[00:00:15] Yippie, build fixed! [00:00:15] Project browsertests-commons.wikimedia.beta.wmflabs.org-linux-firefox build #146: FIXED in 10 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-linux-firefox/146/ [00:04:55] Project browsertests-en.wikipedia.beta.wmflabs.org-linux-firefox build #148: ABORTED in 18 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-linux-firefox/148/ [00:04:55] Project browsertests-en.wikipedia.beta.wmflabs.org-linux-chrome build #182: ABORTED in 1 min 58 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-linux-chrome/182/ [00:19:40] Change merged: jenkins-bot; [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/71091 [00:33:57] Project browsertests-en.wikipedia.beta.wmflabs.org-linux-firefox build #149: STILL FAILING in 27 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-linux-firefox/149/ [00:35:43] Project browsertests-en.wikipedia.beta.wmflabs.org-linux-chrome build #183: STILL FAILING in 29 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-linux-chrome/183/ [00:43:16] New patchset: PleaseStand; "Deprecate a couple one-line global functions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [00:43:23] New patchset: PleaseStand; "Deprecate a few one-line global functions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [00:43:26] New patchset: PleaseStand; "Remove calls to wfArrayMerge" [mediawiki/extensions/CentralNotice] (master) - https://gerrit.wikimedia.org/r/72236 [00:43:32] New patchset: PleaseStand; "Remove call to wfArrayMerge" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/72237 [00:48:30] New patchset: Bsitu; "Remove the feedback link from the flyout - no longer needed" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/69961 [00:50:32] New review: PleaseStand; "Changes to extensions to kill wfArrayMerge() are Idece5562 and I5c60c70d." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [00:57:33] New review: Reedy; "Needs rebasing again.." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/16419 [01:04:47] Change merged: jenkins-bot; [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/72232 [01:06:51] New patchset: PleaseStand; "Deprecate a few one-line global functions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [01:08:46] New review: PleaseStand; "Out of an abundance of caution, I'm keeping wfArrayMerge() the way it was (except for adding a wfDep..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [01:13:33] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/63108 [01:50:52] New patchset: CSteipp; "Basic MediaWiki integration, library cleanup" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70316 [02:05:29] New patchset: Ori.livneh; "Sample debugging configuration for MediaWiki" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/63240 [02:09:43] New review: Ori.livneh; "Krinkle: thanks for sharing your debug settings. They're very useful. As you can see, Erik incorpora..." [mediawiki/vagrant] (master) C: 2; - https://gerrit.wikimedia.org/r/63240 [02:09:45] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/63240 [02:17:27] New patchset: Brian Wolff; "Allow listing old files in Special:ListFiles. Add Special:AllMyUploads" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60088 [02:19:04] New review: Brian Wolff; "PS12: rebase. In order to rebase with the recent tabindex changes, this also changes the tabindex on..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60088 [02:19:13] bawolff: do you remember the 'enable case sensitive title' setting you told me about earlier? [02:19:24] grep doesn't seem to find it yet [02:19:36] Its in Namespace.php [02:20:09] ah [02:20:10] ok [02:20:16] $wgCapitalLinks, $wgCapitalLinkOverrides; [02:20:35] yeah, so I'l have to set wgCapitalLinkOverrides to false [02:20:39] ty [02:20:40] (Sorry, I originally only remembered what file it was in, not what the variable actually was) [02:20:47] yeah, the file was good enough [02:20:54] in retrospect i should've been able to look there automatically [02:21:15] To be honest, I would actually expect the code to be in Title [02:21:41] hmmm, true. [02:21:47] I... am not making sense. must be the sleep. [02:22:02] however, it did makes sense, since this is not per-title [02:22:11] but that's probably stretching the meaning of the word 'sense' a fair bit [02:22:35] Well, count are blessings that its not in WikiPage for some reason :) [02:22:38] *our [02:22:59] :P [02:23:01] true [02:26:53] * bawolff tries to beg for review for https://gerrit.wikimedia.org/r/#/c/69259/ <-- Its easy to review [02:27:38] New review: Catrope; "Yeah, RSS needs to use canonical URLs" [mediawiki/extensions/GoogleNewsSitemap] (master) C: 2; - https://gerrit.wikimedia.org/r/69259 [02:27:42] Change merged: jenkins-bot; [mediawiki/extensions/GoogleNewsSitemap] (master) - https://gerrit.wikimedia.org/r/69259 [02:28:04] RoanKattouw: thanks :) [02:28:32] bawolff: your begging works, mine doesn't! no fair! :P [02:28:56] I don't know what's happening today, but everyone is listening to my begging. Normally it doesn't work [02:29:07] :D [02:29:24] also any idea which user I am to use when doing 'imports'? [02:29:45] I've had three patches merged today, one was from april, and another was over a month old [02:29:53] * bawolff feels good :) [02:30:02] :D [02:30:03] nice [02:30:22] bawolff: I've one patch which is logically over 40 commits, but according to gerrit just one :( [02:30:26] has had lots of review, at least [02:30:35] YuviPanda: that is perhaps your problem :P [02:30:42] the logically 40 commits part [02:30:54] blame marktraceur [02:31:02] and to an extent brion [02:31:04] and also gerrit [02:31:13] bawolff: merging any of them individually would've left master in a broken state [02:31:25] which I'm fine with, but apparently is bad. [02:31:41] so we're essentially going from release to release in one patch [02:31:43] (almost) [02:31:47] If you're importing directly in the extension, you can make your own username and add to $wgReservedUsernames (or use one already there) [02:32:24] oh dear. [02:32:33] can't I just import as 127.0.0.1 and get away with it? :) [02:32:51] Wasn't that the point of this fancy git thing? to have really easy to make branches? Yet so far the only branch I've ever actually seen is Liangent's, and the use of branch there seemed to confuse people [02:33:03] bawolff: gerrit doesn't like branches [02:33:15] bawolff: gerrit is really svn in git clothing [02:33:24] lol [02:33:42] there's irony there [02:33:55] no seriously. using --amend for what they are using should be a punishable crime [02:34:04] punishable with 'forced to use cvs' [02:34:12] or 'barred from using version control for anything' [02:36:32] binasher: If you have some free time, I was wondering if you could give me a second opinion on the sql query in my patch at https://gerrit.wikimedia.org/r/#/c/67468/ ? [02:36:49] YuviPanda: Feature !