<{^_^}>
[nixpkgs] @xeji pushed commit from @spacekitteh to master « Fix sc2-headless map hash (#56320) »: https://git.io/fhb2N
<samueldr>
fpletz: :/
o1lo01ol1o has quit [Ping timeout: 245 seconds]
<samueldr>
right the PR is seemingly in a confus(ed|ing) state
jluttine has quit [Ping timeout: 255 seconds]
_d0t has quit [Quit: Konversation terminated!]
Zer000 has left #nixos ["Leaving"]
<hyperfekt>
Is there a trick to running the tests? I always get 'error: file `/boot/grub/i386-pc/normal.mod' not found.' when a virtual machine tries to boot :<
ng0 has quit [Quit: Alexa, when is the end of world?]
ddellacosta has joined #nixos
o1lo01ol1o has joined #nixos
<{^_^}>
[nixpkgs] @matti-kariluoma opened pull request #56328 → candle: init at 1.1 → https://git.io/fhbaL
MichaelRaskin has quit [Quit: MichaelRaskin]
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<ddellacosta>
I cannot for the life of me figure out how to get icons to show up for qt5 apps on my system (no display manager + xmonad). This is all I could find so far--is there any other documentation on this anywhere? https://github.com/NixOS/nixpkgs/issues/13715
<{^_^}>
#13715 (by flosse, 2 years ago, closed): running kde5 applications without kde as desktop manager
jluttine has joined #nixos
vidbina has joined #nixos
<{^_^}>
[nixpkgs] @Infinisil merged pull request #55792 → Allow duosec to be used in nixos as a pam module. → https://git.io/fh7SH
<ddellacosta>
ottidmes: looks like my network dropped out right when you responded (I checked the logs)--sorry! And thanks for that--I'll poke at your config and see if it helps me fix my busted icons
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
eadwu has quit [Quit: WeeChat 2.4]
<rcshm>
hi, how do i reinstall new vscode on my nixos? the package got corrupted after restart and i tried removing it from the configuration.nix, nixos-rebuild switch, nix-collect-garbage. but that didn't help
<rcshm>
when i put the vscode back in and nixos-rebuild switch, the old broken package is still there.
<rcshm>
thanks for your help.
eadwu has joined #nixos
<alex_giusi_tiri>
rcshm: what do you mean by the package being broken?
<slack1256>
Check out the config files it wrote (maybe on $HOME/.config)
o1lo01ol1o has joined #nixos
<ldlework>
trying to write a new module for my system to install browsersync from npm
<ldlework>
apparently I've forgotten how to use nix
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
ym555 has quit [Quit: leaving...]
<ldlework>
I have this line: nodeEnv = (import ./node-env.nix) { nodejs = pkgs."nodejs-8_x"; inherit (pkgs) stdenv python2 utillinux runCommand writeTextFile; };
<ldlework>
And it say I'm calling that node-env.nix module with an unexpected argument `lib`
<growpotkin>
I feel like I've been on such a bad roll with Haskell Packages lately. When I try to use them from nixpkgs they fail for really common dependencies like `containers` or `semigroups`.
<growpotkin>
I can usually use them if I pull the repo but i have to put them next to any project I want to use them in.
<nDuff>
If I have a snapshot of a backup of my Nix store mounted on a subtree, is there a way to merge its contents into the local store that (1) preserves metadata, and (2) avoids completely rewriting the files? (The backup was restored to an alternate btrfs subvol and has had block-level deduplication run; copying with reflinks will take much less time/space than anything that involves a *real* copy).
matix has joined #nixos
xkapastel has quit [Quit: Connection closed for inactivity]
o1lo01ol1o has joined #nixos
Ralith__ has quit [Remote host closed the connection]
Supersonic has quit [Disconnected by services]
Supersonic112 has joined #nixos
Supersonic112 is now known as Supersonic
<matix>
nDuff: does `cp -a --reflink=always` work for you
<matix>
?
<nDuff>
@matix, yes.
vk3wtf has quit [Ping timeout: 258 seconds]
<nDuff>
my concern isn't about the filesystem layer but nix's sqlite datastores
<nDuff>
copying files into place where they don't exist is easy enough, but places where files have different, conflicting values is the more interesting case.
<matix>
well, if the builds are pure then the contents of these files that exist in both places should be the same
leotaku has joined #nixos
<matix>
and if they are not you're basically free to pick any version
<nDuff>
matix, I'm not referring to derivation results, or even, .drv files
<nDuff>
I'm referring to /nix/var/nix/db/
<clever>
nDuff: you want `nix copy`
<clever>
nDuff: `nix copy --from local?root=/mnt --all` i think will do what you want
<matix>
but it won't reflink
<matix>
i think i understand the problem now
<clever>
any copy that doesnt involve nix commands, will just result in a GC deleting the files
<matix>
i wonder if this would work for you: cp with reflinks first, then try `nix copy` and see what happens when nix notices files already exist
<nDuff>
Right. Anything that reads the bytes from point A and writes the bytes to point B undoes the deduplication, unless it uses the reflink-aware syscalls
<clever>
matix: `nix copy` will delete things that are invalid and in the way
matix has quit [Quit: WeeChat 2.4]
matix has joined #nixos
infinii has joined #nixos
Mateon1 has quit [Ping timeout: 245 seconds]
drakonis has quit [Quit: WeeChat 2.3]
<matix>
nDuff: just go with `nix copy`; it will be faster than whipping up a custom one-off script to merge the sqlite databases
<nDuff>
matix, ...I'm really not sure about that. We're talking multiple TB.
