[10:40:34] errand+lunch [13:00:40] \o [13:05:41] o/ [14:15:26] dcausse well, crap. I completely lost track of time again. Do you have anything for pairing? [14:15:39] inflatador: nothing, np! [14:15:55] cool, sorry about missing it again. I'll try to be on-time next week ;( [14:44:58] For the Semantic Search meeting later today, are there any specific phab/asana/google docs I should look at? I'm going down the workboard now, just wanna make sure I don't miss anything [14:59:13] not that i know of [15:06:29] nothing yet, part of the goal is to create those [15:49:42] We're talking about the failure modes of Semantic Search (from a product perspective - so more search provide unexpected, unwanted, offensive results, than scaling issues). Who would like to be part of this conversation? [15:50:06] I think Search Platform has quite some experience here, so we should share it. [15:50:45] Ah! There go all my questions ;P [17:18:22] lunch [17:18:27] dinner I mean [17:39:11] realized why the graph of the 10 user agents looks like it's only 1.6M requests per day, they are so correlated that all 10 lines are drawn on top of each other [17:39:16] so it's really the 16M [21:45:57] ebernhardson thanks for digging thru T438238 , do you have the UAs in superset or somewhere? I was gonna take a stab at a requestctl rule [21:45:58] T438238: Cirrussearch: investigate unexpected rise in fulltext web searches - https://phabricator.wikimedia.org/T438238