') and description can map the qqq language text [21:55:03] <mwalker> actually; you'd probably need 'title'/ [21:55:13] <YuviPanda> yeah, if you look at the JSON Schema for EventLogging, https://github.com/wikimedia/mediawiki-extensions-EventLogging/blob/master/schemas/schemaschema.json [21:55:15] <mwalker> 'qqq-description'/'description' [21:55:32] <YuviPanda> it has 'title' and 'description' fields. Instead of hard-coding them in english, you just use prefixed message names [21:56:03] <YuviPanda> mwalker: why would you need that? Wouldn't that just be Mediawiki:<key>/qqq ? [21:56:25] <YuviPanda> hmm, right, then *you* would have to fill that in in the appropriate Mediawiki: namespace. and not too elegant. [21:57:23] <mwalker> yep -- gotta try and fill in as much metadata automatically as possible; and you need three, because presumably on the form you would have the title of the field, and then a longer description of exactly what the heck it is (e.g. help text) [21:58:04] <GorillaWarfare> Hmm [21:58:05] <YuviPanda> mwalker: right. qqq-message would actually be the description to the translators of what that is... [21:58:07] <mwalker> though; maybe you could get away with 2 and just prefix the title qqq with 'Title for field consisting of ...' [21:58:10] <YuviPanda> makes sense. [21:58:27] <YuviPanda> mwalker: well, you can have as many fields as you want, so not really a problem :) [21:58:39] <YuviPanda> GorillaWarfare: are we making sense? [21:59:05] <gerrit-wm_> Change merged: jenkins-bot; [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/71133 [21:59:18] <GorillaWarfare> A little bit, I think :P [21:59:57] <GorillaWarfare> I'm looking at the EventLogger schema [22:02:34] <GorillaWarfare> So you're saying that, in the schema, you'd have something like 'bookmanager-custom-author' [22:03:03] <GorillaWarfare> Which would then pull from Mediawiki:bookmanger-custom-author or something? [22:03:08] <YuviPanda> GorillaWarfare: pretty much :) [22:03:28] <YuviPanda> GorillaWarfare: this code for 'pull from Mediawiki:bookmark-custom-author' is already part of our i18n framework too :) [22:03:59] <GorillaWarfare> I see [22:04:12] <GorillaWarfare> And would the Mediawiki:bookmark-custom-author be on each wiki using this extension? [22:04:20] <GorillaWarfare> Or centralized? [22:04:31] <YuviPanda> GorillaWarfare: can be either. [22:04:40] <GorillaWarfare> Ah, I see [22:04:46] <GorillaWarfare> Seems like centralized might be cleaner [22:04:49] <YuviPanda> GorillaWarfare: so how our i18n works is, if you define it in code, it is 'centralized', and goes to translatewiki.net for translation [22:04:59] <mwalker> (code being PHP) [22:05:03] <YuviPanda> right [22:05:44] <GorillaWarfare> Yeah, I've seen it in the context of .i18n.php files, just not in the context of actually drawing the info from a wiki page [22:06:11] <YuviPanda> GorillaWarfare: yeah, so the wiki pages can override the contents of the .i18n files [22:06:14] <YuviPanda> or create new ones [22:06:31] <YuviPanda> In fact, I think Mediawiki: pages are usually called 'interface messages' than 'wiki pages' [22:06:33] <YuviPanda> the right is named similarly too, IIRC [22:06:51] <mwalker> 'edit-interface' [22:06:54] <mwalker> is the right [22:07:04] <YuviPanda> yeah [22:09:17] <GorillaWarfare> Hmm, they can create the .i18n files? [22:09:35] <GorillaWarfare> That seems quicker than querying for the contents of each 'interface message', I guess [22:09:56] <mwalker> anyways -- to actually use a message you call wfMessage (or if you're within a page context $this->getContext()->msg()) [22:10:10] <mwalker> GorillaWarfare: users cannot [22:10:14] <mwalker> *create i18n files [22:10:21] <YuviPanda> GorillaWarfare: i18n files go through gerrit, translatewiki.net, etc. Takes a while :) [22:10:30] <YuviPanda> but *you* can :) [22:10:46] <YuviPanda> so the idea is to put the most common required messages in your .i18n.php file [22:10:53] <YuviPanda> and the community *can* override it if they want to [22:10:59] <YuviPanda> but if you put all they wanted there, they don't need to [22:11:11] <GorillaWarfare> Ohhh, okay [22:11:19] <GorillaWarfare> That makes sense! [22:11:21] <YuviPanda> :) [22:11:28] <mwalker> *thumbs up* [22:11:38] <YuviPanda> In short, it's a solved problem. Don't worry about it :) [22:11:47] <YuviPanda> there could/should be docs somewhere on mw.org [22:11:58] <GorillaWarfare> Hah, I'll try.. I may need some clarification when it gets down to it, but I'll wait until then [22:12:13] <YuviPanda> :D [22:12:16] <YuviPanda> we're all here! [22:13:11] <mwalker> for a huge rabbit hole -- http://www.mediawiki.org/wiki/Localisation and http://www.mediawiki.org/wiki/Localisation/Tutorial [22:13:45] <mwalker> but it might be wise to put that on a backburner whilst you play with contenthandler [22:13:51] <GorillaWarfare> Indeed :) [22:14:20] <GorillaWarfare> I'm still trying to decide if a namespace or an extension is the best approach.. [22:14:58] <mwalker> well; back up a step on that -- what is your relationship with PP? [22:15:06] <GorillaWarfare> mwalker: User only :P [22:15:32] <mwalker> and, more importantly, how do you want users of bookmanager to interact with your software with/without PP [22:15:41] <GorillaWarfare> One of the other GSoC students is working on a refactor [22:16:14] <GorillaWarfare> That was actually what I was going to bring up next. I think it's important that the two extensions aren't _completely_ independent of each other, as that would mean the users would have to do a lot of duplication [22:17:00] <GorillaWarfare> I was wondering if perhaps, instead of trying to inject more fields into the PP Index page like we discussed yesterday, we could just auto-populate some of the JSON metadata from the Index pages [22:17:35] <mwalker> that could work [22:17:46] <mwalker> but you would need to set up a mapping schema [22:17:49] <mwalker> or something similar [22:18:37] <GorillaWarfare> Yeah [22:18:45] <mwalker> ... so what you're thinking is still having two forms per book if PP is installed? [22:18:45] <GorillaWarfare> But the fields are pretty one-to-one [22:18:58] <GorillaWarfare> I guess so... do you think that's unreasonable? [22:19:20] <mwalker> not unreasonable -- but probably not the most friendly user thing in the whole world [22:19:33] <GorillaWarfare> Yeah :S [22:19:47] <GorillaWarfare> Well, do you think injecting fields into the Index page is doable? [22:21:00] <mwalker> absolutely -- especially if it's being refactored already [22:21:31] <mwalker> have two hooks in PP -- PreFormDisplay, and EditFormHandle or something like that [22:21:45] <mwalker> (caveat -- you would both have to be using the same type of edit form generation) [22:22:59] <GorillaWarfare> Both BookManagerv2 and ProofreadPage, you mean? [22:23:01] <mwalker> yep [22:23:10] <GorillaWarfare> Hmmm [22:23:18] <GorillaWarfare> So they'd need to move to a JSON schema too [22:23:54] <marktraceur> Damn it YuviPanda all of your talk of UploadWizard makes for so many pings [22:24:04] <mwalker> GorillaWarfare: nooo; not at the PHP level -- the way I see something like this working is JSON schema node -> HTMLFormElement -> Page [22:24:19] <mwalker> so, if they're at least using HTMLForm in PP; then you just have to modify that array [22:24:30] <YuviPanda> UploadWizard UploadWizard UploadWizard UploadWizard [22:24:34] <YuviPanda> sorry, marktraceur :) [22:24:44] <GorillaWarfare> Ohh, I see [22:24:49] * mwalker sings UuuuuuuupupppppppLoooooaaaaaaaddddd WIZARD! :) [22:25:49] <mwalker> GorillaWarfare: yarp -- something to try and coordinate early; but you don't need a technical implementation -- you can work on the standalone side of things first [22:25:55] <GorillaWarfare> I can speak with Aarti and see what their plans are [22:26:08] <GorillaWarfare> Yeah [22:26:18] <GorillaWarfare> And we'll need a standalone form anyway, for wikis that don't use PP [22:26:22] <mwalker> exactly [22:27:10] <GorillaWarfare> I'll try to get an email out to her today, though, just to start the conversation [22:27:11] <marktraceur> mwalker: Joke's on you, my IRC client isn't smart enough to interpret your singing! Muahaha. [22:27:24] <mwalker> thinking about it more -- I would just have an extension -- because managing two extensions in the Index NS would be painful [22:28:14] <mwalker> wow; that's a terrible sentence; I would use NS_MAIN:*.book for your metadata, so that PP and BM are not fighting in NS_INDEX [22:28:17] <GorillaWarfare> Just have an extension? [22:28:20] <GorillaWarfare> Ohh [22:28:40] <GorillaWarfare> Yeah, I was thinking that the form would save the Index content but also save content to *.book or similar [22:28:48] <mwalker> yes [22:28:48] <GorillaWarfare> (that's possible, right?) [22:28:52] <mwalker> with hooks [22:29:27] <GorillaWarfare> Yeah, that makes sense. Having the two try to share the Index namespace sounds like it would just be asking for trouble [22:30:37] <gerrit-wm_> New review: CSteipp; "Looks like the recent patches addressed Aaron and Tim's comments." [mediawiki/extensions/CentralAuth] (master) C: 1; - https://gerrit.wikimedia.org/r/68199 [22:32:49] <mwalker> GorillaWarfare: so... for my understanding, how are you going to get this metadata to included pages in books? I've seen the info bar mockup; but how are you actually going to get that on the page? [22:33:02] <mwalker> (and what if a page is part of more than 1 book!?) [22:34:17] <GorillaWarfare> Regarding the first question, the navbars should be added to all pages that are included as a section in a book [22:34:31] <GorillaWarfare> Regarding the second... I'm still working on that [22:35:16] <gerrit-wm_> New patchset: Katie Horn; "Removing currencies that were supported with WorldPay, but are not with Wells Fargo" [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71142 [22:35:37] <Amgine_> <blinks> [22:35:49] <mwalker> GorillaWarfare: so... during page render -- how does the page know if it's part of the book or not? [22:37:34] * YuviPanda blinks with Amgine_ [22:38:00] <GorillaWarfare> mwalker: Hmm, let's see [22:38:15] <GorillaWarfare> mwalker: So each section of a book should be a subpage of the book's main page [22:38:50] <GorillaWarfare> And the books main page should have a way of connecting the book with its *.book page [22:39:04] <mwalker> ah! so you will always have a classic NameOfBook\NameOfChapter\NameOfPage kind of page layout? [22:39:21] <mwalker> (which also means that a page cannot be in two books at once) [22:40:19] <Amgine_> "Right, so, since the pa'anga is not supported by Wells Fargo, WMF will remove its existence from the planet with the exception of a footnote in Wikipedia." [22:40:48] <GorillaWarfare> That's how Wikisource and Wikibooks currently do it, and I think it makes sense to stick with that approach [22:41:20] <GorillaWarfare> Although, in terms of making it easy to move around sections and subsections, I think they should maybe all just be one level deep [22:41:22] <Raylton> just to check, there will be a list of pages saved somewhere? [22:41:45] <GorillaWarfare> So instead of Book/Section and Book/Section/Subsection, we'd have Book/Section and Book/Subsection [22:41:50] <GorillaWarfare> Raylton: A list of what pages? [22:42:05] <GorillaWarfare> Raylton: Each page in a specific book, you mean? [22:42:20] <Raylton> pages in a book [22:42:56] <GorillaWarfare> Raylton: Yes, the JSON block I'm envisioning looks something like this: https://meta.wikimedia.org/wiki/Book_management/JSON [22:42:56] <Amgine_> [[My title/1]] [22:43:16] <GorillaWarfare> And in the "sections" object, there is a "pages" array/number [22:43:49] <Raylton> it will be saved where? [22:45:07] <DanielFriesen> GorillaWarfare: If you're doing smart input of the title I might suggest you split title into the ns number and title text. [22:46:03] <GorillaWarfare> Raylton: In the JSON block that will be used by ContentHandler... we've discussed putting them all in pages with a .book extension [22:46:14] <GorillaWarfare> DanielFriesen: Sorry, "smart input of the title"? [22:46:35] <AzaToth> Krinkle: http://paste.debian.net/13233/ [22:46:51] <DanielFriesen> GorillaWarfare: Sorry I just mean giving the user an actual title input (maybe even autocomplete) it would probably be better to store the ns id [22:47:35] <DanielFriesen> Then again I don't know.. [22:47:50] <GorillaWarfare> DanielFriesen: Won't these pages only ever be in the mainspace, though? [22:48:27] <DanielFriesen> Not an assumption you code into the storage format. [22:48:36] <GorillaWarfare> Fair [22:49:07] <Krinkle> AzaToth: I haven't read the full extend of it, but basically if a template has parameters that are not easy to use and/or document, you can't document. Fix it, or wrap it in another template that is more userfriendly, or fork it and make this template use the better one. [22:49:08] <GorillaWarfare> It might be worth, instead of having "title" as a field, just forcing the name of the page storing the JSON to be the same as the name of the main book page [22:49:15] <Raylton> GorillaWarfare, like [bookname].book? Sorry ... did not quite understand [22:49:22] <GorillaWarfare> Raylton: Yes [22:49:25] <Krinkle> AzaToth: I have no intention of allowing every single possible thing that wikitext enables, because that is by nature limitless. [22:49:56] <Krinkle> AzaToth: VE allows insertion of arbitrary wikitext as well as freeform parameters that aren't in templatedata. [22:50:59] <Krinkle> Users that don't know the template itself will likely stay away from that though, which is a good reason to make the templates understandable for new users and the first step would be to not use numbered parameters or only use them if the order is simple and without complicated gaps or alternative signatures. [22:51:23] <DanielFriesen> GorillaWarfare: The ns is just on my mind because one of my plins is a namespace registration system and a UI to make namespace changes possible. [22:52:04] <Krinkle> AzaToth: Note that for things lie foo1, foo2, foo3 you can use "inherits" property to avoid duplication [22:52:07] <Krinkle> like* [22:52:20] <AzaToth> Krinkle: where are all props documented? [22:52:30] <Krinkle> And for things like Unsigned|1= and Unsigned|username= (e.g. {{{username|{{{1|}}}}}}) you can use "alias" [22:52:31] <AzaToth> the spam mail from Okeyes didn't show much info [22:52:37] <Krinkle> AzaToth: In git, and on mw.org [22:52:38] <AzaToth> ok [22:52:50] <Krinkle> https://www.mediawiki.org/wiki/Extension:TemplateData#Defining_a_TemplateData_block [22:53:07] <Krinkle> and https://raw.github.com/wikimedia/mediawiki-extensions-TemplateData/master/spec.templatedata.json if you prefer reading json spec [22:53:20] <mwalker> DanielFriesen: ambitions -- especially given the slowly more prolific usage of namespaces by extensions [22:53:24] <mwalker> *ambitious [22:53:28] <Raylton> GorillaWarfare, then you want to keep the structure book name/chapter name? [22:53:41] <GorillaWarfare> Raylton: Yes, I think so [22:54:00] <DanielFriesen> mwalker: Actually this originally came out of solving that very problem [22:54:21] <DanielFriesen> mwalker: https://bugzilla.wikimedia.org/show_bug.cgi?id=31063 [22:55:01] <Raylton> GorillaWarfare, then it will not be compatible with wikipedia (as collection extension?) [22:55:09] <Raylton> ? [22:55:12] <DanielFriesen> Giving extensions a way to simply use a unique key to get MW to register an NS id for it. Instead of having to co-ordinate the ns ids used by different extensions. [22:55:38] <GorillaWarfare> Raylton: What? [22:55:39] <mwalker> DanielFriesen: I like it [22:56:38] <mwalker> GorillaWarfare: if the book structure is hard coded into the name of the page -- you cannot have arbitrary pages added to books (as in how the collection extension allows you to create books from arbitrary collections of wiki pages) [22:56:47] <AzaToth> Krinkle: can you explain what Sets are? [22:57:19] <AzaToth> I understand what params are, but can't figure out what a set is in the context [22:57:33] <Raylton> exactly mw-jenkinsbot [22:57:45] <Raylton> mwalker* [22:58:11] <DanielFriesen> mwalker: As a bonus I think I'm going to embed that key data into the export XML. ;) Importing pages from another wiki could setup new namespaces automatically instead of putting pages in the main ns. Avoid NS collisions using temp ns names even. [22:58:14] <AzaToth> Krinkle: also, can multiple <templatedata> be used on the same template (for example one main from transclusion, and one specialized on the template) [22:58:30] <GorillaWarfare> Oh, I suppose that is a good point. So do you think it's best to allow both Title/Section pages as well as pages that are more arbitrarily-titled? [22:59:32] <Raylton> GorillaWarfare, yes, of course [23:00:00] <GorillaWarfare> So that brings us back to the question of how to know which page is part of a specific book for navbars [23:00:01] <Krinkle> AzaToth: No. It's not meant to be used to generate a pretty page but to make the data available to the API and database. If you want to elaborate further on the wiki page (more elaborate than what you want VisualEditor to use in the dialog) then you'll have to use regular ==headings== and stuff to add those pieces of information. [23:00:28] <Krinkle> AzaToth: If you prefer to not show it on the page, you can wrap it in <div style="display: none;"> [23:00:33] <AzaToth> Krinkle: uh? [23:00:44] <AzaToth> I don't think I've asked such a question [23:00:57] <Krinkle> "can multiple <templatedata> be used on the same template (for example one main from transclusion, and one specialized on the template)" [23:01:01] <AzaToth> yes [23:01:10] <Krinkle> What do you mean? [23:01:28] <AzaToth> can one <templatedata> be combined with a transcluded <templatedata>? [23:01:47] <Krinkle> I'm not sure what you mean, what would that look like? [23:01:47] <Raylton> GorillaWarfare, maybe: https://pt.wikibooks.org/w/index.php?title=pagename&navbar=listofpages [23:02:01] <AzaToth> Krinkle: like one <templatedata>Z [23:02:24] <AzaToth> "The TemplateData extension works through having a <templatedata> tag defined in the wikitext of a template page (optionally it can be transcluded into the template page as well)." [23:02:40] <James_F> AzaToth: Hey, I can answer questions. [23:02:42] <mwalker> GorillaWarfare: or maybe having a category for it -- I think the Article class allows modifications based on categories [23:02:48] <James_F> AzaToth: Right now Krinkle is rather busy fixing VisualEditor. :-) [23:02:51] <AzaToth> James_F: nice ツ [23:03:04] <Raylton> GorillaWarfare, but you can use bookname as default listofpages value [23:03:07] <AzaToth> James_F: I blame Okeyes for pointing me to Krinkle [23:03:09] <James_F> AzaToth: Best person to answer != best person to answer right now. :-) [23:03:16] <AzaToth> heh [23:03:19] <GorillaWarfare> mwalker: A category for each book? [23:03:24] <GorillaWarfare> Raylton: What do you mean? [23:03:32] <James_F> AzaToth: Anyway. What are you trying to do, and how can I help? [23:03:36] <AzaToth> James_F: will all <templatedata> entries on a template be used? [23:04:11] <James_F> AzaToth: Yes; that is indeed what you're distracting my developers from doing. :-) [23:04:20] <mwalker> GorillaWarfare: no; more just a category that says a page is part of a book; and then the somewhat expensive metadata lookup is limited to those pages [23:04:22] <James_F> AzaToth: Used by VisualEditor, you mean? [23:04:24] <AzaToth> i.e. <templatedata>...some data...</templatedata> blablabla <templatedata>...some more data...</templatedata> [23:04:28] <AzaToth> James_F: yes [23:04:40] <mwalker> GorillaWarfare: but maybe yes; a prefixed category for each book [23:05:07] <mwalker> GorillaWarfare: like BookManager2-<TitleOfBook> [23:05:08] <AzaToth> James_F: for example one of the templatedata is a generic templatedata taken from a common place [23:05:33] <aude> James_F: i was pleasantly surprised when visual editor properly added <nowiki> tags when i entered {{#property:p30}} with some text on mediawiki.org [23:05:34] <AzaToth> and the template in question adds some specialization [23:05:38] <James_F> AzaToth: Oh, you mean, if you have multiples? No, I think it takes the first one on the page. [23:05:39] <aude> very nice [23:05:39] <Raylton> GorillaWarfare, If you can add in a url, where the list of pages/metadata is saved, there will not be much problem navigating through it [23:05:44] <AzaToth> James_F: ok [23:05:52] <James_F> aude: You're welcome, though a lot of the credit for <nowiki>s goes to Parsoid, not VisualEditor. :-) [23:05:57] <GorillaWarfare> Raylton: Add the URL to what? Each subpage/ [23:05:58] <GorillaWarfare> *? [23:06:00] <aude> :) [23:06:24] <AzaToth> James_F: second question, in templatdata, what are "sets"? [23:06:36] <James_F> AzaToth: Groupings of parameters that make sense together. [23:06:54] <James_F> AzaToth: E.g. year/month/day parameters - no point setting "day" if you haven't got a year and a month. [23:07:05] <AzaToth> ok [23:07:13] <Raylton> GorillaWarfare, yes ... but it certainly did not need more than once [23:07:35] <AzaToth> are they all required if they are in a set? [23:07:44] <James_F> AzaToth: Though not all are required... ha, no. [23:07:56] <James_F> AzaToth: It just means that the user would be encouraged to add them as a block. [23:08:00] <AzaToth> ok [23:08:02] <GorillaWarfare> Raylton: I'm not sure I understand where the URL would be added... [23:09:22] <AzaToth> James_F: are there any template on enwiki atm that has a proper templatedata made? [23:09:43] <James_F> AzaToth: Think the demonstration ones I've done so far have mostly been on MW.org [23:09:46] <James_F> AzaToth: One moment. [23:09:52] <AzaToth> ok [23:10:34] <James_F> AzaToth: https://en.wikipedia.org/wiki/Template:Tracked has it. [23:11:05] <Raylton> GorillaWarfare, in extension? [23:11:59] <GorillaWarfare> Raylton: Right, but for the extension to know which pages it should add a navbar to, it needs to know where the JSON is stored. Are you suggesting we add the URL to each subpage of a book? [23:12:29] <James_F> AzaToth: https://www.mediawiki.org/wiki/Template:Wikimedia_engineering_project_information/doc is the most comprehensive one I've made so far. [23:12:43] <AzaToth> James_F: where will the templatedata been shown? [23:12:45] <James_F> AzaToth: BTW, if you were going to start, I'd suggest the Cite templates as ripe for improvement. [23:13:01] <Krinkle> AzaToth: The description you cited earlier means that you can either have one <templatedata> on the Template page itself, *or* you can have it on e.g. another page (most commonly a subpage like /doc) and transclude it into the Template page (likely in a <noinclude> section to avoid it from messing up stuff) [23:13:20] <James_F> AzaToth: The block will be shown where you write it (and wherever that's transcluded); the hints are used for the VisualEditor's template dialog, and whoever else finds it useful. [23:13:29] <AzaToth> is templatedata doing to be used by VE in the future? [23:14:08] <Raylton> yes... but when you click next or prev the place where is stored need not disappear from the url. [23:14:26] <James_F> AzaToth: It's used in VE right now. [23:14:34] <Raylton> GorillaWarfare, ^ [23:14:35] <AzaToth> on the one in enwiki as well? [23:14:44] <James_F> AzaToth: Yes. [23:14:51] <James_F> AzaToth: https://www.mediawiki.org/wiki/Help:VisualEditor/User_guide#Editing_templates [23:15:28] <GorillaWarfare> Raylton: Oh, do you mean hanging on to it as part of a query string in the URL? [23:15:44] <AzaToth> James_F: http://i.imgur.com/Ub7vtPt.png [23:16:00] <AzaToth> didn't get any list of parameters for "Tracked" ヾ [23:16:23] <James_F> AzaToth: Improvements to use more of the TemplateData block coming, as mentioned - developers working on it right now. [23:16:50] <Raylton> GorillaWarfare, sorry ... I do not know what that means [23:16:51] <James_F> AzaToth: ""<James_F> AzaToth: Yes; that is indeed what you're distracting my developers from doing. :-)" [23:16:59] <GorillaWarfare> Raylton: http://en.wikipedia.org/wiki/Query_string [23:17:11] <AzaToth> James_F: sooorrryy ツ [23:17:23] <James_F> AzaToth: No worries. :-) [23:17:28] <mwalker> GorillaWarfare: I would be worried about caching if it was part of the query string [23:17:30] <James_F> AzaToth: Sorry we've not got it entirely plugged-together yet. [23:17:35] <AzaToth> hehe [23:17:38] <GorillaWarfare> mwalker: Plus there's the issue of finding it in the first place [23:17:38] <AzaToth> no probs [23:17:48] <Raylton> GorillaWarfare, yes [23:18:04] <AzaToth> James_F: and that "Page notice" is also something they are worknig on right? [23:18:09] <mwalker> GorillaWarfare: well; if you entered the book from the index and navigated it soley with the navbar; it's possible [23:18:14] <gerrit-wm_> New patchset: Krinkle; "ApiFancyCaptchaReload: Replace use of PHP shorttag" [mediawiki/extensions/ConfirmEdit] (master) - https://gerrit.wikimedia.org/r/71146 [23:18:39] <James_F> AzaToth: In slower time; that's an enwiki-only problem because of the broken way enwiki decided to implement that. [23:18:46] <GorillaWarfare> mwalker: Right, but that's not exactly a common use-case :P [23:18:54] <GorillaWarfare> I think I like the category idea best [23:18:56] <AzaToth> James_F: always blame enwiki ツ [23:19:07] <GorillaWarfare> Although it doesn't do a whole lot to solve the one page, multiple books issue [23:19:21] <James_F> AzaToth: We'll put editing the edit notice inside the "page settings" dialog in time, but not tonight. :-) [23:19:48] <gerrit-wm_> Change merged: jenkins-bot; [mediawiki/extensions/ConfirmEdit] (master) - https://gerrit.wikimedia.org/r/71146 [23:20:07] <mwalker> GorillaWarfare: well; if it was a prefixed category you could at least assign it to multiple books -- and then it's a UI issue in the navbar (which you could make easier by setting a cookie which told the navbar what book it was and then it could preferentially navigate within that book) [23:20:23] <AzaToth> James_F: due to the rollout at july 1th, I assume you are too busy [23:20:35] <Raylton> mwalker, how works 'category idea'? [23:20:39] <GorillaWarfare> Hm, that's true [23:20:43] <James_F> AzaToth: Yes. :-) [23:21:01] <GorillaWarfare> Raylton: mwalker spoke of adding a category like [[Bookmanagerv2-booktitle]] to each page that is part of a book [23:22:11] <mwalker> Raylton: I'm thinking the bookmanager would add the category to the page when the user adds the page to the book [23:22:30] <mwalker> Raylton: and then, I think, on page render; you have access to the page categories [23:22:52] <AzaToth> James_F: is this a known bug? http://imgur.com/4ZJTKOm [23:22:57] <AzaToth> minor though [23:23:08] <James_F> AzaToth: Known and fixed today. :-) [23:23:10] <AzaToth> k [23:23:29] <James_F> AzaToth: https://bugzilla.wikimedia.org/show_bug.cgi?id=49818 specifically. [23:24:49] <AzaToth> James_F: that's not the same [23:25:21] <James_F> AzaToth: It occured in more places than just the template selection process. [23:25:31] <AzaToth> ok [23:25:53] <Raylton> GorillaWarfare, mwalker This seems to be even better ... but categorize books seems to be something that was not in the scope. But if you think you will not overload... ok [23:26:23] <GorillaWarfare> Raylton: I think if we include adding the category as part of something the manager does automatically, it shouldn't be too much additional work [23:26:30] <GorillaWarfare> I definitely don't plan on trying to do it by hand :P [23:27:14] <AzaToth> James_F: I can't edit interwiki links in VE? [23:27:37] <YuviPanda> GorillaWarfare: so, if you are looking to track extra data, you can also just create new database tables and store data in them :) [23:28:01] <GorillaWarfare> Heh.. [23:28:27] <Reedy> AzaToth: Because they're from Wikidata? [23:28:57] <YuviPanda> Reedy: can you +2 https://gerrit.wikimedia.org/r/#/c/70780/ [23:28:57] <YuviPanda> ? [23:28:59] <AzaToth> Reedy: a manual interwiki link on https://www.mediawiki.org/wiki/User:AzaToth [23:29:05] <mwalker> given the dynamic nature of the data -- I don't know if it'll be worth it to create additional database objects... e.g. if it will be just as fast to request the metadata through page objects [23:29:08] <Reedy> YuviPanda: Noope [23:29:17] <AzaToth> damn, I fnucked VE up [23:29:24] <Reedy> gj [23:29:25] <YuviPanda> Reedy: can you +2 ops repos in general? [23:29:31] <gerrit-wm_> Change merged: Mwalker; [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71128 [23:29:33] <YuviPanda> I mean, not this particular one. [23:29:40] <Reedy> Just mediawiki-config AFAIK [23:29:44] <gerrit-wm_> Change merged: Mwalker; [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71142 [23:29:49] <YuviPanda> ah, okay [23:30:17] <James_F> AzaToth: Not yet, sorry; we didn't get around to that. [23:30:35] <AzaToth> Reedy: I had a template marked, and clicked on the "Link" button, first a blank small box was shown where the jigsaw piece usually shown for templates, and now I can't edit the template at all [23:30:38] <James_F> AzaToth: Bug https://bugzilla.wikimedia.org/show_bug.cgi?id=49316 [23:30:52] <AzaToth> James_F: I see [23:31:01] <YuviPanda> mwalker: well, without tables you can't really do any queries [23:31:06] <YuviPanda> mwalker: but that might be okay [23:31:07] <James_F> Reedy: Wikidata does langlinks, not interwikis, silly. ;-) [23:31:13] <James_F> Reedy: (Yeah, everyone switches those two around.) [23:31:20] <Reedy> Pffffffft [23:31:22] <Reedy> Close enough [23:31:25] <YuviPanda> mwalker: on UploadWizard, I have a minimal database with just the properties I want to 'query', and update it whenever the page is saved, deleted, etc [23:31:37] <YuviPanda> mwalker: so the primary source is still the Content, but I've tables for querying too [23:33:36] <AzaToth> Reedy: even after making a new template I cant edit it now [23:34:14] <AzaToth> TypeError: focusedNode is null [23:34:48] <AzaToth> can't find a bug for it [23:35:21] <AzaToth> unless you recently fixed it [23:35:49] <legoktm> csteipp: could i get you to re-review https://gerrit.wikimedia.org/r/#/c/48608/ ? [23:36:10] <mwalker> YuviPanda: what are you querying in UW? [23:36:22] <YuviPanda> mwalker: right now, enabled/disabled status of campaigns [23:36:32] <mwalker> gotcha [23:37:57] <Raylton> GorillaWarfare, mwalker. I have to get going now. then you guys can let me know what you decide? [23:38:22] <Raylton> by email? [23:38:43] <GorillaWarfare> Raylton: Definitely [23:39:47] <Raylton> Thank you and good night everyone :) [23:41:15] <gerrit-wm_> New patchset: Katie Horn; "fundraier submodule update Some last-minute changes before we start running campaigns again on 7/1. DI: adding gateway header to antimessages, and disabling GC currencies that Wells Fargo does not support" [mediawiki/core] (fundraising/1.22) - https://gerrit.wikimedia.org/r/71148 [23:42:02] <csteipp> legoktm: I will try to get to it. Maybe ping me monday if I don't manage to do it this weekend? [23:42:12] <gerrit-wm_> Change merged: Katie Horn; [mediawiki/core] (fundraising/1.22) - https://gerrit.wikimedia.org/r/71148 [23:42:16] <legoktm> csteipp: thanks, will do. [23:42:28] <AzaToth> filed https://bugzilla.wikimedia.org/show_bug.cgi?id=50406 for that James_F as I was unable to detect any bugs for that [23:43:12] <James_F> AzaToth: Hmm. [23:44:29] <mwalker> GorillaWarfare: what YuviPanda is getting at is performance -- a DB is queriable where a page is not; one of the things to be looking for when designing this will be page render time so if you see an opportunity to not query a page, but to query a backend DB you should probably take it. I am not entirely sure you'll be able to given that your nav bar needs all the content in the metadata page -- but you might think of something. [23:44:54] <AzaToth> James_F: for me, testing is not monkey testing, it's gorilla testing [23:45:57] <GorillaWarfare> mwalker: Indeed, I will keep that in mind [23:48:28] <James_F> AzaToth: :-) [23:53:12] <AzaToth> James_F: bugs/feature requests for templatedata, should it be filed under it's extension or under visualeditor? [23:53:45] <James_F> AzaToth: Is the request for VE to implement something TD does already? Then VE. Otherwise, TD. [23:54:07] <AzaToth> k [23:56:52] <GorillaWarfare> mwalker: Alright, so I guess for now I'll get looking at ContentHandler with the intention of using *.book to identify the JSON pages [23:57:07] <GorillaWarfare> mwalker: And I'll send an email to Aarti and her mentors to discuss the HTMLForm and ProofreadPage [23:57:36] <mwalker> *thumbs up* [23:57:38] <GorillaWarfare> And I guess update the RfC/documentation pages with any relavent info from this conversation

