[00:06:03] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Reduce title font size and footprint of topic header - https://phabricator.wikimedia.org/T108263#1582551 (10DannyH) My urgency about this came from the conversations I had at Wikimania. There was a sizeable number of people... [00:07:50] 6Collaboration-Team-Backlog, 10Flow: Change the location of Flow's timestamps within posts - https://phabricator.wikimedia.org/T108096#1582558 (10DannyH) @PGiner-WMF, just a ping about my question above, and Nick's comments. What do you think about the timestamp placement? [00:11:29] 6Collaboration-Team-Backlog, 5Patch-For-Review: Make the VE/source toggle more discoverable - https://phabricator.wikimedia.org/T101316#1582566 (10DannyH) The issue is that there are experienced editors who are specifically looking for the switch to wikitext, and can't find it. I've heard this a number of time... [00:11:42] (03PS1) 10Catrope: Fix positioning of resolved check icon in RTL [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234447 [00:11:45] (03PS1) 10Catrope: Hackily move the check icon a bit so it appears to align better [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234448 (https://phabricator.wikimedia.org/T108263) [00:16:58] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Reduce title font size and footprint of topic header - https://phabricator.wikimedia.org/T108263#1582572 (10Catrope) The tick mark is actually aligned, DOM-wise; it's just that the SVG itself has blank space all around it, w... [00:20:49] (03PS1) 10Mooeypoo: [wip] Use viewmodel in Echo's front end [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234451 [00:22:46] (03CR) 10jenkins-bot: [V: 04-1] [wip] Use viewmodel in Echo's front end [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234451 (owner: 10Mooeypoo) [00:28:32] 3Collaboration-Team-Current, 10Wikimedia-Extension-setup, 5Patch-For-Review, 7WorkTypeNewFunctionality: Re-enable Flow on fawikibooks - https://phabricator.wikimedia.org/T109816#1582580 (10Krenair) Why has this been re-opened? [00:29:57] 6Collaboration-Team-Backlog, 10Flow: Confusing error message in Flow when there is a conflict of comments being posted (needs to be reproduced) - https://phabricator.wikimedia.org/T78718#1582584 (10Quiddity) 5Open>3Resolved a:3Quiddity Ok, I'll close for now, and we can re-open if it appears again. Thanks! [00:30:11] 6Collaboration-Team-Backlog, 10Flow: Confusing error message in Flow when there is a conflict of comments being posted (needs to be reproduced) - https://phabricator.wikimedia.org/T78718#1582589 (10Quiddity) a:5Quiddity>3None [01:03:40] (03CR) 10Catrope: [C: 04-1] "Looks good, minor comments inline. In the future we probably also want to move in the anon warning and a few other things?" (032 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234423 (https://phabricator.wikimedia.org/T109000) (owner: 10Mattflaschen) [01:56:25] 6Collaboration-Team-Backlog, 10Flow: DATA LOSS: press [Esc] to dismiss a VE dialog and Flow discards your entire reply. - https://phabricator.wikimedia.org/T110632#1582733 (10Spage) 3NEW [01:59:22] 6Collaboration-Team-Backlog, 10Flow: DATA LOSS: press [Esc] to dismiss a VE dialog and Flow discards your entire reply. - https://phabricator.wikimedia.org/T110632#1582741 (10Mattflaschen) [01:59:44] 3Collaboration-Team-Current, 10Flow, 10WMF-Server-Backports, 7Regression, 7WorkTypeNewFunctionality: Pressing ESC to close VE inspector also cancels your post - https://phabricator.wikimedia.org/T110146#1569838 (10Mattflaschen) This is worth cherry-picking to production. [02:03:37] 3Collaboration-Team-Current, 10Flow, 10WMF-Server-Backports, 7Regression, 7WorkTypeNewFunctionality: Pressing ESC to close VE inspector also cancels your post - https://phabricator.wikimedia.org/T110146#1582747 (10Mattflaschen) This is in OOjs UI, though, and I don't know if cherry-picks are usually done... [02:05:03] Sorry, skierpage, agree it's a bad bug and sounds like you hit it at the worst time. [03:05:30] Hey I have a small doubt that how will we enable Flow extension [03:05:53] I have installed extension mediawiki [03:37:13] adisha, you may want to use https://www.mediawiki.org/wiki/MediaWiki-Vagrant if you're not already. [03:42:52] Actually my system configration is poor [03:43:08] If I use vagrant my system hangs [03:43:24] So that's why I used manual installion [03:51:40] adisha, alright. [03:51:59] adisha, our documentation on this is out of date. Let me update it. [03:52:07] ok [03:52:39] In 'Special:Version' it is showing Flow [03:53:11] Is extension:Echo is required [03:53:16] for flow [03:53:18] ? [03:55:30] 6Collaboration-Team-Backlog, 10Echo: Echo should warn users who tries to mention more than 50 other users that they won't be notified - https://phabricator.wikimedia.org/T108293#1582806 (10Mattflaschen) Oddly, there is a separate config for Flow ($wgFlowMaxMentionCount) (not hard-coded). [03:57:09] adisha, yes, Echo is documented as required. [03:58:12] I want to ask one more ques how to enable or switch to flow [03:58:43] adisha, for the bug you're currently working on it may not matter, but I would also (especially if you plan to keep working on Flow) recommend you install all the dependencies listed as required at https://www.mediawiki.org/wiki/Extension:Flow#Dependencies , plus Parsoid. [03:59:04] ok [03:59:33] adisha, to enable it on a namespace (which suffices for the bug you're working on), see https://www.mediawiki.org/wiki/Extension:Flow#Configuration . I just fixed those docs. [03:59:46] Thanks [04:01:24] I want to participate in upcoming round of outreachy and want to contribute for wikimedia. [04:01:25] So do Extension:Echo has a project that can I choose for outreachy internship? [04:05:42] adisha, I don't see any currently listed at https://www.mediawiki.org/wiki/Outreach_programs/Possible_projects . [04:06:40] Actually that is outreachy round 10 list [04:07:05] adisha, it's a general list, used for multiple outreach programs. When does round 11 start, December? [04:07:13] yes [04:07:44] But application submission will began in september [04:08:20] adisha, okay, so you might have time to come up with a project idea if you stay in close communication with us. That doesn't guarantee the idea will be accepted, though. [04:08:40] ok thanks [04:10:40] I am facing a small problem during installation [04:10:47] of Extension Echo [04:11:27] when I include 'require_once "$IP/extensions/Echo/Echo.php";' [04:11:44] it gives MediaWiki internal error. [04:12:02] Error: 1146 Table 'my_wiki.echo_target_page' doesn't exist (localhost) [04:12:32] Is this my local database problem? [04:15:37] adisha, you have to run update.php: https://www.mediawiki.org/wiki/Manual:Update.php [04:15:48] BTW, if you mention my username, I will see your comment faster. [04:15:57] ok [04:16:03] yes got it [04:16:14] that it was my update issue [04:17:30] And thanks for the help.. :) [04:17:51] matt_flascen [04:18:54] You're welcome, adisha [04:49:33] (03CR) 10Mattflaschen: [C: 04-1] "Works great (tested internal, external, and categories)." (039 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/233432 (https://phabricator.wikimedia.org/T94569) (owner: 10Matthias Mullie) [06:39:55] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 5WMF-deploy-2015-09-01_(1.26wmf21), 7WorkTypeNewFunctionality: Opt-in for Flow on your own user talk page - https://phabricator.wikimedia.org/T98270#1582980 (10Pginer-WMF) It would be great