[00:04:56] #136: (MH) http://en.wikipedia.beta.wmflabs.org/wiki/Light... DTS:1F2|O:MH|S:IT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/136 [00:04:56] #136: (MH) Tested, works DTE:1F2|S:RfS -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/136 [00:04:56] #136: (MH) S:A -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/136 [00:05:44] (03PS2) 10Gilles: Hide mouse cursor when buttons are faded out in fullscreen [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 [00:06:58] (03CR) 10MarkTraceur: [C: 04-1] "Hm." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 (owner: 10Gilles) [00:07:55] #210: (MH) S:R -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/210 [00:11:33] (03CR) 10MarkTraceur: [C: 04-1] "Small issue, will fix and test and mebbe merge" (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113929 (owner: 10Gilles) [00:17:05] (03PS2) 10MarkTraceur: Make metadata animation more subtle [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113929 (owner: 10Gilles) [00:17:20] (03PS3) 10MarkTraceur: Make metadata animation more subtle [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113929 (owner: 10Gilles) [00:19:21] (03CR) 10MarkTraceur: [C: 032] "Great patch, merging" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113929 (owner: 10Gilles) [00:19:52] (03Merged) 10jenkins-bot: Make metadata animation more subtle [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113929 (owner: 10Gilles) [00:19:55] #213: (MH) DDE:1F2|S:RfT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/213 [00:21:55] #214: (MH) DDE:1F2|S:RfT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/214 [00:21:55] #213: (MH) http://en.wikipedia.beta.wmflabs.org/wiki/Light... DTS:1F2|S:IT|O:MH -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/213 [00:22:56] #214: (MH) http://en.wikipedia.beta.wmflabs.org/wiki/Light... DTS:1F2|O:MH|S:IT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/214 [00:22:56] #214: (MH) Tested, seems good DTE:1F2|S:RfS -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/214 [00:22:56] tgr: Are you able to select the image in the lightbox with e.g. ctrl-a? If so, should #214 still be in dev? [00:24:34] In other news https://gerrit.wikimedia.org/r/113529 has an open request for review from you [00:25:21] ctrl-a only selects the panel for me (current Chrome) [00:25:28] Hm. [00:25:29] that's a good thing, isn't it? [00:25:32] Maybe it's Firefox, dunno [00:25:37] I can select the image with ctrl-a [00:26:07] as far as i remember, there is absolutely no way to make an element unselectable in firefox [00:26:33] Heh [00:26:34] Awesome. [00:26:38] apart from horrible stuff like js polling for where the selection starts/ends on every tick [00:27:09] user-select:none just disables selection start/end on that element [00:27:37] you can still create a selection that completely contains it by e.g. double clicking [00:28:58] i spent a lot of time once trying to make sure that double-clicking on a section title doesn't get the "edit this" text into the clipboard, but as far as i could see it is impossible on firefox [00:29:19] it was a long ago though [00:30:30] Heh. [00:30:34] OK, thanks [00:30:41] As long as that's sane I'll mark it accepted [00:31:55] #214: (MH) S:A -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/214 [00:37:58] (03CR) 10Gergő Tisza: [C: 032] Use provider XHR information in performance metrics + several fixes [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113529 (owner: 10Gergő Tisza) [00:38:00] (03CR) 10jenkins-bot: [V: 04-1] Use provider XHR information in performance metrics + several fixes [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113529 (owner: 10Gergő Tisza) [00:38:28] (03PS3) 10Gergő Tisza: Use provider XHR information in performance metrics + several fixes [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113529 [00:38:38] (03CR) 10Gergő Tisza: [C: 032] Use provider XHR information in performance metrics + several fixes [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113529 (owner: 10Gergő Tisza) [00:38:46] Sweet. [00:38:56] I'll be getting the data: URI one in next [00:39:33] (03Merged) 10jenkins-bot: Use provider XHR information in performance metrics + several fixes [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113529 (owner: 10Gergő Tisza) [00:41:27] rdwrer: can you merge https://gerrit.wikimedia.org/r/#/c/113047/ too? super simple [00:41:38] i think gi11es does not have +2 on that repo [00:41:57] tgr: Will do next, sure [00:45:09] (03PS5) 10MarkTraceur: Avoid double requests when measuring performance of image load [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113130 (owner: 10Gilles) [00:45:43] Holy crap this is getting fast. [00:46:24] (03CR) 10MarkTraceur: [C: 032] "LGTM! Thanks gilles." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113130 (owner: 10Gilles) [00:46:55] (03Merged) 10jenkins-bot: Avoid double requests when measuring performance of image load [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113130 (owner: 10Gilles) [00:47:55] #207: (MH) DDE:1F2|S:RfT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [00:49:55] #207: (MH) https://integration.wikimedia.org/ci/view/Beta/... DTS:1F2|O:MH|S:IT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [01:06:55] #211: (MH) DDE:1F2|S:RfT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/211 [01:10:30] tgr: Well one cause of me not seeing the Permission field was that I wasn't running CMD locally -.- [01:10:35] But I stopped being stupid today [01:11:16] (03CR) 10MarkTraceur: [C: 032] Add Information template Permission field to API output [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/113047 (owner: 10Gergő Tisza) [01:11:22] (03Merged) 10jenkins-bot: Add Information template Permission field to API output [extensions/CommonsMetadata] - 10https://gerrit.wikimedia.org/r/113047 (owner: 10Gergő Tisza) [01:11:33] Right, who's next [01:11:36] I'LL MERGE YA BUDDY [01:15:26] (03CR) 10Aarcos: Avoid double requests when measuring performance of image load (034 comments) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113130 (owner: 10Gilles) [01:18:37] (03PS1) 10MarkTraceur: Add mwe-upwiz-upload-error-duplicate-archive to RL def [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/114117 [01:18:38] bawolff: ^^ be my favorite? [01:19:10] (03CR) 10Brian Wolff: [C: 032] Add mwe-upwiz-upload-error-duplicate-archive to RL def [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/114117 (owner: 10MarkTraceur) [01:19:17] (03Merged) 10jenkins-bot: Add mwe-upwiz-upload-error-duplicate-archive to RL def [extensions/UploadWizard] - 10https://gerrit.wikimedia.org/r/114117 (owner: 10MarkTraceur) [01:23:32] Too bad there couldn't be some magic check jenkins bot ran to make sure all messages are included [01:24:15] Well, it's not useful for most extensions [01:24:23] But...well, hashar wrote one for qqq messages [01:24:30] Maybe we can do it for this [01:25:01] But there'd need to be some level of marking in the i18n file [01:25:04] So pretty useless [01:25:14] People who know to add the marking will know to add it in the RL module [01:25:53] So basically we just need to solve the strong AI problem in order for it to be automatic [01:27:47] I was actually thinking of something that greps js files for "mw.message( 'foo' )" and tries to match those against all RL messages declared [01:28:11] Hm, could do [01:28:27] The issue with UW is it does something like mw.msg( 'mwe-upwiz-upload-error-' + errorCode ) [01:28:40] So we wind up with crazy errors every time an error happens that nobody expected [01:28:45] Which is TOTALLY USELESS [01:29:08] Hmm, maybe we need some thing like eventlogging but for missing messages on the client side [01:29:34] Every so often someone could go through some log and look for what the top missing messages for js are on wikimedia wikis [01:30:38] Maybe [01:30:59] bawolff: We could also just check for message existence in UW, and return an unknown-error message in that case instead [01:31:12] That might be better [01:31:36] "Hey, we don't have an error message for what you did wrong. Can you tell us? Give us this code: _____" [01:32:44] (03PS2) 10Gergő Tisza: Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 [01:33:18] (03CR) 10jenkins-bot: [V: 04-1] Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 (owner: 10Gergő Tisza) [01:33:43] (03CR) 10Aarcos: Lazy-load mmv (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113786 (owner: 10Gilles) [01:40:36] I wonder how hard it is to install swift et al [01:41:44] I bet AaronSchulz could help [01:41:51] He said he was running it locally IIRC [02:08:34] Now that everything is merged (not true, but lots is) I'm writing a GRAND UNIFYING DOCUMENTATION PATCH [02:08:39] Because Krinkle is getting antsy [02:08:49] About us doing things like @method and @class in weird ways [02:24:14] Grumble grumble code conventions [02:24:20] I'll have to run through the files again I'm sure [02:27:59] (03PS1) 10MarkTraceur: Grand Unifying Documentation Patch [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114122 [02:32:41] (03CR) 10Aarcos: [C: 031] Preload fullscreen image (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 (owner: 10Gergő Tisza) [07:06:54] (03CR) 10Gergő Tisza: Preload fullscreen image (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 (owner: 10Gergő Tisza) [07:53:54] (03CR) 10Gilles: Lazy-load mmv (032 comments) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113786 (owner: 10Gilles) [07:56:43] I'm getting js error: Validation error against schema UploadWizardTutorialActions: Unknown schema: UploadWizardTutorialActions [07:57:10] on upload wizard on testwiki. Is that a test wiki is broken thing, or should I report it? [07:57:55] #202: (GD) S:ID -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/202 [07:59:58] oh, I guess that isn't really stopping things [08:00:24] I think these warnings have been turned on recently [08:00:54] it's for EventLogging [08:01:02] the schema exists, though: https://meta.wikimedia.org/wiki/Schema:UploadWizardTutorialActions [08:01:37] maybe it's not declared properly on the PHP side of things? [08:01:57] yeah, the schema thing was unrelated to the issue I was seeing [08:02:16] Basically I fail at reading and didn't notice one of the sections wasn't fulled [08:02:34] for some reason my brain went to must be bug as opposed to must be me being stupid [08:02:49] so I looked in the js console and saw that [08:02:51] these warnings are kind of noisy, though [08:03:12] especially since they repeat for each call, warning once per pageload would be enough [08:05:31] (03PS3) 10Gergő Tisza: Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 [08:06:09] (03CR) 10jenkins-bot: [V: 04-1] Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 (owner: 10Gergő Tisza) [08:09:56] #211: (GD) http://en.wikipedia.beta.wmflabs.org/wiki/Light... O:GD|DTS:1F2|S:IT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/211 [08:09:56] #142: (GD) http://en.wikipedia.beta.wmflabs.org/wiki/Light... S:IT|DTS:1F2 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/142 [08:10:55] #207: (GD) Verified on beta by inspecting the page. Things... DTE:1F2|S:RfS -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [08:10:55] #207: (GD) O:GD|S:A -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [08:18:55] #211: (GD) Verified locally DTE:1F2|S:RfS -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/211 [08:18:55] #211: (GD) S:A -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/211 [08:57:38] (03CR) 10Gilles: "Faking progress is worthless. People can sense when it's fake, particularly when they are on a low bandwidth/high latency environment and " (035 comments) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [09:01:55] #215: (GD) P:Mh|S:RfD|R:#Q|T:Td Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/215 [09:12:08] (03CR) 10Gilles: "https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/215" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [09:12:55] #216: (GD) S:RfD|P:Sh|R:#Q|T:S Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/216 [09:17:02] (03CR) 10Gilles: [C: 04-1] "This already happened in https://gerrit.wikimedia.org/r/#/c/113786/" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114098 (owner: 10Aarcos) [09:25:10] (03CR) 10Gergő Tisza: "Yeah, I didn't mean most suggestions to be part of this ticket, sorry for not being clear about that (and thanks for the cards!). My main " (032 comments) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [09:25:26] (03CR) 10Gilles: [C: 032] Grand Unifying Documentation Patch [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114122 (owner: 10MarkTraceur) [09:25:58] (03Merged) 10jenkins-bot: Grand Unifying Documentation Patch [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114122 (owner: 10MarkTraceur) [09:28:07] (03CR) 10Gilles: "Thanks a lot for doing that!" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114122 (owner: 10MarkTraceur) [09:30:03] (03CR) 10Gilles: Blurred thumbnail preview + progress bar (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [14:24:29] (03PS3) 10Gilles: Blurred thumbnail preview + progress bar [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 [14:24:31] (03CR) 10jenkins-bot: [V: 04-1] Blurred thumbnail preview + progress bar [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [14:24:50] (03CR) 10Gilles: "Still doesn't have any test coverage, but your concerns should be addressed now." (034 comments) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 (owner: 10Gilles) [14:26:26] (03PS3) 10Gilles: Hide mouse cursor when buttons are faded out in fullscreen [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 [14:27:22] (03CR) 10Gilles: "I couldn't reproduce the issue with the cursor, but I saw the buttons reappear when using the keyboard. I fixed that, and since both are t" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 (owner: 10Gilles) [14:43:04] (03PS1) 10Gilles: Missing dependency [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 [14:43:06] (03CR) 10jenkins-bot: [V: 04-1] Missing dependency [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 (owner: 10Gilles) [14:45:03] (03CR) 10Gilles: "recheck" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 (owner: 10Gilles) [14:47:01] (03PS4) 10Gilles: Blurred thumbnail preview + progress bar [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113963 [14:55:55] #210: (GD) O:(s -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/210 [16:47:51] (03Abandoned) 10Aarcos: Get rid of mw.mmvTestHelpers.resetViewer() [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114098 (owner: 10Aarcos) [16:48:55] #164: (F) Hi Gilles, we may want to start the Media Viewe... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/164 [17:02:02] (03CR) 10Aarcos: Preload fullscreen image (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 (owner: 10Gergő Tisza) [17:04:09] gi11es: If you're in the Kaltura thing, I just replied "no" in Calendar, running pretty late today [17:04:24] I'm in there [17:04:57] Just so people know, I didn't want you to wait or anything [17:13:34] Ah shit [17:13:38] Broken in beta [17:13:41] On Wednesday [17:13:43] Tiiiiiming [17:15:10] (03CR) 10Aarcos: [C: 031] Missing dependency [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 (owner: 10Gilles) [17:16:08] https://gerrit.wikimedia.org/r/#/c/114149/ this I know is causing errors locally [17:16:16] might help it it gets merged [17:16:36] although doesn't seem to be what's causing metadata to be missing on beta [17:28:26] actually I've just triggered it on beta, so... [17:32:17] rdwrer: can you review that changeset? [17:41:19] Sure [17:42:32] (03PS1) 10MarkTraceur: Grand Unified Documentation Patch Jr. [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114167 [17:42:43] (03CR) 10MarkTraceur: [C: 032] Missing dependency [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 (owner: 10Gilles) [17:42:45] Simple enough [17:43:08] (03CR) 10jenkins-bot: [V: 04-1] Grand Unified Documentation Patch Jr. [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114167 (owner: 10MarkTraceur) [17:43:39] (03Merged) 10jenkins-bot: Missing dependency [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114149 (owner: 10Gilles) [17:53:48] error gone on beta but metadata still missing [17:59:28] found the cause, it's GlobalUsage taking a crap: http://en.wikipedia.beta.wmflabs.org/wiki/Lightbox_demo?debug=1&action=query&prop=globalusage&titles=File%3AMultimedia_Roundtable_5_Photo_2%2Ejpg&guprop%5B%5D=url&guprop%5B%5D=namespace&gufilterlocal=1&gulimit=100&format=json [17:59:50] going to the wrong URL, it seems [18:02:31] I see, it seems to be a typo [18:02:47] using mw.Api instead of mw.mmv.Api [18:06:37] (03PS1) 10Gilles: GlobalUsage should use the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 [18:06:48] please review :) [18:07:35] the breakage wasn't happening locally because GlobalUsage is off locally by default [18:11:38] (03CR) 10Aarcos: [C: 032] GlobalUsage should use the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:12:13] (03CR) 10jenkins-bot: [V: 04-1] GlobalUsage should use the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:14:30] segfault in the doc check, great... [18:14:44] how do we merge this now? [18:14:58] (03CR) 10Gilles: [V: 032] GlobalUsage should use the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:15:14] (03CR) 10Gilles: "recheck" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:17:40] aarcos: maybe try to downgrade it to +1 then back to +2? [18:18:09] (03CR) 10Aarcos: [C: 031] GlobalUsage should use the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:18:21] Rebase it or change the commit message [18:18:30] I think there were Jenkins problems yesterday, though... [18:19:00] (03PS2) 10Gilles: GlobalUsage should utilize the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 [18:19:16] (03CR) 10Gilles: [C: 032] GlobalUsage should utilize the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:19:32] (03CR) 10jenkins-bot: [V: 04-1] GlobalUsage should utilize the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:19:47] segfault again, ffs [18:20:02] (03Merged) 10jenkins-bot: GlobalUsage should utilize the right Api class [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114174 (owner: 10Gilles) [18:20:22] what the... [18:20:33] try again? [18:20:54] Luck of the draw? [18:20:56] #217: (F) S:RfD|P:Mh|ES:#MVF|R:#Q|I:#F2-S2|T:S Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [18:20:56] #8: (e) PC:0 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/8 [18:22:29] ok, seems it is in now. The e2e test needs to be changed anyways