[00:00:19] New patchset: Kaldari; "Adding support for primary links in notification" [mediawiki/extensions/GettingStarted] (master) - https://gerrit.wikimedia.org/r/70945 [00:12:07] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70435 [00:30:50] New review: Mattflaschen; "Right now, if Vagrant reassigns the port, it won't be 8080. Then, files like puppet/modules/mediawi..." [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/63783 [00:31:48] New patchset: Ori.livneh; "Enable VisualEditor on NS_USER by default when role is enabled" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/70950 [00:33:46] does anyone talk on here? [00:34:15] yes [00:34:20] I checkout #wikimedia on freenode, no one is responding there either. Where's the best play to go for development questions? [00:34:53] #mediawiki and #wikimedia-dev are probably best [00:35:20] Well maybe you can help me... I'm trying to do a rollback using the MediaWiki API. I have the rollback token and am doing a POST with jQuery, but I'm not authenticated, which must be required [00:35:28] [http://www.mediawiki.org/wiki/API:Rollback] doesn't cover this [00:36:02] can you pastebin your code somewhere? [00:36:35] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [00:36:46] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/70950 [00:37:12] http://pastebin.com/dFrGj3uG [00:37:59] going by [http://www.mediawiki.org/wiki/API:Rollback], but you can't just rollback, surely the user (me) must be authenticated? have to pass in my token as well? [00:38:15] I don't see this covered anywhere in the documentation [00:39:05] also that's CoffeeScript by the way, just think of it as pseudocode, no syntax errors there [00:40:02] question is more about authentication, I must be able to make a POST to some URL as shown in the pastebin, and pass my token to perform the rollback, since I have rights [00:40:16] I'm assuming that's how it would work [00:42:48] New patchset: Ori.livneh; "Make FORWARDED_PORT a Facter fact" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/63783 [00:43:33] test [00:43:45] MusikAnimal_: I'll take a look; be patient :) [00:43:55] ori-l: sure thing thanks [00:43:56] Yippie, build fixed! [00:43:56] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_10 build #20: FIXED in 36 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_10/20/ [00:44:10] MusikAnimal: It looks to me like authentication is handled via session cookies. You need to hit the login endpoint, then pass the session cookie to the query endpoint to get a rollback token and then make the actual rollback endpoint call. [00:44:28] ......he left [00:44:34] nope I'm here [00:44:35] bd808: thanks anyhow :/ [00:44:38] oh [00:44:44] that was me on IRC client [00:44:49] went to the same room lol [00:45:20] ok let me find the doc on login endpoint [00:45:25] MusikAnimal: start at https://www.mediawiki.org/wiki/API:Login [00:45:56] MuskiAnimal: then https://www.mediawiki.org/wiki/API:Rollback#Token [00:46:27] cool so I can just make that call as in my pastebin and tag "&enwiki_session=my_token" at the end? [00:46:55] I should just be able to look at my cookies right now, since I'm logged in through the website [00:51:36] any documented examples you know of? [00:51:43] I'm still getting a parse error [00:52:21] I got the rollback token with a normal query to a revision, as with [http://en.wikipedia.org/w/api.php?format=json&action=query&titles=User:MusikAnimal/sandbox&prop=revisions&rvprop=timestamp|user|size|parsedcomment|tags|sha1|ids&rvtoken=rollback&rvdiffto=prev] [00:52:26] ^ goes to my sandbox [00:54:05] instead of success I get the same response as [http://en.wikipedia.org/w/api.php?action=rollback&title=User:MusikAnimal/sandbox&user=FJkjf&token=2424&enwiki_session=2424] [00:54:26] ^ jibberish parameters [01:03:11] MusikAnimal: do you have a pastebin of the error somewhere? [01:04:18] wait the header does say badtoken [01:05:00] Let me try logging in using the API, I just copied the cookie values from when I was logged in through the website [01:05:44] bd808: this is the rollback token I want right? [http://en.wikipedia.org/w/api.php?format=json&action=query&titles=User:MusikAnimal/sandbox&prop=revisions&rvtoken=rollback] [01:06:47] MusikAnimal: Yes, assuming you're not getting '{"warnings":{"revisions":{"*":"Action 'rollback' is not allowed for the current user"}}' in response [01:07:00] nope [01:09:53] New patchset: Ori.livneh; "Allow command-line arguments to be read from a file" [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70956 [01:10:54] then yes. You'd want to pass query.pages.38854251.rollbacktoken on as the "token" param in your call to action=rollback [01:14:16] bd808: yep I used that token, I look at the request cookies, all is as it should be, I'm clearly authenticated, but I still get "MediaWiki-API-Error:badtoken" in response header [01:15:07] MusikAnimal: are you url encoding the token properly? [01:15:43] Your pastebin just shows a literal being inserted [01:16:18] bd808: yes I'm hardcoding it in [01:16:50] db808: e.g. #{rollback_token} would be abcd1234 (some hex code) [01:17:03] no weird characters or anything [01:18:18] ok. I was just noticing that the example token on https://www.mediawiki.org/wiki/API:Rollback#Rolling_back_pages contains a '+' that needed to be escaped [01:19:01] I was about to ask about that, forgot about the + [01:19:18] I took it out earlier, I just tried escaping it with \+ [01:19:52] if I get back the token as JSON, it ends in "+\\" and with XML I see just "+\" [01:20:08] Nope. You need to URL encode it the plus (+) as "%2B" [01:21:00] The "\\" in the raw json is just a single literal backslash encoded to fit in a javascript string [01:21:47] So change your hardcoding to "abcd1234%2B\\" (with the real hex obviously) [01:22:21] yeah just did... still no dice... badtoken :-/ [01:22:46] backslash doesn't need to be decoded? [01:23:06] %5C [01:23:34] nope still doesn't work [01:23:35] ahhh! [01:23:43] You can encode if you'd like, but it's a safe character for a query string parameter [01:24:53] Change merged: jenkins-bot; [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/70666 [01:25:07] say, per my pastebin, the "user" param should be the author of the edit I'm trying to rollback, right? [01:26:05] MusikAnimal: "user: The author of the last revision." [01:26:25] ok [01:27:12] lastly I'm using JSONP, which I guess is required because of cross-origin control [01:27:17] New review: Mattflaschen; "Looks fine. You may want to add descriptions to the ones without it, but it's not related to this c..." [mediawiki/extensions/EventLogging] (master) C: 2; - https://gerrit.wikimedia.org/r/70956 [01:27:40] Change merged: jenkins-bot; [mediawiki/extensions/EventLogging] (master) - https://gerrit.wikimedia.org/r/70956 [01:27:41] sometimes jQuery.ajax is flaky, I'll try jQuery.post, the shorthand [01:28:50] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [01:28:59] you might try passing your parameters in as a data object too. Php may be ignoring query string parameters on a POST request [01:29:16] * bd808 time for dinner. Later folks. [01:30:20] thanks for the help! [01:30:23] dah he's gone [01:32:57] New patchset: Mwalker; "Add some tests to cover do_transaction" [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70958 [01:32:58] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70959 [01:32:58] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70960 [01:32:58] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70961 [01:32:58] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70962 [01:32:59] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70963 [01:32:59] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70964 [01:32:59] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70965 [01:32:59] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70966 [01:33:00] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70967 [01:33:00] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70968 [01:33:01] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70969 [01:33:01] New patchset: Mwalker; "Localisation updates from http://translatewiki.net." [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70970 [01:33:02] New patchset: Mwalker; "mv donationinterface.php DonationInterface.php" [mediawiki/extensions/DonationInterface] (deploy-payments_1.22) - https://gerrit.wikimedia.org/r/70971 [01:33:03] gah [01:33:07] that's not what I meant to do [01:33:09] freaking gerrit [01:36:53] wheeeeeeeeeeee [01:37:26] * YuviPanda hands mwalker a gerrit [01:37:33] it's gotta be a unit of something :) [01:37:43] it sort of sounds like ferrit [01:37:59] *ferret [01:38:02] which apparently make good pets [01:38:12] grey ferret [01:38:58] why would he want a gay ferret? [01:39:30] well, why would we want a gerrit? [01:39:48] that ones easy -- because it's better than MW:CodeReview and SVN [01:40:08] I'm not sure why a gay/grey ferrit would be any better than a normal ferret [01:40:26] however! according to the wiki page -- a group of ferrets is actually called a 'business' [01:40:38] so... if you get enough of them they might form like a ferret mafia [01:41:14] hahaha [01:41:40] well, only if they are always... underground :) [01:43:20] mwalker: I'm going to buy you three ferrets and then get ALL UP IN YOUR BUSINESS [01:43:33] office ferrets! [01:43:56] chillin on the hammock, pettin a ferret [01:44:12] With some champagne and dollar bills [01:44:41] I need to stop reading this page: "Male ferrets are called hobs; female ferrets are jills. A spayed female is a sprite, a neutered male is a gib, and a vasectomised male is known as a hoblet. Ferrets under one year old are known as kits. A group of ferrets is known as a "business",[35] or historically as a "fesnyng".[36]" [01:45:10] I feel like someone took far too much time thinking about these things [01:45:21] * YuviPanda gives mwalker a fesnyng of kit hoblets [01:45:52] * mwalker attempts to teach them to sing [01:46:39] * marktraceur starts a restaurant serving ferret so when someone asks for sprite he can charge them for an entree [01:46:49] btw marktraceur; I'm deploying the renamed DonationInterface.php now [01:46:56] mwalker: Ah, good luck [01:47:02] ori-l: ^^ [01:51:10] New patchset: Mwalker; "Updating all the extensions!" [mediawiki/core] (fundraising/1.22) - https://gerrit.wikimedia.org/r/70975 [01:51:36] Change merged: Mwalker; [mediawiki/core] (fundraising/1.22) - https://gerrit.wikimedia.org/r/70975 [01:53:18] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [02:00:41] ori-l: zeljkof have you run qa/browsertests under vagrant but with OS X? [02:00:47] without xquartz? [02:00:52] * YuviPanda is trying to set them up for UW [02:01:32] YuviPanda: I am not sure I understood the question [02:01:40] what are you trying to do? [02:01:52] zeljkof: I'm trying to run the selenium tests for UploadWizard [02:02:05] using Vagrant [02:02:45] YuviPanda: on mac? [02:02:50] zeljkof: yes, on mac [02:03:30] hmm, let me just download xquarts [02:03:31] * YuviPanda starts [02:03:39] YuviPanda: why don't you just install xquartz? :) [02:03:52] zeljkof: well, was hoping for a way to just turn on X in vagrant [02:03:57] but installing xquartz now :) [02:04:32] YuviPanda: there is probably a way to do it, I remember seeing something at wikitech, but ori-l is the one to ask how to do it [02:04:53] zeljkof: hmm, if it works with zquartz... [02:04:54] err [02:04:55] xquartz [02:05:00] should be good enough [02:05:04] assuming that is path of least resistance [02:05:19] YuviPanda: it should would, that should be the easiest way to set everything up [02:05:38] yeah, running puppet on vagrant now, xquartz also downloading [02:05:42] let's see how this goes :) [02:06:09] YuviPanda: if you did not use mw-vagrant before, it takes some time to download and configure everything [02:06:22] zeljkof: nah, I've been using Vagrant for a week now [02:06:27] just added the browsertests role [02:06:38] YuviPanda: that should take just a few minutes then [02:06:41] yeah [02:06:50] let me know if it does not work [02:07:01] puppet runs are fairly slow on my laptop, because it is a rather old Air [02:07:09] zeljkof: will do :) [02:07:32] anybody knows if the testing event started already? http://www.mediawiki.org/wiki/Meetings/2013-06-27 [02:07:55] I do not see Chris or Quim here [02:09:03] qgil: did the testing event start? [02:09:19] youtube link available? :) [02:09:28] 1 min http://www.youtube.com/watch?v=0dg6B5i0RM8 [02:10:36] qgil: thanks! [02:14:07] and we are live! :) [02:15:00] I lost backscroll, can you paste the link again? [02:18:53] YuviPanda: for the testing event? http://www.mediawiki.org/wiki/Meetings/2013-06-27 [02:19:08] ty [02:20:37] I'm really curious what a/b testing wikilove content would lead to. [02:21:26] New patchset: Yaron Koren; "Changed "text replace" to the more grammatical "text replacement"" [mediawiki/extensions/ReplaceText] (master) - https://gerrit.wikimedia.org/r/70976 [02:24:32] qgil: thanks, good work as a sound engineer too. It's harder than it looks. :-) [02:24:54] chip makes it look easy [02:27:24] New patchset: Yaron Koren; "Removed parsing of search and replacement strings within edit summary" [mediawiki/extensions/ReplaceText] (master) - https://gerrit.wikimedia.org/r/70977 [02:28:01] Change merged: Yaron Koren; [mediawiki/extensions/ReplaceText] (master) - https://gerrit.wikimedia.org/r/70976 [02:28:03] qgil: chrismcmahon is hard to understand, can you check the microphone? [02:28:17] Change merged: Yaron Koren; [mediawiki/extensions/ReplaceText] (master) - https://gerrit.wikimedia.org/r/70977 [02:31:23] hey, anyone got the details for the meetup livestream? [02:31:40] I wish, http://www.mediawiki.org/wiki/Meetings/2013-06-27 doesn't have details [02:31:51] New patchset: Anomie; "Add 'viewmyprivateinfo', 'editmyprivateinfo', and 'editmyoptions' rights" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67876 [02:33:02] http://www.youtube.com/watch?v=0dg6B5i0RM8 [02:33:21] Create an account, visit http://en.wikipedia.beta.wmflabs.org/wiki/User:Selenium_user [02:33:27] or some other user page [02:36:03] The Google Hangout is something like https://plus.google.com/hangouts/_/535c8c46ce6afe2974b275dc57404f2d97f66038 , but can't get there [02:36:14] what's the link to the etherpad? [02:36:55] dunno. qgil, please update the meeting page and this channel with the hangout, the etherpad, etc. [02:37:08] 1 min sorry [02:37:44] kadoo: http://etherpad.wikimedia.org/currentmeeting [02:37:51] thanks! [02:38:07] http://etherpad.wikimedia.org/currentmeeting [02:39:36] wiki page and meetup page update [02:39:53] can you hear now chris better? [02:41:19] is there an easy way to slow down the selenium test so we can get a better sense of each of the scenarios? [02:41:44] grantbow, our technitian was not available and I only knew 10 minuts before starting... It's the first time I do all this. :) [02:43:22] kadoo, I asked, and of course the idea of the software is to run those tests automatically as fast as possible... [02:43:27] qgil: :-) [02:43:46] qgil's voice sounds a little fuzzy on my end - any one else getting that? [02:44:18] kadoo, what you see in the testing are those four tests you can read now in the screen [02:44:26] qgil: chrismcmahon is still hard to understand, maybe he is holding the microphone at the angle it does not pick up correctly [02:44:47] arad000: same here, Chris is hard to understand [02:45:12] im on youtube, is the goog hangout clearer? [02:45:22] I bet the sound is the same [02:46:04] qgil: the sounds is way better now, thanks! :) [02:46:12] ah! was better