<samueldr>
I think it might be wise to look into defaulting to using the ext4 partition for extlinux.conf
<muvlon>
hmm, seems to not boot
<muvlon>
i'll try to find an hdmi screen lol
<samueldr>
muvlon: did you `mv /boot /_boot`?
<muvlon>
yes
<samueldr>
ha, sorry!
sanscoeu_ has quit [Ping timeout: 246 seconds]
<clever>
then it cant find /boot/ on p2
<samueldr>
power if off, mount the card on your computer and mv it
<muvlon>
right, i can just mount the sdcard here
<muvlon>
no need for a screen :)
<clever>
~2 days ago, i did some even more crazy things
<samueldr>
right, u-boot checks only at the root and in the /boot/ directories, so it won't know about /_boot :)
Ariakenom has quit [Read error: Connection reset by peer]
<clever>
where did i drop it...
<samueldr>
clever: this sounds ominous
<clever>
NIX_REMOTE=local?root=/mnt/img/ nix-env --profile /mnt/img/nix/var/nix/profiles/system --set -f '<nixpkgs/nixos>' -A system -I nixos-config=/mnt/img/etc/nixos/configuration.nix --option system aarch64-linux
<clever>
samueldr: that effectively does `nixos-rebuild build` in /mnt/img, without chroot, even if the arch is mismatched (relies on qemu-user-aarch64)
endformationage has joined #nixos
<muvlon>
whoop, it's back! (but with the old initramfs)
<muvlon>
i have very mixed feelings about the idea of having nixos manage partitions and filesystems
<muvlon>
on one hand, it would mean stuff like this just works
<clever>
muvlon: the only time i have nix manage partitions, is via a bash script at install time
<muvlon>
but on the other hand, dear god if it goes wrong
phaebz has quit [Remote host closed the connection]
<pbb>
Hi, I am writing a networking module, that takes an attribute set as configuration. Can I somehow create an option inside one of the attributes, that uses its name as default?
<pbb>
for example if I were to do `networking.xyz.interfaces.myinterface = {}`, that `networking.xyz.interfaces.myinterface.foobar` would default to `myinterface`?
<pbb>
I have interfaces defined as an option of `attrsOf (submodule interfaceOpts)`, but can I somehow access the name of that interface in those options, i.e. give it as an argument to interfaceOpts?
<clever>
pbb: the function you give to that system, will take something in the form of { name, config, ... }: i think
<adisbladis>
janders223: You mind pasting your expressions?
goibhniu has quit [Ping timeout: 268 seconds]
danbst has quit [Ping timeout: 256 seconds]
<janders223>
adisbladis: I can see them in /nix/store... however according to the directions I should be able to add either /run/current-system... or .nix-profile/sw... to the load-path
goibhniu has joined #nixos
<janders223>
the emacs site-lisp directories in both of those locations is empty
<adisbladis>
janders223: If you use something like emacsWithPackages from nixpkgs you should end up with a wrapper script setting up the load path for you
<muvlon>
one more raspi-related question
<janders223>
adisbladis: I did that, and assume that it is symlinking that wrapper to ~/Applications but still no dice
<janders223>
Tried to glean some wisdom from jwiegley's nix config but that wizardy is beyond my novice nixing at this point
<adisbladis>
janders223: If you would paste your nix expressions it would help a lot
<muvlon>
is there a nice way to modify my device tree?
cyounkins has quit [Remote host closed the connection]
cyounkins has joined #nixos
aleph- has quit [Ping timeout: 246 seconds]
cyounkins has quit [Ping timeout: 240 seconds]
janders223 has quit [Remote host closed the connection]
silver has quit [Read error: Connection reset by peer]
<fresheyeball>
why is enableLibraryProfiling enabled by default?
<clever>
fresheyeball: it builds libraries twice, with and without profiling, which saves you from having to rebuild the world with profiling when you decide to profile something
<clever>
and typically, the libs are pre-build by hydra for you
ng0_ has joined #nixos
<fresheyeball>
I see --enable-library-profiling --profiling-detail=all-functions
<fresheyeball>
AND
<fresheyeball>
--disable-profiling
<fresheyeball>
isn't that contradictory?
xkapastel has quit [Quit: Connection closed for inactivity]
<clever>
fresheyeball: i believe --disable-profiling refers to executables, not libraries
<clever>
so the binaries cabal produces wont have the profling rts, and wont use the profiled variants of libs
ng0 has quit [Ping timeout: 256 seconds]
<fresheyeball>
ok do my debugging confirms that enableLibraryProfiling = false;
<fresheyeball>
does not remove the flag
<fresheyeball>
and that shit just cost me like 2 days
<fresheyeball>
dammit
cyounkins has joined #nixos
Supersonic112 has joined #nixos
Supersonic has quit [Disconnected by services]
Supersonic112 is now known as Supersonic
AveryLychee has quit [Ping timeout: 246 seconds]
janders223 has joined #nixos
<pbb>
is it possible to nix-install a 18.09 system from an unstable one?
chessai has quit [Ping timeout: 272 seconds]
janders223 has quit [Remote host closed the connection]
<blumenkranz>
Okay, found the issue. It is some incompatibility issue with the sandbox, it seems.
<pbb>
ey, now I definitely changed the channel to 18.09 before I installed the install tools and it still generates configurations.nix with stateVersion 19.03!
<jasongrossman>
pbb: "stateVersion" is a very misleading name.
jmgai has joined #nixos
<jasongrossman>
pbb: You can leave it at "19.03".
<pbb>
okay.. but that means I still have nixos-generate-config from unstable, right?
<jasongrossman>
pbb: No.
<pbb>
I'm so much confused, but if you say so I'll just leave it at 19.09 and not worry about it. I was afraid that it would install an unstable system.
<pbb>
but probably it'll be fine
<simpson>
,stateVersion
<{^_^}>
Setting stateVersion to the latest release doesn't upgrade anything and can only break your setup at best. To actually upgrade NixOS see https://nixos.org/nixos/manual/#sec-upgrading. If you want to update the stateVersion option regardless, Ctrl-F for "stateVersion" in https://nixos.org/nixos/manual/release-notes.html to see things that need to be manually migrated with the new value.
<pbb>
yeah I know
<simpson>
pbb: Hm. I guess that that doesn't directly address your concern. `stateVersion` is a version for any stateful data on your system which might be versioned in a way that you don't want to have changed suddenly without your knowledge.
<pbb>
At least I think I understand the concept of stateVersion, but it's still confusing that after downgrading my installation tools to 18.09 they still generate new configuration files with stateVersion=19.03
<pbb>
because the new installation would have a fresh state, starting at 18.09, right?
<simpson>
Yeah, that's a little interesting; how do they know that? Maybe there's something readable in /etc.
<pbb>
no, it's in the nixos-generate-config package
<pbb>
/nix/store/92z4h130xidy60zlhzqczzal8b96wyy8-nixos-generate-config/bin/nixos-generate-config: system.stateVersion = "${\(qw(19.03))}"; # Did you read the comment?
<clever>
pbb: what rev of nixpkgs are you using?
