[13:29:07] <ebernhardson1>	 \o
[13:35:56] <dcausse>	 o/
[14:02:45] <inflatador>	 <o/
[14:05:19] <dcausse>	 o/
[14:51:21] <pfischer>	 o/
[15:05:47] <ebernhardson1>	 dcausse: thanks for the reviews! certainly a few things i missed.  I dunno if you are interested, but there are a few patches i had already merged for other plugins (opensearch-ml-extra, opensearch-innerhits) and lucene libs (hebmorph).  Don't have to, but it seems i did miss a thing or two
[15:12:36] <dcausse>	 ebernhardson1: I think I had a quick look at opensearch-ml-extra & opensearch-innerhits, will check hebmorph I might have missed it since it's a dep of the plugin itself
[16:14:01] <dcausse>	 clearing popularity_score_weekly & transfer_to_es_weekly dags to unblock them 
[16:14:09] <ebernhardson1>	 thanks!
[16:51:44] <ebernhardson1>	 random thought, not sure i actually want to bother, but plausibly claude could swap the gradle builds to maven builds, so we could have consistency across all opensearch support repos.
[16:53:11] <dcausse>	 last time I checked (creating a pom for opensearch-ml-extra) I did not bother setting all the integ test suite and the like and so have no clue how painful that is vs existing gradle builds
[16:54:02] <dcausse>	 if it's not a mess with maven why not?
[16:56:29] <dcausse>	 but not sure that's worth the effort... perhaps having claude take a quick look at opensearch-innerhits to see how it looks and ponder?
[16:57:10] <dcausse>	 opensearch-innerhits is the one I found relatively painly to upgrade with all the dep versions that were not necessarily available as build variables
[16:57:22] <dcausse>	 s/painly/painful
[16:59:05] <ebernhardson1>	 hmm, yea hard to say
[16:59:46] <dcausse>	 or if gradle is the way to go we could try to template some of the gitlab ci bits you created for them?
[17:00:36] <ebernhardson1>	 hmm, yea we probably could.  I guess gradle is the upstream way, i don't know enough about gradle though, i suspect we would need a plugin or some such that recreates wmf-maven-parent-pom
[17:00:52] <ebernhardson1>	 hopefully easy-ish
[17:01:19] <ebernhardson1>	 but is it worth it? I'm thinking now because we keep upgrading these plugins and the variation is kinda annoying.  But at 3.8 we've caught up
[17:02:12] <ebernhardson1>	 but long term it might all be easier if it was the same
[17:03:06] <dcausse>	 if long term gradle is for some reasons the way to go we should perhaps not unify to maven yet :)
[17:03:42] <ebernhardson1>	 indeed!
[17:04:28] <dcausse>	 personally I used maven for opensearch-ml-extra because I was lazy, trying to remember how things were when actively worked on the ltr plugin I barely remember that gradle was a bit less painful (esp. regarding the integ tests)
[17:16:37] <dcausse>	 dinner