== Release [02:36:56] YuviPanda: A release is comprised of multiple features [02:37:08] patches this big are wrong. [02:37:14] bawolff: in SpecialMIMEsearch.php? [02:37:17] However, appropriately enough, we'll probably need to bump the feature number for the backwards-breaking changes. [02:37:25] we definitely will [02:37:45] i've already started using 1.4 for the 'since' stuff [02:37:55] binasher: yes, its an attempt to make it use the index on (img_media_type, img_major_mime, img_minor_mime) instead of just being all out disabled for not using any index [02:38:06] marktraceur: ideally this should've been ~10-15 commits on a branch which would've then gotten merged to master [02:38:13] in this case, even svn would've been better than gerrit [02:38:42] binasher: For me it seems to perform most of the time, and all of the time if I have use index set. Tyler thinks I shouldn't put the use index there [02:39:09] YuviPanda: Bring it up in a [bikeshed] thread on wikitech-l, I'm only concerned with the existing workflow and how it applies to the patch at hand - I'm too tired to deal with the big picture right now [02:39:37] marktraceur: yeah, me too. I just want this merged so I can move on. The maintenance script is ready, I'm just double, triple testing it before commiting [02:39:45] *nod* [02:40:22] I'm going to, like, glance at the code in a cursory fashion, I won't tie us up in more whitespace finnicking [02:40:42] It will definitely cause me to itch all over, but I'll suffer. [02:40:50] Take one for the team, etc. [02:41:16] marktraceur: if only we had something that'll tell me when I'm using ~~sane spacing~~non-standard spacing... [02:41:38] marktraceur: I've tried to make sure that it is all if ( and not if(, and I promise to fix any that I find by greps [02:41:39] *nod* [02:42:01] YuviPanda: Don't worry about it, I'll probably just run a series of cleanup patches for different files later in the year [02:42:05] again, just because I think the process is stupid doesn't mean I'll not follow them. [02:42:13] :) [02:42:36] I think this is the purest bikeshedding conversation we could have :) [02:43:01] marktraceur: heh :) [02:43:24] marktraceur: also I'm testing the maintenance script by running it on a dump of the campaigns tables from production :D [02:43:30] so should be fairly accurate. [02:43:32] Eeeeexcellent [02:43:57] it was also a very good idea (from the usual producer of such ideas, ori-l) - caught at least 4-5 bugs [02:44:33] Bugs in the maintenance script, or in the original patch? [02:44:56] marktraceur: both. [02:45:00] Wow craziness. [02:45:36] YuviPanda: Have we decided that we're probably not losing any performance on this, or that it's negligible? [02:46:05] marktraceur: brion told me that caching of pages should be done automatically and hence should not be something I'm worreid about [02:46:06] *worried [02:46:20] Ah cool [02:46:22] bawolff: i'd concur with parent5446 in this case. commonswiki is on mariadb and the new img_media_mime index is getting used consistently [02:46:35] ok [02:46:47] Thanks :) [02:46:56] bawolff: plus indexes on image may change without the code changing, and force index may disable new optimizer improvements down the line [02:47:48] so i'd generally only suggest using it in the special cases where the optimizer often does the wrong thing for a query type. this is simple though, no risk of that here [02:48:22] marktraceur: maintannece script + bugfixes pushed, should be on gerrit shortly [02:48:31] Coolio [02:48:39] you can also at least look at the github pull request which has a summary of commits to figure out why things went where [02:48:57] YuviPanda: I'm sure this is a silly question, but I can edit the campaigns, at least through the normal JSON editor, right? [02:49:07] yes yes you can [02:49:08] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [02:49:16] marktraceur: it is validated, and will tell you if you do something wrong [02:49:32] Wexcellent [02:49:44] marktraceur: it needs documentation, but currently it is 'documented' on UploadWizard.config.php [02:49:59] marktraceur: it gives 'friendly'*cough* error messages like Invalid node: expecting "array", got "string". Path: "Root node -> autoAdd -> categories" [02:49:59] I will go find some caffeine, make myself stink less, and then attempt to read through it all [02:50:20] you have nobody but yourself, brion and gerrit to blame that this is so big :P [02:50:26] Oh totally. [02:50:29] i'm looking for spacing issues now [02:50:37] YuviPanda: I recognize my part in this and you shouldn't feel bad about it [02:50:43] :) [02:50:44] Anyway, BBLs. [02:51:00] marktraceur: it also means I get credit for only 1 commit rather than 50+ on GitHub, which means legoktm will still be beating my handsomely [02:51:11] Victory! [02:51:15] so far! [02:51:49] I've been working on my github streak [02:51:54] Trying to push at least one commit per day [02:52:11] legoktm: i could see that [02:52:33] legoktm: github doesn't seem to count pushes to non master branches :( [02:52:42] Oh what is this [02:52:48] It says I didn't push on July 3rd. [02:52:50] Noooo [02:52:50] so despite the last few days being very productive for me, it's counting me as zero [02:52:55] yeah me too [02:53:00] 29 days :( [02:56:03] New patchset: Brian Wolff; "Make Special:MIMESearch a non-expensive special page." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67468 [02:56:50] New review: Brian Wolff; "PS3: I asked Asher, and he agreed with Tyler, so I removed the USE INDEX." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67468 [02:57:23] New review: Ladsgroup; "Main writer is Hooman-Mehr (writer of Iranian Sans) and IranianSansZ is just little improvement for ..." [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71658 [02:59:23] New patchset: Brian Wolff; "Make Special:MIMESearch a non-expensive special page." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67468 [03:00:01] bawolff: so I am to create a user (Programatically?) and then set it to reserved with UserGetReservedNames and then use that for import? and also create a user page for it? [03:00:27] YuviPanda: Is this an import via a maintinance script? [03:00:35] yes [03:00:40] Create the user page manually [03:00:58] If its a one time thing that is only going to happen on a single wiki, you could probably get away with just creating the user normally [03:01:14] yeah, just gonna happen on Commons, I guess. [03:02:07] but in general yes, you would reserve the username with $wgReservedUsernames . Historically I don't think people created user pages for such users, but its not a bad idea [03:02:30] so should I manually create it? and then pass it as a paremeter to the script? [03:02:52] You could also just re-use User:Maintenance_script which is used by a bunch of maintenance scripts [03:03:15] Umm, I don't know if you have to actually create a "real" user for it [03:03:29] hmmm [03:03:32] ah [03:04:59] Other maintinance scripts don't seem to [03:05:32] ack Maintenance_script doesn't give me anything [03:05:33] New review: Brian Wolff; "PS4: rebase" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67468 [03:05:34] * YuviPanda looks more [03:05:57] No underscore [03:06:05] Usernames are weird and don't use normal title conventions [03:06:37] I see it in rollbackEdits.php [03:06:41] binasher: Thanks again. I've now updated my patch appropriately [03:06:53] and a couple more [03:07:07] sounds good, bawolff [03:07:09] thanks! :) [03:07:39] bawolff: np! [03:16:10] Is it possible to have an extension change the dependencies for a resource loader module in core? [03:17:37] bawolff: Why on Earth would you want to d othat? [03:19:27] RoanKattouw_away: Basically, I want to make some javascript (for pretty display of images) that runs immediately on page load, I also want extensions to be able to register javascript (for