<clever>
pbb: and nixos-generate-config will never overwrite configuration.nix
<pbb>
I know, I don't have a configurations.nix because my current system is not NixOS
<pbb>
and I used `nix-env -iE "_: with import <nixpkgs/nixos> { configuration = {}; }; with config.system.build; [ nixos-generate-config nixos-install nixos-enter manual.manpages ]"` to install the tools
jmgai has quit [Ping timeout: 246 seconds]
<clever>
pbb: what does `nix-instantiate --find-file nixpkgs` return?
<pbb>
/home/pbb/.nix-defexpr/channels/nixpkgs
<clever>
pbb: and realpath /home/pbb/.nix-defexpr/channels/nixpkgs
<clever>
pbb: which in your case, is /nix/store/vrz9gkwk82c8dp83798nyhlsw862xzaj-nixpkgs-18.09.1922.97e0d53d669/nixpkgs/.version
<pbb>
okay, so it will actually install a stable system, even though it generates a config file with stateVersion 19.03
<clever>
did you re-run the nix-env command, after changing the channel?
<pbb>
yes
<pbb>
nix-env -u --always
<clever>
-u cant update nixos-generate-config
<pbb>
why not?
<clever>
it doesnt exist in the main pkgs set, so it never finds an update
<pbb>
o..okay?
<clever>
anything you install using -E is imposible to update in an automated way
<pbb>
I see
<clever>
because it never saves that expression
blumenkranz has quit [Quit: WeeChat 2.3]
<clever>
-u is very dumb, it just searches all packages for something with a matching name
<clever>
and you needed -E because its not in the main package set
sheth has quit [Ping timeout: 272 seconds]
<clever>
re-run the nix-env -iE you gave above, and it will actually "update" to the older nixos-generate-conifg
<pbb>
okay, after re-running it it gives me the right config
<pbb>
as we say in germany "1000 dank" (1000 times thank you) :-)
<clever>
you can also do clever ++ (without the space)
<pbb>
clever++
<{^_^}>
clever's karma got increased to 72
<pbb>
{^_^}
<pbb>
I remember you've helped me before
<pbb>
clever++
<{^_^}>
clever's karma got increased to 73
<simpson>
clever++ for being faster and more useful than me.
<{^_^}>
clever's karma got increased to 74
<clever>
part of it is just having OCD about how things work, and reading source code non-stop :P
<clever>
part of it is being able to memorize exactly where nixos-generate-config comes from
<clever>
and remembering all the quirks of all the tools
<pbb>
now I'll try my luck with NIX_LUSTRATE
<pbb>
*NIXOS_LUSTRATE
<clever>
fun fact, i think i was the first one to do that, before it was automated
<clever>
and the script in nixos was written from the story of how i did it
<pbb>
that's pretty cool
<pbb>
and I think I'll try to add something to the nix manual about this sysctl that's needed for installation on some systems that have user namespaces built-in but disabled
<clever>
i believe i first did it on a netbook, andthen had to debug all the breakage from /lib and /etc and /bin having bits of gentoo in them
<pbb>
sounds awful :D
<pbb>
totally my kink
<clever>
and the basic solution was just rm -rf /usr /lib /bin /etc
<pbb>
yeah, who needs 'em anyways
<clever>
NIXOS_LUSTRATE is safer, in that it moves everything to /old-root/
ng0_ has quit [Quit: Alexa, when is the end of world?]
<clever>
pbb: another option, is kexec
<pbb>
yeah actually I would prefer that, but it's not documented, so I will stick with lustrate for now
<clever>
currently, you must unpack the tarball to /, then run /kexec_nixos
<clever>
but, arx can turn the tar into a self-extracting executable
<clever>
though unlike vps2arch, the tar contains the entire closure, including the nixos rootfs
<clever>
so its ~200mb, rather then only 350 lines
<pbb>
wish me luck that this works
<pbb>
now rebooting
cyounkins has quit [Ping timeout: 246 seconds]
cyounkins has joined #nixos
<pbb>
ok I f-ed up the same way as the last time I tried this
<pbb>
probably some typo in the network config
petersjt014 has quit [Remote host closed the connection]
<clever>
yeah, network is where i had the most trouble too
<pbb>
but that's why I do this stuff in my free time, to not make the mistakes later :D
<clever>
there is a test script in my kexec dir, to generate qemu VM's with/without uefi, luks, sata/nvme, and include a kexec image inside thats configured for that
<clever>
ive yet to find coreboot support for any hardware i have
<pbb>
I like my x230t :D
<pbb>
so .. I don't start grub from the disk, but it's part of my coreboot image
<muvlon>
hey, i could some more help again -.-'
eadwu has quit [Ping timeout: 252 seconds]
<pbb>
what's the matter muvlon
<clever>
pbb: as long as the modules nixos uses, are in your grub2 payload, it should just work, as long as the payload reads the grub.cfg nixos generates
<pbb>
yeah, I chainload the grub configuration
<pbb>
but something funky happens
<pbb>
if I boot the default entry, I don't get graphics
<muvlon>
i have patched a device tree blob (took dtb from /boot/nixos, turned it into dts, edited it manually, turned back into dtb and copied over the original one in /boot)
<muvlon>
is there a way to make this persistent (i.e. do it in a .nix file)?
<clever>
pbb: i'm guessing thats more likely a coreboot issue then a grub issue, since the bios is typically responsible for bringing the gpu into a sane state
<pbb>
yeah, but when I go to all configurations and choose the latest one it doesn't happen
<pbb>
even though it's exactly the same configuration
<muvlon>
as ist stands, i think this doesn't survive a nixos-rebuild (but it does work when rebooting)
<clever>
pbb: thats weird
<pbb>
just found that to be funny, but I fixed it by adding the i915 module to the initramfs
<clever>
muvlon: youll want to find out what script is copying it to /boot/ and then setup an override, is this on an rpi?
<clever>
muvlon: then youll want to read what samueldr linked
goibhniu has quit [Ping timeout: 268 seconds]
muvlon_ has joined #nixos
cyounkins has quit [Ping timeout: 240 seconds]
<muvlon_>
i'm still very new to nix though, so i'm having a bit of a hard time reading some of these things
cyounkins has joined #nixos
muvlon has quit [Ping timeout: 264 seconds]
<muvlon_>
if i add a different script that also copies stuff to that destination, will they clash?
<muvlon_>
will the functional gods smite me for destroying purity?
<clever>
muvlon_: only one script can be setup for installing a bootloader, so it will just run one
<muvlon_>
hmm
<fresheyeball>
why would nix-build be touching /usr/local/bin anyway?
pie__ has quit [Read error: Connection reset by peer]
<fresheyeball>
that doesn't make sense
<muvlon_>
so no other scripts can even touch /boot ?
<clever>
fresheyeball: is /usr in the cabal file?
pie_ has joined #nixos
<clever>
muvlon_: no other scripts are ran at the right phase in nixos-rebuild
<fresheyeball>
clever: what do you mean?
<clever>
fresheyeball: does the cabal file or Setup.hs maybe have a hard-coded command to copy things to /usr/ ?
<muvlon_>
what if one runs later? it could overwrite the dtb, no?
<muvlon_>
or is /boot immutable later on?