to have some screenshots here of the current status (or som... [07:16:30] (03CR) 10Legoktm: Use db-replicated objectcache for storing last seen time (032 comments) [extensions/Echo] - 10https://gerrit.wikimedia.org/r/233639 (https://phabricator.wikimedia.org/T95839) (owner: 10Legoktm) [07:59:54] 6Collaboration-Team-Backlog, 5Patch-For-Review: Make the VE/source toggle more discoverable - https://phabricator.wikimedia.org/T101316#1583110 (10Pginer-WMF) > The icon doesn't mean code to me -- you could easily see it as two arrows pointing in opposite directions, so people assume it's a navigation icon... [08:00:27] (03CR) 10Legoktm: [C: 032] Put back notification_timestamp for web. [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234126 (owner: 10Mattflaschen) [08:02:51] (03Merged) 10jenkins-bot: Put back notification_timestamp for web. [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234126 (owner: 10Mattflaschen) [09:29:59] 6Collaboration-Team-Backlog, 10Flow: Change the location of Flow's timestamps within posts - https://phabricator.wikimedia.org/T108096#1583336 (10Pginer-WMF) Regarding the layout, it is dependent on how frequent people look for dates only versus username + timestamp, but considering the pros and cons I'd recom... [10:12:47] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Reduce title font size and footprint of topic header - https://phabricator.wikimedia.org/T108263#1583435 (10Pginer-WMF) > There comes a time in every product manager's life when you have to say, you know what? Maybe the head... [10:15:32] (03CR) 10Krinkle: Use db-replicated objectcache for storing last seen time (031 comment) [extensions/Echo] - 10https://gerrit.wikimedia.org/r/233639 (https://phabricator.wikimedia.org/T95839) (owner: 10Legoktm) [12:16:41] (03PS5) 10Matthias Mullie: Fix removal/addition of categorylinks [extensions/Flow] - 10https://gerrit.wikimedia.org/r/233432 (https://phabricator.wikimedia.org/T94569) [12:17:19] (03CR) 10Matthias Mullie: Fix removal/addition of categorylinks (039 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/233432 (https://phabricator.wikimedia.org/T94569) (owner: 10Matthias Mullie) [12:47:41] (03PS1) 10Matthias Mullie: Make sure flow entries are created with correct content [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234515 (https://phabricator.wikimedia.org/T109768) [12:55:07] (03CR) 10Matthias Mullie: "The thing about incorrect page content was indeed not related to this patch: https://gerrit.wikimedia.org/r/#/c/234515/" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/230648 (https://phabricator.wikimedia.org/T98270) (owner: 10Sbisson) [12:57:10] (03CR) 10Sbisson: "Make sure you don't have NS_USER_TALK content model === FLOW_BOARD when you test." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/230648 (https://phabricator.wikimedia.org/T98270) (owner: 10Sbisson) [13:22:29] 6Collaboration-Team-Backlog, 10WikiLove, 10Browser-Tests: Update WikiLove repository to mediawiki_selenium Ruby gem 1.x - https://phabricator.wikimedia.org/T99660#1583972 (10zeljkofilipin) [13:46:48] (03PS1) 10Zfilipin: Update repository to mediawiki_selenium Ruby gem 1.x [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) [13:47:02] (03CR) 10jenkins-bot: [V: 04-1] Update repository to mediawiki_selenium Ruby gem 1.x [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) (owner: 10Zfilipin) [13:51:06] (03PS2) 10Zfilipin: Update repository to mediawiki_selenium Ruby gem 1.x [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) [13:52:35] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to userspace in the Hebrew Wikipedia - https://phabricator.wikimedia.org/T110676#1584040 (10Amire80) 3NEW [13:53:41] (03CR) 10Sbisson: "I know I'm going against the grain here but have you considered an approach based on composition instead of inheritance?" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234423 (https://phabricator.wikimedia.org/T109000) (owner: 10Mattflaschen) [13:57:23] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to userspace in the Hebrew Wikipedia - https://phabricator.wikimedia.org/T110676#1584068 (10Amire80) For what it's worth, it does work when editing a non-Flow page in VisualEditor. See https://he.wikipedia.org/wiki/%D7%9E%D7%A9%D7%AA%D7%9E%D7%A9:Amire80/T1... [14:03:30] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to userspace in the Hebrew Wikipedia - https://phabricator.wikimedia.org/T110676#1584082 (10IKhitron) [14:05:11] (03CR) 10Zfilipin: "Jenkins is happy:" [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) (owner: 10Zfilipin) [14:16:08] (03PS1) 10Zfilipin: Updated RuboCop to the latest released version [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234533 [14:19:10] (03PS1) 10Zfilipin: Fixed Style/StringLiterals RuboCop offense [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234534 (https://phabricator.wikimedia.org/T110672) [14:19:31] 6Collaboration-Team-Backlog, 10Flow: MediaWiki error after preview of deleted flow page - https://phabricator.wikimedia.org/T110679#1584142 (10Luke081515) [14:20:14] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Opt-in: Guided tour on user talk for first visit to new Flow board - https://phabricator.wikimedia.org/T108266#1584149 (10Pginer-WMF) It would be great to have some screenshots here of the current status (or some instruction... [14:23:16] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted flow boards - https://phabricator.wikimedia.org/T110680#1584152 (10Luke081515) 3NEW [14:23:33] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted flow boards - https://phabricator.wikimedia.org/T110680#1584161 (10Luke081515) See also T110679 [14:23:42] 6Collaboration-Team-Backlog, 10Flow: MediaWiki error after preview of deleted flow page - https://phabricator.wikimedia.org/T110679#1584133 (10Luke081515) See also T110680 [14:38:42] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted flow boards - https://phabricator.wikimedia.org/T110680#1584210 (10Luke081515) Sometimes there is also {F2473615} [15:30:02] (03CR) 10Sbisson: [C: 031] Make sure flow entries are created with correct content [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234515 (https://phabricator.wikimedia.org/T109768) (owner: 10Matthias Mullie) [15:41:13] 6Collaboration-Team-Backlog, 10Flow, 7Wikimedia-log-errors: MediaWiki error after preview of deleted flow page - https://phabricator.wikimedia.org/T110679#1584368 (10Krenair) [15:47:59] 3Collaboration-Team-Current, 10Flow, 7WorkTypeNewFunctionality: Links from Flow topics to special pages are incorrectly included in link tables - https://phabricator.wikimedia.org/T109814#1584389 (10matthiasmullie) a:3matthiasmullie [16:04:02] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 5WMF-deploy-2015-09-01_(1.26wmf21), 7WorkTypeNewFunctionality: Editing an empty Flow board description (that has previously contained a category?) gives "Previous revision identifier is missing" - https://phabricator.wikimedia.org/T110358#1584446 (10... [16:21:13] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to user talk subpage, when part before / contains full stop, in the Hebrew Wikipedia - https://phabricator.wikimedia.org/T110676#1584490 (10Mattflaschen) [16:24:50] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted flow boards - https://phabricator.wikimedia.org/T110680#1584496 (10Mattflaschen) Can you explain what happened after the Flow board was created? And what determines which error you get? Does