| 8 Dec 2025 |
|  Vodros joined the room. | 14:36:16 |
| Vodros | Hi, since support for download progress was merged the day after the last release, are there any plans to make a new release soon? | 14:37:48 |
 maralorn | I will have to look into that. | 14:43:59 |
| Vodros | Awesome thanks! | 14:45:36 |
| maralorn | Vodros: The problem is that we currently have kind of an experimental build finished detection on main and I don’t feel comfortable releasing that. | 14:47:16 |
| maralorn | So potentially I have revert that change before the release. | 14:47:33 |
| maralorn | * So potentially I have to revert that change before I can cut a release. | 14:47:44 |
| maralorn | Other problem is that I would love to polish the download progress stuff even more before release. But I probably won’t get to that in a while. | 14:48:39 |
| maralorn | So I probably should just drop that plan. | 14:49:23 |
| maralorn | I think a release until end of year sounds realistic. | 14:50:44 |
| Vodros | I understand. No pressure then. If reverting is manageable then that would be cool, otherwise we'll just wait ig. | 14:51:23 |
| Vodros | Yeah that's reasonable | 14:51:32 |
| Vodros | And above all thanks for your work :) | 14:52:05 |
 hexa | it doesn't look comfortable either 😄 | 14:54:37 |
| hexa | nix-output-monitor 2.1.8
nix (Lix, like Nix) 2.95.0-pre20251201-dev_2ae7480
| 14:54:52 |
| hexa | 
Download image.png | 14:55:05 |
| maralorn | Yeah, I know. 😄 | 14:55:56 |
| maralorn | Although I have to say that this whole misery started with a behavior change in lix. | 14:56:20 |
| hexa | right, but at least lix people are here who offered better signaling, no? | 14:57:12 |
| maralorn | Yeah, sure. | 15:06:53 |
| maralorn | The problem is I kinda feel that this should be fixed on the nom side. | 15:07:17 |
| maralorn | I am just trying to reflecting blame for the problem not yet being fixed.^^ | 15:07:49 |
| hexa | if the signaling isn't clear and needs to be guessed that is bad | 15:15:24 |
| maralorn | Sure. But that is a structural problem which won’t go away for nom unless all nix implementations fix it. | 15:22:06 |
| maralorn | Yes, I guess for this case specifically we could really do the capabilities approach where lix signals correctly why a job finishes. | 15:23:36 |
| maralorn | Then I keep the old detection for everything else and do the better detection for lix. | 15:23:58 |
| maralorn | Otoh the fix could be easy when I look at the logs by Emma. But I am actively trying not do that because I don’t have time for this. 😄 | 15:25:00 |
| hexa | one implementation adding better signaling can force another to follow suit 😄 | 15:25:02 |
| maralorn | The last week I developed on nom was super fun, I am really enjoying the result, but the ensuing sleep deprecation triggered a medium mental health crisis. i.e. nom development is a health hazard for me. 😄 | 15:26:16 |
| maralorn | * The last week I developed on nom was super fun, I am really enjoying the result, but the ensuing sleep deprivation triggered a medium mental health crisis. i.e. nom development is a health hazard for me. 😄 | 15:26:26 |
