[07:25:46] <dcausse>	 o/
[08:48:41] <pfischer>	 o/ dcausse: thanks for the review, IIUC, the pipeline requires an approval, and obviously self-approving does not do the trick 🤷 https://gitlab.wikimedia.org/repos/data-engineering/airflow-dags/-/merge_requests/1514
[08:49:43] <dcausse>	 pfischer: sure lemme click on this button :)
[08:54:16] <pfischer>	 dcausse: thanks, that worked. I’ll go ahead and merge it.
[09:53:56] <pfischer>	 dcausse: if you have a sec’ before lunch: https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1182503 (SUP flink update, helmfile)
[09:54:05] <dcausse>	 pfischer: sure
[09:55:00] <dcausse>	 pfischer: just checking, do you know if the new flink-k8s-operator was deployed on wikikube?
[09:57:50] <dcausse>	 checking helmfile.d/admin_ng/flink-operator/values.yaml I see 1.4.0-wmf1-20241013 which seems a bit old
[10:03:39] <pfischer>	 Hm, I remember having seen tickets related to that but I’d have to check.
[10:05:33] <dcausse>	 hm I don't see a task to actually deploy the operator :/
[10:05:53] <dcausse>	 only T377134
[10:05:53] <stashbot>	 T377134: Create and distribute a flink base image with flink 1.20.0 - https://phabricator.wikimedia.org/T377134
[10:06:22] <pfischer>	 https://gerrit.wikimedia.org/r/c/operations/deployment-charts/+/1090430 - there’s another stale CR
[10:07:59] <dcausse>	 yes I think we forgot to the last steps :/
[10:09:28] <dcausse>	 or it's me not reading values properly, T398162#11064520 says that it's done
[10:09:29] <stashbot>	 T398162: Flink: Update k8s operator to 1.12.0 - https://phabricator.wikimedia.org/T398162
[10:10:08] <dcausse>	 lunch
[12:34:13] <inflatador>	 <o/
[12:48:17] <dcausse>	 o/
[13:37:27] <ebernhardson>	 \o
[13:40:56] <dcausse>	 o/
[13:58:05] <inflatador>	 .o/
[13:59:06] <brouberol>	 ~o~
[13:59:50] <brouberol>	 FYI, we noticed that the flink operator hadn't been upgraded to 1.12, so we're doing this atm w/ dcausse, cf https://phabricator.wikimedia.org/T398162
[16:06:12] <inflatador>	 quick break, back in ~20
[17:15:01] <inflatador>	 sorry, been back awhile
[18:28:45] <ebernhardson>	 Trey314159: numbers much ore sane,  12.8k with \r\n, 13.2k with \r, 539M with \n, 602M total (still no \v or \f)
[18:30:50] <Trey314159>	 ebernhardson: still ~70M with no line breaks! Still unexpected but less unbelievable.
[18:31:10] <Trey314159>	 0 \v and \f I totally believe
[19:02:30] <ebernhardson>	 Trey314159: turns out we can support \r \n and \t with about 5 lines of code. Are there others we should support too?
[19:04:37] <Trey314159>	 Good news! I couldn't think of any others, and I looked up a list and didn't see anything else worth including that we don't already cover.
[19:04:57] <ebernhardson>	 i suspect i could also implement unicode escapes in another 5 lines..
[19:07:11] <Trey314159>	 Ooo.. shiny! That would be especially useful for the very occasional character that the browser/UI won't let you input or which get normed away (even regexes get some normalization applied).
[19:09:15] <ebernhardson>	 almost embarasing how little it took to support, for how long we haven't supported it: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/CirrusSearch/+/1182630
[19:10:39] * ebernhardson should probably write a browser test or something that proves it works though :P
[19:13:15] <Trey314159>	 "works" is so subjective. I'm sure it'll do *something*
[19:13:40] <ebernhardson>	 hey i tested on my machine, it works here. It must work everywhere else too :P
[19:47:58] <ebernhardson>	 i suppose in a way this is odd that it has escape sequences, but doesn't require \\, because it's not real escape sequences
[20:22:31] <Trey314159>	 I saw "\\r" => "\r" and I thought it was too obvious and clear. Shouldn't it be "\\\\\\\\\\\\\\r" => "\\\\\\\\\r", or something? If I don't have to carefully count slashes is it even a real programming language?
[20:25:36] <ebernhardson>	 sigh, the more i think about edge cases the more it's odd :P   For example \u0023 (.) would be expanded and then interpreted as a match-all.  Could turn it into "." but that needs to know if it's in a char class or not
[20:28:23] <ebernhardson>	 or maybe we declare people that use \u expansion for regex syntax get what they deserve :P
[21:05:02] <inflatador>	 rebooting for security updates, hopefully back in ~30
[21:24:30] <inflatador>	 back