<clever>
muvlon_: there is currently no way to tell nixos-rebuild to run 2 scripts
<fresheyeball>
clever: I doubt it, this just started happening
<fresheyeball>
I will check though
<clever>
fresheyeball: maybe your missing the --prefix=$out now?
<clever>
so cabal has fallen back to defaults
<fresheyeball>
ok I grepped for /usr in my cabal file
<fresheyeball>
ooooo
<fresheyeball>
ooOooooOoooooOooooo
<fresheyeball>
maybe
<muvlon_>
clever: so exlinux-config-builder.sh is the only shell script that is ever executed during my nixos-rebuild?
<Ralith>
is it just me or is the rust overlay broken on nightlies again
<muvlon_>
sorry if these are silly questions
<clever>
muvlon_: the only script being ran impurely, when updating /boot/
<ottidmes>
I removed the Nix config from my NixOS config that included substituter 'https://nix-linter.cachix.org', but it still seems to check it, checked /etc/nix/nix.conf and it no longer contains a reference to it, I restarted nix-daemon too, how do I purge it for real, because I can now not build because it is down
<clever>
muvlon_: and thats based on which boot.loader.X.enable you have set
lord| has quit [Quit: WeeChat 2.3]
<muvlon_>
i see
<clever>
ottidmes: what about the nix.conf in $HOME ?
<muvlon_>
is there a way to impurely run a script after everything else?
<Ralith>
yep
<clever>
muvlon_: you could just manually run it after every nixos-rebuild, but your bound to forget
<muvlon_>
right
<clever>
muvlon_: it would be simpler to clone the nixpkgs repo at the current rev, and just edit the script
<muvlon_>
but then i'd have to keep that up to date somehow
<ottidmes>
clever: thanks! did not even knew it existed, must have been generated by cachix
<ottidmes>
clever++
<{^_^}>
clever's karma got increased to 75
<clever>
ottidmes: there is also `nix show-config`
slack1256 has quit [Remote host closed the connection]
<ottidmes>
clever: yeah, but I assumed it would give me /etc/nix/nix.conf because I did not know about the existance of ~/.config/nix/nix.conf
<{^_^}>
haskell/cabal#5516 (by hvr, 22 weeks ago, closed): new-install fails with empty store: ghc-pkg: .../package.db: getDirectoryContents:openDirStream: does not exist (No such file or directory)
<fresheyeball>
so far this is all I can find
<clever>
pbb: i still dont trust grub with zfs support though
<pbb>
I wouldn't either
<pbb>
I'd rather make a small ext4 boot partition in front of the zfs
<clever>
pbb: i usually make /boot 1gig, for my rescue system
<clever>
pbb: this puts the same kind of kernel/initrd uses for kexec into /boot/
<pbb>
nice
<clever>
and if you can access grub, you can then boot into it at any time, to handle reinstalls
<clever>
or other repair
<pbb>
I really hope my server starts to respond in a few seconds D:
<pbb>
but probably I just f-ed up again
o1lo01ol1o has joined #nixos
<pbb>
grub installation finished this time at least
mounty has joined #nixos
o1lo01ol1o has quit [Ping timeout: 250 seconds]
<fresheyeball>
I figured it out
<fresheyeball>
sorry for rubber ducking so much
<clever>
fresheyeball: what was it?
<fresheyeball>
well I manually tested each of the compiler flags the generic builder was imposing on me
<fresheyeball>
by over riding the configure phase
<fresheyeball>
and in doing so I pasted the filled version of the $packageConfDir
<pbb>
oohh: in the lustrate script /boot is hardcoded as an exception, but I mounted my boot partitions in /boot1 and /boot2 because I use grub mirroredBoot
<fresheyeball>
which is right for my machine
<fresheyeball>
but not ci
<fresheyeball>
maybe instead of trying to duplicate the flags work done in the builder
<fresheyeball>
I should try and filter out the toxic flag, since the option doesn't work
<fresheyeball>
yeah replicating all this is going to be bad
<clever>
fresheyeball: now, when every single package in the set tries to call mkDerivation, it calls this variant
<clever>
which will // an attr in, before calling the real one
<fresheyeball>
gcc: error trying to exec '/nix/store/imfm3gk3qchmyv7684pjpm8irvkdrrkk-gcc-7.3.0/libexec/gcc/x86_64-unknown-linux-gnu/7.3.0/collect2': execv: Argument list too long
<fresheyeball>
`cc' failed in phase `Linker'. (Exit code: 1)
<fresheyeball>
build failed with exit code 1 (ignored)
<fresheyeball>
this. is. never. ending
<clever>
fresheyeball: which channel?
<fresheyeball>
it builds locally
<fresheyeball>
only ci does that
<fresheyeball>
clever: I am using this hash 153d32cd9fac7e885979426b0e86b560a661a1ac of nixpkgs
<clever>
fresheyeball: run `xargs --show-limits` on both CI and local
<clever>
Maximum length of command we could actually use: 2082552
<clever>
fresheyeball: i believe master has had ghc gixed involving command line length, though it may only be in angerman's branch
<angerman>
that one will stop nix from throwing everything into the LD_FLAGS. ghc will do that on its own just fine.
<clever>
fresheyeball: what happens if you switch to that commit of nixpkgs?
<fresheyeball>
you mean the fork?
<clever>
fresheyeball: yeah, that commit, on the fork
<angerman>
the response file fix does help, but might not help with older gccs, as they expand the response_file. And then hands those over to collec2 internally without using a response_file.
<angerman>
Don't ask.
<fresheyeball>
clever: how can I up the limit?
<fresheyeball>
wont that be easier?
<clever>
fresheyeball: i dont think its easy to up the limit
<clever>
fresheyeball: changing nixpkgs is easyer, -I nixpkgs=https://github.com/input-output-hk/nixpkgs/archive/d11dbab6ae9f15fa2bd97906f09eaec934021f6d.tar.gz
<angerman>
I doubt it. It's been in GCC last year when I checked. It's still expanding the reponse-file even if you hand gcc one, when it hands the stuff down to collect2 internally.
<angerman>
fresheyeball: you *really* want to stop nix polluting your LD_FLAGS with junk.
<fresheyeball>
angerman: I am just confused
<fresheyeball>
I have worked on haskell+nix projects much much larger than this, and never seen this before
<angerman>
fresheyeball: your underlying issue is that the size to pass arguments, and env vars is limited. Now if you have a lot of dependencies in your transitive dependnecy tree, you end up with a lot of -L flags passed to your compiler/linker.
<angerman>
ghc will generate them when invoking the compiler/linker, and nix wants to be smart and injects them all as well into the LD_FLAGS. (Because it completely disregards ghc being smart).
<angerman>
so now as your transitive dependency closure grows, you end up injecting everything twice. (once from ghc, once from nix).
<fresheyeball>
what do you mean be transitive dependency tree?
<angerman>
The only thing I'm confused about is, I though this was fixed in upstream nix, but I haven't looked. so *shrug*
<angerman>
fresheyeball: lib a depends on lib b depends on lib c, now you have a,b,c
<fresheyeball>
angerman: I am trying jump bumping my rev as well
<fresheyeball>
ci just timed out
<angerman>
fresheyeball: basically all dependencies of all dependencies, ...
