!UNVBThoJtlIiVwiDjU:nixos.org

Staging

401 Members
Staging merges | Running staging cycles: https://github.com/NixOS/nixpkgs/pulls?q=is%3Apr+is%3Aopen+head%3Astaging-next+head%3Astaging-next-25.11 | Review Reports: https://malob.github.io/nix-review-tools-reports/129 Servers

Load older messages


SenderMessageTime
2 Jul 2026
@k900:0upti.meK900Also bisect ran out of disk space02:19:41
@k900:0upti.meK900Also I'm up at 5AM don't ask02:19:46
@k900:0upti.meK900Oh well, freed up some more space, hopefully it's done by normal person hours02:24:49
@k900:0upti.meK900Oh great10:11:31
@k900:0upti.meK900There's multiple issues10:11:34
@k900:0upti.meK900At least one failure is 3.14 specific10:11:44
@k900:0upti.meK900But there are also other failures that are not10:11:52
@k900:0upti.meK900So it bisected to the 3.14 update10:13:11
@k900:0upti.meK900But now I guess I have to pin 3.13 and rerun everything10:13:18
@hexa:lossy.networkhexa😬10:13:44
@k900:0upti.meK900I think there are also tests that just hard lock up on 3.1410:15:05
@k900:0upti.meK900Anyway here we go again I guess10:15:59
@k900:0upti.meK900At least everything should be cached...10:16:03
@k900:0upti.meK900Redacted or Malformed Event10:16:05
@k900:0upti.meK900OK great10:27:03
@k900:0upti.meK900All the google-cloud stuff is broken on 3.13 now10:27:09
@hexa:lossy.networkhexamaybe the one update I merged https://github.com/NixOS/nixpkgs/pull/53686410:31:54
@hexa:lossy.networkhexadunno, they generally want to be in sync10:32:04
@k900:0upti.meK900It's broken inside grpc somewhere10:32:56
@k900:0upti.meK900I think I'm just going to stub it out10:33:04
@k900:0upti.meK900It's only pulled in via celery10:33:19
@k900:0upti.meK900OK yes these are different10:42:41
@hexa:lossy.networkhexalooking at the google-cloud-shitfest10:47:51
@hexa:lossy.networkhexait's probably just that they haven't been bumped in a while10:48:07
@hexa:lossy.networkhexabecause guess what… nobody wants to10:48:12
@k900:0upti.meK900OK so on 3.13 nothing locks up and there's two tests that fail and they're both in the same place and something about Django signal handling10:49:33
@k900:0upti.meK900https://github.com/paperless-ngx/paperless-ngx/commit/5e1202a4168fbc8e36f816f36eb16dd7636e9d9c10:56:35
@k900:0upti.meK900OK cool10:56:37
@k900:0upti.meK900
diff --git a/pkgs/by-name/pa/paperless-ngx/package.nix b/pkgs/by-name/pa/paperless-ngx/package.nix
index 641da02e2848..8f6a53ea7034 100644
--- a/pkgs/by-name/pa/paperless-ngx/package.nix
+++ b/pkgs/by-name/pa/paperless-ngx/package.nix
@@ -3,10 +3,12 @@
   stdenv,
   fetchFromGitHub,
   fetchPypi,
+  fetchpatch,
   callPackage,
   nixosTests,
   gettext,
-  python3,
+  # tests fail and eventually lock up on 3.14
+  python313,
   ghostscript_headless,
   imagemagickBig,
   jbig2enc,
@@ -40,9 +42,21 @@ let

     # tesseract5 may be overwritten in the paperless module and we need to propagate that to make the closure reduction effective
     ocrmypdf = prev.ocrmypdf_16.override { tesseract = tesseract5; };
+
+    # these are broken on 3.13
+    google-cloud-firestore = null;
+    google-cloud-iam = null;
+    google-cloud-kms = null;
+    google-cloud-monitoring = null;
+    google-cloud-pubsub = null;
+    google-cloud-storage = null;
+
+    # these depend on google-cloud stuff in tests
+    celery = prev.celery.overridePythonAttrs { doCheck = false; };
+    kombu = prev.kombu.overridePythonAttrs { doCheck = false; };
   };

-  python = python3.override {
+  python = python313.override {
     self = python;
     packageOverrides = lib.composeManyExtensions [
       defaultPythonPackageOverrides
@@ -83,6 +97,15 @@ python.pkgs.buildPythonApplication (finalAttrs: {
     hash = "sha256-Czh4Knel0IIHsTc3kEnp1153Kv+3721GRCbTYTkeCDg=";
   };

+  patches = [
+    # fix tests with latest filelock
+    (fetchpatch {
+      url = "https://github.com/paperless-ngx/paperless-ngx/commit/5e1202a4168fbc8e36f816f36eb16dd7636e9d9c.diff";
+      includes = [ "src/*" ];
+      hash = "sha256-ZDC+T4DyOBBV8SCw8xyeYGua1XOhiP7eoZthnSE/Fkk=";
+    })
+  ];
+
   postPatch = ''
     # pytest-xdist with to many threads makes the tests flaky
     if (( $NIX_BUILD_CORES > 3)); then
@@ -114,6 +137,7 @@ python.pkgs.buildPythonApplication (finalAttrs: {
     # requested by maintainer
     "imap-tools"
     "ocrmypdf"
+    "filelock"
   ];

   dependencies =
11:07:35
@k900:0upti.meK900Doing one last run of this on current staging-next11:08:04

Show newer messages


Back to Room ListRoom Version: 6