it depend what steps you do, or is it random? [16:25:10] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted Flow boards using revision undeletion - https://phabricator.wikimedia.org/T110680#1584497 (10Mattflaschen) [16:27:23] 6Collaboration-Team-Backlog, 10Flow, 7Regression, 7Wikimedia-log-errors: MediaWiki error after preview of deleted flow page - https://phabricator.wikimedia.org/T110679#1584499 (10Mattflaschen) [16:27:45] 3Collaboration-Team-Current, 10Echo, 5Patch-For-Review, 5WMF-deploy-2015-08-25_(1.26wmf20), and 3 others: Missing database prefix for echo_event - https://phabricator.wikimedia.org/T105890#1584502 (10Etonkovidova) Since in betalabs ``` SELECT * FROM information_schema.tables WHERE table_name = 'bw_echo... [16:27:55] 6Collaboration-Team-Backlog, 10Flow, 7Regression, 7Wikimedia-log-errors: MediaWiki error for deleted Flow board, when clicking 'Show changes' - https://phabricator.wikimedia.org/T110679#1584503 (10Mattflaschen) [16:28:15] 6Collaboration-Team-Backlog, 10Flow, 7Wikimedia-log-errors: MediaWiki error for deleted Flow board, when clicking 'Show changes' - https://phabricator.wikimedia.org/T110679#1584133 (10Mattflaschen) [16:30:15] 3Collaboration-Team-Current, 10Flow: Can no longer view deleted board in Special:Undelete - https://phabricator.wikimedia.org/T110696#1584507 (10Mattflaschen) 3NEW [16:30:28] 3Collaboration-Team-Current, 10Flow, 7Regression: Can no longer view deleted board in Special:Undelete - https://phabricator.wikimedia.org/T110696#1584516 (10Mattflaschen) [16:30:56] 6Collaboration-Team-Backlog, 10Flow, 7Wikimedia-log-errors: MediaWiki error for deleted Flow board, when clicking 'Show changes' - https://phabricator.wikimedia.org/T110679#1584133 (10Mattflaschen) [16:36:50] (03CR) 10Catrope: "What is the argument for which? Remember that OOjs also offers a third way of doing this, which is using a mixin." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234423 (https://phabricator.wikimedia.org/T109000) (owner: 10Mattflaschen) [16:41:10] 6Collaboration-Team-Backlog, 10Echo, 7I18n: Split notification-mention and notification-mention-flyout messages - https://phabricator.wikimedia.org/T110699#1584543 (10matej_suchanek) 3NEW [16:42:19] Interesting. "The revision #3712 of the page named "Adrines" does not exist." (caused on my local by switching back and forth between External Store) is cached in parser cache. [16:42:27] action=purge fixes it. [16:43:16] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to user talk subpage, when part before / contains full stop, in the Hebrew Wikipedia - https://phabricator.wikimedia.org/T110676#1584554 (10Mooeypoo) This also happens in English, as well. A link to "Somepage./Subpage" leads to "SomepageSubpage". [16:44:59] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584558 (10IKhitron) [16:45:55] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7Tracking, 5WMF-deploy-2015-09-01_(1.26wmf21): Writing any posts through the mobile Flow interface is completely impossible - https://phabricator.wikimedia.org/T110485#1584562 (10Etonkovidova) > Start a new topic is broken, too: you can type the sub... [16:47:28] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584574 (10IKhitron) Indeed, it's not just in English, see [[ https://en.wikipedia.org/wiki/Topic:Snuwl6fgrv7uk2ks | here]]. And it's in any namespace, not j... [16:49:05] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584577 (10IKhitron) [16:50:07] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584040 (10IKhitron) [16:50:44] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584040 (10IKhitron) [16:51:27] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584582 (10IKhitron) But if you open the post for editing you can see in preview/VE the right link. Exit will destroy it again. [16:55:56] (03CR) 10Dduvall: [C: 032] "Excellent!" [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) (owner: 10Zfilipin) [16:56:10] (03Merged) 10jenkins-bot: Update repository to mediawiki_selenium Ruby gem 1.x [extensions/WikiLove] - 10https://gerrit.wikimedia.org/r/234527 (https://phabricator.wikimedia.org/T99660) (owner: 10Zfilipin) [16:56:23] 6Collaboration-Team-Backlog, 10WikiLove, 10Browser-Tests, 5Patch-For-Review: Update WikiLove repository to mediawiki_selenium Ruby gem 1.x - https://phabricator.wikimedia.org/T99660#1584588 (10dduvall) 5Open>3Resolved [16:58:50] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584607 (10Mooeypoo) @IKhitron yes, the problem is found, we're fixing now. Thank you for helping narrow this down! [16:59:06] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584610 (10Mooeypoo) a:3Catrope [17:00:07] 6Collaboration-Team-Backlog, 10Flow: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584611 (10IKhitron) :-) [17:00:34] (03PS1) 10Catrope: Anchor the regex for ./ stripping [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234571 (https://phabricator.wikimedia.org/T110676) [17:01:15] (03CR) 10Mooeypoo: [C: 032] Anchor the regex for ./ stripping [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234571 (https://phabricator.wikimedia.org/T110676) (owner: 10Catrope) [17:05:11] (03Merged) 10jenkins-bot: Anchor the regex for ./ stripping [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234571 (https://phabricator.wikimedia.org/T110676) (owner: 10Catrope) [17:05:55] (03PS1) 10Catrope: Anchor the regex for ./ stripping [extensions/Flow] (wmf/1.26wmf20) - 10https://gerrit.wikimedia.org/r/234574 (https://phabricator.wikimedia.org/T110676) [17:06:05] (03CR) 10Catrope: [C: 032] Anchor the regex for ./ stripping [extensions/Flow] (wmf/1.26wmf20) - 10https://gerrit.wikimedia.org/r/234574 (https://phabricator.wikimedia.org/T110676) (owner: 10Catrope) [17:06:12] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Allow user to create new board at Special:EnableFlow without description - https://phabricator.wikimedia.org/T108348#1584628 (10Mattflaschen) The patch will allow all content to be zero-length with external store, so even if... [17:06:20] (03PS1) 10Mattflaschen: Allow zero-length descriptions in both S:EF and standard edits [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234575 (https://phabricator.wikimedia.org/T108348) [17:07:26] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase: Switch Flow to RESTBase - https://phabricator.wikimedia.org/T94574#1167306 (10cscott) [17:07:32] matt_flaschen: So there was no reason we were enforcing the >0 length thing, it was just paranoia? [17:08:20] RoanKattouw, it's not an external store requirement (this is how standard zero-length pages are implemented with External Store when I test locally). It is a product requirement at a different level for certain revision types. [17:08:42] (03Merged) 10jenkins-bot: Anchor the regex for ./ stripping [extensions/Flow] (wmf/1.26wmf20) - 10https://gerrit.wikimedia.org/r/234574 (https://phabricator.wikimedia.org/T110676) (owner: 10Catrope) [17:08:48] (different Flow revision types) [17:10:48] (03CR) 10Catrope: [C: 031] Allow zero-length descriptions in both S:EF and standard edits [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234575 (https://phabricator.wikimedia.org/T108348) (owner: 10Mattflaschen) [17:12:35] 6Collaboration-Team-Backlog, 10Flow, 5Patch-For-Review: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584665 (10Catrope) Fix is live now, and confirmed working. Thanks for reporting this. Sorry it took 3 hours; most of us were asleep whe... [17:12:55] 6Collaboration-Team-Backlog, 