[02:05] Yippie, build fixed! [02:05] Project windmill build #44: FIXED in 52 min: https://hudson.wedontsleep.org/job/windmill/44/ [02:12] Erm. [02:13] StevenK: It failed to detect the testrunner breakage :( [02:55] Whee, subunit bug [04:23] woo [04:23] * 367 Time Outs [04:24] Er? [04:25] Oh, I read the wrong one this morning :( [04:25] I saw 1225 and was pretty disappointed. [04:26] We had less than 2000 hard OOPSes/timeouts. Not bad. [04:27] And about 400 exceptions are fixed by stuff in the deploy queue. [04:27] we're getting there [05:06] lifeless: Are you happy with the teams progress? [05:49] StevenK: I think we're making fantastic progress on maintenance [05:50] StevenK: recipes are looking awesome, but I think the transition off them has been a little fuzzy [05:50] StevenK: message shared/bug subscriptions seem to be coming along well [05:50] StevenK: and derived distros has only had one week now, so no opinion [06:23] Project windmill build #45: FAILURE in 1 hr 10 min: https://hudson.wedontsleep.org/job/windmill/45/ [06:43] lifeless: the transition off recipes has been fuzzy? What do you mean? [06:44] * StevenK wonders if he can fetch an SPR by ID using the API [06:45] StevenK: SPRs aren't exposed, so no. [06:45] :-( [06:45] You're wanting to check the changelogs? [06:45] Just one or two, yes [06:46] I ran it for a while on DF and it seemed sane there. [06:46] so I've already qa-ok'd it. [06:48] So on qastaging it looks to be doing around 500 in 15 minutes. [06:49] And we have what, 600,000 that are unset? [06:49] I don't recall the number, but that seems plausible. [06:49] Which means 1200 runs to finish [06:50] On overloaded asuka. [06:50] Yes, we should silence garbo-hourly on Tuesday [06:50] I was thinking when it starts running on production. [06:51] 1200/23 is 52, so, almost 2 months? [06:51] Am I doing that right? [06:51] Yes. [06:51] But !asuka should be faster. [06:51] At least a bit. [06:52] But we're still only running at 25% of what I had initially planned, so it's going to take a while. [06:52] I don't restrict the archive? [06:52] It is only running for 15 minutes out of every 60. [06:52] Ah well. [06:53] And garbo-hourly loses an hour to garbo-nightly doesn't? [06:53] Er, doesn't it [06:53] I'm not sure. [06:53] LFA is exposed, isn't it? [06:53] No. [06:53] Some files are exposed, but not LFA itself. [06:54] Why? [06:54] Can I trivially go from LFA ID to URL? [06:54] If you know the filename. [06:54] They're all "changelog" [06:54] http://staging.launchpadlibrarian.net/$ID/changelog [06:55] Er, qastaging, but yeah. [06:55] lifeless: When you care, "SELECT id, changelog from SourcePackageRelease WHERE id IN (63623, 63624, 63625);" on qastaging [06:56] StevenK: You could also guess. [06:56] Add an attachment to get an idea of where the IDs are, then search around there. [06:56] That doesn't sound nearly as much fun as bugging lifeless. [06:57] It is more fun when it leads to security holes :) [06:57] * StevenK ponders when to start on the Poulet Sauté Au Vinaigre [06:58] It sounds better in French [06:58] "Chicken in Vinegar" doesn't sound very appealing. [06:59] wgrant: Current plan is to average six runs on production and work out length from there. [07:01] StevenK: You might also consider a graph. [07:01] I believe they're pretty easy to add. [07:01] StevenK: id | changelog [07:01] -------+----------- [07:01] 63623 | 64295224 [07:01] 63624 | 64295225 [07:01] 63625 | 64295226 [07:03] StevenK: lfa isn't exposed because lfa is just a join [07:03] files don't have standalone existence, you need a containing object [07:03] like spr, bug, etc [07:03] it would be like exposing message [07:05] lifeless: Uh... [07:05] https://api.launchpad.net/+apidoc/devel.html#message [07:05] wgrant: bug [07:08] Right, those 3 look good [07:09] lifeless: And thank you [07:09] de nada [07:13] StevenK: do my answers on team progress vs happiness make sense? [07:15] [17:43] < StevenK> lifeless: the transition off recipes has been fuzzy? What do you mean? [07:29] its dragging on [07:29] at bit [07:29] s/at/a/ [07:38] lifeless: I think that might be a little unfair, since it is the first transition since the move to squads. [07:39] I think it's inevitable. [07:39] For a feature like that. [07:39] I think transitions need to handled with an iron fist -- finish off anything you have in progress, then pick up criticals [07:40] That requires completion to be declared at a reasonable point. [07:40] And that's difficult. [07:40] I think recipes can be considered delivered, but not done. [07:54] StevenK: I didn't say anyone had done anything wrong [07:54] StevenK: I think its reasonable to assess how we do dispassionately- haven't dug into causes yet [10:09] as far as I'm concerned, recipes aren't delivered as long as they have the beta flag up [10:13] Well, that's a matter of LOSA. [10:23] jml: Isn't that waiting for you to review it again? [10:37] Project windmill build #46: STILL FAILING in 1 hr 9 min: https://hudson.wedontsleep.org/job/windmill/46/ [19:07] so in Launchpad API's bug representation "owner_link" is used for "Reported by" [19:08] or is it represented by some other field. Cant find it === almaisan-away is now known as al-maisan [19:25] owner_link, yes === al-maisan is now known as almaisan-away