[00:17:38] <quiddity>	 jeblad, I've fixed it.
[00:18:34] <jeblad>	 Thanks! :)
[09:08:47] <Alchimista>	 g'morning humans
[09:09:14] <Alchimista>	 Reedy, jetbrain licences are for all ide's? pycharm is included?
[09:19:47] <andre__>	 Alchimista: https://lists.wikimedia.org/pipermail/wikitech-l/2016-April/085332.html
[09:33:37] <Alchimista>	 thanks andre__ , seems to be included. do you know if the licence shows up for all, or if i must contact someone?
[10:22:39] <andre__>	 Alchimista: I know nothing, sorry.
[10:23:43] <Alchimista>	 thanks anyway, i've emailed yuri :D
[14:40:04] <andre__>	 Anyone around who's knows their basics of Git and rebasing? https://www.mediawiki.org/wiki/Gerrit/Tutorial#Squash_several_commits_into_one_single_commit_via_rebase welcomes some help.
[14:40:30] <andre__>	 There are two sentences on that page which feel contradictive to me: "It is important to leave the Change-Id for each commit in the message, otherwise Gerrit will not know how to handle the patch." and
[14:40:34] <andre__>	  "In regards to which Change-Id you want to use, squashing a commit into an existing commit (one that's already in Gerrit), you need to pick the Change-Id that belongs to the one you meant to submit a new patchset for (the surviving commit). If your commits are new and are not in Gerrit, it does not matter which Change-Id you choose."
[14:40:58] <andre__>	 If anyone knows better and if there's something to correct, please edit https://www.mediawiki.org/wiki/Gerrit/Tutorial#Squash_several_commits_into_one_single_commit_via_rebase - thanks in advance!
[14:41:04] <bawolff>	 The second one is correct
[14:41:58] <bawolff>	 at least I think having multiple change-ids would not work
[14:42:01] <bawolff>	 never actually tried it
[14:46:20] <Harri>	 hmm
[14:47:18] <Harri>	 in extension.json, I have hooked several hooks with functions from my handler extension
[14:47:33] <Harri>	 however, they seem to run on every filetype, not just the type i am interested in
[14:48:29] <Harri>	 "Hooks": {     "ImageOpenShowImageInlineBefore": [       "TypeHandler::onImageOpenShowImageInlineBefore"     ],
[14:48:51] <bawolff>	 yeah, that would run every time someone views an image page
[14:49:27] <Harri>	 whats the way to fix it
[14:49:51] <Harri>	 my idea was that my handler extension was only loaded when its of a certain type
[14:50:23] <bawolff>	 if ( $imagePage->displayImg->getMimeType() === 'image/<type here>' ) 
[14:50:35] <bawolff>	 where $imagePage is the first argument given to your hook
[14:50:52] <bawolff>	 or actually no
[14:50:54] <bawolff>	 that won't work
[14:52:25] <bawolff>	 since displayImg is private
[14:52:29] <Harri>	 yeah but
[14:52:33] <Harri>	 thats a good idea actually
[14:52:52] <Harri>	 to change the bodu.php file and leave hooks as is
[14:53:03] <bawolff>	 really most of the per-image type stuff should be in the MediaHandler subclass, and not from the ImageOpenShowImageInlineBefore hook
[14:53:42] <Harri>	 well its not an image actually
[14:55:43] <Harri>	 what do you mean displayimage is private?
[14:55:46] <Harri>	 $full_url = $imagepage->getDisplayedFile()->getFullURL();
[14:55:54] <Harri>	 that way getting the url works
[14:56:20] <Harri>	 ah, i misread, displayimg is private
[14:59:31] <bawolff>	 Oh, I totally didn't see that method
[15:00:29] <Harri>	 yeah, jus ttested, Call to a member function getMimeType() on a non-object (null) Backtrace:
[15:00:55] <bawolff>	 $imagepage->getDisplayedFile()->getMimeType() should work
[15:01:37] <bawolff>	 unless there is no image, then $imagepage->getDisplayedFile() might be false
[15:02:08] <Harri>	 Call to a member function getDisplayedFile() on a non-object (null)
[15:02:35] <bawolff>	 well that probably means that $imagepage has a typo in it
[15:02:46] <Harri>	 is there a
[15:02:49] <bawolff>	 or the first argument to the hook was named something else
[15:03:14] <Harri>	 anyway
[15:03:18] <Harri>	 is there a page somewhere
[15:03:37] <Harri>	 that just contains full_volume audio of "YOU ARE STUPID!"?