10Flow, 5Patch-For-Review: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584668 (10Catrope) 5Open>3Resolved [17:14:03] 6Collaboration-Team-Backlog, 10Flow, 5Patch-For-Review: Flow breaks links to subpage, when part before / finishes with full stop - https://phabricator.wikimedia.org/T110676#1584685 (10IKhitron) Indeed working. Thanks a lot! [17:16:22] (03PS4) 10Jforrester: Migrate to extension.json and deprecate PageTriage.php [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/230016 (https://phabricator.wikimedia.org/T87954) [17:18:38] (03CR) 10Sbisson: "My general argument against inheritance is that is creates complexity (coupling, cognitive overhead) that far outweights the code deduplic" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234423 (https://phabricator.wikimedia.org/T109000) (owner: 10Mattflaschen) [17:22:01] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 5WMF-deploy-2015-09-01_(1.26wmf21), 7WorkTypeNewFunctionality: Editing an empty Flow board description (that has previously contained a category?) gives "Previous revision identifier is missing" - https://phabricator.wikimedia.org/T110358#1584716 (10... [17:22:15] 3Collaboration-Team-Current, 10Echo, 5Patch-For-Review, 5WMF-deploy-2015-08-25_(1.26wmf20), and 3 others: Missing database prefix for echo_event - https://phabricator.wikimedia.org/T105890#1584717 (10DannyH) 5Open>3Resolved a:3DannyH [17:28:16] 6Collaboration-Team-Backlog, 10Collaboration-Team-Sprint-A-2015-05-06, 10Collaboration-Team-Sprint-B-2015-05-20, 10Echo, and 2 others: A6. Notifications resulting from a conversation in flow do not disappear after visiting the page. - https://phabricator.wikimedia.org/T95860#1584772 (10DannyH) [17:28:20] 3Collaboration-Team-Current, 10Collaboration-Team-Sprint-T-2015-04-08, 10Echo, 7Schema-change, 7WorkTypeNewFunctionality: Perform schema change to echo_target_page changing from a 1 to 1 mapping between pages and user/notification to a 1 to many. - https://phabricator.wikimedia.org/T94427#1584769 (10Danny... [17:30:31] (03PS4) 10Jforrester: build: Use npm for jshint, jscs, jsonlint, csslint and banana [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/230968 (https://phabricator.wikimedia.org/T63618) [17:32:27] (03PS2) 10Jforrester: build: Enable jscs rule 'requireSpacesInsideBrackets' and make pass [extensions/Flow] - 10https://gerrit.wikimedia.org/r/232423 [17:32:40] 3Collaboration-Team-Current, 10Flow, 10WMF-Server-Backports, 5Patch-For-Review, and 2 others: addurl CAPTCHA broken (HTML is rendered as text and CAPTCHA always fails you even if you fill in the form right) - https://phabricator.wikimedia.org/T109554#1584790 (10DannyH) 5Open>3Resolved [17:34:58] (03CR) 10jenkins-bot: [V: 04-1] build: Enable jscs rule 'requireSpacesInsideBrackets' and make pass [extensions/Flow] - 10https://gerrit.wikimedia.org/r/232423 (owner: 10Jforrester) [17:36:10] Sigh, somehow my browser closed [17:36:14] 6Collaboration-Team-Backlog, 10Flow, 10MobileFrontend, 7Tracking: Mobile web - Flow issues (tracking) - https://phabricator.wikimedia.org/T93430#1584798 (10DannyH) [17:36:16] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7Tracking, 5WMF-deploy-2015-09-01_(1.26wmf21): Writing any posts through the mobile Flow interface is completely impossible - https://phabricator.wikimedia.org/T110485#1584797 (10DannyH) 5Open>3Resolved [17:40:47] mooey|away, https://gerrit.wikimedia.org/r/#/c/233851/ and https://gerrit.wikimedia.org/r/#/c/234402/ [17:43:50] 3Collaboration-Team-Current, 10Flow: Don't switch preference for VE/wikitext if your device doesn't support your preferred editor - https://phabricator.wikimedia.org/T110706#1584835 (10DannyH) 3NEW [17:45:34] 6Collaboration-Team-Backlog, 10MassMessage, 5Patch-For-Review, 7Wikimedia-log-errors: Invalid parameter for message "logentry-massmessage-failure" - https://phabricator.wikimedia.org/T93110#1584850 (10bd808) [17:48:07] (03CR) 10Legoktm: [C: 032] Make sure flow entries are created with correct content [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234515 (https://phabricator.wikimedia.org/T109768) (owner: 10Matthias Mullie) [17:50:09] (03Merged) 10jenkins-bot: Make sure flow entries are created with correct content [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234515 (https://phabricator.wikimedia.org/T109768) (owner: 10Matthias Mullie) [17:51:22] (03CR) 10Legoktm: [C: 031] "This is reverting 686959b368c296f01ab6fd199bc2c3425af132a6, which was a safety net in response to a data loss bug. I think it's a good ide" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234575 (https://phabricator.wikimedia.org/T108348) (owner: 10Mattflaschen) [17:51:54] (03CR) 10Legoktm: "This is being reverted in Ie409962f27e2cd5b8d1f8fde579c7536195a8020" [extensions/Flow] - 10https://gerrit.wikimedia.org/r/203479 (https://phabricator.wikimedia.org/T95580) (owner: 10EBernhardson) [18:01:07] (03CR) 10Legoktm: Allow converting entire LQT namespaces (031 comment) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234089 (https://phabricator.wikimedia.org/T109931) (owner: 10Sbisson) [18:06:37] (03CR) 10Mattflaschen: "Yes. It's not a full revert, though. I'm keeping the part that checks for null (which will also catch if rev_content is totally missing," [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234575 (https://phabricator.wikimedia.org/T108348) (owner: 10Mattflaschen) [18:48:49] (03CR) 10Legoktm: [C: 04-1] Migrate to extension.json and deprecate PageTriage.php (036 comments) [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/230016 (https://phabricator.wikimedia.org/T87954) (owner: 10Jforrester) [19:08:23] (03PS5) 10Jforrester: Migrate to extension.json and deprecate PageTriage.php [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/230016 (https://phabricator.wikimedia.org/T87954) [19:08:28] (03CR) 10Jforrester: Migrate to extension.json and deprecate PageTriage.php (036 comments) [extensions/PageTriage] - 10https://gerrit.wikimedia.org/r/230016 (https://phabricator.wikimedia.org/T87954) (owner: 10Jforrester) [19:17:26] (03PS1) 10Sbisson: [WIP] Edit post widget [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234603 (https://phabricator.wikimedia.org/T109000) [19:20:00] RoanKattouw_away, mooeypoo: do you see something obviously wrong in the way the editor is initialized inside the widget that would cause not to behave like every other editor we have? The browser tests are complaining that the editor is disable. It never happens with the description widget. https://gerrit.wikimedia.org/r/234603 [19:22:08] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase: Switch Flow to RESTBase - https://phabricator.wikimedia.org/T94574#1585185 (10Mattflaschen) Okay, sounds like next step (soon, and before conversion to RESTBase) is to convert to VRS and Parsoid v3. [19:24:07] 6Collaboration-Team-Backlog, 10Flow: Switch Flow to use Parsoid v3 API with VirtualRESTService - https://phabricator.wikimedia.org/T110721#1585194 (10Mattflaschen) 3NEW [19:24:32] 6Collaboration-Team-Backlog, 10Flow: Switch Flow to use Parsoid v3 API with VirtualRESTService - https://phabricator.wikimedia.org/T110721#1585194 (10Mattflaschen) [19:24:35] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase: Switch Flow to RESTBase - https://phabricator.wikimedia.org/T94574#1585204 (10Mattflaschen) [19:30:50] 6Collaboration-Team-Backlog, 10Flow: Search for localized useliquidthreads as well - https://phabricator.wikimedia.org/T110723#1585238 (10Mattflaschen) 3NEW [19:30:58] 6Collaboration-Team-Backlog, 10Flow: Search for localized useliquidthreads as well in find/replace - https://phabricator.wikimedia.org/T110723#1585246 (10Mattflaschen) [19:42:24] (03CR) 10Mattflaschen: [C: 04-1] "Looks good except small issue with the Flow description." (033 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234089 (https://phabricator.wikimedia.org/T109931) (owner: 10Sbisson) [19:48:44] (03CR) 10Sbisson: Allow converting entire LQT namespaces (032 comments) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234089 (https://phabricator.wikimedia.org/T109931) (owner: 10Sbisson) [19:48:49] (03PS3) 10Sbisson: Allow converting entire LQT namespaces [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234089 (https://phabricator.wikimedia.org/T109931) [19:51:51] stephanebisson: If the editor really stays disabled for too long (verify in a real browser), that usually means an unreached popPending() call [19:56:12] RoanKattouw: it becomes enabled after a very short time, like the other ones I assume, but the tests trip on this one [19:56:37] Strange [19:56:41] How do the tests know how long to wait for? [19:57:40] RoanKattouw: active waiting, they keep retrying until it is there but they dont retrying until it is enabled [19:58:00] I'd like to understand why it's only a problem with this one [19:58:31] the description editor should start disabled as well, right? [20:00:29] mooeypoo, reviewing your split alert/messages patch now. [20:01:09] (03PS1) 10Cscott: Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T94574) [20:03:18] (03CR) 10jenkins-bot: [V: 04-1] Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T94574) (owner: 10Cscott) [20:03:32] (03PS2) 10Mattflaschen: Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T110721) (owner: 10Cscott) [20:04:19] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review: Switch Flow to use Parsoid v3 API with VirtualRESTService - https://phabricator.wikimedia.org/T110721#1585407 (10Mattflaschen) a:3cscott [20:06:02] (03CR) 10jenkins-bot: [V: 04-1] Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T110721) (owner: 10Cscott) [20:06:11] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase, 5Patch-For-Review: Switch Flow to RESTBase - https://phabricator.wikimedia.org/T94574#1585414 (10cscott) @Mattflaschen Once Flow uses VRS, switching to RESTBase will happen magically, since that's the default VRS configuration for production. [20:09:06] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase, 5Patch-For-Review: Switch Flow to RESTBase - https://phabricator.wikimedia.org/T94574#1585418 (10Mattflaschen) >>! In T94574#1585414, @cscott wrote: > @Mattflaschen Once Flow uses VRS, switching to RESTBase will happen magically, since that's th... [20:09:28] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase, 5Patch-For-Review: Switch Flow from ExternalStore to RESTBase - https://phabricator.wikimedia.org/T94574#1585419 (10Mattflaschen) [20:09:50] 6Collaboration-Team-Backlog, 10Flow, 10Parsoid, 10RESTBase, 5Patch-For-Review: Switch Flow from ExternalStore to RESTBase - https://phabricator.wikimedia.org/T94574#1585420 (10cscott) Ah, ok. [20:15:06] Does anybody know of an existing accepted term for "wikis that aren't as big and established as English"? [20:15:58] non-English Wikipedias ? [20:16:35] I'm trying to figure out a general term that doesn't include "non" in the name. What I have so far is "emerging wiki" but that isn't very good [20:17:07] "smaller wikis" would be simplest, if that's not considered negative. [20:17:10] I guess what I want is the current accepted euphemism [20:17:25] dannyh, I don't know if there is one. There is "sister projects", but that refers to non-Wikipedias. [20:17:47] (03PS3) 10Cscott: Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T110721) [20:17:56] Or I guess Wikipedia is one of the sisters, so it depends on the context. [20:18:34] "sister projects" does (or at least, should) include wikipedia [20:18:54] (03CR) 10jenkins-bot: [V: 04-1] Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T110721) (owner: 10Cscott) [20:19:13] 6Collaboration-Team-Backlog, 10Echo: Echo's 'seen time' should be replaced with marking specific notifications as seen - https://phabricator.wikimedia.org/T110731#1585438 (10Mooeypoo) [20:21:09] (03PS2) 10Mooeypoo: Use viewmodel in Echo's front end [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234451 [20:21:27] mooeypoo, how does 'seen' affect Special:Notifications? [20:21:42] I thought it only affected the badge. [20:21:53] matt_flaschen, 1) it should... the notifications should appear as "new" / "unseen" just like in the badge. [20:22:20] 2) You're right about this particular moment, I meant more to say that the user can miss unseen notifications when they look at the full list in Special:Notifications [20:22:52] (03CR) 10jenkins-bot: [V: 04-1] Use viewmodel in Echo's front end [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234451 (owner: 10Mooeypoo) [20:22:56] As in, you still haven't seen the 75 notifications -- until you go to Special:Notifications are realize you haven't seen them yet -- but at that point, your badge claims you have none. That makes sense? [20:22:56] matt_flaschen thanks :) [20:23:27] dannyh, don't know that I really answered your question. Sister projects does not refer to smaller Wikipedias. [20:23:35] mooeypoo, 2) meaning they'll see unread notifications, but there could be too many, so they should be able to zero in on unseen? [20:24:14] matt_flaschen, yeah, I think operationally speaking, there should be no difference between seen and read in that aspect. They should both be flags on the notification. [20:24:38] (03PS4) 10Cscott: Use the VirtualRESTService to configure Flow [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234667 (https://phabricator.wikimedia.org/T110721) [20:25:00] matt_flaschen, another example: You have 50 unread/unseen notifications. You get emails, you ignore all except for *one* which you click and go to some reply on your board. Is that notification read? (should be) is it seen? (should be) [20:25:24] but the other notifications in your queue are still unseen even if that notification was the latest. [20:25:27] matt_flaschen, does that make sense? [20:26:18] If both seen and read are flags on the notification, doing that is trivial. [20:26:35] Right now, only 'read' is a flag, and 'seen' is going by a user config timestamp [20:27:03] mooeypoo, yes, but there are some issues (and not just technical). For example, once you go to Special:Notifications, now all should be seen, so you're really saying, "Here's what was unseen immediately prior to this page load". [20:27:13] Yes [20:27:17] Not all, though. All in that PAGE. [20:27:23] You still have pagination in Special:Notification [20:27:28] True [20:27:45] what I'm saying is that "seen" should be actually seen, and not "you've seen notification up to this date, even if you haven't seen all notifications prior" [20:27:51] (03PS2) 10Sbisson: [WIP] Edit post widget [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234603 (https://phabricator.wikimedia.org/T109000) [20:28:11] For seen to be helpful, it should actually mean 'seen' which should be per notification [20:28:21] matt_flaschen, actually, I"m worried this will be compounded with global notifications [20:28:51] There's a lot more risk of users having notifications they haven't actually seen being marked as "Seen" [20:30:30] mooeypoo, yeah, it