now that we are using data: URIs. Sending that soon... [18:23:06] Yeah [18:23:20] chrismcmahon: The data: URI stuff is a feature and not a bug, FYI [18:24:02] Though brion brings up good points. [18:24:05] rdwrer: yeah, it seemed deliberate but that is a heck of a long string. [18:24:12] Well, yeah [18:24:21] rdwrer: thus the BZ is marked 'minor' [18:24:41] We're just not used to seeing that data in the DOM; maybe it's a bug after all but there's not much we can do about it [18:24:45] At least not now [18:24:53] But then, I didn't write that patch [18:25:16] we can move the sampling check higher up to avoid doing the XHR read when not measuring anything [18:25:24] so it would only happen when we measure the performance of the image load [18:27:33] one in X image displays would use the data URI instead of every image display [18:28:15] and if we want to get rid of data URI altogether, that means anyone getting sampled would load the image twice, once as XHR for sampling (has to happen first) and then the real image load [18:28:44] which isn't great, especially on beta where we want every load to be measured [18:30:13] an alternative would be to load the data URI into a canvas, I think the canvas just "absorbs" it without the URI ending up in the DOM [18:30:27] is the concern that it's bloating the DOM on inspection? [18:30:36] gi11es: Then we're left with the issue of not-copiable [18:30:56] "I love this image let me right click, copy image location, and share with my friends!" [18:31:02] "Oh god what have I done" [18:31:14] True story [18:31:35] sneak-inject the image with the proper src on right click? :) [18:31:38] rdwrer: I had that reaction running the browser test locally :-) [18:31:52] Is there a way to make sure the XHR result goes into the browser cache, maybe? [18:31:59] rdwrer: the URI string blew out the buffer on my shell [18:32:07] Hahaha [18:32:14] doesn't go to the browser's cache for images, we don't know why [18:32:27] Bugs reported? [18:32:54] could just be that browsers have a cache keys include whether it's an XHR or an image [18:33:05] not sure if that is a bug, since it is an xdomain request [18:33:15] tgr thought it might be the difference in CORS headers in the request [18:33:31] so we send origin header in xhr, but not on image src requests [18:33:38] I don't think there's a workaround [18:34:08] would have to dig up to spec to see whether requests with different origin are cacheable [18:34:35] also, not sure how consistent the behavior is between browsers, i have only looked at chrome [18:34:56] I think firefox behaves the same [18:36:37] Guess I could check [18:36:38] maybe we could just start loading the right src after the data URI is displayed? [18:36:53] Eh, we'd still be using the bandwidth [18:37:04] sure, but at least the person isn't waiting twice as long to see the content [18:37:30] True I suppose [18:38:01] and in production it would only happen once every 1000 requests or something, not that bad [18:38:09] we could load the src first, use it in the page, then load the xhr for timing [18:38:10] you're unlikely to have it happen to you twice in a row [18:38:19] tgr: nope, because caching [18:38:25] on the varnish end [18:38:36] oh, true [18:38:39] if we do that we can't tell if the thumbnail had to be re-generated [18:39:19] gtg [18:39:33] Fun times [18:39:51] Should we leave the repo in a semi-perilous state for tomorrow's deploy? :/ [18:39:54] what do you specifically need the XHR for? [18:40:33] brion: Timing data, for performance analysis [18:40:54] and you need more than just a raw time-to-load? [18:41:16] for the headers [18:41:44] and we need the headers to tell apart request loading time and thumbnail generation time on the servers [18:41:55] #111: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/111 [18:42:12] hmm [18:42:41] * brion considers something crazy, like XHR a HEAD request so you don't transfer the body, then fetch the (now rendered) image with a regular :) [18:43:42] anyway if you do stick with the XHR, i strongly recommend blob + createObjectURI instead of data URI [18:43:51] it'll still be a "surprise" result but it won't be SUPER_LONG [18:44:11] one thing we could test is html cors attributes, maybe that would fix the caching [18:44:55] img.crossOrigin = 'Anonymous' [18:45:05] worth a shot [19:01:55] #217: (MH) R:(s|I:(s|T:B|ES:(s -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [19:01:55] #8: (e) PC:0 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/8 [19:02:13] didn't work [19:02:19] I'll take the uploaded-by thing [19:02:28] Oh, wait, I know why that's broken already [19:02:34] Maybe this is a story card after all [19:02:37] but at least we now its not the origin header, since both requests send it that way [19:02:55] #217: (MH) R:#Q -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [19:02:55] #217: (MH) R:#Q|I:#F2-S2 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [19:02:58] (03PS1) 10Aarcos: Fix e2e tests to support data:URIs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 [19:03:30] the only other difference in headers is Accept, and there is no way to fix that since it would turn the request into preflighted [19:03:55] #217: (MH) O:MH|I:#F2-S2|S:ID -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [19:12:58] #147: (P) - Select the whole text seems a reasonable appr... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/147 [19:13:44] (03PS1) 10MarkTraceur: Get article path for foreign DB repos too [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114188 [19:18:55] #207: (GT) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [19:22:00] (03CR) 10Cmcmahon: [C: 04-1] "Not passing for me (and blows out the buffer on my shell):" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [19:27:55] #207: (GT) Tried to set img.crossOrigin = Anonymous. Tha... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [19:29:17] chrismcmahon: This is the error you currently get without the fix. I ran the test many times with the fix and works fine. Are you sure you are running the test with the fix? [19:31:09] aarcos: yes: $ git branch [19:31:09] master [19:31:09] * review/aarcos/fix-e2e-tests-to-match-data-uris [19:31:09] cmcmahon@ubuntu:~/mmv/MultimediaViewer/tests/browser$ [19:32:33] aarcos: it always fails for me [19:33:08] chrismcmahon: Umm, let me check the values, maybe it is a race condition. [19:41:55] #217: (MH) https://gerrit.wikimedia.org/r/114188 S:ACR -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/217 [19:45:19] Then the image and metadata of the previous image should appear # features/step_definitions/basic_mmv_navigation_steps.rb:75 [19:45:19] expected [19:45:29] chrismcmahon: Can you send me the first 2 lines after [19:45:45] chrismcmahon: the lines above, ;-). [19:46:06] I want to see what hex values are you getting? [19:46:08] aarcos: maybe if I pipe the output through less() or something [19:46:30] just redirect to a file >& res [19:46:44] and then look for "expect" [19:47:06] it should start with something like: "data:image;base64,/9j/4AAQSk... [19:47:36] aarcos: Then the image and metadata of the previous image should appear # features/step_definitions/basic_mmv_navigation_steps.rb:75 [19:47:37] expected "data:image;base64,/9j/4AAQSkZJRgABAQEAtAC0AAD//gBeRmlsZSBzb3VyY2U6IGh0dHA6Ly9jb21tb25zLndpa2ltZWRpYS5vcmcvd2lraS9GaWxlOlN1bnJpc2Vfb3Zlcl9maXNoaW5nX2JvYXRzX2luX0tlcmFsYS5qcGf/4gxYSUNDX1BST0ZJTEUAAQEAAAxITGlubwIQAABtbnRyUkdCIFhZWiAHzgACAAkABgAxAABhY3NwTVNGVAAAAABJRUMgc1JHQgAAAAAAAAAAAAAAAAAA9tYAAQAAAADTLUhQICAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA [19:48:06] tgr, aarcos: Scrum notes (don't worry, I forgot too) [19:48:56] yeap, will do in a sec, ;-). [19:49:00] bd [19:49:56] #148: (P) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [19:50:56] #148: (P) Added design notes about sizes and their relati... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [19:52:02] I...am regretting the whole put-it-all-on-Mingle thing [19:56:57] #163: (F) Gilles, I agree with your recommendation to do ... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/163 [19:59:56] #207: (GT) Uhh, apparently the whole double request thing ... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [20:01:30] (03PS2) 10Aarcos: Fix e2e tests to support data:URIs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 [20:02:12] (03CR) 10jenkins-bot: [V: 04-1] Fix e2e tests to support data:URIs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [20:02:47] chrismcmahon: Changed RE to match stuff that is closer to the beginning of the string. Can you try and see if it works for you this time? Thanx ! [20:03:45] tgr: caching works for me http://jsfiddle.net/BvNU8/ [20:04:05] in fact, I took your word for it last time :) how are you testing it? [20:05:41] works even without the crossOrigin property, on Chrome stable + Firefox (OS X) [20:07:45] (03CR) 10Aarcos: "Chris: Changed RE to match stuff that is closer to the beginning of the string. Can you try and see if it works for you this time? Thanx !" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [20:08:29] gi11es: now i am testing with wireshark [20:08:52] there is a 304 request, but the image is loaded from cache anyway [20:08:53] i was tripped by cache disabling in chrome devtools, it seems [20:09:00] (03CR) 10Cmcmahon: [C: 032] Fix e2e tests to support data:URIs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [20:09:10] in chrome there is a single request [20:09:17] aarcos: much better. that is still a ridiculously long string :-) [20:09:26] in firefox there are two requests, i havent looked at the contents yet [20:09:41] (03Merged) 10jenkins-bot: Fix e2e tests to support data:URIs [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [20:09:54] two requests, but look at the time, the second onload fires in a fraction of the time [20:10:48] maybe I'm seeing a bigger difference because I'm tethering, but basically it's 2.5-3 seconds versus 200-400ms [20:11:39] chrismcmahon: Thanx for the prompt review Chris ! It seems we will need to address this sooner rather than later but this fixes the tests for the time being. Cheers [20:12:37] yeah, i see the difference [20:13:01] on the other hand, if i read wireshark correctly, the full image is downloaded twice [20:13:36] so it wastes bandwidth, not time. I can personally live with that, considering how rare the sampling hits will be in production [20:14:33] no, wait [20:14:50] the jsfiddle does not download twice, but mediaviewer does [20:14:56] let me double-check [20:16:24] yeah, without setting crossOrigin it downloads twice, but if you set it, it just sends a 304 [20:17:03] and Chrome does not even send a second request [20:17:19] cool [20:17:28] would be nice to have a cross-browser test for this [20:17:49] the issue is that the time it takes for the 304 is dependent on network conditions [20:18:00] but i have no idea how, apart from manually [20:18:03] for chrome you can check relatively reliably based on timing that you're hitting the cache [20:18:43] we can always whitelist the browsers that are used for detailed sampling [20:19:10] everything else just gets "$.now() - start" [20:19:47] already with the binary functions condition we're only sampling a handful of modern browsers [20:20:36] do you have time today to rip out data URI support and add crossOrigin? or should I do it tomorrow morning my time? [20:20:54] i can do it [20:21:13] fantastic, thank you [20:21:21] how do we get the whitelist? is there a cleverer way than just trying whatever browser is around? [20:21:22] I'm glad we've found a loophole [20:21:38] I don't think there's any way to avoid manual testing [20:22:09] for our purpose, though, I think if we just sampled chrome and firefox we'd have valuable data already... then we can file a separate card to do the manual investigation and add more [20:22:39] the config changeset is still waiting anyway, afaik, nothing's getting measured on beta nor live [20:22:44] (yet) [20:22:57] #119: (AA) O:AA -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/119 [20:22:57] #119: (AA) I:#F2-S2 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/119 [20:22:59] gi11es: it might be possible to get volunteers from the QA mail list for some of that manual investigation. [20:23:56] #119: (AA) S:RfD -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/119 [20:24:34] really afk now [20:24:58] #119: (AA) S:ID -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/119 [20:25:19] gi11es: about the config change, probably asking on the ops channel is the fastest way of getting a reviewer [20:25:29] i don't know who exactly has +2 [21:14:56] #79: (P) The attribution text looks good (shorter would ... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/79 [21:16:56] #207: (GT) After some more wireshark testing on the fiddle... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/207 [21:17:53] FYI, since we have been introducing a few regressions with the refactorings that are taking place, I think investing in e2e tests is a good idea. I plan to work on https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/119. [21:26:00] #79: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/79 [21:27:59] #79: (F) Thanks, Pau. I tightened the plain text copy a ... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/79 [21:34:50] (03CR) 10Gergő Tisza: "We are retiring data URIs, they are a lot of trouble (see bug 61542) and seems like they aren't be needed for logging after all." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [21:38:53] (03CR) 10Aarcos: "I just wanted to fix the tests. I will revert the changes once data: URIs are retired." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114186 (owner: 10Aarcos) [21:39:57] #205: (P) Regarding the email info: - Can we have the ... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/205 [21:45:57] #193: (P) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/193 [21:48:57] #200: (P) If open in new tab goes to the file page (usi... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/200 [21:51:56] #218: (F) S:IA|P:Mh|R:#Q|T:S Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/218 [21:56:57] #219: (F) S:IA|P:Ch|R:#Q|T:S Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/219 [21:57:59] #219: (F) I:#F2-S3 -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/219 [21:57:59] #219: (F) P:Mh -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/219 [21:58:58] #219: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/219 [22:01:57] #147: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/147 [22:01:57] #147: (F) Linked this feature to card #219: 'Use this fil... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/147 [22:01:57] #218: (P) This will not work for other languages due to d... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/218 [22:12:56] #147: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/147 [22:17:59] #148: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [22:24:18] (03PS4) 10Gergő Tisza: Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 [22:25:01] (03CR) 10jenkins-bot: [V: 04-1] Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 (owner: 10Gergő Tisza) [22:30:57] #148: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [22:34:00] #148: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [22:45:01] #148: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [22:46:04] #148: (F) Description changed -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/148 [22:59:22] * rdwrer grumbles about lack of info in repodata constantly [23:14:25] (03PS5) 10Gergő Tisza: Show permission [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113312 [23:14:27] (03PS1) 10Gergő Tisza: Fix image provider unit test [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 [23:14:29] (03PS1) 10Gergő Tisza: Fix fileUsage test breaking behavior [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114361 [23:16:58] #118: (GT) There are fixes for some unrelated test bugs in... -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/118 [23:18:00] #118: (GT) O:(s -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/118 [23:23:03] (03PS5) 10Gergő Tisza: Preload fullscreen image [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 [23:23:04] (03CR) 10jenkins-bot: [V: 04-1] Preload fullscreen image [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 (owner: 10Gergő Tisza) [23:24:26] (03PS6) 10Gergő Tisza: Preload fullscreen image [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/112753 [23:36:52] (03CR) 10Gilles: [C: 04-1] Fix image provider unit test (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:37:22] gi11es: Bit late for you, innit? [23:37:50] it's never too late for a -1 ;) [23:39:04] Heh [23:39:15] And as we proved on Monday, never too late for a +2 [23:39:27] Or rather, Tuesday morning [23:40:17] gi11es: FWIW I'm merging and deploying your change in 20 minutes [23:42:12] (03CR) 10Gergő Tisza: Fix image provider unit test (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:42:17] the settings thing? [23:42:27] Yeah [23:45:22] (03CR) 10Gilles: [C: 032] Fix image provider unit test (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:45:27] cool [23:46:00] (03CR) 10jenkins-bot: [V: 04-1] Fix image provider unit test [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:47:56] (03CR) 10MarkTraceur: [C: 032] "You did indeed! Thanks." [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 (owner: 10Gilles) [23:48:33] (03Merged) 10jenkins-bot: Hide mouse cursor when buttons are faded out in fullscreen [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/113921 (owner: 10Gilles) [23:52:07] (03CR) 10Gergő Tisza: Fix image provider unit test (031 comment) [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:52:53] (03PS2) 10Gergő Tisza: Fix image provider unit test. [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 [23:53:13] (03CR) 10Gergő Tisza: [C: 032] "Resubmitting, doc test segfaulted" [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:53:28] libel is down again [23:53:52] Sigh [23:54:23] (03Merged) 10jenkins-bot: Fix image provider unit test. [extensions/MultimediaViewer] - 10https://gerrit.wikimedia.org/r/114360 (owner: 10Gergő Tisza) [23:54:25] Annoying [23:54:33] I should really daemonize that somehow [23:54:38] It's currently just running in a screen [23:55:39] #210: (GD) http://en.wikipedia.beta.wmflabs.org/wiki/Light... O:GD|DTS:1F2|S:IT -- https://wikimedia.mingle.thoughtworks.com/projects/multimedia/cards/210 [23:57:39] rdwrer: puppet has an ensure => running if you implement the init.d commands for it [23:57:55] Yes, but I'm lazy [23:57:58] So [23:58:01] Maybe someday