[15:03:50] <Harri>	 id like to put this on repeat
[15:04:21] <bawolff>	 You know, phabricator has a feature where you can make things do auto-playing repeat audio
[15:04:35] <bawolff>	 or at least it used to
[15:04:55] <bawolff>	 if you wrote the word seal on a line by itself, it used to make seal noises
[15:05:18] <CoolCanuck>	 LOL
[15:05:31] <Harri>	 look what happens when people have too much free time
[15:06:16] <CoolCanuck>	 I want to try it :( how long ago did it work?
[15:06:59] <Harri>	 serious tho, can i get mime type in a similar way in onbeforepagedisplay hook
[15:07:00] <bawolff>	 last time I tried I think was a year ago
[15:07:14] <Harri>	 cause its parameter is outputpage, not imagepage necessarily
[15:09:19] <bawolff>	 Harri: Maybe you could do $out->getWikiPage()->getFile() 
[15:09:37] <bawolff>	 CoolCanuck: I think i last tried like a year ago
[15:10:21] <bawolff>	 CoolCanuck: Tested right now. still works (You can even do it in a preview so not to make a real comment)
[15:10:29] <CoolCanuck>	 OMG 
[15:10:38] <CoolCanuck>	 what site is this under?
[15:10:56] <bawolff>	 https://phabricator.wikimedia.org
[15:11:19] <bawolff>	 (Please only test in preview, I don't want to be responsible for spamming seal sounds on random bugs)
[15:12:42] <CoolCanuck>	 not owrking for me xD
[15:13:19] <CoolCanuck>	 https://phab.wmfusercontent.org/file/data/wyugt2amcwnt32w35ult/PHID-FILE-e5nmjq2xrzg6fmons77u/2048px-Mirounga_leonina.jpg
[15:13:21] <CoolCanuck>	 LOL
[15:13:31] <CoolCanuck>	 wait a second
[15:13:40] <CoolCanuck>	 is it capturing our keylogs?!
[15:14:00] <bawolff>	 image shameless stolen from commons (Without crediting the original author!)
[15:15:16] <CoolCanuck>	 ooohhh.. Not good
[15:15:20] <CoolCanuck>	 also there MUST be others ;)
[15:15:34] <CoolCanuck>	 Imma find the trigger words
[15:15:42] <Harri>	 bawolff: i cant really, because onbeforepagedisplay runs on every page
[15:16:00] <bawolff>	 CoolCanuck: oh you can make your own trigger words
[15:16:05] <bawolff>	 one of the stupider features
[15:16:08] <CoolCanuck>	 O_O
[15:16:16] <bawolff>	 I made the seal thing to prove a point about how stupid an idea that feature is
[15:16:17] <CoolCanuck>	 that's very stupid
[15:16:19] <Harri>	 for example, Call to undefined method WikiPage::getFile() comes on main page
[15:17:48] <bawolff>	 Harri: Yeah, you'd need to do checks like $out->canUseWikiPage() && $out->getWikiPage() instanceof ImagePage
[15:17:53] <bawolff>	 or something
[15:18:01] <Harri>	 tried to do it with isset, but it didnt wqork
[15:19:05] <Harri>	 that eunexpected T_VARIABLE tho
[15:19:21] <Harri>	 if ($out->canUseWikiPage() && $out->getWikiPage() instanceof ImagePage $out->getWikiPage()->getFile()->getMimeType() ===
[15:20:05] <bawolff>	 If you really want to be paranoid, you can also check $out->getWikiPage()->getFile() instanceof File
[15:20:37] <bawolff>	 I'm not sure what you're doing, but honestly, it sounds like there should be a better hook for it then this one
[15:21:01] <Harri>	 yeah, i am loading a module from my extension
[15:21:05] <Harri>	 in onbeforepagedisplay
[15:21:23] <Harri>	 so that a script from the module can be called in imageopenshowinlinebefore
[15:21:32] <bawolff>	 Hmm, TimedMediaHandler uses an OuputHook to do that
