[14:13:01] <kindrobot>	 I think these are on beta cluster. My understanding is that these should be automatically renewed using ACME. James_F is that true? If so are these errors expected?
[14:17:09] <kindrobot>	 florent: would you feel OK with Nikhil61 working with you to bring https://gerrit.wikimedia.org/r/c/mediawiki/extensions/WikiLambda/+/868787 to completion?
[14:18:39] <kindrobot>	 I don't want to take it away from you, rather, I think because it's almost there, it'd be a good way for Nikhil61 to pick up and follow the patterns, and have someone to ask for help/context.
[14:53:01] <florent>	 <kindrobot> "florent: would you feel OK..." <- Hi!
[14:53:01] <florent>	 Hmm, it's not just almost there, it's blocked by another issue. Please take a look at my latest comment
[14:56:28] <florent>	 I updated the patch, but it won't work
[14:57:37] <kindrobot>	 > blocked by another issue
[14:57:42] <kindrobot>	 You mean the performance issue?
[14:57:47] <florent>	 Yep
[14:57:56] <kindrobot>	 I think that's been fixed.
[14:58:05] <florent>	 I think not totally
[14:58:16] <florent>	 Let me check with a new rebase
[14:58:22] <kindrobot>	 https://phabricator.wikimedia.org/T326312
[14:58:31] <kindrobot>	 OK, thanks. :)
[14:58:55] <kindrobot>	 I'll be in a meeting for a bit, so probably won't be too responsive.
[14:59:12] <florent>	 no problem :)
[15:45:27] <florent>	 kindrobot: OK, there were a second culprit and it was a missing `await` in my own code 🤦. So I made some progress
[15:46:01] <florent>	 I'd like to finish this patch if that's OK for Nikhil61?
[15:48:22] <florent>	 I have this error at the end of the run: User does not have permission to edit. It's shown in the publication popup
[15:59:15] <kindrobot>	 Are you running with a logged in user?
[15:59:28] <florent>	 No, indeed!
[15:59:50] <florent>	 Is there a loggin I can use, or does it require to create one?
[16:00:32] <kindrobot>	 Hmmm, interesting question.
[16:01:06] <kindrobot>	 On the k8s e2e test runs you can use the standard docker username/password (so secure!)
[16:02:08] <kindrobot>	 zeljkof: is there a pattern around creating or passing in a user for the selenium tests?
[16:05:43] <kindrobot>	 florent: maybe setting MEDIAWIKI_USER and MEDIAWIKI_PASSWORD will automatically get you logged in? https://www.mediawiki.org/wiki/Selenium/How-to/Set_environment_variables
[16:06:19] <kindrobot>	 If that's the case, we can set up a known good user for beta cluster and use the default user for k8s.
[16:09:19] <florent>	 Yay! But that means supporting the logging scenario
[16:14:14] <kindrobot>	 Oh, it seems like there's a LoginPage.js in the wdio-mediawiki package.
[16:15:52] <kindrobot>	 LoginAdmin is probably what you want as it will use the MEDIAWIKI_USER and MEDIAWIKI_PASSWORD. I'm not sure if those are set for beta cluster, but if they're not, we can set them, and I can make sure they're set for the k8s clusters.
[16:16:22] <kindrobot>	 While you're developing it, you can set it either to your account, or a test account that you create.
[16:22:07] <florent>	 I could login using the MEDIAWIKI_USER and MEDIAWIKI_PASSWORD environment variables 😁
[16:55:14] <florent>	 Ahah, now I get an issue with the aliases not persisted
[16:55:28] <florent>	 I reproduce the issue when I create a new function
[16:55:42] <florent>	 Should I report it? Or is it a known bug?
[16:59:30] <James_F>	 florent: Please report that.
[16:59:34] <James_F>	 I think that's new. :-(
[17:00:05] <James_F>	 florent: Oh! Unless you're not pressing return after typing in the alias candidate.
[17:00:19] <James_F>	 In which case this is a known UX failure, but technically working-as-specified. :-(
[17:55:23] <florent>	 James_F: Oh! yeah, that's the reason!
[17:55:49] <James_F>	 florent: But "we" (Wikimedia, not necessarily Abstract) should fix that too. :-(
[18:26:38] <florent>	 James_F: Do you need me to create a ticket for that?
[18:27:02] <florent>	 When you say "we", that means it's a codex component behind the hood?
[18:38:20] <florent>	 If I may suggest another thing, we may warn the user about their inability to create a function before they fill any field ^^
[18:40:15] <florent>	 I tried to fill the form another browser and forgot to log in. I would have lost my changes
[18:54:01] <florent>	 I pushed another progress. The funny-but-not-so-funny thing now is that I fill the label of the first argument but when webdriver clicks on publish, the fields get cleared.
[18:55:26] <florent>	 I cannot reproduce manually (i.e. not through WebDriver), that makes that bug really "funny-but-not-so-funny" °_°'
[19:06:31] <James_F>	 florent: I'll make sure we have a ticket for that, thanks though!
[19:06:50] <James_F>	 That's very odd that the fields get cleared.
[19:07:27] <James_F>	 Agreed re. stopping you before you make edits; we'll be working on that as part of "Goal 6" of the current phase right now.
[21:19:11] <wm-bb>	 <Hogü-456> The https://zobject-decoder.toolforge.org/ Z-Object-Decoder is from my point of view a helpful tool. This helps me understanding more of the structure of the JSON-Definition of an Z-Object.  I hope that I will understand it better if I look more on it . Is the JSON-Format the format in which a Z-Object is saved or is it only a format used to be more user friendly as the format in <clipped message>
[21:19:11] <wm-bb>	 <Hogü-456>  which Z-Objects are saved. In Wikidata there are diffent formats available in which a Q-Item can be displayed. https://commons.wikimedia.org/wiki/File:Wikidata_URIs_for_different_formats.svg Is this also possible in Wikifunctions.
[21:30:27] <wm-bb>	 <vrandecic> It's a subset of JSON, yes