is potentially a bigger issue then. Not sure how big an issue it actually is now. (Background: This decision was made in https://phabricator.wikimedia.org/T94634 to simplify the implementation.) [20:30:37] matt_flaschen, and the badge itself can only reflect what it has, so if you have 25 in the popup, and all of those are seen, but there are 75 taht aren't seen yet, the badge will show you "no unseen messages". We can fix that with infinite scroll (that's relatively trivial in this commit, like ToC does) and suddenly you do have unseen stuff. [20:30:53] matt_flaschen, aye, that's why I added as a general bug and not an urgent one [20:31:13] I think we need to have a good conversation about Echo in general in the next quarter when we work on it more thoroughly [20:31:33] there are a lot of issues that we should fix and rethink, especially with global notifications coming up [20:35:46] matt_flaschen, I'm about to +2 both of your GuidedTour fixes now that the underlying one has tests. They both work perfectly. [20:35:54] Any objections? Is it waiting on anything? [20:36:02] mooeypoo, nope. Thanks for reviewing. [20:36:52] matt_flaschen, done [20:41:11] dannyh, "smaller wikis" is good. Some of them have been small for over a decade, so "emerging" wouldn't be appropriate. :) [20:41:22] (03PS3) 10Mooeypoo: Use viewmodel in Echo's front end [extensions/Echo] - 10https://gerrit.wikimedia.org/r/234451 [20:44:05] mooeypoo: can you do me a favor, please? Can you check you local vagrant, something like - cat /Users/etonkovidova/Documents/gerrit/vagrant/puppet/modules/role/templates/flow/conf.php.erb [20:44:16] mooeypoo: and check if you can see: define( 'NS_FLOW_TEST', 190 ); [20:44:36] mooeypoo: it's a short file - you should see/not see it immediately... [20:44:37] 3Collaboration-Team-Current, 10Flow, 5Patch-For-Review, 7WorkTypeNewFunctionality: Opt-in: Guided tour on user talk for first visit to new Flow board - https://phabricator.wikimedia.org/T108266#1585539 (10Mooeypoo) Here are screenshots of the guided tour: Step 1: {F2475821} Step 2 (optional, depending on... [20:45:37] etonkovidova, hang on [20:45:57] mooeypoo: kk [20:50:24] 6Collaboration-Team-Backlog, 10Flow: Search for localized {{#useliquidthreads}} as well in find/replace in LQT→Flow conversion - https://phabricator.wikimedia.org/T110723#1585543 (10Legoktm) [21:03:16] etonkovidova, negative. I don't have that. But I am not sure when I updated my vagrant repo [21:05:03] (03CR) 10Mattflaschen: [C: 04-1] "Looks good." (0310 comments) [extensions/Echo] - 10https://gerrit.wikimedia.org/r/231200 (https://phabricator.wikimedia.org/T108190) (owner: 10Mooeypoo) [21:06:04] etonkovidova, yes, it should be there if you have updated MWV: https://git.wikimedia.org/blob/mediawiki%2Fvagrant.git/243f2d50503328c4471d48bc7b1d52d310fa078a/puppet%2Fmodules%2Frole%2Ftemplates%2Fflow%2Fconf.php.erb [21:10:48] mooeypoo: matt_flaschen: - thanks . Yes, I see it but somehow I cannot update my vagrant - with vagrant git-update. Hmm... [21:11:06] etonkovidova, to update MediaWiki-Vagrant itself, i believe you have to do: [21:11:09] git pull [21:11:09] I'm out for a cpl hours. bbl. [21:11:11] I don't think git-update covers that. [21:11:24] etonkovidova, while you're in the top-level MWV directory. [21:12:21] git-update updates the extension repos and mw repo, not vagrant itself [21:12:31] yeah, git pull in vagrant [21:12:39] I just did that too [21:13:01] And then vagrant provision. [21:14:20] matt_flaschen, hm. I just saw your comment about mw.trac timing (it wasn't on purpose, thanks for catching it) -- but that also made me notice that we have mw.hook( 'ext.echo.overlay.beforeShowingOverlay' ) in the old version -- and not in the new.. should I re-add that too? [21:14:27] matt_flaschen: mooeypoo - oh - will do it [21:14:43] mooeypoo: that hook is necessary for nav popups to work in the generated flyout [21:14:45] If so, is the full popup what I need to output with that event, or just the notification list select widget [21:14:53] nav popups? [21:14:59] que? [21:15:35] mooeypoo: bottom of https://en.wikipedia.org/wiki/MediaWiki:Gadget-popups.js [21:17:42] mooeypoo, https://en.wikipedia.org/wiki/Wikipedia:NAVPOPS . It's not just for them, though. [21:17:55] I added it as a general user script entry point (I think originally so I could hook into something that annoyed me). [21:17:59] The docs say the overall overlay, but as long as it includes the full list of notifications, it's probably good enough. [21:18:41] Still might break some user scripts, but that's an occupational hazard. [21:18:45] I'll check where it's being used. [21:20:41] (03PS3) 10Sbisson: [WIP] Edit post widget [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234603 (https://phabricator.wikimedia.org/T109000) [21:22:14] quiddity: I showed mooeypoo the amazingness that is navpopups [21:22:31] matt_flaschen: last time I mwgrep'd a few weeks ago nothing was using it outside of popups [21:23:45] legoktm, there's one person now on ru. I apparently removed my listener user script: https://phabricator.wikimedia.org/P1948 [21:24:00] I didn't include the private wikis, but those are just popups. [21:26:21] mooeypoo: matt_flaschen: vagrant is updated - no problems Just needed ./setup.sh :) [21:28:06] mw.hook('ext.echo.overlay.beforeShowingOverlay').add(reduce_crap); [21:28:08] hahaha [21:28:15] https://ru.wikipedia.org/wiki/%D0%A3%D1%87%D0%B0%D1%81%D1%82%D0%BD%D0%B8%D0%BA:Incnis_Mrsi/%D0%BD%D0%B5%D0%9E%D0%B1%D1%81%D1%83%D0%B6%D0%B4%D0%B5%D0%BD%D0%B8%D0%B5_%D1%83%D1%87%D0%B0%D1%81%D1%82%D0%BD%D0%B8%D0%BA%D0%B0.js [21:28:26] :) [21:29:46] lol [21:30:56] mooeypoo, there is also https://ps.wiktionary.org/wiki/%D9%85%D9%8A%DA%89%D9%8A%D8%A7%D9%88%D9%8A%DA%A9%D9%8A:Main.js . Shouldn't need anything special though. [21:31:38] that's popups too [21:32:48] (03CR) 10Mattflaschen: [C: 032] "Testing confirms this space is not required." [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234212 (owner: 10Amire80) [21:33:40] 6Collaboration-Team-Backlog, 5Patch-For-Review: Make the VE/source toggle more discoverable - https://phabricator.wikimedia.org/T101316#1585664 (10Catrope) I've been meaning to change the WT->VE button too (not just the icon, more aspects of it). Would you say that changing that to be identical to the proposed... [21:33:55] Doh [21:35:29] (03Merged) 10jenkins-bot: Remove extra space from {{#time:}} calls [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234212 (owner: 10Amire80) [21:38:38] (03PS35) 10Mooeypoo: Split alerts and messages in Echo [extensions/Echo] - 10https://gerrit.wikimedia.org/r/231200 (https://phabricator.wikimedia.org/T108190) [21:40:04] matt_flaschen, isn't that a copy of popups ? [21:40:15] mooeypoo, hence Doh. ^ [21:40:53] matt_flaschen, also, I just squashed the change of the viewmodel onto the original change. I'm going to go over your comments and implement what's relevant (most are) but when you re-review, take into account things are different. I'm using a view model for all the operations and to keep the badge and notification list updated now. [21:41:12] Will do [21:49:10] matt_flaschen, okay, regarding your comment about "mw-echo-unread-notifications" -- the reason I didn't touch it is because it's also affecting tests. I could switch it up completely, though (didn't want to risk issues + touching too many files) [21:49:23] do you think we can do it in another commit or should I just get this over with