[15:21:43] <bawolff>	 which sort of doesn't really work that great either
[15:22:03] <bawolff>	 There's a bug somewhere about how there is no good interface for media files to add scripts
[15:22:25] <bawolff>	 Some people have also suggested the idea, of totally loading complex formats in an iframe, and handling all the scripts that way
[15:25:38] <CoolCanuck>	 .... not the best idea. lol
[15:25:49] <CoolCanuck>	 iframes are for content.. not for scripts :P
[15:27:35] <Harri>	 alternatively, i could just let the code as is and load the module on every page
[15:27:58] <Harri>	 it is not perfect and a waste of resources, but at least its something ican fall back on
[15:28:45] <bawolff>	 CoolCanuck: Well if you load the video + javascript in an iframe, you can separate the security context (if its loaded cross-origin), so then you get a security win that if there's something wrong with the video javascript, it can't compromise the rest of your site
[15:28:59] <CoolCanuck>	 :|
[15:29:10] <CoolCanuck>	 Load the video in <video> or the player lol
[15:29:31] <bawolff>	 CoolCanuck: There's talk (or used to be) about eventually doing that for videos on Wikimedia
[15:29:54] <CoolCanuck>	 there's an option to use pure html5 for videos and audio on commons
[15:30:12] <CoolCanuck>	 but it's not default. it uses an html5 player.. (not browser built in)
[15:30:36] <CoolCanuck>	 eg: We can't add a "add subtitles" link to the browser's built in player..
[15:33:10] <bawolff>	 CoolCanuck: Well it kind of uses both
[15:33:21] <bawolff>	 It uses the browsers player, and then modifies it significantly
[15:33:37] <CoolCanuck>	 but I mean you can choose to use the browser's built in
[15:33:40] <CoolCanuck>	 :P
[15:35:38] <Harri>	 hmm
[15:35:46] <Harri>	 so timedmediahandler has the code
[15:35:47] <Harri>	 if ( $title->isSpecialPage() ) {            list( $name, /* subpage */ ) = SpecialPageFactory::resolveAlias( $title->getDBkey() );            if ( stripos( $name, 'file' ) !== false || stripos( $name, 'image' ) !== false                || $name === 'Search' || $name === 'GlobalUsage' ) {
[15:36:39] <Harri>	 tried to adapt that code as
[15:36:39] <Harri>	 if ($out->getTitle()->isSpecialPage()){ 			list( $name, /* subpage */ ) = SpecialPageFactory::resolveAlias( $title->getDBkey() ); 			if ( stripos( $name, 'file' ) !== false){
[15:47:04] <Harri>	 think i fixed it for good now
[15:47:05] <Harri>	 if (strpos($out->getTitle()->getPrefixedText(),"file")){ //is file page 				if( $out->getWikiPage()->getFile()->getMimeType() === 
[16:00:10] <Harri>	 but i do have another qquestion however
[16:00:22] <Harri>	 what might be the problem if mw.loadder.getstate says "executing"
[16:03:33] <Harri>	 extension is working in vagrant, but not in live testing environment
[16:09:23] <andre__>	 bawolff: (a belated) thanks!
[16:53:03] <bd808>	 brion: should we start a go-fund-me for a Phabricator contracting budget? ;)
[16:56:33] <brion>	 bd808: srsly :D
[16:56:44] <brion>	 ah i see it getting bumped up on their prio board already
[16:56:46] <brion>	 money talks \o/
[16:59:11] <mutante>	 http://wikicoin.wikia.com/wiki/WikiCoin_Wiki
[17:00:00] <mutante>	 no mining in labs :p
[17:02:39] <Nemo_bis>	 bd808: yes please
[17:21:22] <paravoid>	 bd808/brion: what about? :)
[17:21:42] <paravoid>	 what was this about I mean?
