<ajs124>
anybody got any idea what this might be about? "src/libexpr/attr-set.hh:55: void nix::Bindings::push_back(const nix::Attr&): Assertion `size_ < capacity_' failed."
<aterius>
I don't think that's what I mean, I'm aware you can apply an overlay to the channel within the expression that is called via the flake, or within the derivation itself, but I don't think it's possible to set the NIX_PATH to point to a nixpkgs set with the overlay already applied
<nbp>
aterius: otherwise, you can override the overlays on the command line with: nix-shell -I nixpkgs-overlays=<dir>
<aterius>
Yes, I tried doing that, but it doesn't seem to work from within a flake
<nbp>
aterius: no, the script which overrides it is responsible for honoring the nixpkgs-overlay path.
<aterius>
Right, but the nixpkgs-overlays path can't point to something defined within a flake I believe, at least not easily
<nbp>
aterius: don't use flake or maybe use a --impure flag or similar …
<nbp>
I am no expert in flakes, and I do not intend to be …
<aterius>
It works without flakes no problem, I've just been trying to switch everything over
<Raito_Bezarius>
is it the recent call for packging funded by NGI0 ?
<teto>
I've been trying to convert things to flakes but have some issues when referencing local flakes, some times nix picks up changes, some times not I am not sure why.
<aterius>
I'm still unclear when `nix flake update` works it seems like sometimes I've had to call it with --update-inputs for it to actually update
<teto>
Raito_Bezarius: that should be it, they announced it on the discourse
<Raito_Bezarius>
teto: interesting
<Raito_Bezarius>
because there is a package for mangaki.fr
<Raito_Bezarius>
which is arguably a package useful only for the core team of Mangaki (i.e. me at the moment)
AlwaysLivid has quit [Ping timeout: 260 seconds]
disasm has quit [Ping timeout: 265 seconds]
disasm has joined #nixos-dev
teto has quit [Ping timeout: 260 seconds]
cole-h has joined #nixos-dev
<clever>
deleting '/nix/store/add-15398-2'
<clever>
creating file '/nix/store/add-15398-2/x/hackage/amazonka-route53-1.4.1-r0-a4b6939395fbc0dac4bbdf9c07d931e3dbac2b44acdd5ef0a392463da6b62c60.nix': No such file or directory
<clever>
auto-gc deleted a temp dir in-use by another process
<manveru>
is there any way to have a flake input that has a branch with a slash in it?
<manveru>
(github type, that is)
<manveru>
why do people use branch names like `foo/bar` anyway...
<gchristensen>
just wait until you see the total list of valid branch names.
<michaelpj>
it's a perfectly sensible name-separator. And then git actually stores the refs in folders on the filesystem (not that this exactly matters...)
<manveru>
well, then nix probably shouldn't complain about it :P
<manveru>
both `github:org/repo/foo/bar` and `github:org/repo?ref=foo/bar` don't work, so i have to specify the rev instead, which can't be updated as nicely :|
<manveru>
oh, and submodules are broken... so i guess i have to use fetchGit instead anyway :|
<cransom>
i thought / as a valid character was fixed. at least 4-5months ago? at least when i poked flakes and saw that it couldn't fetch any of our branches
<cole-h>
manveru: I thought support was added for submodules recently-ish?
<manveru>
cole-h: well, in theory, but in practice you cannot specify `submodules = true` or `?submodules=true` for a flake input
tokudan has quit [Remote host closed the connection]
rajivr has quit [Quit: Connection closed for inactivity]
tokudan has joined #nixos-dev
tokudan has quit [Remote host closed the connection]
tokudan has joined #nixos-dev
orivej has joined #nixos-dev
eyJhb has quit [Quit: Clever message]
eyJhb has joined #nixos-dev
eyJhb has joined #nixos-dev
eyJhb has quit [Changing host]
AlwaysLivid has joined #nixos-dev
saschagrunert has joined #nixos-dev
cole-h has quit [Quit: Goodbye]
alp has quit [Ping timeout: 272 seconds]
orivej has quit [Ping timeout: 246 seconds]
saschagrunert has quit [Remote host closed the connection]
alp has joined #nixos-dev
MichaelRaskin has joined #nixos-dev
alp has quit [Ping timeout: 272 seconds]
__monty__ has quit [Quit: leaving]
<abathur>
could use a little input on how the installer handles rc/profile files, nix#3608 "meant" to make the installer edit /etc/zshenv instead of zshrc (apparently to avoid issues with the file getting clobbered by OS updates?), but the installer checks [ -e "$profile_target" ] before editing the file--so it won't work for anyone who doesn't already have /etc/zshenv
<abathur>
the current list is PROFILE_TARGETS=("/etc/bashrc" "/etc/profile.d/nix.sh" "/etc/zshenv")
<abathur>
I could make it create the file(s) if they don't exist, yeah? anyone aware of problems it would cause?
<samueldr>
I'll ask again in more european-friendly hours, and in work hours next week, but I think we really need to figure out a strategy to help the big aarch64 builds that commonly fail pass
<samueldr>
since AFAICT the main reason they fail is they're sent to a build machine that is doing too much
<samueldr>
thinking firefox, when there is no queue, a restart and it's done in ~1 hour
<samueldr>
when there is a queue, so misc. builds running it generally will time out
peelz has quit [Remote host closed the connection]