| 28 Feb 2024 |
 @domenkozar:matrix.org | actually it's already paying for a community darwin machine | 14:35:43 |
| @domenkozar:matrix.org | if someone can coordinate with infra team, then we just submit an expense | 14:36:01 |
 hexa | I think that one is being well received | 14:36:04 |
 @delroth:delroth.net | I mean, more capacity will help, but I can't help but think that there's a bug somewhere and it's not just a capacity issue | 14:39:50 |
| @delroth:delroth.net | The Hydra macs run through a ton of builds every day | 14:40:07 |
| @delroth:delroth.net | Probably at least an order of magnitude more than what ofborg should be using imo | 14:40:32 |
 @adam:robins.wtf | do we know the specs of the existing machines? | 14:41:01 |
| hexa | the hydra macs don't have low timeouts | 14:41:13 |
| hexa | they don't have to waste time retrying llvm all day | 14:41:22 |
| @adam:robins.wtf | do the hydra macs have to evaluate everything to know whether it needs to be built on darwin? | 14:41:59 |
| @delroth:delroth.net | Can we just stop running ofborg for staging, to start with? :) | 14:42:10 |
| hexa | no, they're pure builders | 14:42:14 |
| @delroth:delroth.net | (do we already?) | 14:42:16 |
| hexa | hydra itself does evaluations | 14:42:18 |
| @adam:robins.wtf | that's what i figured | 14:42:23 |
 Alyssa Ross | OfBorg on staging is very useful | 14:42:33 |
| hexa | but ofborg macs also don't do evals 😛 | 14:42:37 |
| Alyssa Ross | Except for Darwin | 14:42:38 |
| @delroth:delroth.net | Can we make it lower priority? | 14:42:55 |
| Alyssa Ross | If we had no CI for staging staging-next cycles would take forever | 14:43:41 |
| @adam:robins.wtf | In reply to @hexa:lossy.network
but ofborg macs also don't do evals 😛 then what are they doing that makes them wait even on packages that don't support darwin? | 14:43:42 |