now? [21:49:39] mooeypoo, if it's just find/replace shouldn't break tests, right? [21:49:43] Though I admit they are delicate. [21:49:45] *shouldn't*. [21:50:00] Yeah mostly browser tests. The other test I should get rid of, as it's no longer valid anyways. [21:50:08] I would go for it, but up to you. [21:50:09] Also, now that we have a viewmodel, I should create a few tests for it. [21:56:16] matt_flaschen, what's the convention in php documentation with optional default params? Is it like jsduck, "* @param int [$dbSource=DB_SLAVE] Use master or slave database to pull count" [21:56:47] 6Collaboration-Team-Backlog, 10Flow: Can not undelete deleted Flow boards using revision undeletion - https://phabricator.wikimedia.org/T110680#1585714 (10Luke081515) OK, so at first, this page was a normal talkpage, I deleted the page, after that, I created a new talk page, and replaced it by flow, so the act... [21:57:13] legoktm, you probably will also know (?) [21:57:35] doxygen doesn't have one afaik [21:57:41] so it usually gets put in the text [21:57:52] @param type $var Blah blah blah (default false) [21:58:02] Any way to mark "optional" ? [21:58:21] well, I guess it's implied [21:58:40] mooeypoo, I think that 'optional' to be done in text too in PHP. [21:59:04] * @param boolean $cached Optional. Set to false to bypass the cache. (Defaults to true) [21:59:05] * @param int $dbSource Optional. Use master or slave database to pull count (Defaults to DB_SLAVE) [21:59:07] ? [21:59:16] I'm going to copy/paste these all over the file so I want to make sure [21:59:23] That looks ok? [21:59:34] yeah, I'd probably put optional at the end though [21:59:58] Looks fine (either there or with optional at end). [22:00:02] @param boolean $cached Set to false to bypass the cache. (Optional. Defaults to true) [22:00:07] I'll pepper those in as needed [22:01:10] 6Collaboration-Team-Backlog, 10Flow, 7Database: Add artificial primary key to flow_wiki_ref and flow_ext_ref - https://phabricator.wikimedia.org/T109676#1585718 (10Mattflaschen) [22:03:29] 6Collaboration-Team-Backlog, 10Flow, 7Database: Add artificial primary key to flow_wiki_ref and flow_ext_ref - https://phabricator.wikimedia.org/T109676#1585731 (10Mattflaschen) We'll probably go with an artificial primary key (which I think should be a UUID), both to reduce index size and avoid collisions (... [22:03:54] matt_flaschen, eek, "There is no way of knowing whether they have *ever* had them, since if they get 2000 alerts their message history will be forgotten." <-- REALLY?? If someone has 2000 alerts, their messages are forgotten? Is that intended? sounds extremely dangerous, especially when dealing with global notifications, no? is it coming on on the expense of the other? [22:05:08] mooeypoo, according to the docs, $wgEchoMaxUpdateCount is also the maximum number of notifications period. It is set to 2000. [22:05:22] meh. We might want to reconsider that. [22:05:30] mooeypoo, maybe. I'm sure it's for performance reasons. [22:05:49] Let me take a look. [22:06:27] matt_flaschen, but if we're separating them now, it might be worth considering to separate the max-count too [22:06:50] 1000 each, even... something. Think about extreme cases where a prolific user gets tons of alerts and few messages [22:07:04] they can wake up one day to find their "inbox" empty of messages [22:07:13] Sounds dangerous to me. [22:07:47] "performance reasons" [22:08:07] Can we performance-reason it differently? [22:08:15] 1000 each? 500 messages, 1500 alerts? [22:08:32] It would probably actually perform better split as well. [22:08:56] Especially since that would not only cause their message flyout to be empty, but to completely vanish [22:09:30] Since it works by deleting everything before a certain event ID. To get that event ID, it does ORDER BY ... DESC OFFSET 2000 LIMIT 1, and there is a warning about having a large offset which suggest something table scan-ny. [22:11:13] mooeypoo, do you want to file it? If not, I can. [22:13:57] matt_flaschen, if you can, that'll be awesome, I'm in the middle fo grepping for hasHadMessages and changing clearHasMessages [22:14:06] a little confused here with namings of things [22:16:57] Okay, that was mind bending a bit. [22:17:42] 6Collaboration-Team-Backlog, 10Echo: Investigate maximum echo notification count - https://phabricator.wikimedia.org/T110745#1585744 (10Mattflaschen) 3NEW [22:18:00] 6Collaboration-Team-Backlog, 10Echo: Investigate maximum echo notification count - https://phabricator.wikimedia.org/T110745#1585752 (10Mattflaschen) [22:18:09] matt_flaschen, I was about to delete the unused LESS mixins (truncated-text()) in exho.mixins.less, but there's a comment above that makes me weary: "// Use in Flow, Echo and MobileFrontend extensions." <-- I grepped the Echo repo and couldn't find any use of them, and Flow and MobileFrontEnd *shouldn't* be using mixins from a different repo (right??) but... do you know anything about this? [22:18:14] is it safe to remove or am I missing something? [22:18:58] MobileFrontend/resources/skins.minerva.base.styles/ui.less [22:18:58] 199:.truncated-text { [22:19:11] So, reimplemented in MF, not used in Flow at all. [22:19:13] so they have their own [22:19:16] (anymore) [22:19:28] I can delete them, then, no? [22:19:33] yes [22:20:36] Excellent. I can now rename it to .mixin- too. \o/ [22:20:37] thanks [22:24:07] (03CR) 10Mooeypoo: "Answers to comments on ps 34. Corrections coming up in the next patchset." (0310 comments) [extensions/Echo] - 10https://gerrit.wikimedia.org/r/231200 (https://phabricator.wikimedia.org/T108190) (owner: 10Mooeypoo) [22:28:38] (03CR) 10Mattflaschen: [C: 032] VisualEditorWidget: Guard against setting selection to -1 [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234343 (https://phabricator.wikimedia.org/T110393) (owner: 10Catrope) [22:31:24] (03Merged) 10jenkins-bot: VisualEditorWidget: Guard against setting selection to -1 [extensions/Flow] - 10https://gerrit.wikimedia.org/r/234343 (https://phabricator.wikimedia.org/T110393) (owner: 10Catrope) [22:45:42] (03CR) 10Mattflaschen: [C: 04-1] "Looks good, except you can use the flow-link-to-archive class now. After that, I'll do final testing." (031 comment) [extensions/Flow] - 10https://gerrit.wikimedia.org/r/230924 (https://phabricator.wikimedia.org/T108266) (owner: 10Mooeypoo) [22:46:58] hmmmmmm [22:48:08] 6Collaboration-Team-Backlog, 10Flow: Change title (document.title) when the user has an editor open - https://phabricator.wikimedia.org/T107356#1585883 (10Mattflaschen) >>! In T107356#1581796, @Quiddity wrote: > I think we should keep the same "Editing" terminology, and not invent a new "Posting" term. (Especi... [22:48:20] matt_flaschen, legoktm brainstorming a second about the ext.echo.overlay.beforeShowingOverlay ? So, in the 'old' system, it fired right before the overlay was created, so after the button was clicked. In the current code, though, the flyout is created on click, *but* it can actually change ... the model populates itself from the api and the widgets listen to "add" / "remove" and "clear" events and respond accordingly in terms of dis [22:48:20] playing the list of notifications [22:49:05] That said, I am trying to decide where and when to add this hook now. [22:49:43] If I add it to the end of the operation of the button click, then it might not update correctly when/if the list is updated while closed (which can happen, some day, if/when we do push notifications or