[12:36:08] * c bugs legoktm [19:04:55] Reedy: Want to re-run your find for T179624? [19:04:56] T179624: Cleanup remaining Job::insert() calls in extensions - https://phabricator.wikimedia.org/T179624 [19:44:06] DanielK_WMDE_: it would be nice to have a list of failed old RfCs so people do not propose the same thing again and again [19:44:16] so just removing the rfc tag is not ideal IMO [19:45:05] maybe replace it with an #archive-rfc tag? [19:49:35] tgr: most of the rfcs I'm looking at right now are simply outdated or orphaned. these are not reasons to not suggest these things again. [19:49:46] removing the rfc tag keeps the task findable [19:49:56] it'S still there, just not on the work board [19:50:30] when an rfc is actually closed as decline, it can also still be found since it's still tagged [19:50:40] but i doubt people search for closed rfcs [19:51:54] but i'll float the idea of a hidden archive column. could be an alternative to just dropping. [19:52:17] DanielK_WMDE_: I tend to, for example [19:52:29] (although mostly on-wiki and search actually works there :) [19:53:14] finding an ex-RfC in a project that has thousands of tasks is not really feasible, so I think there is value in marking up high-level things [19:53:46] maybe the #rfc tags can be used for that, it's not really clear what that's intended for currently [19:54:04] (plain rfc as opposed to techcom-rfc, I mean) [19:54:15] we use it for internal wikidata rfcs. things to discuss in the team [19:54:16] but a hidden column would be even nicer [19:54:20] other teams seem to do the same [19:54:36] but how do i move something to a hidden column? [19:54:42] i'm not sure that's possible [19:55:01] maybe it would need to bea sub-workboard?... [19:55:08] this is a bit arcane. [19:55:21] cogwheel icon, show hidden columns [19:55:59] oh, but that column can grow big. and once there are a few hundred things in a column, the workboard page breaks and stops responding... [19:56:05] or inside the task there is a move on workboard option in the action dropdown [19:56:31] no idea if that shows hidden columns though [19:56:39] i don't think it does [19:56:47] but i'll check [19:56:55] thx [19:58:52] Hmmm [19:59:06] Is there a was to tell if a DB object is a master or replica? I see nothing in IDatabase [20:03:57] DanielK_WMDE_: I'm thinking of killing some of revisionstore before writing new verisons of some of the stuff, basically all of the stuff that says @todo kill me soon [20:04:58] basically all of the loadRevisionFrom* methods in RevStore [20:06:17] addshore: there is a way to tell, but if you have a need to do that, something is probably very wrong :) [20:09:13] addshore: i'm not sure they can easily be removed - there are few callers, but the ability to pass in a concrete database connection may be essential, to provide a transaction context. Well... if they pass in a master connection. If they pass in a replica connection, it probably doesn't matter. But do think hard, this kind of thing tends to introduce nasty creeping inconsistencies, and is hard to find. get a +1 from aaron, if you can - [20:09:15] for from Tim or Roan. [21:24:25] hmmmmm [21:39:57] Yeah why do you need to know [23:44:44] * TimStarling is imagining a future, 50 years from now, where we're all senile but still work on the same MW code base [23:45:11] and we keep going round and round in code review, asking for a workaround for some bug, and nobody can convince us that it was fixed in 2005 [23:46:55] you know it's the early memories which are the last to go [23:47:48] If I'm here at 95...