[17:21:56] <bd808>	 paravoid: brion pledged to pay for a fix for a missing feature out of his own pocket
[17:22:10] <paravoid>	 which feature?
[17:22:57] <paravoid>	 https://secure.phabricator.com/T5187 I suppose
[17:23:09] <bd808>	 T165
[17:23:10] <stashbot>	 T165: Provide a way to upload a file if drag'n'drop is not available or not wanted - https://phabricator.wikimedia.org/T165
[17:23:25] <bd808>	 yeah
[17:23:43] <andre__>	 Erm, doesn't that way exist already? https://phabricator.wikimedia.org/file/upload/
[17:24:42] <andre__>	 On that page I see a big "Browse" button that opens a file chooser dialog on my desktop machine running a boring non-mobile OS.
[17:24:53] <MatmaRex>	 it's super inconvenient, on a separate page, and not linked from anywhere
[17:25:06] <MatmaRex>	 but yes, it technically exists
[17:25:11] <paravoid>	 bd808: doesn't sound like "out his own pocket" :)
[17:25:35] <bd808>	 https://phabricator.wikimedia.org/T165#2289766
[17:25:50] <andre__>	 MatmaRex, that is also very very true indeed :)
[17:25:53] <paravoid>	 oh heh
[17:26:05] <paravoid>	 I totally think we should be paying them and other companies like them for stuff
[17:26:30] <paravoid>	 put our money where our mouth (fl/oss) is
[17:26:41] <bd808>	 *nod* The problem is basically finding a budget bucket that allows for it
[17:26:50] <andre__>	 Hmm, good price. Let's start the internal fundraiser. :P
[17:27:33] <paravoid>	 if this was a paid-for service we wouldn't be even discussing this, budget would have probably been found already
[17:27:42] <MatmaRex>	 i can chip in for this too. we can fund it from my expense report for the last offsite that i wasn't going to bother submitting. :P
[17:28:12] <paravoid>	 https://secure.phabricator.com/T8092 would be my own pet peeve
[17:28:41] <paravoid>	 or https://secure.phabricator.com/T10691#167705 really
[17:35:34] <bd808>	 Phabricator is a pretty magical upstream. Its hard to fault their logic for being choosy about what features they build out.
[17:36:17] <bd808>	 Over the last 2 weeks Evan has magically unblocked a project I was working towards so I"m pretty happy with them right now. :)
[17:45:41] <brion>	 paravoid: yeah should have no problem using the contracting budget :)
[17:45:45] <brion>	 but i would if i had to
[17:45:49] <brion>	 it's been pissing me off ;)
[17:46:54] <brion>	 and always remember with open source that writing and maintaining our own patch costs *us* money in opportunity cost
[17:47:27] <brion>	 andre__: yeah the good news is all the pieces are there, they're just not connected in a user-friendly way
[17:47:36] <brion>	 when you don't have anything to drag-drop from cause you're a tablet or phone
[17:47:55] <brion>	 so it'll be a quick patch once they're in
[18:06:06] <The_Photographer>	 Hi
[18:06:22] <The_Photographer>	 How I could transfer files using ftp to my project?
[18:12:50] <The_Photographer>	 maybe using sftp?
[18:14:30] <CoolCanuck>	 what's wrong with ssh?
[18:15:17] <The_Photographer>	 I can't become with sftp
[18:24:43] <Nemo_bis>	 Duplitron 5000, heh http://blog.archive.org/2016/05/12/the-tech-powering-the-political-tv-ad-archive/
[18:25:51] <Nemo_bis>	 audfprint, interesting
[18:39:10] <The_Photographer>	 well I using simply git again, no problem
[18:39:16] <The_Photographer>	 I forked the project