for a sec [02:46:17] and then it reverted back :P [02:46:30] arad000: it is better for me [02:46:44] better now :) thanks all! [02:46:50] zeljkof, glad to hear. :) [02:47:07] and hey, anyone else getting failures when you run bundle exec cucumber features/wikilove.feature ? [02:47:16] when I run the tests, it looks like it fails when it tries to login as the user [02:48:15] so... on ubuntu; Ruby 1.9.1 is telling me that to override the debian default behavior I should set REALLY_GEM_UPDATE_SYSTEM when I run ruby system --update. Setting that to true however doesn't seem to do anything but give me the same error -- any immediate solutions before I investigate further? [02:48:53] mwalker, I just ignored that step and that seemed to work fine [02:48:54] arad000: can you paste the error message somewhere and provide the link? [02:48:55] please ping me when you want me to ask a question to Chris, as opposed to just commenting here. Thank you. [02:49:12] mwalker: on ubuntu ignore updating rubygems [02:49:40] dreamnid: ah; so if I do that and install bundler via apt; then it complains about missing rake -- which is not in apt [02:49:45] feel free to ping me, I am remote and I should be able to help you with the browser tests [02:50:25] mwalker: can you post your terminal output somewhere [02:50:51] it's just one line, so I'll do it here: "Could not find rake-10.1.0 in any of the sources" [02:51:10] bundle exec cucumber features/wikilove.feature Using the default profile... secret.yml file at /private/wmf/ or config/ is required for tests tagged @login *** DEPRECATION WARNING *** You are calling a method named wait_until_present at /home/arthi/code/wiki/wikilove-test/qa-browsertests/features/support/pages/login_page.rb:22:in `login_with'. *** This method does not exist in page-object so it is being passed to the driver. * [02:51:34] mwalker: I need the command that causes the error too [02:51:48] zeljkof: bundle exec cucumber features/wikilove.feature [02:52:03] arad000: please read the error message, what does it say? :) [02:52:16] mwalker: you have to run "bundle install" first [02:52:29] mwalker: (remote quotes) [02:52:35] mwalker: (remove quotes) [02:52:40] zeljkof: oh; duh -- thanks [02:53:05] that's what I get for skipping steps [02:53:21] arad000: let me give you a hint :) "secret.yml file at /private/wmf/ or config/ is required for tests tagged @login" [02:53:32] arad000: let me know if you have further questions [02:53:48] ha ha ha -_- thanks :) [02:54:20] qgil: please remind Chris to hold the microphone correctly, sometimes he is hard to understand [02:54:31] ok [02:55:31] he is not talking now :) [02:55:52] better? [02:56:00] qgil: and also please make sure to repeat the question, we can not hear the question sometimes [02:56:32] sound ok? [02:56:40] qgil: youtube stream is a few minutes late, he just started talking, sounds great, thanks :) [02:57:31] New review: CSteipp; "Tyler: I would actually say the User class handles a good bit of controller logic (dnsblacklisting?)..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67876 [02:59:00] I run `bundle exec cucumber features/wikilove.feature`, Firefox comes up, it enters my username, *two* characters of my password (it has 4 chars) and then fails to login. [02:59:34] Spagewmf, Does your pass have any special characters? [02:59:56] spagewmf: try running the test a few times, does it behave the same every time? [03:00:07] Specifically # characters make the rest of the line a comment. [03:00:43] nope, just a-z. Every single time I see my username perfectly, but my password just two dots. Weird [03:01:39] Yah, that's super weird. Can you dump the sanitized contents of your secret.yml here? [03:01:43] It should just be ~1 line [03:02:07] (i.e. Mine looks like "mediawiki_password: selenium" in it's entirety. [03:02:36] $ cat config/secret.yml [03:02:36] mediawiki_password: abcd [03:03:21] thx for repeating question! [03:03:55] :) [03:04:24] chrome dev toolkit is another one for dom [03:04:32] op, he already said that:P [03:06:15] `bundle exec cucumber features/pdf_readonly.feature` works perfectly. [03:07:39] any questions? [03:08:49] i get a lot more deprecation warnings than chris does [03:10:01] also, hi zeljkof! :) [03:13:09] judytuna, I think he must have the depreciation warnings turned off or an older version of cucumber. [03:13:44] yeah, that's what i figured too estsauver xD do you see a bunch of 'em as well? they don't cause test failures or anyhing [03:14:07] great to see you zeljkof!!! [03:14:11] judytuna: please ignore deprecation warnings, we are working on it :) https://bugzilla.wikimedia.org/show_bug.cgi?id=49867 [03:14:16] okay =) [03:14:25] rfarrand: hi :) [03:14:29] 'm judytuna Yah that's what I'm seeing. [03:15:25] judytuna: sorry, this is the correct bug https://bugzilla.wikimedia.org/show_bug.cgi?id=46893 [03:15:45] spagewmf: are you at the meeting? you should ask chris for help [03:16:58] He's busy. It's OK. I blame teh Ubuntu :) [03:17:55] is there a way to lint the scenario before you run it? [03:18:42] ... (and/or lint the step file) just to make sure everything exists [03:19:21] mwalker: I like to just run it and it'll tell me what i'm missing [03:19:45] ah; but that takes time :) [03:19:59] bugs for Ubuntu encouraged by group ubuntu-california.org ;-) [03:20:06] judytuna: will it at least tell you ALL the steps/things you're missing, or will it die at the first oen? [03:21:11] it'll tell you all of the pending ones! so you can copy and paste [03:21:24] at least i'm pretty sure D: i will be very embarrassed if i'm wrong [03:21:45] spagewmf, I can help and I'm not very busy. I'm the guy who's about to stand off by the side. [03:22:04] judytuna: I'll take your initial word for it -- my enviornment is still a little wonky so I wont test it now [03:23:10] cucumber has pretty nice error messages; catch is you gotta read em :P [03:23:44] that's not really a catch -- I feel a bigger catch is that if you have a HUGE test file; you have to run it first before it tells you there's a failure [03:25:41] question break, any questions? i will read them over the mic [03:25:51] hah -- apparently the lint tool for all of this is called foodcritic http://acrmp.github.io/foodcritic/ [03:28:06] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6 build #77: FAILURE in 11 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_6/77/ [03:30:59] New patchset: Cmcmahon; "test steps written at WMF with volunteers!!" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/70979 [03:32:12] New review: Cmcmahon; "these tests are pretty good! I'll clean up the extra white space later" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/70979 [03:32:46] handy link to see the code: https://gerrit.wikimedia.org/r/#/c/70979/ [03:33:18] last chance for qs! [03:33:52] what can we do to get involved/contribute? [03:34:18] arad000: there's a triage, i'll pull it up [03:34:38] arad000: http://www.mediawiki.org/wiki/Quality_Assurance/Browser_testing/Test_backlog [03:34:46] arad000: also, http://www.mediawiki.org/wiki/QA/Browser_testing [03:34:51] arad000: join your qa list and introduce yourself https://lists.wikimedia.org/mailman/listinfo/qa [03:35:00] arad000: that is the best way to get started [03:35:05] the qa list is super friendly!!!!!! =D [03:35:09] sweeeet! [03:35:12] woooo! [03:35:25] yayaya thanks judy! [03:35:35] mwalker: to lint the file, you can run it with -c option, like this [03:36:10] $ ruby -c features/step_definitions/wikilove_steps.rb [03:36:11] Syntax OK [03:36:37] some stranger figured out my problem! [03:36:40] secret_yml_locations = ['/private/wmf/', 'config/'] [03:37:09] spagewmf: what was the problem? [03:37:19] turns out I had a /private/wmf/secret.yml from my previous attempt to write tests back in February. [03:37:34] `locate secret.yml` was my friend [03:37:58] spagewmf: :) [03:38:29] Hi zeljkof/spagewmf, I'm earl and I'm the secret friend. [03:39:09] qgil, rfarrand: thanks for hosting [03:39:28] hi estsauver and thank you :) [03:39:28] its all qgil! :) [03:39:45] thanks