<angerman>
anything you transitively depend on.
<fresheyeball>
ah
<fresheyeball>
I still don't get it. My deps list is not huge
<fresheyeball>
it seems like I should be running into this with massive projects, and not medium ones
<fresheyeball>
angerman: moving to a newer hash made me upgrade from 843 to 844
<fresheyeball>
when did you upstream?
<angerman>
I don't know if this ever got upstreamed. there were some discussions about the right approach on the issuetracker somewhere.
<angerman>
We just needed a solution in our forked nixpkgs, as we ran into this issue pretty fast on macOS where the limit is even smaller.
<angerman>
fresheyeball: and yes, you should not run into this with small projects.
<angerman>
fresheyeball: but I don't really know what's in your dependency closure and what that might pull in.
<fresheyeball>
true
<fresheyeball>
angerman: and this fork worked for you?
<fresheyeball>
it still building
<sphalerite>
Anyone familiar with gpg complaining "error searching keyserver: No data" when you try to search for a key?
<angerman>
fresheyeball: we use our fork with a pretty large haskell project on Linux, and macOS.
<fresheyeball>
angerman: are you in Denver?
<angerman>
no
glasserc_ has quit [Ping timeout: 246 seconds]
<fresheyeball>
angerman: do you have a nix cache?
<fresheyeball>
It looks like this could run for a few hours
<clever>
fresheyeball: hydra.iohk.io has most of this branch cached
<etu>
Yeah, libraries can't be installed that way using nix-env or so
<rcshm>
do you know how to resolve this issue?
<etu>
,libraries rcshm
<{^_^}>
rcshm: Don't install libraries through nix-env or systemPackages, use nix-shell instead. See https://nixos.wiki/wiki/FAQ/Libraries for details.
<{^_^}>
[nixpkgs] @FRidh pushed commit from @r-ryantm to staging « zeromq: 4.3.0 -> 4.3.1 »: https://git.io/fhEXr
hedning has joined #nixos
<rcshm>
etu: declaring the in packages in stack/nix does not resolve it though.
<rcshm>
can you tell me more about nix-shell?
<etu>
you can spawn a temporary environment using nix-shell. For example: "nix-shell -p name-of-packages..." gives you a bash-shell with that package available
<etu>
And that works for libraries as well
<etu>
For example, I sometimes use ipython, but I don't have it installed. But I launch it with nix-shell like this: $ nix-shell --run ipython -p python3Packages.ipython
<etu>
And have that as an alias in my shell
<{^_^}>
[nixpkgs] @FRidh merged pull request #53674 → Limit parallel building of Haskell packages. [Fixes #53665] → https://git.io/fhZki
<{^_^}>
[nixpkgs] @FRidh pushed commit from @ElvishJerricco to staging « Limit parallel building of Haskell packages. [Fixes #53665] »: https://git.io/fhEX9
<etu>
And you can do the same with lzma as a lib. You probably want something like: $ nix-shell -p lzma
<etu>
and then use that shell to build your project
alex`` has joined #nixos
<rcshm>
ok. i think i got what you are saying now.
alex`` has quit [Read error: Connection reset by peer]
rcshm has quit []
alex`` has joined #nixos
<jluttine>
i'm trying to write a new package derivation. nix-build managed to build but when i run the binary, i get an error: "symbol lookup error ... undefined symbol: XXX". any ideas what's wrong? does some package provide XXX and i should add it as a dependency..? :o
init_6 has quit []
<{^_^}>
[nix-mode] @etu pushed to master « Move the logic for indentation of first line of file »: https://git.io/fhEMq
fendor has joined #nixos
<{^_^}>
[nixpkgs] @jluttine opened pull request #54307 → [Help needed] energyplus: init at 9.0.1 → https://git.io/fhEMc
orivej has quit [Ping timeout: 245 seconds]
xkapastel has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed 7 commits to haskell-updates: https://git.io/fhEMX
MichaelRaskin has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed 2 commits to haskell-updates: https://git.io/fhEMS
<{^_^}>
[nixpkgs] @peti pushed commit from @ElvishJerricco to haskell-updates « Limit parallel building of Haskell packages. [Fixes #53665] »: https://git.io/fhEMH
<{^_^}>
[nixpkgs] @Profpatsch pushed commit from @r-ryantm to master « rabbitmq-server: 3.7.9 -> 3.7.10 »: https://git.io/fhEy5
o1lo01ol1o has joined #nixos
<Laalf>
hello! environment.systemPackages = with pkgs; [ ] ++ builtins.filter stdenv.lib.isDerivation (builtins.attrValues kdeApplications); installs all kdeApplications into my systemPackages. is it possible to exclude packages from there? i thought about lib.subtractLists [ kdeApplications.random_package ] config.environment.systemPackages but was not able to get it working
o1lo01ol1o has quit [Ping timeout: 268 seconds]
<Profpatsch>
Laalf: I’d advise to do it the other way around and list all packages you want to include instead.
<andi->
Laalf: you probably need an intermediate list or maybe priorities but I am not sure about the later
<andi->
But yeah, there is no good reason (for me) to install "everything"
<Laalf>
that is more for learning nix than anything else. i wouldnt use it on kdeapplications
Makaveli7 has joined #nixos
Makaveli7 has quit [Remote host closed the connection]
<lesh>
hey people, I'm trying to setup wee-slack plugin, which seems to be missing websockets module when installed via nixos. if I ask for python2Packages.websockets I'm getting a 'websockets-6.0 not supported for interpreter python2.7' error, yet weechat only has 2.7 interpreter embedded. how would you get around this? tnx..
<{^_^}>
[nixpkgs] @adisbladis opened pull request #54315 → mpv: Add javascript support → https://git.io/fhEQj
<lesh>
disasm: slack protocol is significantly more complex then irc protocol (threads etc) so ideally I was hoping to get a slack support in my client and avoid the irc 'downgrading' but yeah I have this in mind as a last resort
<lesh>
yeah going through your repo now, awesome, tnx
<disasm>
lesh: yeah, I have it setup to launch 3 weechats on boot via a systemd service, 2 for slack and 1 for freenode to separate things out. It works pretty well.
<disasm>
although if you have team chats across multiple slack accounts, I found that crashed weeslack faster than a tortoise in a race with a hare.
vidbina has joined #nixos
<lesh>
disasm: haha damn, ok, good to know. and I see I was just installing the wrong module, websockets vs websocket_client
<lesh>
disasm: actually googled out your repo before asking, but figured from the 'irc-gateway' name that it's some different project, some service that listens on 6667 and not weechat plugin
<disasm>
so... I feel like an idiot, but in copying files from my old -> new laptop I accidentally whacked my configuration.nix/hardware-configuration.nix. Is there any nix tool (or somewhere in the store I could copy from) to regenerate from the running config?
<{^_^}>
[nixpkgs] @kampka opened pull request #54316 → Configure sysconfdir to point to /etc → https://git.io/fhE7W
<disasm>
lesh: yeah, clever originally was using the slack gateway I think before slack dropped support for it and then he added his patches to weeslack in the same repo.
