Sender | Message | Time |
---|---|---|
30 May 2025 | ||
16:18:20 | ||
Hi all, I've got a flake that builds fine on all but one system. On that system I get a hash mismatch in the cache.
How to I debug this? I've updated the flake lock, I've tried unstable, 24.11 and 25.05 as the input source and nix store repair and more. I'm a little list. | 16:19:30 | |
* Hi all, I've got a flake that builds fine on all but one system. On that system I get a hash mismatch in the cache.
How to I debug this? I've updated the flake lock, I've tried unstable, 24.11 and 25.05 as the input source and nix store repair and more. I'm a little lost. Flake can be found here: https://gist.github.com/ordnungswidrig/55120535ab4a1e99cccdc76decac10c5 | 16:19:46 | |
Thanks dramforever sorry for the slow reply. Just running the command now | 17:19:17 | |
Download stacktrace.txt | 17:20:31 | |
I am looking through it but don't see anything interesting so far | 17:21:56 | |
In reply to @ordnungswidrig-test:matrix.orgtry nix-hash --type sha256 /nix/store/9545lc12nznckmbrsgsjy5vhvgrj5i4n-getty on your various systems and see what they have? | 17:22:10 | |
From my understanding of it it is basically just stating that it is evaluating the openjdk-17 package and that it doesn't support i686 | 17:23:24 | |
but the problem is I don't know why it thinks it needs to be. I need to track down what is specifying we need i686. | 17:23:52 | |
18:11:33 | ||
31 May 2025 | ||
okay. granted, it's not the most easy to read, but what you can do is pare it down a bit. there are lines inserted in the trace like "… while evaluating". if we grep while it looks much more managable | 03:14:09 | |
we can get it down even further while evaluating derivation | 03:14:25 | |
* we can get it down even further with while evaluating derivation | 03:14:30 | |
this paints a pretty clear picture of what's going on:
| 03:14:56 | |
heroic depends on 32-bit versions of various stuff, which eventually goes to ffmpeg, which depends on libbluray | 03:15:25 | |
but an overlay of yours made libbluray depend on java | 03:15:40 | |
i don't know enough about your use case but what i would do is to override whatever you actually use for bluray to use the overridden libbluray | 03:17:13 | |
In reply to @dramforever:matrix.orgI will temporarily disable that and see if it helps. The reason for the Java is the Blu-ray menus | 06:09:35 | |
I disabled the overlay for the libbluray and this is the new stack trace | 06:40:04 | |
Download stacktrace.txt | 06:40:11 | |
seems to be still getting hung up on libbluray | 06:40:55 | |
20:52:27 | ||
21:29:38 | ||
1 Jun 2025 | ||
07:52:23 | ||
07:52:29 | ||
23:33:09 | ||
23:38:14 | ||
3 Jun 2025 | ||
03:42:48 | ||
09:25:48 | ||
What exactly will be updated when running the auto update and using flakes? | 09:27:43 |