| 2 Apr 2026 |
 raitobezarius | (if you want to open an issue with details, that'd be welcome) | 16:31:03 |
 tebriel | I have the nix log but haven’t found the subtest details. I’ll do so thanks. | 16:31:15 |
| tebriel | #1173 still trying to wrangle the nix log to see the subtest failures if you have any advice there I’d love it. | 16:37:32 |
| raitobezarius | what if you try to look at all processes running | 16:47:12 |
| raitobezarius | normally you will see pytest process names | 16:47:18 |
| raitobezarius | perhaps you will see the "stuck" process name in the process name | 16:47:24 |
| tebriel | Also interesting the output seems to be misaligned with things like 7709stests passed and 248.19s exit status 1ssed | 17:02:59 |
| tebriel | Never say a pytest proc that had stuck in the name, will keep searching | 17:13:18 |
| tebriel | * Never saw a pytest proc that had stuck in the name, will keep searching | 17:13:35 |
| raitobezarius | darn | 17:21:06 |
| raitobezarius | I'm asking the F2 experts on this | 17:21:09 |
 Rutile (rootile) | Cab you build with --keep-failed and send the meson testlog.txt? | 17:26:02 |
| Rutile (rootile) | Also, we should re-open the discussion about tap, as this hasnt been the first time people need to dig deep to figure out what failed :/ | 17:26:35 |
 Yureka (she/her) | which package are you building specifically? there was a recent issue with the nixpkgs packaging on mac os | 17:29:24 |
 nightuser | I'm getting errors like error: NAR info file '4dh8j6q43apjc62yjp1fwxh55l2qrzhx.narinfo' is corrupt: expecting '\n' at line 2. The issue disappears when I specify --option substituters https://cache.nixos.org, and indeed curl https://cache.nix.systems/4dh8j6q43apjc62yjp1fwxh55l2qrzhx.narinfo gives the correct output while curl https://cache.lix.systems/0z54cs3y9xif48f4zvgnfgsnzva0fv8y.narinfo outputs some binary data (but still with status code 200). | 17:40:25 |
| raitobezarius | Yureka (she/her) ^ | 17:41:12 |
| raitobezarius | it's indeed a corrupted narinfo | 17:41:39 |
| raitobezarius | hmm it seems it was corrupted on the original garage as well anyway | 17:42:01 |
| raitobezarius | i think we should nuke it | 17:42:29 |
| nightuser | The same issue happens with some other packages like bash. | 17:43:34 |
| nightuser | * The same issue happens with some other packages, including bash, for example. | 17:44:05 |
| raitobezarius | thank you for the report | 17:44:10 |
| raitobezarius | we should probably review all the narinfo in the cache then | 17:44:17 |
| Yureka (she/her) | this is returning zstd compressed data | 17:57:54 |
| Yureka (she/her) | the decompressed content is valid | 17:58:25 |
| raitobezarius | shouldn't decompression happen transparently with that curl invocation? | 17:59:03 |
| Yureka (she/her) | only if the server returns the appropriate header | 17:59:44 |
| raitobezarius | response contains content-encoding: zstd | 18:00:49 |
| raitobezarius | am I missing something | 18:00:58 |
| Yureka (she/her) | is it maybe double encoded? | 18:01:07 |