media handler extensions) which modify the way the original javascript works (currently via mw.hook). [03:19:46] Problem is that since the original js runs immediately on page load, it runs before the extension js could hook into it [03:21:03] Hmm, maybe I could delay it by making it hook into a real on load event instead of ondomready or whatever jQuery constructor thing hooks into [03:21:34] but it seems like the best solution would be for resource loader to handle dependencies [03:21:52] I think anyways, not really all that familiar with the js side of things. [03:23:08] bawolff: Ideally this would become a hook in the core JS [03:23:12] See mw.hook [03:23:52] marktraceur: javascript seems full of 'if( ' [03:23:55] marktraceur: yes, I'm using that. The problem is that the code that calls the hook runs on page load, which seems to happen before the extension code runs [03:23:55] is that accepted [03:23:56] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [03:24:01] marktraceur: code I've never touched... [03:24:11] YuviPanda: Accepted in that the people who maintain this repo are BARBARIANS [03:24:15] Except for me [03:24:24] Yippie, build fixed! [03:24:24] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6 build #93: FIXED in 10 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6/93/ [03:24:42] marktraceur: :P I'm going to fix the ones I introduced. [03:25:10] YuviPanda: git diff gerrit/master HEAD [03:25:36] Then search that for "^\+.*if(" or something like it [03:25:41] yeah, that's what I'm doing [03:25:46] YuviPanda: Up until 6 months ago, if( was considered acceptable in core php code :P [03:26:01] bawolff: i think it still should be. if ( is an abomination [03:26:11] bawolff: imagine , if we started using all forms of punctuation that way [03:26:34] I like 'if (' better, 'if(' makes me claustrophic [03:26:53] i'm waiting for someone to start saying 'let us put spaces between [ too!' [03:26:58] (i don't think we put them there now' [03:27:02] ) [03:27:03] But I also feel its a space, and neither way is very ugly so I don't understand why we care :) [03:27:13] indeed [03:27:33] I've actually seen both $some_array[foo] and $some_array[ foo ] [03:27:35] bawolff: I can't make up my mind about which I hate more, the spacing or gerrit [03:27:47] rofl [03:28:03] YuviPanda: Also don't get me started on the spacing in gerrit :P [03:28:13] :) [03:28:24] YuviPanda: but you missed your chance to object: https://www.mediawiki.org/wiki/Manual_talk:Coding_conventions/PHP#Control_structures [03:28:41] bawolff: I saw that email fly by. It was during my mandatory exam period :( [03:28:50] Yippie, build fixed! [03:28:51] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7 build #93: FIXED in 10 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7/93/ [03:30:15] I think it was during the month of 100's of bikeshed "lets be anal about trivialities in the coding convention" threads [03:30:19] yeah [03:32:58] * YuviPanda spends time productively fixing 'spacing issues' [03:33:49] For php spacing issues, we have a script that can do that for you (stylize.php) [03:34:08] where it is I have no idea though. Its not in the main repo [03:34:39] Yippie, build fixed! [03:34:39] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8 build #131: FIXED in 10 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8/131/ [03:34:52] hand fixed the handful that I had missed anyway [03:35:32] marktraceur: fixed spacing wherever I had touched :) [03:35:49] Woot [03:36:06] and added option to use whatever username to import to maintenance script [03:36:17] pretty mergable state, methinks [03:36:28] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [03:38:26] marktraceur: I'll probably head off to sleep shortly, though unsure 'how' shortly. [03:40:49] OK [03:40:54] YuviPanda: I'm scared, but I'll test it [03:41:19] I'll stick around then :) [03:42:15] marktraceur: you alrso realize that this won't go onto commons right after you merge? will head to beta labs where the WLM folks will test it a fair bit [03:42:36] Yeah. [03:45:06] YuviPanda: EventLogger, right? [03:45:17] marktraceur: EventLogging, I think [03:45:21] there's a vagrant role [03:45:27] Aha, you are correct [03:45:29] EventLogging [03:45:51] * marktraceur installs to his own machine [03:47:39] OK! Looks like it runs, doesn't affect the default UW, and will work once we run the maint script on Commons, probably [03:47:40] * YuviPanda kills his vagrant for a while [03:48:23] I think I'd trust JeroenDeDauw or multichill to test the actual functioning of the campaigns, but from what I saw of your code it will work fine [03:48:36] marktraceur: :) [03:48:46] marktraceur: do you want a copy of the production campaigns so you can test the maintenance script? [03:49:17] marktraceur: havent' heard anything at all from JeroenDeDuaw, though [03:49:28] marktraceur: but it's mostly going to be odder and other folks from WLM who'll be testing it [03:49:50] Actually... [03:50:04] I migrated an existing one, and it looks like the default categories don't get applied [03:50:13] marktraceur: hmm? [03:50:13] And it also looks like the header label isn't shown [03:50:37] YuviPanda: http://paste.marktraceur.info/103 [03:50:43] * YuviPanda clicks [03:50:55] marktraceur: 'headerLabelPage' [03:51:08] marktraceur: do you have a page named 'Something something blargh side'? [03:51:18] Oh, OK [03:51:20] That was my bad. [03:51:23] But the category! [03:51:32] marktraceur: so, was this produced by the maintenance script? [03:51:36] and are you on the latest patchset? [03:51:47] New patchset: Ori.livneh; "include role::eventlogging in role::uploadwizard" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/72240 [03:52:09] * ori-l is EventLurking [03:52:12] :P [03:52:17] uYup yup [03:52:35] marktraceur: ah, dammit, this was just me being stupid [03:52:41] bd [03:52:46] I'm a fan of you realizing it [03:53:14] marktraceur: fix pushed. [03:53:31] marktraceur: basically, they should be 'lat', 'lon', 'categories', not 'defaultLat', 'defaultLon', 'defaultCategories' [03:54:03] marktraceur: so the validation is rather permissive - it tells you if you get a type wrong, but silently ignores extra values. Do you think that is okay? [03:54:05] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [03:54:07] *nod* [03:54:17] ori-l: ^ thoughts on permissive validation that way? [03:54:36] YuviPanda: I would say it would be nice to *warn* but not *block*. Not sure if there's a way to do that. [03:54:39] marktraceur: so you can change the json manually, but the new update script sohuld do that nice. [03:55:33] *nod* [03:56:24] Great, works with the manual change [03:56:26] YuviPanda: that is strictly in adherence with draft 3 of the spec, which permits additional properties unless there is an { 'additionalProperties': false } key/val set on the object. This is a major pain in the ass if you want to use JSON Schema for validation, so after deliberating on it for a while I decided to deviate from the specs. [03:56:45] ori-l: indeed, and I decided to deviate from your deviation :P [03:57:00] two wrongs can't make a right [03:57:01] ori-l: so I merge the json into the campaign's defaults and then check for validity [03:57:04] unless additionalRights is set [03:57:29] sounds sane, I don't know the specs of what you're trying to do in detail [03:58:33] https://en.wikipedia.org/wiki/Nulla_poena_sine_lege [03:58:59] YuviPanda: btw, how long have you been up? [03:59:09] ... not sure. [03:59:10] :D [03:59:38] actually just about 16 hours [04:07:06] OK, YuviPanda, I think I'll breeze through your source code one last time and maybe merge it if nobody objects by tomorrow morning. [04:07:13] \o/ [04:07:22] but now I'm still wondering if I should deviate from ori-l's deviation from the spec [04:07:24] or not [04:07:42] YuviPanda: What files should I ignore until you make that change? [04:07:54] well, it works fine as of now [04:08:03] it's a design decision, sortof [04:08:04] Ah hm [04:08:16] should we let people have arbitrary JSON fields without having to modify / deploy code? [04:08:16] or not [04:08:18] * marktraceur could either go through it all now or wait 'til you've decided [04:08:29] I think it's mostly harmless [04:08:35] And in compliance with ori-l's spec [04:08:36] marktraceur: oh go through it, the code to change it, if required, is trivial [04:08:40] Or someone's spec [04:08:46] Ah, K [04:15:21] marktraceur: okay, I think current behavior is desirable for us, for now, rather than EventLogging's [04:15:54] marktraceur: since that means that people don't have to have lots of empty properties - just have properties for the ones you want to override from the default [04:15:56] so... no changes [04:15:57] to make [04:23:51] marktraceur: I'm about to head to sleep. Can you leave a comment on the patchset, so brion sees it? [04:32:40] Yuuuup [05:20:12] Project browsertests-en.wikipedia.org-windows-internet_explorer_8 build #134: FAILURE in 6 min 0 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_8/134/ [05:31:18] marktraceur: Shouldn't mw.hook use jQuery's stopOnFalse so it behaves like hooks on the php side? [05:32:56] I have no idea! [05:32:57] Maybe! [05:33:21] bawolff: Whine at Krinkle|detached slash submit some patches [05:33:32] * bawolff doesn't really know either [05:36:16] bawolff: krinkle's explanations in https://gerrit.wikimedia.org/r/#/c/56762/8 address that indirectly [05:39:29] I haven't yet looked at the mw.hook stuff in detail [05:39:34] Because I haven't needed to yet [05:47:54] New patchset: Spage; "Don't show position arrow if no attachTo found" [mediawiki/extensions/GuidedTour/guiders] (master) - https://gerrit.wikimedia.org/r/72241 [05:48:18] Project browsertests-test2.wikipedia.org-linux-firefox build #394: FAILURE in 17 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-linux-firefox/394/ [05:54:22] New patchset: Shirayuki; "Fix message names and spellings" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/72242 [05:56:45] New patchset: Shirayuki; "Fix message names and spellings" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/72242 [05:57:40] New patchset: Aaron Schulz; "Merged some of the access responsibility into the DAO object" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/72243 [06:23:14] New patchset: Aaron Schulz; "database: added DBConnectionPool wrappers to manage calling reuseConnection()" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72244 [06:35:20] New patchset: Aaron Schulz; "Added log_search tag support to ManualLogEntry" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72245 [06:43:48] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/72244 [06:45:26] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/72245 [06:47:39] New review: Aaron Schulz; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72245 [06:49:24] New review: Aaron Schulz; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72244 [07:08:30] New review: Rtdwivedi; "The second regex that we have included to consider the cases where there is no footer should be /^ New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [07:18:55] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/72240 [07:25:46] New review: Hashar; "> Are you merely asking to host the list added by this commit on a wiki page, to be linked from here?" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68381 [07:37:41] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [07:39:30] New review: Nemo bis; "Hashar, ok, can we link it directly from Special:Version then? It makes little sense to have it be j..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68381 [07:43:59] New review: Brian Wolff; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71997 [07:46:31] Nemo_bis: I kind of like the link idea too. Otherwise the CREDITS file becomes super long [08:09:46] good bawolff went to bed [08:10:19] link is ok but maintaining MediaWiki credits should be a MediaWiki's responsibility, not someone else's... in theory [09:22:54] hello [09:23:10] do we have graphs on the number of open commits in gerrit? [09:23:19] and if we don't, who could make some? :) [09:32:48] no we don't [09:32:58] and they were made by analytics but later died [09:33:30] at http://gerrit-stats.wmflabs.org/ [09:41:15] MatmaRex: it looks rather stable though, 636 now down from about 650 in the last few weeks (https://www.mediawiki.org/wiki/Gerrit/Navigation#Commits_lists ) [09:42:15] Nemo_bis: but there are 1190 open commits right now? [09:42:21] according to Gerrit/Reports/Open changesets by owner [09:42:31] and it's been steadily rising since at least february [09:42:37] (ijust made the graph myself :) ) [09:43:03] http://i.imgur.com/ztlNH00.png [09:43:32] yes, that's not a particularly useful one though [09:44:03] the "official" number to consider is what I said above, used on the monthly report too [09:44:57] hm, or maybe those use another one now [09:45:30] i don't see how the official numbe can be twice lowre thn the realone [09:45:38] ah, that limits to mediawiki/* [09:45:45] and not -1'd [09:45:49] or -2'd [09:46:08] tjere are 939 total in mediawiki/* and 361 in core itself [09:46:14] which are also darn huge [09:46:19] yep [09:46:25] ( https://www.mediawiki.org/wiki/Gerrit/Reports/Open_changesets_by_owner ) [09:46:34] there isn't any incentive to do CR nowadays [09:47:31] no fixme tag, no CR stats etc. [09:48:57] oh, only 71 % of unreviewed commits are from outside WMF, it was 80 [10:17:02] Change merged: Kipcool; [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71867 [10:23:05] New patchset: Nikerabbit; "Accessibility attributes for ULS trigger" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71784 [10:23:58] New patchset: Nikerabbit; "Accessibility attributes for ULS trigger" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71784 [10:24:58] New review: Nikerabbit; "Only tested in Chrome so far." [mediawiki/extensions/UniversalLanguageSelector] (master) C: -1; - https://gerrit.wikimedia.org/r/71784 [10:33:31] New patchset: Nikerabbit; "Use /** for method documentation blocks" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/72248 [10:41:34] New patchset: Nikerabbit; "Use little bit longer GPL-2.0+ instead of GPL2+" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/72249 [10:49:35] MatmaRex: http://p.defau.lt/?QsrZLUIhbaU5aCLJX_R6wA [11:00:08] Nemo_bis: hmmm. i forgot that's in commit notes [11:01:09] New patchset: Amire80; "Fix indentation and tweak comments a bit" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/72250 [11:01:40] MatmaRex: I'm now fetching them for all extensions, will take a while [11:02:21] Nemo_bis: they only contain reviews for latest patchset of eachchangeset, right? [11:05:12] MatmaRex: dunno, I only checked +2 https://gerrit.googlesource.com/plugins/reviewnotes/+/6aa54e7848d78640c773fdc8c266e55f44090735/src/main/resources/Documentation/refs-notes-review.md [11:07:06] All review labels and scores present on the change at the time of submit are [11:07:07] included. [11:07:16] so yeah, just the latest patchset. fair enough [11:07:17] thanks [11:15:05] New review: Nikerabbit; "(1 comment)" [mediawiki/extensions/GeoData] (master) C: -1; - https://gerrit.wikimedia.org/r/72218 [11:17:45] New patchset: TheDJ; "colSpan: do proper accessing of the property/attribute colSpan" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72251 [11:20:44] New patchset: TheDJ; "Tablesorter: colSpan property was incorrectly accessed" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [11:21:59] MatmaRex: well, I guess it can (only) be a way to see the review(er)s agreeing with the +2'er [11:23:05] more funny is the opposite :) http://p.defau.lt/?HrZNTagN1JGH_XhTMkt26A [11:23:36] heh [11:24:22] New review: TheDJ; "I split of the testcase changes into I192a95d2eeca40265156f7a05893ab3006f65ace" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [11:24:39] hmm, git log --grep won't grep over thenotes? [11:27:57] New review: Nikerabbit; "(1 comment)" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/63240 [11:31:21] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72251 [11:31:26] New review: Matmarex; "Merged those as trivial. This is really begging for a test case, though." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [11:31:32] thedj: ^ [11:40:58] New review: Nikerabbit; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72245 [11:42:44] New review: Matmarex; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72245 [11:45:37] MatmaRex: actually, i'm doing a rewrite of that now. [11:46:10] the code is just incorrectly named, and i'm doing some repositioning at the same time. [11:47:11] aw, tablesorter rewrite? [11:47:15] well, good thing it has a lot of tests [11:47:27] :) [11:48:35] the colspan there always resolves to "1". so if you just rename this.column to this.headerIndex then there is 0 change in behavior and it all works just fine [11:59:27] New patchset: TheDJ; "Tablesorter: colSpan property was incorrectly accessed" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [12:04:14] New patchset: Hashar; "Changed the

of galleries for
and
" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [12:04:28] New review: Hashar; "Triggered tests :D" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [12:09:16] New review: Matmarex; "It doesn'twork on IE6 unless we use the JS HTML5 shiv." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [12:10:40] MatmaRex: i still need to rebase that TS patch btw. takes a bit of time [12:11:20] i was just looking at it myself to see why it wouldn't merge [12:11:38] won't even automerge [12:11:52] manually rebasing it is [12:13:04] New patchset: Rtdwivedi; "Separation of rendering part from ProofreadPage.body.php." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/72253 [12:13:42] thedj: umm, are you doing it just with a text editor? [12:13:49] get meld [12:14:03] it makes merges and rebases just a pain in the ass, instead of a royal pain in the ass [12:14:21] it seems like here it's just git bitchin' in general, since the other patch changed the code you killed [12:14:51] oh lol [12:15:02] half of this I did before [12:15:10] righ, this will be nice. [12:15:44] meld? [12:15:51] hmm [12:19:09] New patchset: Matmarex; "CSSJanus: Support text-shadow and box-shadow flipping" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/52611 [12:19:20] New patchset: Matmarex; "CSSJanus: Fix handling for border-radius" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67264 [12:20:47] Change merged: jenkins-bot; [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/72250 [12:22:36] Change merged: jenkins-bot; [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/72248 [12:26:44] Change merged: jenkins-bot; [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71764 [12:26:52] New patchset: Santhosh; "Log ime-enable and ime-disable events" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71008 [12:29:00] New patchset: TheDJ; "Tablesorter: colSpan property was incorrectly accessed" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [12:31:16] wow, that was quite a rework [12:35:20] New review: TheDJ; "the followup link in the commit msg is broken..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69257 [12:40:44] anomie: you there? [12:41:44] MatmaRex: Not really, but I can answer a quick question or two. [12:41:44] anomie: https://gerrit.wikimedia.org/r/#/c/67876/11/includes/specials/SpecialWatchlist.php this looks a little weird to me [12:41:52] the "$user->isAllowed( 'editmyoptions' )" part [12:41:59] i mean, it does technically edit user's options [12:42:15] but only technically, the tokenshouldn't be an option anyway [12:42:25] i came upon this when rebasing my own https://gerrit.wikimedia.org/r/#/c/64565/ (Refactor watchlist token handling) [12:42:48] which abstracts out that token generation bit, and i'm wondering what to do with the permissionscheck now [12:43:18] shouldn't it be in User->saveSettings() instead? [12:43:20] Since it only fills in the token if it's not already set, it would be a good case for ignoring the permission. [12:43:25] or should it just be killed? [12:44:23] MatmaRex: I placed it under https://toolserver.org/~nemobis/crstats/ [12:44:38] New patchset: Matmarex; "Refactor watchlist token handling" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64565 [12:45:56] New review: Matmarex; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64565 [12:46:03] anomie: thanks. ^ [12:46:32] Nemo_bis: nice [12:47:18] heh, i have second most merged commits last month in MW core. [12:47:26] and i only started merging them ten days ago. [12:47:40] I did it just for you ;P [12:47:49] * Nemo_bis liar [12:48:14] Nemo_bis: can you include thecommands used there for posterity? [12:51:01] they are world readable in my home [12:51:39] MatmaRex: do you really think something like this is worth putting anywhere else? :) http://p.defau.lt/?75oYkikQ7vRRiaQl6fLNnw [12:51:56] the main added value is that one doesn't have to clone the whole repo [12:52:55] Nemo_bis: i'd put that somewhere in https://toolserver.org/~nemobis/crstats/ [12:54:03] wonders what happens if you nowrap an inline-block [12:55:21] MatmaRex: makes sense :) done [12:55:41] "posterity" is a pompous word though, given TS will be killed soon [12:55:47] heh [12:56:08] thedj: i'm looking now [12:56:27] it still doesn't have tests, btw. :P [12:57:11] MatmaRex: i didn't change any behavior [12:57:17] MatmaRex: does "git remote update" + "git submodule update --init" update the review notes like "git fetch all", considering I added said refs to my global config? [12:57:37] thedj: but wasn't the behavior broken in a subtle way? [12:57:50] No, only if you fixed the colspan to colSpan [12:57:59] heh [12:58:13] Nemo_bis: no idea, sorry. ask ^d or qchris [12:58:29] or Reedy [12:58:33] the old code simply always evaluate to colspan == 1, which made it behave like a headerIndex [12:58:38] submodules suck [12:59:08] and the one use of it used it as a headerIndex as well [12:59:12] (they just suck slightly less than periodically copying over the files) [13:00:09] thedj: my eyes glazed over at 'no'. :) [13:00:41] so the two wrongs made a right? bleh [13:00:55] i hope that wasn't my code [13:01:13] exactly [13:04:56] MatmaRex: https://bugzilla.wikimedia.org/show_bug.cgi?id=49648 you have an example page for that ? [13:05:49] thedj: nope, sorry [13:06:15] actually, you might be able to use the Notifications preferences checkboxes [13:06:18] :) [13:09:55] rev undelete probably would use it. [13:12:53] MatmaRex: it's readOnly [13:13:17] oh poop. [13:14:13] also, hm [13:14:21] apparently readonly doesn't do anything for checkboxes anyway [13:14:52] yeah, i was just reading the same :D [13:15:32] or rather, it varies per browser apparently [13:16:17] New patchset: Matmarex; "jquery.checkboxShiftClick: don't toggle if disabled" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68961 [13:17:34] http://www.whatwg.org/specs/web-apps/current-work/multipage/common-input-element-attributes.html#attr-input-readonly [13:17:38] (For this reason, only text controls can be made read-only: it wouldn't make sense for checkboxes or buttons, for instances.) [13:17:42] i'll go with the spec. :P [13:18:15] New patchset: Matmarex; "jquery.checkboxShiftClick: Don't toggle disabled checkboxes" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68961 [13:22:04] Change merged: TheDJ; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68961 [13:27:05] Change merged: TheDJ; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/68949 [13:29:14] Change merged: TheDJ; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71303 [13:31:49] Change merged: TheDJ; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71608 [13:45:42] New review: TheDJ; "(1 comment)" [mediawiki/extensions/WikiEditor] (master) C: -1; - https://gerrit.wikimedia.org/r/69896 [13:50:05] thedj: that's a lot of merges, thanks [13:54:44] MatmaRex: no way i can catch up with you : [13:57:49] New review: Nemo bis; "Later worth investigating whether "article message box" is really the best term here (it's rather ob..." [mediawiki/extensions/GettingStarted] (master) C: 1; - https://gerrit.wikimedia.org/r/70798 [14:00:30] New review: Nemo bis; "Passive is equally ok I guess, in context." [mediawiki/extensions/AJAXPoll] (master) C: 1; - https://gerrit.wikimedia.org/r/70594 [14:02:53] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [14:03:34] ori-l: around? Point of clarification about JSONSchema validation. There is no way to have it complain when there are extra fields? [14:04:16] at least with the current JsonSchema implementation? [14:04:21] looks that way to me, just confirming [14:05:20] marktraceur: uploaded a new patchset, but that is a trivial, typo fix. do not panic [14:06:12] so still on course to getting merged. just needs commit message update befoer tho [14:20:23] New review: Tpt; "Some points" [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) C: -1; - https://gerrit.wikimedia.org/r/71172 [14:26:38] uhh, thedj [14:27:13] am i wrong, or is tablesorter directly assigning stuff to properties of the s in the table? [14:30:05] "globals are magical structures, like little winged monkeys that fly your data from wherever it is generated to wherever it is needed" [14:30:46] my eyeeeeees are bleeeeding [14:30:50] MatmaRex: you are not wrong, it does that at quite a few places yes. [14:31:03] man, i did something like that once [14:31:12] when iwas in goddamn grade school [14:32:47] actually i think diebuche once looked at using .data for some of that stuff, and it was soo slow that it wasn't a better idea. [14:33:18] but some of that stuff could be kept in script state. [14:34:27] also that until recently we actually tried to keep that once diffable with the original script. but we gave up on that I think :D [14:34:34] s/once/one [14:34:55] if you really want to cry, look at mathjax :D [14:35:29] heh [14:36:49] which reminds me that I should still add role button and tabIndex to the headers. [15:04:20] New patchset: TheDJ; "Tablesorter: colSpan property was incorrectly accessed" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [15:15:06] New patchset: TheDJ; "tablesorter: Keyboard accessibility for sort headers." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72259 [15:16:04] New patchset: Ori.livneh; "Add missing qqq messages" [mediawiki/extensions/GettingStarted] (master) - https://gerrit.wikimedia.org/r/70798 [15:16:28] Change merged: jenkins-bot; [mediawiki/extensions/GettingStarted] (master) - https://gerrit.wikimedia.org/r/70798 [16:02:20] New patchset: Brian Wolff; "Follow-up 42333412833a - Fix behaviour $wgVerifyMimeType = false;" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69257 [16:02:49] New review: Brian Wolff; "Gerrit was stupid and didn't think it was a commit hash as it only had decimal numbers." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/69257 [16:31:32] Bonjour Somebody can help me to find the good command lines to add a patch in gerrit ? [16:32:53] add? [16:33:03] What do you mean by add a patch [16:33:13] karima: Have you read https://www.mediawiki.org/wiki/Gerrit/Tutorial#How_to_submit_a_patch [16:33:24] do you want to apply the patch to your current branch? [16:34:55] For this commit https://gerrit.wikimedia.org/r/#/c/71926/ I want test the patching on my project [16:35:29] karima: Your project, meaning a local test wiki? [16:36:33] you can download it using git review -d 71926 [16:36:37] if you have git review installed [16:37:02] git review -d 71926 -> move the branch ? [16:37:03] otherwise you can do git fetch https://gerrit.wikimedia.org/r/mediawiki/extensions/LinkedWiki refs/changes/26/71926/2 && git checkout FETCH_HEAD [16:37:23] git review -d 71926 will automatically check out the right branch [16:38:42] where have you see the number 71926 ? [16:40:35] git status => # Your branch is ahead of 'origin/master' by 2 commits. [ I need to do a think before todo git review ? [16:42:38] 71926 is the gerrit change id number (from the url on gerrit you gave me) [16:43:11] marktraceur: yes [16:43:18] You don't neccesarily need to fix the git status warning before doing the git review thing (It won't stop git review from working) [16:43:24] but its easy to fix [16:43:58] easy when it's not the first time... [16:43:59] Do git log to make sure the last two commits are *not* things you need (the following will delete them). And then run git reset --hard origin/master [16:44:43] For reference, that type of warning comes up if you development on the master branch, instead of a feature branch [16:45:46] Generally you never want to run git commit when you're on the master branch, and instead always create a new branch via git checkout -b MyNewFeatureName , develop your feature, git add, git commit, git review, git checkout master, git checkout -b MyNextFeature I work on [16:47:50] my last commit are good and in gerrit why I have this line : our branch is ahead of 'origin/master' by 2 commits. [16:48:16] I need to synchronize ? [16:48:32] Are the last two commits merged in gerrit? [16:48:55] the first yes [16:49:32] the second... I play with today to learn to patch it [16:49:43] If they were both merged, git pull would probably update your branch [16:49:59] since they are both in gerrit, and you don't have to worry about loosing them, I would reccomend [16:50:09] git reset --hard origin/master && git pull [16:51:03] pff... wait I do a backup