<{^_^}>
[nixpkgs] @peterhoeg opened pull request #54317 → pythonPackages.zm-py: init at 0.3.0 → https://git.io/fhE7B
slyfox has quit [Ping timeout: 245 seconds]
<{^_^}>
[nixpkgs] @peterhoeg opened pull request #54318 → home-assistant: missing a parse-requirements.py run → https://git.io/fhE7D
gwosix has quit [Quit: WeeChat 2.3]
<{^_^}>
[nixpkgs] @danbst opened pull request #54319 → postgresql: reorganize package and it's extensions → https://git.io/fhEdI
<{^_^}>
[nixpkgs] @FRidh merged pull request #54094 → treewide: use ${stdenv.shell} instead of /bin/sh where possible → https://git.io/fhlrH
<{^_^}>
[nixpkgs] @FRidh merged pull request #54225 → gn: use python2 as gn is incompatible with python3 → https://git.io/fh4yc
<{^_^}>
[nixpkgs] @FRidh pushed commit from @Ekleog to master « gn: use python2 as gn is incompatible with python3 »: https://git.io/fhEds
<{^_^}>
[nixpkgs] @dotlambda merged pull request #54318 → home-assistant: missing a parse-requirements.py run → https://git.io/fhE7D
<{^_^}>
[nixpkgs] @dotlambda pushed commit from @peterhoeg to master « home-assistant: missing a parse-requirements.py run (#54318) »: https://git.io/fhEdZ
dsiypl4_ has joined #nixos
joehh has joined #nixos
phaebz has quit [Ping timeout: 250 seconds]
slyfox has joined #nixos
aleph- has quit [Ping timeout: 240 seconds]
hamishmack has joined #nixos
<{^_^}>
[nixpkgs] @adisbladis opened pull request #54320 → youtube-dl: Change to ffmpeg_4 → https://git.io/fhEFm
<{^_^}>
[nixpkgs] @primeos pushed to master « tdesktop: Drop the patch for a GCC bug »: https://git.io/fhEF4
jmgai has joined #nixos
Soo_Slow has joined #nixos
phaebz has joined #nixos
<{^_^}>
[nixpkgs] @vbgl opened pull request #54321 → ocamlPackages.lablgtk: make glade support optional → https://git.io/fhEFS
knupfer has quit [Ping timeout: 264 seconds]
<Soo_Slow>
is there way to ask for package updates? For example - eduke32 is 20150420 in repo, while latest is 20190113. Same for gzdoom - 3.5.0 in repo, while latest is 3.7.1
phaebz has quit [Ping timeout: 246 seconds]
orivej has quit [Ping timeout: 246 seconds]
<{^_^}>
[nixpkgs] @adisbladis pushed to master « mpv: Use docutils from python3 »: https://git.io/fhEFx
<adisbladis>
Soo_Slow: You can try to report an issue, though the best is to make a PR :)
<adisbladis>
Typically it's not very hard to perform a version bump. If you need any help just ask here.
<Soo_Slow>
adisbladis, PR - e.g ask for updates directly?
moet has joined #nixos
Diagon has quit [Quit: Leaving]
linkrage has joined #nixos
<adisbladis>
Soo_Slow: PR as in a pull-request on github
<Soo_Slow>
aaaah
erictapen has joined #nixos
amir has quit [Ping timeout: 264 seconds]
jmgai has quit [Ping timeout: 245 seconds]
amir has joined #nixos
b has joined #nixos
hedning has quit [Quit: hedning]
thc202 has quit [Ping timeout: 240 seconds]
cheriimoya has quit [Remote host closed the connection]
thc202 has joined #nixos
dermetfan has joined #nixos
moet has quit [Quit: leaving]
Soo_Slow has quit [Remote host closed the connection]
<yorick>
infinisil: I didn't know that feature, thanks
marsam has joined #nixos
joehh has quit [Ping timeout: 240 seconds]
<joko>
Hey, speaking of nixops, I am setting up some new entries on fstab, but the fstab is not getting updated to the target host. Any ideas what's wrong?
ixxie has quit [Ping timeout: 268 seconds]
<yorick>
joko: how are you setting them?
<nh2[m]>
yorick: in the case of my PR, it's actually blocked on me (as I put into the title, "DO NOT MERGE"). But I agree in general it would be nice if nixops reviews could be quicker
<yorick>
nh2[m]: should be a trivial fix
<joko>
yorick: in filesystems, as attribute sets
<{^_^}>
[nixpkgs] @jtojnar opened pull request #54324 → Vapi setup hook → https://git.io/fhEAb
<yorick>
nh2[m]: maybe it needs more maintainers?
<{^_^}>
[nixpkgs] @dotlambda opened pull request #54325 → docutils: use python3Packages → https://git.io/fhEAx
<nh2[m]>
yorick: perhaps; maybe more people checking out PRs and posting "I have reviewed and run this and it works" would also do a lot. I do not know yet. I would still encourage you to to contribute, as overall the process works and nixops improves
fendor has quit [Ping timeout: 244 seconds]
marsam has quit [Ping timeout: 252 seconds]
phaebz has quit [Remote host closed the connection]
ilya-fedin has joined #nixos
selfsymmetric-mu has joined #nixos
dsiypl4_ has quit [Ping timeout: 246 seconds]
<selfsymmetric-mu>
Hi! A minor contribution of mine just got merged into nixpkgs. Where can I see the nixpkgs master build, so that I know once it the revision appears in the unstable channel?
<gchristensen>
nixpkgs-unstable or nixos-unstable?
<gchristensen>
,unstable = nixpkgs-unstable or nixos-unstable?
<__monty__>
selfsymmetric-mu: You might be looking for this ^
<selfsymmetric-mu>
Hmmmm. I actually didn't know those were different things. I mean `nixos-unstable` in my case.
<selfsymmetric-mu>
The PR is merged but I'm still getting an `attribute missing` error, so I imagine there's a failing build somewhere.
<selfsymmetric-mu>
Since `nixos-unstable` I think is supposed to be the last _passing_ build? I may be missing some nuance here.
<ilya-fedin>
Hello... I'm still trying to get Plasma from Nix worked on non-NixOS. I got to solve the OpenGL related errors, but plasmashell still crashes... I decided to try create the system-path package using packageOverrides. Like this: https://pastebin.com/CeeRtQgP I get `error: cannot convert a function to JSON`. Like this: https://pastebin.com/KnfYqHRb - `error: while evaluating the attribute 'passAsFile' of the derivation
<ilya-fedin>
/nix/store/q6x3d1p3snpl1xd382ps77bs8j6lvwla-nixpkgs-19.03pre165281.7d864c6bd63/nixpkgs/pkgs/stdenv/generic/make-derivation.nix:185:11: value is a function while a set was expected, at /etc/nix/nixpkgs-config.nix:22:12`. What am I doing wrong?