something) [22:50:27] On the other hand, even if that happens, it happens in the background, and the gadget that listens to it now (popups) seems to only care about the user actually *seeing* it.... [22:50:54] mooeypoo, yeah, the goal is that everything is updated before the user sees it. It doesn't really matter how long it's been sitting hidden. [22:51:04] So, is it safe to just fire this hook (with the list of notifications) only after click, and if the widget changes -- either while it is open (can happen, theoretically) or while closed, it won't fire. [22:51:07] mooeypoo, but I'm not sure if it's same to re-run the hook on the same content. [22:51:25] matt_flaschen, well, every time you open the popup it loads notifications for you [22:51:29] so we're not sure it's the same content [22:51:41] also, it's all being rebuilt, so.. it's not technically the same content :\ [22:52:00] mooeypoo, well, I was also thinking about the add/remove events. [22:52:16] yeah, the add/remove are good if I emit with the SPECIFIC $elements of the options [22:52:19] If there can be an add event that happens while it's opened, it should preferably handle that. [22:52:21] that actually might be my best bet [22:52:38] the widget adds itself on add event [22:53:00] matt_flaschen, okay, but that means the hook is fired multiple times while we build. [22:53:26] mooeypoo, yeah, that (firing on the specific elements) might be a good bet. It's slightly different from the hook name since it's not really the overlay, but that could probably just be handled by updating the docs. [22:53:43] Actually, no, it fires once from the API with a list of model elements -- BUT theoretically (and practically) we can add individual notifications to the model, and they will be "added" individually, causing that hook to fire again. [22:54:42] So, every time we open the popup, it will fire once. If we add individual notifications (not done today, but may be in the future) regardless of whether the popup is open or closed, it will fire again. [22:54:45] mooeypoo, so initially it fires with e.g. max 25, then you might get a few notifications gradually on the same page which will be caught by later API requests and added bit-by-bit? [22:54:47] I don't think we should constrain ourselves to keep compatability with the old hook. We just need to provide the same functionality: a hook that gadgets can use to modify the notification content before/after the user sees it [22:54:53] matt_flaschen, yes [22:55:10] legoktm, if that's the case, then on "add" is best [22:55:28] we just need to be aware it might be fired a coupleof times -- every time something is added to the popup [22:55:36] hopefully we add things in batches [22:55:50] but we don't necessarily HAVE to... so.. there's a "risk" of multiple hook fires. [22:55:59] I can see how re-running on the same content could cause issues, I don't think firing e.g. 10 times on a page will actually cause perf issues. [22:56:11] And it's not even 10 times in close succession. [22:57:36] matt_flaschen, as I just demonstrated to legoktm (to his chagrin) there's nothing stpping the user from click-click-click-clicking the badge. I'm caching the promises right now, so at least we won't get insane amounts of api requests while one is running, but we absolutely COULD have tons of api+hook runs if users are abusing this [22:58:16] mooeypoo, but I thought it would only fire the hook when actually adding a notification for the first time? [22:58:29] (Maybe adding in a large batch, maybe a batch of one, but only when new) [22:58:37] matt_flaschen, yes. But when you reload, right now at least, it empties + refetches from the API [22:58:46] because we don't have an API way of knowing "show me only what i didn't already get" [22:58:55] so it's mimicking current behavior -> clear + refetch [22:59:16] mooeypoo, so if it's the same as the current hook behavior, I don't see a big issue. [22:59:32] I thought you were saying it would only add the new ones. [22:59:38] matt_flaschen, yeah, I guess so. We potentially wil have issues with non batched adds, but that doesn't happen YET anyways. [23:00:07] technically it's in the "onAdd" handler, so we *could* potentially have it respond to multiple adds. It doesn't happen now, but I am slightly concerned it may happen in the future. [23:00:16] I guess it's not relevant yet, though. [23:00:54] Also, we really REALLY should fix the API so that we can only query new stuffs. [23:01:02] mooeypoo, as long as the same DOM element is not contained in two hook firings I don't think it will cause wrong listener behavior. Popups actually apparently has a check to avoid running twice on the same DOM element, but I don't really think listeners should have to handle that themselves. [23:02:13] matt_flaschen, hm. Okay, I'm sending it "this.$element" (the whoel of the notifications widget)... so I guess I shouldn't do that. Instead I should only send the widgets that pop up... But those are separate jQuery elements, and the gadget is doing $whatIget.find( ... ) [23:02:25] will .find() work on a group ... [23:02:28] * mooeypoo tests [23:02:32] Never mind, I don't think Popups actually has such a check. It seems it always assumes it's new content. [23:03:07] it uses .find( ... ) to set up the popups on links [23:04:12] Right, I'm wondering if it will break if it gets the same mw-echo-state more than once. Preferably, listeners should not have to worry about that. [23:04:36] I think they probably wont, it should override or no-op if it is attached twice [23:04:41] but I'm trying to see if I can prevent that too [23:14:55] 6Collaboration-Team-Backlog, 10Echo: Investigate maximum echo notification count - https://phabricator.wikimedia.org/T110745#1585965 (10Legoktm) IMO we should consider making these time based rather than a raw count. [23:25:21] matt_flaschen, so, when I add widgets, I am grouping their $element together "$elements = $elements.add( widget.$element );" and then the hook fires the entire group. Since that responds to .find( ... ) it should work, and will not overwhelm anything, since it only sends stuff that are new to whatever listens to that hook. [23:25:32] I just tested the $() group thing and it works. [23:25:57] Sounds good to me [23:26:28] I assume we don't need to notify anything when/if content is deleted/removed [23:28:23] Nope [23:42:55] 3Collaboration-Team-Current, 10Wikimedia-Extension-setup, 5Patch-For-Review, 7WorkTypeNewFunctionality: Re-enable Flow on fawikibooks - https://phabricator.wikimedia.org/T109816#1586036 (10Mattflaschen) It's in QA review. [23:48:01] 6Collaboration-Team-Backlog, 10Echo: Investigate maximum echo notification count - https://phabricator.wikimedia.org/T110745#1586052 (10Mooeypoo) I actually think this is a bit more complicated than that, potentially. I am not sure how much traffic there is on average, but consider a prolific user that is act... [23:57:44] 6Collaboration-Team-Backlog, 10Flow: {{FlowMention|Someone}} displays a link that connects with the following characters after it, in case there's no space between them (but only if the text is not capitalized) - https://phabricator.wikimedia.org/T110757#1586088 (10Mooeypoo) [23:58:09] 6Collaboration-Team-Backlog, 10Flow: {{FlowMention|Someone}} displays a link that connects with the following characters after it, in case there's no space between them (but only if the text is not capitalized) - https://phabricator.wikimedia.org/T110757#1586093 (10Mooeypoo)