| 28 Oct 2024 |
 adamcstephens | In what capacity? | 22:54:09 |
| 29 Oct 2024 |
 @rick:matrix.ciphernetics.nl | Possibly for ofborg and no for hydra, I think | 21:45:22 |
| 30 Oct 2024 |
 @philip4g:matrix.org | In reply to @adam:robins.wtf
In what capacity? I heard that there is a shortage of Mac hardware in OfBorg, causing auto builds to take a very long time. I have some hardware that OfBorg could use remotely, if such a decentralized solution is feasible. | 18:38:47 |
| adamcstephens | Is it Apple Silicon? Would you be willing to give over administrative control? (not sure if this is a requirement or not) | 18:50:04 |
| 2 Nov 2024 |
 aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Ofborg eval seems to be blocked again? (How do we check queue status?) | 15:56:05 |
 hexa | url in th e topic | 16:01:55 |
| hexa | * url in the topic | 16:01:58 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | In reply to @hexa:lossy.network
url in th e topic The evaluation queue, not build queue | 16:45:58 |
| hexa | the upper graph | 16:46:14 |
| hexa |
ofborg_queue_evaluator_waiting
| 16:46:21 |
| aleksana 🏳️⚧️ (force me to bed after 18:00 UTC) | Oh I see | 16:50:58 |
| adamcstephens | 
Download image.png | 17:57:53 |
| adamcstephens | this doesn't seem sustainable | 17:58:01 |
 Artturin | We need nuclear ofborg | 17:58:23 |
| adamcstephens | the many hours wait just to get the ofborg eval is starting to become annoying | 21:30:05 |
 Tristan Ross | I've just been running nixpkgs-review on my system but yeah, waiting for Ofborg isn't great. | 21:36:31 |
| 3 Nov 2024 |
 Bot_wxt1221 | Could we move some aarch64-linux builder as evaluator | 09:20:04 |
| hexa | no, we can't magic resources into existance | 15:04:32 |
| Tristan Ross | In reply to @bot-wxt1221:matrix.org
Could we move some aarch64-linux builder as evaluator I think this has already been talked about. | 15:20:34 |
| Tristan Ross | Looking at the Prometheus graph, we really are lacking that bad with aarch64-darwin? | 15:39:47 |
| adamcstephens | Have been for a long time | 17:23:27 |
|  Matteo Pacini changed their profile picture. | 21:50:27 |
| Bot_wxt1221 | In reply to @rosscomputerguy:matrix.org
I think this has already been talked about. So what, we have more aarch64-linux builder that we need. There's no tasks wating for aarch64-linux. Anyway, evaluator is more meaningful than builder. We can run nixpkgs-review locally but without ofborg. It's really slow. It takes about 1 hour on my computer. | 23:48:33 |
| Tristan Ross | In reply to @bot-wxt1221:matrix.org
So what, we have more aarch64-linux builder that we need. There's no tasks wating for aarch64-linux. Anyway, evaluator is more meaningful than builder. We can run nixpkgs-review locally but without ofborg. It's really slow. It takes about 1 hour on my computer. Idk, using aarch64-linux to eval for aarch64-darwin doesn't feel right lol. Ideally it should evaluate the same but what if it doesn't? | 23:52:29 |
| Tristan Ross | The better solution would probably be to get rid of some aarch64-linux machines replace them with Darwin ones. | 23:53:48 |
| Bot_wxt1221 | In reply to @rosscomputerguy:matrix.org
Idk, using aarch64-linux to eval for aarch64-darwin doesn't feel right lol. Ideally it should evaluate the same but what if it doesn't? It is better to have it than there's no. It should be a bug in nix if we have different result if we evaluate on x86 and darwin. Anyway, if you are really worried about this. We can just run them in qemu. | 23:55:59 |
| Bot_wxt1221 | In reply to @rosscomputerguy:matrix.org
The better solution would probably be to get rid of some aarch64-linux machines replace them with Darwin ones. I think the most serious problem is about evaluator | 23:56:53 |
| Bot_wxt1221 | In reply to @rosscomputerguy:matrix.org
Idk, using aarch64-linux to eval for aarch64-darwin doesn't feel right lol. Ideally it should evaluate the same but what if it doesn't? * | 23:57:16 |
| Tristan Ross | In reply to @bot-wxt1221:matrix.org
It is better to have it than there's no. It should be a bug in nix if we have different result if we evaluate on x86 and darwin. Anyway, if you are really worried about this. We can just run them in qemu. You can't emulate Darwin without breaking the terms | 23:57:31 |
| Tristan Ross | So that's a no | 23:57:38 |
