[07:01:19] <Krinkle>	 A_smart_kitten: Indeed. It also looks like it breaks the default message by default which would be a regression.
[07:01:27] <Krinkle>	 both are display:none
[07:01:51] <Krinkle>	 the JS refers to WikimediaEvents, which suggests maybe this patch was first tried there but abandoned? There is no such patch linked on the task
[07:02:09] <Krinkle>	 it's not obvious why WikimediaEvents couldn't do this.
[07:03:09] <Krinkle>	 Athough I will say that wgEmailConfirmationBanner is false by default
[07:03:52] <Krinkle>	 it lacks @since or @warning Experimental though, so yeah it certainly suggests that it would work
[07:04:01] <Krinkle>	 and until this patch did
[07:04:17] <Krinkle>	 it also requires JS with no fallback for client-nojs
[07:05:45] <Krinkle>	 Task says it runs for 2 weeks, so they'll presumably revert it out of REL_46
[08:57:59] <Reedy>	 famous last words
[13:06:21] <Neriah>	 Hi :)
[13:06:27] <Neriah>	 does anyone know if I should split this change into two separate changes, one for the table changes and the other for the rest of the change?
[13:06:28] <Neriah>	 https://gerrit.wikimedia.org/r/c/mediawiki/core/+/1254265
[13:10:08] <Krinkle>	 Neriah: yes, and the change should work even without the schema change
[13:11:03] <Krinkle>	 or, the second change wait until the first one is applied to all production databasesa at WMF
[13:12:48] <Krinkle>	 Adding columsn to a table the size of logging is not trivial. It may be worth getting consensus on that first before spending too much time on the details of how the columns w oud be used (=the easy part, relatively speaking)
[13:14:18] <Krinkle>	 it seems that, unlike for edits, there is no bot flag for log actions today. That should probably be done first (https://phabricator.wikimedia.org/T18816) and woudl be less controversial. That woudl be specific to Recent changes.
[13:14:50] <Krinkle>	 I suspect we may end up doing bot flag via a second table like change_bot by log_id/rev_id perhaps, akin to change tags. Or perhaps even as a special kind of change tag.
[13:15:10] <Krinkle>	 rather than inside the table itself, but that's for our DBAs to decide.
[13:24:19] <Krinkle>	 commented at T21322
[13:24:19] <stashbot>	 T21322: Special:Log needs a way to show only activity from humans (Hide bots) - https://phabricator.wikimedia.org/T21322
[13:29:34] <Neriah>	 Krinkle: Thanks for the feedback!!
[13:29:45] <Neriah>	 Do you suggest waiting for the DBA team's response now before deciding how to proceed?
[13:29:53] <Neriah>	 that is, even before I split the change?
[13:29:57] <Krinkle>	 yes
[15:50:52] <A_smart_kitten>	 Krinkle: thanks for the reply re the core/testkitchen patch :)
[15:51:31] <A_smart_kitten>	 fwiw, it looks like that config variable was added relatively recently (after 1.45 was released) in https://gerrit.wikimedia.org/r/c/mediawiki/core/+/1250603
[15:54:31] <A_smart_kitten>	 i'll leave a comment on the patch about the fact that (IIUC) MW Core shouldn't be aware of ext:TestKitchen in this way
[22:14:25] <James_F>	 AaronSchulz: Note T408697 BTW. :-(
[22:14:25] <stashbot>	 T408697: Remove Swagger UI from MW-1.46 (~May 2026) release - https://phabricator.wikimedia.org/T408697
[22:22:33] <AaronSchulz>	 James_F: yeah, we keep playing the game of removing it from the REL_ branches until the magic Codex-based replacement comes.
[22:23:52] <James_F>	 :-(
[22:28:21] <AaronSchulz>	 Krinkle: I was experimenting with using https://github.com/swagger-api/swagger-ui/blob/v5.32.4/dist/swagger-ui-bundle.js but I get "unexpected token" syntax errors on my console. It looks like the problematic line is cut off during minimization (it has << ({xmlns:"http: } ...>> but the source file has the correct << ({xmlns:"http://www.w3.org/2000/svg", ...>>.  It works fine when I use ?debug=1...
[22:32:48] * AaronSchulz keeps running into obstacles with this rest sandbox stuff