qgil! and all who anserwed q's [03:40:11] robla spoke, then ryan lane and now Timo Tijhof [03:40:40] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7 build #77: FAILURE in 12 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_7/77/ [03:40:51] Is there a meeting going on? [03:41:05] Elsie: https://www.youtube.com/watch?v=0dg6B5i0RM8 [03:41:50] also https://www.mediawiki.org/wiki/Meetings/2013-06-27 [03:42:44] Project browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8 build #115: FAILURE in 11 min: https://wmf.ci.cloudbees.com/job/browsertests-commons.wikimedia.beta.wmflabs.org-windows-internet_explorer_8/115/ [03:50:21] heh, in the middle of the meeting, the browsertests are failing [03:52:16] they timed out [03:57:16] New patchset: Zfilipin; "Updated Firefox to version 21" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/70982 [04:04:57] New patchset: Yaron Koren; "Small fix for logic in setArticleHeader(), and improved that method's comments" [mediawiki/extensions/ApprovedRevs] (master) - https://gerrit.wikimedia.org/r/70983 [04:06:49] Change merged: Yaron Koren; [mediawiki/extensions/ApprovedRevs] (master) - https://gerrit.wikimedia.org/r/70983 [04:06:51] New patchset: Catrope; "Update VisualEditor to master" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70984 [04:07:20] New patchset: Catrope; "Update VisualEditor to master" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/70985 [04:08:21] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70984 [04:08:37] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/70985 [04:13:46] New patchset: Zfilipin; "Fix logging in at test2.wikipedia.org" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/70986 [04:22:02] New patchset: Aaron Schulz; "profiler: log slow methods that worsen DB locks in transactions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/70923 [04:33:43] New patchset: Rjain; "Added the db schema" [mediawiki/extensions/Annotator] (master) - https://gerrit.wikimedia.org/r/70794 [04:41:25] New patchset: Catrope; "Add WikibaseDataModel to unbreak wmf8" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70987 [04:41:37] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/70987 [04:53:41] New patchset: Hiong3-eng5; "SpecialPage:OWDownloads" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/70325 [04:54:45] um....anyone working on the fatal exceptions? [04:54:55] Yes. [04:55:01] okay thanks :) [04:55:15] :) [04:55:37] BobTheWikipedian: -tech is usually the more relevant channel :) [04:55:44] ah [05:06:13] great job on getting everything back so quick guys \o/ [05:07:26] sorry for spamming [05:15:31] New patchset: Physikerwelt; "Enabling MathJaX in LaTeXML rendering mode except for Firefox" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [05:19:42] Project browsertests-en.wikipedia.org-linux-chrome build #151: FAILURE in 3 min 33 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-linux-chrome/151/ [05:23:34] Project browsertests-en.wikipedia.org-linux-firefox build #140: FAILURE in 3 min 51 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-linux-firefox/140/ [05:28:34] Project browsertests-en.wikipedia.org-windows-internet_explorer_10 build #15: FAILURE in 4 min 59 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_10/15/ [05:32:30] New review: Physikerwelt; "mh... where do I find the correct jshint config file" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [05:33:48] Project browsertests-en.wikipedia.org-windows-internet_explorer_6 build #114: FAILURE in 5 min 14 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_6/114/ [05:34:19] New patchset: Physikerwelt; "Enabling MathJaX in LaTeXML rendering mode except for Firefox" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [05:36:23] New review: Physikerwelt; "Maybe option 'math' should be marked as deprecated now." [mediawiki/extensions/Math] (master) C: 1; - https://gerrit.wikimedia.org/r/69873 [05:43:46] New patchset: Physikerwelt; "Enabling MathJaX in LaTeXML rendering mode except for Firefox" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [05:43:58] Yippie, build fixed! [05:43:58] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_9 build #99: FIXED in 45 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_9/99/ [05:44:54] Project browsertests-en.wikipedia.org-windows-internet_explorer_8 build #118: FAILURE in 5 min 50 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_8/118/ [05:49:21] New review: Physikerwelt; "(1 comment)" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [05:51:18] Project browsertests-en.wikipedia.org-windows-internet_explorer_9 build #106: FAILURE in 7 min 20 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_9/106/ [08:36:02] New patchset: Pastakhov; "Add while loops (version 0.5.0)" [mediawiki/extensions/Foxway] (master) - https://gerrit.wikimedia.org/r/69836 [08:38:46] Change merged: jenkins-bot; [mediawiki/skins/erudite] (master) - https://gerrit.wikimedia.org/r/69319 [08:41:44] New patchset: Nilesh; "Test commit" [test/mediawiki/extensions/examples] (master) - https://gerrit.wikimedia.org/r/70989 [08:54:14] New patchset: Hashar; "mw doc only triggered for master/release branches" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70990 [08:55:03] qchris: morning :) [08:55:12] seems the bugzilla pluginin Gerrit no more report back in bugzilla :( [08:55:17] hashar: :-] [08:55:22] What? [08:55:23] :-( [08:55:24] qchris: example https://gerrit.wikimedia.org/r/#/c/70990/ , nothing on https://bugzilla.wikimedia.org/show_bug.cgi?id=50325 [08:55:29] i noticed that yesterday already [08:56:00] Cough ... minuscule b in bug. [08:56:01] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/70990 [08:56:06] Did you receive my Email from yesterday? [08:56:32] hashar: https://gerrit.wikimedia.org/r/#/c/70870/ [08:57:52] haven't checked my emails yet [08:58:01] Oh, ok. [08:58:12] So I read your comments from yesterday, but did not cactch you online [08:59:46] hooks-bugzilla onle checks for �Bug� (capital �B�) footers, not for �bug� (minuscule �b�) footers. [08:59:53] The above change should correct that. [09:00:10] * apergos lurks [09:02:59] New review: Hashar; "I am not sure where to mark the deprecation though. I dont think that tag was used anywhere else :D" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69873 [09:03:15] hey qchris. [09:03:24] do you know of any pyhon wrappers to the Gerrit API? [09:03:24] Hi YuviPanda [09:03:37] No. I do not. [09:03:37] YuviPanda: yeah OpenStack has one iirc [09:03:38] * YuviPanda should work on the GitHub sync this weekend [09:03:46] hashar: ooh, name / link? [09:03:48] * YuviPanda searches [09:03:54] https://github.com/openstack-infra/gerritlib [09:04:04] I am not are sure whether it is in use though [09:04:31] you would want to ask in #openstack-infra during US business hours. [09:04:42] hashar: :D ty [09:05:01] hmm, it actually queries the ssh interface tho [09:05:04] not http [09:05:07] Zuul is using Gerrit [09:05:26] and they thought that its Gerrit support was not really robust and could be used by other softwares [09:05:27] Gerrit's REST API is rather fresh, so most tools do not yet use it. [09:05:29] so they wrote the lib [09:05:39] it probably use the SSH api though :( [09:05:41] (If they use a REST API, they typically use the old REST API) [09:05:50] hashar: yeah, it does use the SSH api [09:06:19] YuviPanda: if you are interested, you can talk with them to have that lib support REST :-D [09:06:25] hashar: yes, will do! [09:06:30] or might even write one! [09:06:46] they are very friendly and are really welcoming regarding new user/committers [09:06:48] the API's URL building concepts are similar to github's [09:06:49] and always offer nice code review [09:06:53] whee [09:07:03] for such a change, you might want to get in touch with them first [09:07:23] qchris: I checked but there seems to be no way to get inline comments via the ssh 'api' [09:07:29] jeblair or mordred are probably the nicks you are looking after [09:07:43] just joined :) [09:07:45] ty hashar [09:07:57]