<ilya-fedin>
'system-path' at /nix/store/q6x3d1p3snpl1xd382ps77bs8j6lvwla-nixpkgs-19.03pre165281.7d864c6bd63/nixpkgs/pkgs/stdenv/generic/make-derivation.nix:185:11: value is a set while a list was expected, at /etc/nix/nixpkgs-config.nix:19:26`. Like this: https://pastebin.com/eTLwp5F8 - `error: while evaluating the attribute 'passAsFile' of the derivation 'system-path' at
<aminechikhaoui>
yorick: I'll have a look at those PRs next week. In general I would think having more functional tests would help for the review process for non trivial PRs to be more confident that nothing would break.
iqubic has left #nixos ["ERC (IRC client for Emacs 26.1)"]
<iqubic>
I'll be updating the cockatrice default.nix file, and then having nix load that up, then run the command to start the gui. Which just so happens to be `cockatrice`
<iqubic>
How do I specify that I want nix-shell to use a specific attribute?
<simpson>
However you were doing it before.
<iqubic>
error: getting status of '/home/avi/default.nix': No such file or directory
<iqubic>
So I can't run the nix-shell, as I get that error.
rcshm has quit [Ping timeout: 268 seconds]
<simpson>
Something like `nix-shell -p cockatrice --run cockatrice`? I don't know; your mind is hard to read and I feel like you don't read the manual.
<iqubic>
Yes that is what I'm trying to do.
<iqubic>
But I want to point nix-shell at my local fork of nixpkgs
<ryantm>
When I try to use the Live CD on a new Ryzen 7 Lenovo laptop with Radeon vega graphics, it gets stuck with a white unblinking underscore in the top left corner after selecting any of the install options (regular, nomodeset, debug). Any tips for how to proceed?
<etu>
ryantm: stable or unstable image?
<etu>
I'm thinking that you want a very new kernel and mesa etc. You may need to build your own image for that.
Ariakenom has quit [Read error: Connection reset by peer]
<etu>
afaik most of the modules in the kernel defaults to `m`... But it was a long time ago I built a kernel myself... Back when I used gentoo like 4 years ago.
erictapen has quit [Ping timeout: 246 seconds]
Ariakenom has joined #nixos
<samueldr>
`nix log /nix/store/y5zsk8h32r8x1208xi77a496hcpajkw5-linux-config-4.19.16 | less` and search (/) for DRM_AMD_DC
<samueldr>
(that's a really recent kernel config build from trunk-combined)
<samueldr>
oh, useless use of less as it's already using the pager
<samueldr>
I'm unsure why DRM_AMD_DC_DCN1_0 doesn't show up, maybe it's not part of what would be given as questions?
<samueldr>
I'm not sure why it's not asking for the setting though
<etu>
Beats me...
<etu>
I know that my kernel didn't hit the binary cache on last update due to how long it took to install the update :-)
erictapen has joined #nixos
tikznut_ has joined #nixos
<etu>
(Running the auto updating thingy)
ddellacosta has joined #nixos
ddellaco_ has quit [Ping timeout: 240 seconds]
<ar1a>
i cant find much literature on google/the reddit about nixos. wheres all the cool articles i can learn stuff from :(
martin___ has joined #nixos
<tikznut_>
There is a lot you can find on the nixos website
<ToxicFrog>
ar1a: I generally just use the docs on nixos.org, except where they're missing important things about nixpkgs (which is often) in which case I refer to the source code.
<ToxicFrog>
ar1a: I mean, you know about that site
<ryantm>
Looks like the latest unstable installer with the 4.19 kernel also doesn't work with my laptop
<ToxicFrog>
Not those specific options
<ar1a>
oh, yea
<samueldr>
there are gaps in the *linux* side of the operation, in documentation, where there are assumptions that knowledge for those tools outside Nix/Nixpkgs/NixOS applies
<ar1a>
mm yeah of course but like i found out about home-manager from this irc and thats only because i specifically knew what i wanted, instead of googling "nixos" and finding articles about cool things i never knew i wanted/that it existed
<samueldr>
well, I say *linux*, but I meant *tooling outside nix*; NixOS still is a Linux distro, so there is a bunch of things that hold true, while some are handled in a different way
<samueldr>
and the google juice for NixOS I think is partially hampered by how *nix is often used to talk about unix-like systems :/
tikznut_ has quit [Quit: leaving]
<samueldr>
so if you were to search "nix install gnu make" you might or might not see relevant results (due to darn internet bubbles)
tikznut_ has joined #nixos
<ar1a>
i would probably search "nixos install gnu make" instead
<srhb>
ar1a: I thought you were on nixos-unstable, not nixpkgs-unstable? :)
<ar1a>
i...
<ar1a>
what's the difference ;_;
tikznut_ has joined #nixos
<srhb>
ar1a: nixos tests mainly.
<srhb>
(Incidentally the ones that tend to hold up channels)
<ar1a>
is discord in a channel yet gosh d*rn
erictapen has joined #nixos
<srhb>
ar1a: It'll be in nixpkgs-unstable for sure
<srhb>
ar1a: Here's a trick: nix eval -f channel:nixpkgs-unstable discord.name
<ar1a>
my nix laptop is unavailable for 10-20 minutes so ill keep it in mind thanks
<ar1a>
lol i put it in discord so i would remember
<ar1a>
i wont be able to see it 🙃
<iqubic>
It's not in nixpkgs-unstable yet.
shibboleth has quit [Quit: shibboleth]
<MichaelRaskin>
ar1a: web client!
<obadz>
primeos: do you think fscrypt is ready for show?
<srhb>
iqubic: Yes it is.
<iqubic>
Oh, when did that happen?
<worldofpeace>
I think if we can't contact discord to achieve a contact point for upgrades, maybe it should be in a communtiy maintained channel/overlay.
<srhb>
iqubic: Several days ago I think?
<ar1a>
its in nixpkgs-unstable but not nixos-unstable. are those two channels interchangable and i can just swap to nixpkgs?
<etu>
Maybe one should build an overlay for discord that via travis-cron automatically does builds every 24 hours to update to latest version so people can run that overlay from master or something...
<srhb>
ar1a: All channels are theoretically "interchangable" -- but if you go from nixos to nixpkgs, you're losing out on all of the integrity NixOS integrity tests before your channel bumps
<etu>
But I don't care about that application so I'm a bit meh about it
<ar1a>
literally all i need to do to update is change the version line in the .nix file. is there some way to do that from configuration.nix?
<ar1a>
.override or something
<srhb>
etu: Yes, I'd recommend doing something similar for all sorts of applications that NEED fast updates, really.
<etu>
ar1a: .overrideAttrs
<srhb>
ar1a: overrideAttrs, but you'll need to override at least the entire src.
<srhb>
ar1a: The substitution has already taken place at the time of your override.
<ar1a>
not on my current channel
<ar1a>
(nixos-unstable)
<srhb>
ar1a: As in, ${version} is already substituted into src
<etu>
srhb: Yeah, me and adisbladis did that for exwm, it builds every 24 hours on travis and grabs the latest master: https://github.com/adisbladis/exwm-overlay -- and I run it directly from that master as an overlay :-)
<srhb>
ar1a: If you override just version using overrideAttrs, it won't affect name nor src.