before [16:52:37] probably a good plan. :) [16:53:28] git reset --hard origin/master && git pull => Do nothing "Already up-to-date. [16:54:33] ok the line has disappeared in git status [16:55:09] ok after I do : git review -d 71926 [16:56:18] yep [16:58:03] ok I see the last patch in the log... How I can delete some change of the last patch ? [16:59:12] karima: First make sure its the most recent patch in the log (or this won't work) [16:59:25] Just edit some files, do git add for them [16:59:30] but then instead of running git commit [16:59:36] run git commit --amend [16:59:38] instead [16:59:54] and that will change the top commit instead of creating a new one [17:00:29] Once you make all your changes and do git commit --amend, check with "git show" that everything looks like what you want, and then submit using [17:00:31] git review -R [17:00:59] the last patch change an image... but I dont want this file how remove this file of this patch ? [17:01:25] remove or restore previous version? [17:01:35] remove: git rm path/to/image.jpg [17:01:46] restore previous: git checkout master -- path/to/image.jpg [17:01:55] (and the commit --amend) [17:01:57] then* [17:02:14] hmm ... you are a killer ! [17:02:23] ok i try [17:18:21] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [17:32:41] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [17:32:42] New patchset: Rtdwivedi; "Separation of rendering part from ProofreadPage.body.php." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/72253 [17:36:30] New patchset: Matmarex; "Watchlist: Separate non-form parts out of the
" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64557 [17:38:39] New patchset: Matmarex; "Watchlist: Separate non-form parts out of the
" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64557 [17:38:51] New review: Matmarex; "Redone from scratch." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/64557 [17:43:10] New patchset: Aaron Schulz; "Added log_search tag support to ManualLogEntry" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72245 [17:44:36] New patchset: Karima Rafes; "Patch 3" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72261 [17:45:39] bawolff: it's not goot git has create another review [17:45:55] New patchset: Kipcool; "Coding conventions (sql queries)" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/72262 [17:46:48] karima: When you were making the commit, did you use git commit --amend ? [17:46:59] Did you change the Change-Id: part of the commit message? [17:47:14] New patchset: Matmarex; "Reorder SpecialRecentChanges::webOutput" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/65325 [17:47:39] iwrite ; git commit --amend -m "Patch 3" [17:47:47] New review: Matmarex; "Rebased. Manually because of whitespace changes. What the hell, git." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/65325 [17:47:53] oh, don't use the -m option [17:48:01] It causes things not to work [17:48:16] are you still on that branch? [17:48:38] Do [17:48:42] git commit --amend [17:48:45] else git open VIM and I dont known the keys in vim for git [17:49:02] yes again in the branch [17:49:10] oh, you can change that [17:49:39] do [17:49:40] ok [17:49:43] EDITOR=nano git commit --amend [17:49:45] New review: Kipcool; "@Hiong3-eng5: this should solve the problem with straight_join" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/72262 [17:49:51] (replace nano with your favourite editor) [17:49:57] when the text editor pops up [17:50:07] change the Change-Id line at the bottom to be: [17:50:18] Change-Id: Ib04913a966e5fdbc3b6b1b22189d86f184c15523 [17:51:24] Once you do that, could you run git show, put the results in dpaste.org and give me a link so I can check if everything looks right? [17:52:38] New patchset: Rtdwivedi; "Separation of rendering part from ProofreadPage.body.php." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/72253 [17:54:03] bawolff: https://dpaste.de/CK1Gu/ [17:54:45] New review: Umherirrender; "A idea from a user on de.wp is: allfilesused" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61817 [17:55:10] ok, if you run git review -R it should work [17:55:47] It will replace the commit message with just 'patch 3' [17:56:03] New patchset: Karima Rafes; "Patch 3" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/71926 [17:56:18] Generally when you're making a new version of a commit, you just keep the original message, and comment on the change what you did in the new version of the patch set [17:56:26] New review: Umherirrender; "There is a DeprecatedGlobals.php to get deprecation warnings on a global (used for $wgArticle), but ..." [mediawiki/tools/codesniffer] (master) - https://gerrit.wikimedia.org/r/69090 [17:57:11] hmm... Wait I merge this version and I retry a new commit to A_Z [17:57:31] Did that work like you wanted it to? [17:58:33] wait the last it OK ? [17:58:34] New patchset: Kipcool; "View and edit popup annotations with ajax" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/72263 [17:58:51] I change only the name of this review [17:59:27] I try again to add a other patch [17:59:30] Hmm, you changed wait.png [17:59:39] but i see what you mean [18:00:06] New review: Kipcool; "@Hiong3-eng5: this one is a big change involving many" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/72263 [18:01:07] Maybe best to redownload the patchset ( git review -d 71926 ) and re-do making the changes [18:05:34] bawolff: I have the message : ! [remote rejected] HEAD -> refs/publish/master/71926 (no changes made) [18:05:46] Did you remember to do git add [18:05:59] before doing git commit --amend [18:06:19] Does git diff show any output? [18:06:24] add is only for new file ? [18:07:06] wait I merge this version [18:07:33] No, git add is for any file you change [18:07:42] (git is different from svn in that regard) [18:10:01] ok when the merge is OK I will do : git checkout master ... git pull [18:11:24] the merge is automatically in gerrit ? [18:11:43] I'm not sure what you mean [18:12:24] I m sure I know I dont mean all gerrit... [18:12:26] lol [18:12:49] Any change you submit with git review is in gerrit [18:12:59] I valided the patch... There is another step ? [18:13:39] Which patch are you referring to? [18:14:09] ok but I valid in the interface web ? and the gerrit merge automatically ? [18:14:16] generally you don't set the validated flag (usually only jenkins-bot touches that), you usually set code-review flag to +2 in order to get something merged [18:14:29] Sometimes you have to press a submit button [18:15:00] I click on submit patch Set 3 ? [18:15:11] Change merged: Karima Rafes; [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/71926 [18:15:22] Now its merged :) [18:15:23] lol I found ;) [18:15:54] There's also a Status line in an infobox near the top [18:16:13] It now says "merged" where previously it said "review in progress", which is another way to tell [18:16:35] ok merge is bold [18:16:53] ok now I can do git pull ? [18:17:07] make sure to switch your branch to master [18:17:12] and then you can do git pull [18:17:22] git checkout master [18:17:31] git pull should say something like "fast forwarding changes" if things worked correctly [18:18:36] hmm error : error: Your local changes to the following files would be overwritten by checkout: [18:19:02] I force the checkout ? [18:19:40] no [18:19:59] It means