[07:11:54] o/ [08:12:35] o/ [08:19:00] dcausse I'm catching up with what I missed in the last two weeks. Can I maybe ping you later this morning for some pointers on getting started with search data & modelling? [08:20:08] gmodena: sure! feel free to schedule meetings and/or ask anything here :) [08:22:29] thx :) [08:44:33] o/ [09:19:01] gmodena: if you haven't caught up yet, we have a new year :b [10:31:13] hashar ah right! happy 45^2 ! [10:34:44] gmodena: how are you getting up to speed? Do you need a chat? Should I join your meeting with David? [10:34:56] and happy new year to you all! [10:37:50] happy new year indeed :) [10:37:56] errand+lunch [10:41:28] gehel always happy to chat. I'll extend the invite for a call I have with dcausse later (hope the time works). [10:41:41] the time works for me [10:41:59] ack [10:42:46] today I'm mostly in reading mode - notebooks and wiki. My goal would be get hands on with the dataset and start reproducing some of the existing analysis. [11:24:10] errand+lunch [13:01:24] meet is behaving... [13:40:17] gmodena: updated some broken links on https://wikitech.wikimedia.org/wiki/Search/MLR_Pipeline [13:40:25] dcausse: thanks! [13:40:34] dcausse thank you! [15:32:58] There was a nice comment on my last blog post back in December: https://techblog.wikimedia.org/2024/07/18/how-many-languages-does-wikimedia-search-support/#comment-44843 [15:33:03] The best bit: "Wikimedia’s efforts truly reflect a commitment to inclusivity and accessibility for global users. Bravo to the search platform team!" [15:43:31] Hey hey, in Growth team, we noticed that in our performance metrics, the time to render the suggested edits module on the server got _a lot_ faster after the last deploy year: https://grafana.wikimedia.org/d/vGq7hbnMz/special-homepage-and-suggested-edits?orgId=1&from=now-30d&to=now&viewPanel=10 [15:43:39] I'm trying to understand why. [15:44:32] I have a few candidates for that on our end, but I'm wondering if any changes from your side got deployed then that might also be responsible for that improvement? [16:01:39] Trey314159: nice! [16:59:17] Trey314159: my brother (a linguist) loved your blog post and is attempting to nerdsnipe you with https://scholarspace.manoa.hawaii.edu/server/api/core/bitstreams/6f608290-7a4b-4925-a3fd-2cb5550a3fef/content this linguistics paper about the difficulty of defining what a language is [17:03:25] MichaelG_WMF: not that I'm aware of, if you mean the improvement I see from your graph on dec 20 I don't see a similar improvement on search latencies so I'd guess it's not related to something happening in search [17:04:22] dcausse: Good to know. Thanks for checking on your end! 🙏 [17:04:24] cdanis: nice! That's an interesting looking paper. I read the abstract and conclusion and I see it includes the semi-weasely "linguistic variety", which is also what I use when I want to avoid the entire question [17:05:47] 😂 [18:20:18] dinner