<srhb>
etu: Neat! :)
<srhb>
> rec { version = "1.0"; name = "foo-${version}"; } // { version = "2.0"; } # This is a simplified demonstration of what'd happen, ar1a
<{^_^}>
{ name = <CODE>; version = "2.0"; }
<ar1a>
lol
<srhb>
> (rec { version = "1.0"; name = "foo-${version}"; } // { version = "2.0"; }).name
<{^_^}>
"foo-1.0"
<srhb>
Yeah, sorry, forgot it wouldn't evaluate that deeply :)
o1lo01ol1o has quit [Remote host closed the connection]
<srhb>
The point is, we're changing the _result_
o1lo01ol1o has joined #nixos
<srhb>
Thus we end up not touching name again (nor src)
<clever>
i prefer putting the version= into a let block, so its more obvious it cant be changed
<selfsymmetric-mu>
clever: What a beautiful website!
<clever>
iqubic: and use nix-store --query --roots to find the roots
<srhb>
iqubic: (I prefer the term "delete" here since "installed" tends to imply it's used by some profile)
<selfsymmetric-mu>
clever: That's exactly what I needed.
<iqubic>
srhb: Thanks
<iqubic>
clever: Do I need to give a path to that query command?
<clever>
iqubic: yes
<iqubic>
path to where?
jboy has joined #nixos
_kwstas has joined #nixos
_kwstas has quit [Remote host closed the connection]
<srhb>
iqubic: /nix/store/*yourpackage*
_kwstas has joined #nixos
<srhb>
/nix/store/*yourpackage*/ more like..
_kwstas has quit [Client Quit]
<iqubic>
I see.
<iqubic>
Is there a way to go from attribute name to nix-store path?
<clever>
iqubic: nix repl, "${pkgs.hello}"
Dagger has quit [Excess Flood]
<iqubic>
That's failing with `undefined variable pkgs`
Dagger has joined #nixos
<clever>
iqubic: give it something to import, like nix repl '<nixpkgs>'
<srhb>
Or something like... nix eval nixpkgs.hello.outPath
<iqubic>
what sort of a path is /nix/var/nix/profiles/system-188-link?
<{^_^}>
[nixpkgs] @avdv opened pull request #54339 → installer/tools: Avoid duplicate LUKS device entries in hw config → https://git.io/fhut3
<iqubic>
That's the root of the thing I want to delete from my nix-store.
_kwstas has joined #nixos
<srhb>
iqubic: You probably do NOT want to do that
<iqubic>
Why?
<iqubic>
Why is that?
<srhb>
iqubic: Unless you're not using that system generation
<iqubic>
oh.
<clever>
iqubic: what does `ls -lh /nix/var/nix/profiles/system` return?
<srhb>
iqubic: It is the symlink to your system generation 188. That is, the gc root that keeps it alive
<primeos>
obadz: Not sure... I tried it out just for fun (but only for testing purposes) but unfortunately the development seems pretty slow ATM (not sure when there will be a stable 1.x release)
<srhb>
iqubic: If you're on a later (or older generation) you can.
<ottidmes>
I am trying to port a Haskell implementation of a fast repeat function, which uses quoteRem, but I guess I can just define that with builtins.div alone: quotRem = x: y: rec { quot = builtins.div x y; rem = x - y * quot; }
ilya-fedin has quit [Quit: Connection closed for inactivity]
<infinisil>
Neat, I wonder whether it handles negatives as haskell does
worldofpeace has quit [Remote host closed the connection]
<infinisil>
It totally does
<iqubic>
so I installed a thing using: `nix-env -f /home/avi/nixpkgs -iA cockatrice` where /home/avi/nixpkgs is a local clone of nixpkgs.
<iqubic>
how do I delete that now?
<__monty__>
iqubic: I think nix-env can only delete by name.
<iqubic>
so?
<__monty__>
nix-env -q to find the name nix-env -e name to remove it.
hedning has quit [Quit: hedning]
<iqubic>
Cool.
<infinisil>
,fancy-uninstall
<{^_^}>
Fancy way to uninstall packages, needs fzf installed: nix-env -q | fzf | xargs -I{} nix-env -e {}
rpg has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<iqubic>
unfortunately it's too old for my purposes. So I'm trying to update it to use the 2.6.2 release from here: https://github.com/Cockatrice/Cockatrice/releases but I'm not sure how to do that.
<iqubic>
I have a fork of nixpkgs on my drive already, so I can work with that.
<ixxie>
anybody here use LessPass?
<ixxie>
it seems to fit the spirit of the Nix community somehow...
<{^_^}>
[nixpkgs] @ar1a opened pull request #54342 → vdmfec: init at 1.0 → https://git.io/fhuqi
sanscoeur has joined #nixos
<ar1a>
hey thats me!
<infinisil>
ixxie: Almost, I sometimes do `pass | less` to search for passwords :P
<ixxie>
lol
<worldofpeace>
gopass ftw :P
<iqubic>
kpcli
<iqubic>
keepass
<ixxie>
Less
<ixxie>
LessPass is stateless, i.e. works by having a pure function hash the domain, the username and a master password
<ixxie>
interesting idea but I wonder how it is in practice
<ixxie>
because it could be very annoying
<MichaelRaskin>
And then you need to know which braindead password rule which domain has implemented
<infinisil>
ixxie: Oh, I read about that, doesn't sound too promising
<MichaelRaskin>
I think even length limits are not always intersecting
<infinisil>
Especially when you need to change the password
<ixxie>
MichaelRaskin: they have a database for tracking those rules
<MichaelRaskin>
If noone ever changed their braindead policies to different, still braindead ones
<laas>
has anyone here gotten IBus to work with sway?
<laas>
I can get the process running but it isn't actually used
<disasm>
woohoo! new laptop is mostly setup :) still have to tweak window manager stuff, like the infinitesimally small cursor I can only see if I move to the corner of the screen and slowly move it where I want it :) but I'm back in i3 so mouse isn't as important.
davidak has quit [Quit: Leaving]
<disasm>
laas: are you using sway-beta?
<laas>
yeah
<laas>
of course
<disasm>
how do you start it? just login on console and run sway?
rcshm has joined #nixos
<laas>
wait
<laas>
actually
<laas>
it works now, nvm lol
<laas>
I have no idea why though, perhaps it was some hidden thing I had broken
<laas>
thanks anyway though
Cale has quit [Remote host closed the connection]
<disasm>
I'm more asking because I was thinking of trying it out :)
rcshm has quit [Client Quit]
<laas>
Ah, well I have the input method set up in my configuration.nix file, and I've set the necessary environment variables in my config.fish file (`GTK_IM_MODULE`, `QT_IM_MODULE`, `XMODIFIERS`), and then I just do `ibus-daemon -dr`
spirou1415 has joined #nixos
<laas>
this didn't work for me before though, so it may not work for you either
<laas>
it may just have been me being stupid though
jasongrossman has quit [Ping timeout: 244 seconds]
hphs^ has quit [Ping timeout: 246 seconds]
knupfer has quit [Ping timeout: 250 seconds]
* samueldr
gumbles
<samueldr>
was reviewing a PR and got ninja'd by veprbl
<ar1a>
lol, mine? :P
<samueldr>
right :)
<samueldr>
I'll share the only thing that wasn't part of their review here:
<samueldr>
>> I'm prefacing this by a quick note: sometimes PRs can be left unseen and unanswered for a while, right now I'm looking at it mainly because I've seen you react to it on IRC. If it happens, after a while you can always ask/ping specific maintainers which could know more about the change, ask again on IRC or on the [NixOS community discourse](https://discourse.nixos.org/).
