| 3 Aug 2025 |
raitobezarius | so plausibly, this could be the job you described | 16:42:31 |
raitobezarius | it will still generate more work for x86 which will have to calculate one more job at the nej phase | 16:42:43 |
raitobezarius | but not a compilation job at least | 16:42:48 |
K900 | But what if I add it to the goal loop | 16:44:29 |
aloisw | In reply to @raitobezarius:matrix.org it will still generate more work for x86 which will have to calculate one more job at the nej phase Evaling one Lix should be quite cheap though? | 16:48:19 |
raitobezarius | In reply to @aloisw:julia0815.de Evaling one Lix should be quite cheap though? Yeah, I didn't mean this as this would be much more expensive | 16:48:44 |
raitobezarius | but cost for x86(after) > cost for x86(before) | 16:48:54 |
raitobezarius | maybe by epsilon | 16:48:56 |
K900 | I am fully ready to commit crimes to make this happen and Lix won't even be the worst of them | 16:49:53 |
K900 | I am making a thing to forward logind wakelocks from WSL to Windows | 16:50:06 |
K900 | It is going to be VERY CURSED and I don't care | 16:50:20 |
WeetHet | raitobezarius: how do i read this? https://buildkite.com/lix-project/lix/builds/3673/steps/canvas?sid=01986f84-2c33-4e02-9c8e-6ad9e83250c9 | 18:34:58 |
Rutile (Commentator2.0) feel free to ping | ma27: currenlty rebasing the nix framework rewrite and now your test is failing... there is a chance the code is broken and not the framework but not 100% sure, could you take a look at it? https://gerrit.lix.systems/c/lix/+/3333/ | 18:50:36 |
ma27 | can you link me a ci failure, please? | 19:05:00 |
Rutile (Commentator2.0) feel free to ping | ci currently has issues with some of the changes
assert nar_entries[0]["present"]
assert nar_entries[0]["hashPart"] == hash_part
assert nar_entries[0]["namePart"] == "test-file"
> assert nar_entries[0]["url"] == store.uploaded_nars[hash_part]["URL"]
E AssertionError: assert 'nar/14r07wnk...vxwm08idf.nar' == 'nar/snix-cas...?narsize=f128'
E
E - nar/snix-castore/00000000000000000000000000000000000000000000000000000?narsize=f128
E + nar/14r07wnk5h2bl40v1nknwkwrzh5li7z4bl02vcsqh52vxwm08idf.nar
the relevant error log(?)
| 19:15:21 |
Rutile (Commentator2.0) feel free to ping | ok its starting to build now | 19:16:50 |
Rutile (Commentator2.0) feel free to ping | uhhhh i guess it works on the build machine?! | 19:28:02 |
ma27 | I repeatedly get now
> raise child_exception_type(errno_num, err_msg, err_filename)
E FileNotFoundError: [Errno 2] No such file or directory: 'chmod'
from
def nix(tmp_path: Path, env: ManagedEnv) -> Generator[Nix, Any, None]:
"""
Provides a rich way of calling `nix`.
For pre-applied commands use `nix.nix_instantiate`, `nix.nix_build` etc.
After configuring the command, use `.run()` to run it
"""
yield Nix(env)
# when things are done using the nix store, the permissions for the store are read only
# after the test was executed, we set the permissions to rwx (write being the important part)
# for pytest to be able to delete the files during cleanup
cmd = Command(argv=["chmod", "-R", "+w", str(tmp_path.absolute())], _env=env)
> cmd.run().ok()
(yes, there's a chmod in the shell available and I killed the entire temp dir before)
so something looks pretty off here | 20:04:10 |
Rutile (Commentator2.0) feel free to ping | are you using darwin? otherwise do a just clean & just setup & just install | 20:04:42 |
ma27 | ah and I misread the buildlog from the previous nix build of my checkout, it's not reproducible from within. | 20:05:08 |
ma27 | In reply to @commentator2.0:elia.garden are you using darwin? otherwise do a just clean & just setup & just install nope, linux. let me try... | 20:05:57 |
ma27 | same result for me. | 20:14:10 |
Rutile (Commentator2.0) feel free to ping | can you check if it is a code/implementation issue and one from the functional2 framework? | 20:16:11 |
Rutile (Commentator2.0) feel free to ping | * can you check if it is a code/implementation issue and not one from the functional2 framework? | 20:16:19 |
ma27 | so, the quickfix to make the test green again:
diff --git a/tests/functional2/testlib/fixtures/nix.py b/tests/functional2/testlib/fixtures/nix.py
index fa459c41a..9edb9a60b 100644
--- a/tests/functional2/testlib/fixtures/nix.py
+++ b/tests/functional2/testlib/fixtures/nix.py
@@ -132,5 +132,7 @@ def nix(tmp_path: Path, env: ManagedEnv) -> Generator[Nix, Any, None]:
# when things are done using the nix store, the permissions for the store are read only
# after the test was executed, we set the permissions to rwx (write being the important part)
# for pytest to be able to delete the files during cleanup
+ import os
+ env.path.prepend(os.getenv("PATH"))
cmd = Command(argv=["chmod", "-R", "+w", str(tmp_path.absolute())], _env=env)
cmd.run().ok()
this is obviously not what we want, but this confirms that the Command/Env machinery is misbehaving here. No idea why I saw no complaints in the sandbox build though 🤷 | 20:22:49 |
raitobezarius | In reply to @weethet:catgirl.cloud raitobezarius: how do i read this? https://buildkite.com/lix-project/lix/builds/3673/steps/canvas?sid=01986f84-2c33-4e02-9c8e-6ad9e83250c9 you look at jobs rather | 20:23:25 |
raitobezarius | functional-sandbox-certificate-authorities has broken | 20:23:35 |
Rutile (Commentator2.0) feel free to ping | .... thanks, that could also be the way to stupidly fix the darwin crash... | 20:24:42 |
Rutile (Commentator2.0) feel free to ping | still curious though why the urls mismatch in the narstore test... | 20:25:35 |
Rutile (Commentator2.0) feel free to ping | at least locally for me | 20:25:46 |