| 7 Feb 2022 |
 @grahamc:nixos.org | I have certain feelings of if you don't report obvious bugs and then I fix obvious bugs then it isn't on me, is that unfair? | 18:28:13 |
 das_j | No, it's totally fair | 18:29:06 |
| das_j | And as Hydra is unreleased software, it would be my fault to depend on its current behaviour | 18:29:27 |
| @grahamc:nixos.org | I do hope you see real effort from me to not break its behavior | 18:30:12 |
| das_j | In reply to @grahamc:nixos.org
I do hope you see real effort from me to not break its behavior Yeah I do. I haven't noticed any breakages you introduced that were not necessary to improve Hydra in the long run | 18:30:46 |
| @grahamc:nixos.org | 🙏 | 18:31:03 |
| das_j | (also I think you broke our Hydra only 1 or 2 times in ways that were pretty trivial to fix) | 18:31:28 |
| @grahamc:nixos.org | Bug breaks or intentional breaks? (Thinking about missed column changes or function signature mistakes) | 18:32:11 |
| das_j | In reply to @grahamc:nixos.org
Bug breaks or intentional breaks? (Thinking about missed column changes or function signature mistakes) Intentional breaks. Yeah there was also the function signature mistake but that was easily reverted locally for the time being | 18:32:40 |
| @grahamc:nixos.org | Gotcha | 18:32:50 |
| das_j | I do however think it would be useful to document intentional breakages somewhere (unless it introduces too much work for the developer) | 18:33:26 |
| das_j | something like a changelog but without releases. Just 2021-02-07 fefefe emitted eval_* events are now tab-separated instead of \t separated | 18:34:08 |
| das_j | * something like a changelog but without releases. Just 2021-02-07 fefefe Emitted eval_* events are now tab-separated instead of \t separated | 18:34:32 |
| @grahamc:nixos.org | That does sound good. Especially, for example, the Postgres version change | 18:35:17 |
| @grahamc:nixos.org | Let’s do that and add it to the pr template | 18:35:37 |
| das_j | Yeah, I could then use either the date or the commit hash when updating my Hydra after a long time and see what I can expect to be broken | 18:35:43 |
| @grahamc:nixos.org | Yeah, seems great. | 18:37:02 |
| @grahamc:nixos.org | When I broke your hydra on purpose did you adapt or partially revert, or? | 18:37:25 |
| das_j | well adapt. there's no point in carrying patches for years and having to rebase them over time | 18:43:24 |
| das_j | also your changes did usually improve things so staying with the old working state was not ideal either | 18:43:42 |
| @grahamc:nixos.org | I just wanted to make sure it wasn’t a nuisance you were actively choosing to avoid. That wouldn’t be very nice! | 18:44:22 |
| das_j | nono, the only time we reverted your changes was when the signature stuff broke our hydra | 18:44:59 |
| das_j | I swear I never disliked your changes :D | 18:45:09 |
| @grahamc:nixos.org | whew, thanks :) | 18:48:08 |
| das_j | Btw, is both dynamic runcommand and runcommand logs on master now? | 18:48:31 |
| das_j | We could upgrade then and drop the patches (and my insecure implementation đź‘€ | 18:48:47 |
| das_j | * We could upgrade then and drop the patches (and my insecure implementation đź‘€) | 18:48:48 |
 cole-h | Dynamic RunCommand, not yet. But the logs are on (Hydra) master. | 18:53:03 |
| cole-h | (Dynamic RunCommand PR: https://github.com/NixOS/hydra/pull/1103) | 18:53:16 |
| @grahamc:nixos.org | I'm a bit shy on merging the dynamic runcommand PR given, well, its behavior :P | 18:53:34 |
