[14:26:52] Does anyone have ideas on how to get two links on the same row in the Sidebar? [14:31:32] In the standard MediaWiki sidebar? That's not possible as far as I'm aware [14:44:59] hoo: Is it totally impossible? Or could one do soemthing clever with css or js? [14:45:38] You can always hack that in JS... [14:46:01] but that's probably not going to be nice [14:46:53] I see [14:46:59] But please don't do that on a Wikimedia hosted wiki... we have various skins etc., so such evil hacks tend to fall apart [14:48:28] Oh, we were just about to do that on Swedish Wikipedia [14:59:54] There is consensus for something like ** randompage-url|randompage //tools.wmflabs.org/slumpartikel/|(-bot) But as you said, that does not work. [15:05:07] Ainali: You can either replace the existing link with it, make two lines... or well do the evil JS hack [15:12:07] hoo: Ok, I am not sure the community want to do evil hacks though :) Perhaps this will change their minds. [16:08:59] Could someone review/merge the account throttling patch in https://gerrit.wikimedia.org/r/#/c/184119/ ? [16:09:52] andre__: Today? [16:10:12] hoo: The event is on Monday [16:11:29] :/ [16:11:35] I see [16:11:53] doubt you'll find someone to deploy before 7am UTC [16:12:33] I pinged greg-g about it earlier [16:12:49] What did he say? [16:13:13] see the task [16:13:16] https://phabricator.wikimedia.org/T85773 [16:15:08] Ok, so he doesn't want it before Monday [18:57:08] andre__: it seems ok to me https://www.google-melange.com/gci/task/view/google/gci2014/5828632354226176 , can I upload some more identical tasks for any last-minute invasion of students looking for beginner tasks? [18:57:56] Nemo_bis: Hey. Sure! Though students are only allowed to claim two beginner tasks in total, so whoever already had some won't be able to claim those [18:59:41] That's a feature perhaps. Only the less motivated students will do these easy ones [19:00:11] andre__: should I force the *installation* of a free client, or is a web client good enough? [19:00:43] My theory was that if they install a client they're more likely to reuse it/stick around. I'm not so sure it's true though. [19:00:46] Hmm, good question [19:01:16] I also have that theory but no idea if that's true or just a feeling of old-school IRC folks... [19:05:57] The concrete difference may be that 1) an installed application is permanently listed in their application launcher/manager/whatever; 2) it may even autoboot [19:06:38] Of course with the web frontend one could 1) add to bookmarks and 2) add to tabs opened at browser startup. But no way to verify that [19:08:31] I'll try tweaking and iterating once before bulk-ification [19:14:32] https://www.google-melange.com/gci/task/view/google/gci2014/6413193911992320 [19:16:14] Point 9 not quite clear, does melange *force* adding an attachment or URL? https://www.google-melange.com/gci/document/show/gci_program/google/gci2014/help_page#claim [19:17:33] Nemo_bis: IIRC, yes it does, can't mark the task as done otherwise [19:22:52] Ugh. [19:23:16] I'll ask to link the main page of the client they chose [20:14:35] Shell user needed to run mwgrep and answer Lupo: https://commons.wikimedia.org/wiki/MediaWiki_talk:Gadget-HotCat.js#footer [20:42:51] Nemo_bis, what would be searched for exactly? [20:46:34] Figuring it out [20:52:35] Krenair: see there again [20:59:28] Nemo_bis, I don't think you can mwgrep for multiple things on a line [21:00:15] though I don't remember how exactly it works, will look for the source [21:05:34] Krenair: it just queries elasticsearch [21:05:57] Unless it's forcing a regex search, there should be no problem trying an AND search [21:07:14] yeah I'm not sure if it's a regex search or not [21:20:30] it's regex [21:50:34] Nemo_bis, legoktm[NE] I think we'd want to check for wg* in each HotCat file actually [21:51:29] maybe just mwgrep wg | grep -A 50 or something :) [21:51:43] * grep -A 50 HotCat [21:52:13] that would return an awful lot of results to grep [21:52:37] Can only try [21:52:54] Otherwise it gets easier to google :P [21:54:46] seems to contact the search backend directly... not sure what effects it would have on server load if you tried that [21:56:08] it does set a timeout for the search though