!rGlCMuXgAhgEpdvJUz:nixos.org

NixOS KDE

223 Members
63 Servers

Load older messages


SenderMessageTime
18 May 2026
@k900:0upti.meK900Does that come with custom window decorations?19:46:39
@aktaboot:tchncs.deaktabootit does yea19:46:54
@k900:0upti.meK900Try turning those off maybe19:46:58
@aktaboot:tchncs.deaktabootheh turning them off also crashed19:47:46
@k900:0upti.meK900 That's a sign of something 19:47:57
@k900:0upti.meK900Possibly19:48:00
@aktaboot:tchncs.deaktabootyeah will try to see if i can reproduce with the breeze ones19:48:24
@aktaboot:tchncs.deaktabootalso, how can i check which upstream commit your PR is referencing?19:48:46
@k900:0upti.meK900Upstream commit of what?19:49:10
@k900:0upti.meK900Nixpkgs?19:49:11
@aktaboot:tchncs.deaktabootkwin19:49:23
@k900:0upti.meK900Whatever the 6.6.90 tag is19:49:50
@aktaboot:tchncs.deaktabootokay thanks19:50:04
@aktaboot:tchncs.deaktabootyeah your educated guess was correct! it most definitely seems to be the window decorations!20:02:03
@k900:0upti.meK900OK, so now the next question20:10:48
@k900:0upti.meK900Do you know if those decorations are custom code or just aurorae?20:10:59
@aktaboot:tchncs.deaktabootno idea20:11:30
@aktaboot:tchncs.deaktabootthey seem to be aurorae20:12:08
@aktaboot:tchncs.deaktaboot https://github.com/vinceliuice/Layan-kde/tree/master/aurorae/themes/Layan 20:12:38
@k900:0upti.meK900 Yeah OK then we're probably looking at an aurorae bug 20:12:53
@aktaboot:tchncs.deaktabootI opened an issue on kde.org, will update the issue with this info then!20:16:22
20 May 2026
@grimmauld:m.grimmauld.deGrimmauld (any/all) cc @K900 :
github.com/NixOS/nixpkgs/issues/513970
Turns out plasma-login-manager currently attempts to load pam modules just by name, no from an absolute path... This is horrible and awful and just randomly exploded due to apparmor assertions, but i am not sure what else may be affected.
I am currently building nixosTests.plasma6 with the paths patched, PR will be up soon, but figured i'd give you a poke here too
14:47:22
@grimmauld:m.grimmauld.deGrimmauld (any/all) cc @K900 :
https://github.com/NixOS/nixpkgs/issues/513970
Turns out plasma-login-manager currently attempts to load pam modules just by name, no from an absolute path... This is horrible and awful and just randomly exploded due to apparmor assertions, but i am not sure what else may be affected.
I am currently building nixosTests.plasma6 with the paths patched, PR will be up soon, but figured i'd give you a poke here too
14:47:29
@grimmauld:m.grimmauld.deGrimmauld (any/all) cc @K900 :
https://github.com/NixOS/nixpkgs/issues/513970
Turns out plasma-login-manager currently attempts to load pam modules just by name, not from an absolute path... This is horrible and awful and just randomly exploded due to apparmor assertions, but i am not sure what else may be affected.
I am currently building nixosTests.plasma6 with the paths patched, PR will be up soon, but figured i'd give you a poke here too
14:47:38
@k900:0upti.meK900Uhhh14:48:29
@grimmauld:m.grimmauld.deGrimmauld (any/all)That was my reaction too14:48:47
@k900:0upti.meK900Yeah it does14:48:48
@k900:0upti.meK900Is that a problem?14:48:59
@grimmauld:m.grimmauld.deGrimmauld (any/all)It breaks an assertion we have with apparmor in pam, which is how it was noticed. I didn't dig through pam enough to judge how else it may break. like, custom definitions to pam package? Overrides? does it find the modules at all? (probably does or it'd have exploded earlier, but i don't like this)14:50:36
@grimmauld:m.grimmauld.deGrimmauld (any/all)Like, using modules from an explicitly defined absolute path is objectively the correct thing to do, because it is more predictable. Fix will be up once tests passed, and is trivial.14:53:08

Show newer messages


Back to Room ListRoom Version: 9