10:48] :) [20:11:34] Subfader: if you explain your use case, it's always useful to have the desire written down [20:12:00] But I don't know the context so I can't tell whether there is a chance of your request being satisfied [20:13:29] I want to query recentchanges for pages matching certain strings, e.g. "2014" or exclude such [20:17:20] Nemo_bis, ConfirmAccount is already updated [20:17:26] FYI [20:18:15] marcinlawnik: ok, thanks; please edit the task desscription on phabricator to remove the outdated line :) [20:19:46] wth, /shared/mediawiki/extensions/ConfirmAccount was outdated [20:21:19] dependency hell? [20:22:30] dunno, doing git submodule foreach "git checkout master || :" [20:23:20] marcinlawnik: did you try logging in on tools and grepping that directory? [20:24:11] Nemo_bis, Fetched the repo and did a search there, bu tonly in ConfirmAccount. [20:24:16] What's "tools?" [20:25:00] marcinlawnik: a group of servers https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools [20:25:16] Doesn't matter for now, I was just checking :) [20:27:18] Nemo_bis, I don't have a tools account, could you run this grep for me? [20:27:19] ack-grep --php "(wfMsgGetKey|wfEmptyMsg|wfMsgReal)" [20:27:42] in "/shared/mediawiki/extensions" [20:35:49] that's what I'm updating the repos there for :) [20:35:54] marcinlawnik: did you try PageNotice? [20:36:31] Will do [20:42:45] Nemo_bis, PageNotice has a few of these [20:42:55] Will fix tomorrow. [20:43:01] You're the mentor? [20:44:02] marcinlawnik: yes [20:44:39] there are other reviewers as well, remember to mention the phabricator task in commit message and they'll come https://www.mediawiki.org/wiki/Gerrit/Commit_message_guidelines [20:45:03] So, fixing PageNotice and NewestPages tomorrow then [20:45:42] Nemo_bis, one last question, how many reviewers do I need? [20:46:03] One reviewer is enough [20:46:13] Just whether people actually get to it is the question [20:47:20] So just the mentor's approvement is sufficient to get it merged? [20:47:34] Indeed. But in theory, it shouldn't have to be the mentor [20:49:50] marcinlawnik: don't worry, with my GCI tasks nobody ever got stuck [21:07:38] Nemo_bis is a GCI sensei. [22:22:29] I want to use the json-API with imageusage for all images used on the page (usually 1-3). Do I really have to make multiple calls? [22:28:34] Subfader: you can use a generator [22:28:38] Can a mediawiki.org sysop delete [[User:ShareTwitch]]? I tried tagging it for speedy twice, but they are paying attention and are willing to edit war over the tag. [22:28:56] * legoktm looks [22:29:44] legoktm: Special:ApiSandbox? [22:31:27] Subfader: https://en.wikipedia.org/w/api.php?action=query&generator=images&titles=Main%20Page&prop=info like that? [22:33:37] That would give me all images on a page? But I want all pages that use these images. But imageusage only takes one title to check usage. [22:37:52] ah, now I understand the "generator" section on the bottom of ApiSandbox. Thanks! [22:44:13] Ok, I don't get it. I wants an iutitle from me. I these get generated. [22:44:28] ** I thought [22:45:23] Um [22:45:31] ohhh [22:45:55] What do you want? All images that are on a page? Or all pages that contain a specific image? [22:49:52] An article uses A.jpg and B.jpg. On the article I want to link all other pages which use A.jpg and B.jpg. query.imageusage works for 1 each only. I can grab the image title with $(".floatright:first a.image"). When I make 2 API calls for each image then I furthermore have to join the results to make a unique link list [22:50:22] A link section "Related pages". [22:51:25] You can't do that via the API, you'll have to do a database query. [22:52:55] Yes but thats an expensive query on each article view. The API is much lighter when the link section is toggled :) [23:50:29] If I want to add the documentation on syntax and editing MediaWiki to my self-hosted wiki, what to do? [23:50:57] !exporthelp [23:51:01] Seems someone broke that [23:51:52] petan: wm-bot is broken. [23:51:53] porton: https://www.mediawiki.org/wiki/Project:PD_help/Copying