<worldofpeace>
samueldr: You should define that with infinisil bot :)
<samueldr>
something is already for that
<samueldr>
,pr
<ar1a>
ill post a message here if it hasn't been merged in the next few days, thanks for the headsup samueldr
<{^_^}>
0. Sorry. 1. If you complain about PR being in limbo, always provide a number («I have an unmerged PR (#23924)…»), bot will print PR title and it might bring you an impulse-merge. 2. Please look up who commits similar work and ping these people (or request-review), keeping up with all the PRs is not really possible.
<Church->
Yeah do we have a guide for PR's? Wanna upstream some packages.
<Church->
samueldr: I'm gonna have like 50+ fracking packaged to PR.
<samueldr>
Church-: great, you're gonna face 50 review boards :D
<samueldr>
(mostly joking here)
<samueldr>
since you have multiple packages to PR, you probably have dependencies required for others; what you might do is to one PR per "leaf" package, with their dependencies included in the PR, and always one commit for each package as chapter 14 describes
davidak has joined #nixos
<ar1a>
50? wow i struggled to find one that wasnt in nixpkgs lol. seems to have *everything*
<iqubic>
So I want to update a package that exists on nixpkgs. I want to have a newer version of the package than what's online now.
ddellaco_ has quit [Remote host closed the connection]
<ar1a>
also, my .git is 900MB . 160k commits is insane
<iqubic>
I want to update a thing in my local nixpkgs and then install that to get the binary on my $PATH.
<ar1a>
where is the code that creates the nix isos and whatnot?
<iqubic>
etu: I have updated the default.nix
<iqubic>
How do I test if this builds properly?
<obadz>
primeos: I'm playing with e4 encryption but oddly I'm unable to revoke my access on a directory by dropping the key—it's not letting me create new files but I can read the ones I created (!!)
<obadz>
must be doing something wrong
<iqubic>
etu: I have gone into my clone of nixpkgs and edited the default.nix what now?
<{^_^}>
[nixpkgs] @marsam opened pull request #54343 → z-lua: init at 2019-01-19 → https://git.io/fhuYt
<iqubic>
etu: If this works, should I submit a PR?
hedning has joined #nixos
<etu>
iqubic: yeah, sure :)
<{^_^}>
[nixpkgs] @thoughtpolice opened pull request #54344 → postgresql: minor rearrangements → https://git.io/fhuYm
<ar1a>
samueldr: guess i wont be bothering anyone about that pr anymore :D
<samueldr>
sure, luck of the draw :) the message was for your future contributions
<iqubic>
,tofu
<{^_^}>
To get a sha256 hash of a new source, you can use the Trust On First Use model: use probably-wrong hash (for example: 0000000000000000000000000000000000000000000000000000) then replace it with the correct hash Nix expected.
<iqubic>
Doing that right now.
<samueldr>
(also good execution, ar1a, there wasn't much to say about your PR)
<ar1a>
how does nix hash? nix-prefetch-url and sha256sum produce different things
<iqubic>
do people generally use the TOFU method?
<srhb>
ar1a: Different algorithms, unpacked/not, ...
<Church->
samueldr: All separate packages no shared dependencies except two packages one a dependency of the other.
<srhb>
ar1a: Use TOFU, it's the right thing always :)
<srhb>
iqubic: Yeah.
<Church->
ar1a: Was porting a bunch of Kali utilities.
<samueldr>
Church-: then I sense a bunch of PRs in your future :)
<ar1a>
oh yeah fair nuff
<Church->
Wonder if I can write a meta-package to cover installing them all?
<Church->
pie___: ^ Happy? :p
<iqubic>
Compiling from source takes a fair bit of time.
<samueldr>
ar1a: also, base32 sha256 sums by default
<srhb>
I guess we can transition to prefixing the hash algo soon in most functions..
jasongrossman has joined #nixos
<pie___>
Church-, hehehe
acarrico has quit [Ping timeout: 268 seconds]
<pie___>
Church-, nice ^.^
<pie___>
lassulus, ^
<pie___>
getting $random pentesting tools in nixos
leothrix has quit [Ping timeout: 250 seconds]
<Church->
I mean not random. I'm slowly porting every tool from Kali not already here.
<Church->
Wonder if I can package wordlists as well?
<iqubic>
etu: So I got the package built and there a result symlink..
<iqubic>
What do I do now?
acarrico has joined #nixos
freeman42x has joined #nixos
<Church->
iqubic: nix-env -i ./result
<iqubic>
Alright, I'm really impressed.
<iqubic>
this qt program was able to find out what OS I'm using.
<MichaelRaskin>
Hm, I might be misremembering what they dropped
hphs^ has quit [Ping timeout: 250 seconds]
<MichaelRaskin>
Hm, strange…
ddellaco_ has joined #nixos
<{^_^}>
[nixpkgs] @pacien opened pull request #54345 → epkowa: add s80 series driver → https://git.io/fhu3E
<martin___>
There's a "gd-2.2.5-bin" and a "gd-2.2.5-dev" in my store
<clever>
martin___: split outputs
<clever>
also, nix-env will never install the -dev
<clever>
,libraries martin___
<{^_^}>
martin___: Don't install libraries through nix-env or systemPackages, use nix-shell instead. See https://nixos.wiki/wiki/FAQ/Libraries for details.
lopsided98 has quit [Ping timeout: 246 seconds]
lopsided98 has joined #nixos
<{^_^}>
[nixpkgs] @veprbl opened pull request #54346 → mujs: enable on darwin → https://git.io/fhu3a
ddellacosta has quit [Ping timeout: 245 seconds]
<martin___>
I was installing gd with nix-env just to see what happens to the store
<martin___>
these paths will be fetched (0.03 MiB download, 0.13 MiB unpacked):
<martin___>
installing 'gd-2.2.5'
<martin___>
nix-env -iA nixos.gd ~
<clever>
martin___: the gcc that comes with nix will also never search the location nix-env affects
<clever>
martin___: you must use nix-shell for gcc to find the libs
<primeos>
obadz: Unfortunately I can't help you with that (didn't use it much lately), but IIRC I also experienced such strange behaviour (the attribute has the -experimental suffix for a reason)
<primeos>
If I have time tomorrow I could try to reproduce it
<martin___>
clever: Sorry, I might have accidentally misled what my point is
<{^_^}>
[nixpkgs] @veprbl merged pull request #54346 → mujs: enable on darwin → https://git.io/fhu3a
<{^_^}>
[nixpkgs] @veprbl pushed to master « mujs: enable on darwin »: https://git.io/fhusY
<laas>
how would I go about using a local build of sway in addition to the one I'm currently using?
<laas>
I can see that I can change the package that `programs.sway-beta` uses for sway, but I don't want to change it, I want to add another version of sway
<laas>
I don't think I can install it the normal way, since it then won't have the necessary privileges, right?