| 24 Aug 2021 |
 Vladimír Čunát | I see nothing wrong about it, but it's not like I know this stuff. | 20:55:47 |
| 25 Aug 2021 |
|  trofi joined the room. | 21:59:53 |
| trofi set a profile picture. | 22:19:15 |
| 26 Aug 2021 |
| trofi changed their display name from Sergei Trofimovich to trofi. | 07:41:02 |
| 28 Aug 2021 |
 hexa | the openssl bump was reverted on master | 13:25:31 |
| hexa | and the revert was reverted on staging-next | 13:25:39 |
| hexa | while staging builds had been canceled | 13:26:01 |
| hexa | verpbl wanted to wait until last nights eval of trunk has completed before reverting on master | 13:27:25 |
| hexa | which was cancelled 🎉 | 15:01:27 |
| Vladimír Čunát | ...and restarted by me now :-) Coordination can be hard. | 15:51:47 |
| hexa | so very true 😀 | 16:07:52 |
| 30 Aug 2021 |
 andi- | Is it possible to keep the scope of the next staging-next run smaller and include https://github.com/NixOS/nixpkgs/pull/131618 instead? It is already a comparatively big change (on the risk side) and I'd like to avoid having to debug binutils, glibc, ... issues if things go wrong while already debugging systemd. | 13:15:23 |
| hexa | I don't mind, then we skip staging->staging-next once and you'd retarget on staging-next with base branch master | 13:17:10 |
| hexa | * I don't mind, then we skip staging->staging-next once and you'd retarget on staging-next with base branch master. Also running v249 locally for 48h and didn't notice any obvious breakage 👍️ | 13:18:02 |
| andi- | I'm currently testing the PR rebased on master (with a hydra jobset) so ideally the amount of (additional) rebuilds will be very small anyway. | 13:22:13 |
| Vladimír Čunát | Maybe the systemd jobset should not have more shares than "trunk-combined"? | 14:44:11 |
| Vladimír Čunát | * Maybe the systemd jobset should not have more build shares than "trunk-combined"? | 14:44:19 |
| Vladimír Čunát | * Maybe the systemd jobset should not have (many times) more build shares than "trunk-combined"? | 14:44:26 |
| Vladimír Čunát | I lowered them now. nixos-unstable is still waiting for the first bump with new openssl. | 15:09:36 |
| Vladimír Čunát | (and I cancelled all but the latest builds in the systemd jobset) | 15:29:33 |
| hexa | In reply to @vcunat:matrix.org
Maybe the systemd jobset should not have (many times) more build shares than "trunk-combined"? which is kinda funny, given the shares of the haskell-updaes job | 19:16:09 |
| hexa | In reply to @vcunat:matrix.org
Maybe the systemd jobset should not have (many times) more build shares than "trunk-combined"? * which is kinda funny, given the amount of shares allocated to the haskell-updates job | 19:16:20 |
| Vladimír Čunát | My understanding (from many years past) is that it was intentional there. | 19:16:59 |
| Vladimír Čunát | It was meant that the haskell updates get iterated extremely quickly. | 19:17:18 |
| hexa | yup, they eval 3-4 times a day currently. https://hydra.nixos.org/jobset/nixpkgs/haskell-updates#tabs-evaluations | 19:17:40 |
| Vladimír Čunát | (i.e. the jobset just gets checked and merged within a couple dayes) | 19:17:43 |
| Vladimír Čunát | * (i.e. the jobset just gets checked and merged within a couple days) | 19:17:50 |
| Vladimír Čunát | Ah, well, I didn't mean this iteration speed :-) Maybe it now consumes quite a fraction of Hydra's resources. | 19:18:21 |
| hexa | pretty sure the cycle is bi-weekly now | 19:18:27 |
 sterni | it's at least bi-weekly, it really depends on how much regressions there are to fix | 19:18:46 |
