| 25 Mar 2026 |
 emily | they want you to upgrade :) | 17:29:06 |
 nasso | hell yeah it did
thanks! i thought i did something wrong but it turns out i only really tried on shitty wifis (hotel and now train) | 17:29:22 |
| nasso | lk scared of keeping the bitwig dot com url i dont trust them keeping it up | 17:30:19 |
| nasso | but yeha okay thank you emily once again | 17:30:21 |
| nasso | ill get rid of the brew-nix overlay lmao | 17:30:31 |
| nasso | yeah okay this clears up some things for me cuz i see a lot of projects use overlay when i feel like they shouldn't (e.g. here they're using an overlay to add something to nixpkgs, but they aren't modifying anything in nixpkgs, so like, whats the point) | 17:37:17 |
| nasso | * yeah okay this clears up some things for me cuz i see a lot of projects use overlays when i feel like they shouldn't (e.g. here they're using an overlay to add something to nixpkgs, but they aren't modifying anything in nixpkgs, so like, whats the point) | 17:37:26 |
| emily | could override it with a web.archive.org link | 17:47:00 |
| emily | overlays compose better and can avoid instantiating a bunch of different copies of Nixpkgs (because of flakes' awful transitive dep handling / follows messes etc.), but yeah they also have an awkward shared global namespace like that | 17:47:34 |
| nasso | yeah but also i have no idea how i would do that 😶 | 17:48:00 |
| emily | in particular that specific overlay still uses the nixpkgs flake input rather than even using final as the package set, so it won't "see" other overlays | 17:48:04 |
| emily | probably .overrideAttrs { src = fetchurl …; }, but I'm not sure exactly how the brew-nix derivations are structured :) | 17:48:19 |
| nasso | avoid instantiating a bunch of different copies of Nixpkgs
oooh is it because the overlay can use final (or prev) instead of a separate nixpkgs instance? | 17:51:25 |
| emily | it's because it picks it out of packages which refers to inputs.nixpkgs.legacyPackages | 17:51:43 |
| emily | where it should be using final instead (which would contain changes from other overlays) | 17:51:53 |
| emily | overlays is meant to not use a nixpkgs flake input | 17:52:02 |
| nasso | yeah okay that makes sense to me so that means im getting it | 17:53:39 |
| nasso | great | 17:53:48 |
| 26 Mar 2026 |
 K900 | > : && /nix/store/s7qlr26bmc6n4r607scz8iiwcg6yg4ic-clang-wrapper-21.1.8/bin/clang++ -DNDEBUG -O2 -arch arm64 -isysroot /nix/store/g0vv53n49k0kvnl5p4ag3bygypbvf8sh-apple-sdk-15.5/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk -mmacosx-version-min=14.0 -Wl,-headerpad_max_install_names -fPIE -Xlinker -pie tools/qmltestrunner/CMakeFiles/qmltestrunner.dir/qmltestrunner_autogen/mocs_compilation.cpp.o tools/qmltestrunner/CMakeFiles/qmltestrunner.dir/main.cpp.o -o bin/qmltestrunner -F/nix/store/vp7mzysndh7vmqp4rqmigxl801x1pvn7-qtdeclarative-6.11.0/.build/qtdeclarative-everywhere-src-6.11.0/build/lib -F/nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib -Wl,-rpath,/nix/store/vp7mzysndh7vmqp4rqmigxl801x1pvn7-qtdeclarative-6.11.0/.build/qtdeclarative-everywhere-src-6.11.0/build/lib lib/QtQuickTest.framework/Versions/A/QtQuickTest /nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib/QtTest.framework/Versions/A/QtTest -framework Security -framework ApplicationServices -framework Foundation /nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib/QtGui.framework/Versions/A/QtGui -framework AppKit -framework OpenGL -framework ImageIO -framework Metal lib/QtQml.framework/Versions/A/QtQml /nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib/QtNetwork.framework/Versions/A/QtNetwork /nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib/QtCore.framework/Versions/A/QtCore -framework IOKit -framework DiskArbitration -framework UniformTypeIdentifiers && cd /nix/store/vp7mzysndh7vmqp4rqmigxl801x1pvn7-qtdeclarative-6.11.0/.build/qtdeclarative-everywhere-src-6.11.0/build/tools/qmltestrunner && codesign --remove-signature /nix/store/vp7mzysndh7vmqp4rqmigxl801x1pvn7-qtdeclarative-6.11.0/.build/qtdeclarative-everywhere-src-6.11.0/build/bin/qmltestrunner && codesign --sign - --entitlements /nix/store/0rzzw62rkc2bga3spjs73g0km36i0lda-qtbase-6.11.0/lib/cmake/Qt6/macos/test.entitlements.plist /nix/store/vp7mzysndh7vmqp4rqmigxl801x1pvn7-qtdeclarative-6.11.0/.build/qtdeclarative-everywhere-src-6.11.0/build/bin/qmltestrunner
> --sign is required
| 06:27:59 |
| K900 | Anyone know what is going on here | 06:28:12 |
 viraptor | The first codesign does just --remove-signature - does it work in isolation? Could be from there | 06:36:43 |
| K900 | I genuinely have no clue | 06:37:17 |
| K900 | I think this has happened before, somehow | 06:37:49 |
| K900 | But I still don't understand it | 06:37:57 |
| emily | probably sigtool limitations | 12:36:40 |
| K900 | I think last time this happened it was transient or something? | 12:42:17 |
| emily | race condition with build order? | 12:45:22 |
| emily | undeclared deps? | 12:45:28 |
| K900 | idk | 12:45:35 |
| K900 | But it seems consistent now | 12:45:42 |