you have local changes that you didn't commit [18:20:07] you probably meant to put them in your last patch [18:20:11] run git diff [18:20:15] to see what they are [18:20:26] Basically any file you changed, but didn't run git add [18:20:33] would be included in local changes [18:20:49] To get delete the local changes you can do git reset --hard HEAD [18:23:05] I mean... the precedent patch change a File ... if I want change again the file I have to do git add of this file ? [18:24:24] yes. You have to run git add anytime you change a file [18:24:37] or it won't be included when you do git commit [18:26:03] As a way to check, the output of git diff is all the changes you have yet to run git add on. Compared to the output of git diff --staged which is all the changed you have done git add for, but have not yet comitted [18:29:45] pff... change 71926 is closed [18:30:11] i create another branch ? [18:30:53] yeah, do [18:31:05] git checkout -b newBranchName [18:31:09] and git add all the files [18:31:21] and you can make a new commit with all your local changes [18:33:47] git diff --staged is empty after my add ... I lost my change ? [18:34:35] Possibly [18:34:38] is git diff also empty? [18:34:46] yep [18:35:06] yeah, sounds like [18:35:24] git status would also say if you have any pending changes [18:38:44] bawolff:ok I think I restart my tests in 3 hours... I need to compare each files with my backup Thanks Bye [18:39:03] ok, good luck [18:39:13] ;) [18:45:14] Yippie, build fixed! [18:45:14] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_9 build #143: FIXED in 18 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_9/143/ [18:51:02] New patchset: Aaron Schulz; "Various edit conflict handling fixes" [mediawiki/extensions/FlaggedRevs] (master) - https://gerrit.wikimedia.org/r/72265 [19:10:01] New review: Tpt; "Last points..." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:16:14] New review: Tpt; "And some others" [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) C: -1; - https://gerrit.wikimedia.org/r/71172 [19:25:16] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:29:07] New review: Matmarex; "I have since learned that this is fixing two bugs which were cancelling each other out, and thus the..." [mediawiki/core] (master) C: 2; - https://gerrit.wikimedia.org/r/71530 [19:31:45] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [19:40:42] AzaToth: anything worth a -1 on the new patcht (since you're in here)? [19:42:10] apergos: which patch? [19:42:36] makefile [19:42:56] lemme find it [19:42:59] looking [19:43:18] it's probably on your reivew list now [19:43:27] New patchset: PleaseStand; "Make wfMatchesDomainList not match partial domains" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72312 [19:44:36] apergos: nothing has changed [19:44:52] apergos: difference between patchset 1 and 2 is only the commit message [19:45:03] seems you phailed with your git skillz [19:46:23] brb [19:47:28] ugh [19:47:35] sec [19:48:36] forgot to add the docs directory anyways [19:49:18] New review: Hiong3-eng5; "Looks good. Some comments though" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/72262 [19:55:31] New patchset: Nilesh; "Added Hadoop Streaming scripts to parse wikidata xml.bz2 dumps, generate csv files, build database to test prototype" [mediawiki/extensions/WikidataEntitySuggester] (master) - https://gerrit.wikimedia.org/r/72313 [19:56:45] Change merged: Nilesh; [mediawiki/extensions/WikidataEntitySuggester] (master) - https://gerrit.wikimedia.org/r/72116 [19:57:22] Change merged: Nilesh; [mediawiki/extensions/WikidataEntitySuggester] (master) - https://gerrit.wikimedia.org/r/72313 [20:00:20] New patchset: Umherirrender; "Avoid date()/gmdate() for date formatting" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71121 [20:00:45] apergos: heh [20:00:52] New patchset: Hashar; "stop git recloning!" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/72314 [20:01:19] New review: Hashar; "do not deploy there, I need to double check this." [integration/jenkins-job-builder-config] (master) C: -2; - https://gerrit.wikimedia.org/r/72314 [20:03:46] presumably that actually had the changes in it :-P [20:04:11] * apergos looks at the news and hopes they know no one on that flight [20:07:03] the power of the internet... instant coverage from a local news station, live [20:07:40] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [20:16:20] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [20:20:43] Yippie, build fixed! [20:20:44] Project browsertests-en.wikipedia.org-windows-internet_explorer_8 build #135: FIXED in 7 min 7 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_8/135/ [20:22:25] New patchset: Hashar; "doc additions" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/72315 [20:22:53] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/72315 [20:47:01] Yippie, build fixed! [20:47:02] Project browsertests-test2.wikipedia.org-linux-firefox build #395: FIXED in 17 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-linux-firefox/395/ [21:03:46] New patchset: Matmarex; "Show change tags on diffs" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72317 [21:09:37] New review: Jforrester; "On bug 49602 comment 0, Matt Flaschen suggested that we might want to have tags be configurable as t..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72317 [21:14:43] New review: Matmarex; "Yeah, I disagree with that suggestion. I see no reason to show any tag in history view, but keep it ..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72317 [21:41:02] New patchset: Amire80; "Load messages from a fallback chain" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/72065 [21:53:12] New patchset: Matmarex; "mediawiki.notification: Show the notifications relative to the screen" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72322 [22:03:18] New patchset: SuchABot; "Apply same styles as Gingerbread to Honeycomb" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/72323 [22:08:11] New patchset: Aaron Schulz; "filebackend: normalize paths used in locking" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/72324 [22:41:03] New review: Mattflaschen; "Matmarex, I thought jQuery's createSafeFragment took care of that (if they have JS enabled of course)." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [22:41:40] New review: Mattflaschen; "createSafeFragment call: http://goo.gl/HZjNL" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [22:43:23] New patchset: Karima Rafes; "fix several problem" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72326 [22:46:55] New review: Matmarex; "Hm, yeah, apparently. I didn't know jQuery does that." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62207 [22:51:57] New patchset: Karima Rafes; "fix several problem" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72326 [22:53:04] Change merged: Karima Rafes; [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72326 [23:34:49] New patchset: Karima Rafes; "Fix the tests" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72327 [23:35:41] Change merged: Karima Rafes; [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/72327 [23:53:14] Hi Somebody can explain howto merge the master in the official branch REL1_21 ? [23:53:48] git status => # Your branch is ahead of 'gerrit/REL1_21' by 4 commits. [23:54:10] but I can not push in gerrit