<nDuff>
and moreover, I don't have enough *space* to store it all deduplicated
<nDuff>
so any process that requires redoing the dedup will result in an outright failure
<nDuff>
...which is to say, pretty sure I have a one-off script in my future.
<nDuff>
...anyhow, sqlite's ATTACH may make it straightforward-ish.
<nDuff>
err, to store it all *re*duplicated, rather :)
matix[m] has left #nixos ["User left"]
eadwu has quit [Quit: WeeChat 2.4]
eadwu has joined #nixos
<matix>
good luck then
matix has quit [Quit: WeeChat 2.4]
eadwu has quit [Client Quit]
justanotheruser has quit [Ping timeout: 255 seconds]
eadwu has joined #nixos
wedens has joined #nixos
nDuff has quit [Quit: zzz]
infinii has quit [Ping timeout: 272 seconds]
justanotheruser has joined #nixos
etbe has joined #nixos
<etbe>
I've started work on adding SE Linux support to NixOS. What's a good place on a Wiki for documenting this?
eadwu has quit [Quit: WeeChat 2.4]
eadwu has joined #nixos
<simpson>
etbe: https://nixos.wiki/ but it is unofficial. If your work is headed upstream, then you'd presumably want to add it to the manual instead.
<etbe>
simpson: I want it to go upstream, but we still have to decide how exactly it will work first...
domogled has joined #nixos
ottidmes has quit [Ping timeout: 255 seconds]
<etbe>
The implementation of AppArmor with NixOS is fairly obvious in terms of the high level design. SE Linux is not so obvious.
<etbe>
I think that a Wiki that lists what we agree on, and what the options are for the things we don't have agreement on would be useful.
<etbe>
But it's definitely not a manual.
eadwu has quit [Quit: WeeChat 2.4]
eadwu has joined #nixos
eadwu has quit [Client Quit]
<etbe>
I'll create Workgroup:SELinux
eadwu has joined #nixos
softinio has quit [Quit: Connection closed for inactivity]
<rcshm>
alex_giusi_tiri: this is the error i have got - /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1/bin/.code-wrapped: line 35: /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1/lib/vscode/bin/../code: No such file or directory
<rcshm>
i just don't know how to force reinstall vscode or any package really.
<etbe>
simpson: I don't know if what I'm proposing counts as "substantial" given that anyone who wishes can keep using NixOS without enabling it. An opt-out change is more "substantial" than an opt-in change.
wfranzini has quit [Remote host closed the connection]
infinii has quit [Ping timeout: 245 seconds]
knupfer has joined #nixos
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
domogled has quit [Remote host closed the connection]
<elvishjerricco>
Hm. virtualisation.useBootLoader doesn't seem to work at all...
gerschtli has quit [Quit: WeeChat 2.2]
gerschtli has joined #nixos
lejonet has joined #nixos
infinii has joined #nixos
domogled has joined #nixos
knupfer has quit [Ping timeout: 250 seconds]
pbb has joined #nixos
domogled has quit [Quit: domogled]
jackdk has quit [Ping timeout: 255 seconds]
ddellacosta has quit [Ping timeout: 255 seconds]
PeterHK has joined #nixos
<PeterHK>
how to fix `ld: unknown option: -isysroot` on OSX? i get ld errors with ruby quite often (xcode tooling)
simukis has joined #nixos
iqubic` has joined #nixos
iqubic` has left #nixos [#nixos]
<{^_^}>
[nixpkgs] @peti pushed 2 commits to haskell-updates: https://git.io/fhbrt
endformationage has quit [Quit: WeeChat 2.3]
iqubic has quit [Ping timeout: 264 seconds]
growpotkin has quit [Ping timeout: 246 seconds]
domogled has joined #nixos
palo1 has joined #nixos
kenshinCH has joined #nixos
jethro` has joined #nixos
<jethro`>
anyone here working on C++ projects with NixOS? I'm having trouble getting clang-tools to find stdlib header files, and would appreciate some guidance
tmaekawa has joined #nixos
palo has quit [Ping timeout: 246 seconds]
palo1 is now known as palo
<dminuoso>
jethro`: Can you show how you build your derivation?
<jethro`>
that's my shell.nix file, i enter a nix-shell, and I suppose clangStdenv makes clang-tools available
<kyren>
okay, this might not be a question for #nixos specifically, I'm not totally sure, but I'm following this guide: https://nixos.wiki/wiki/IGVT-g
<kyren>
and the result is that I have a baffling libvirtd error with *no* error output, just: 'internal error: process exited while connecting to monitor', and I'm asking in case either anybody has ever used iGVT-g emulation or even just knows how to *find a better error*
<dminuoso>
jethro`: So for completion, can you also show us how you build your C++ file/files and what errors you get?
<kyren>
there's nothing interesting in dmesg or the logs, and I'm out of ideas
kenshinCH has quit [Ping timeout: 246 seconds]
<dminuoso>
jethro`: And how do you open your nix-shell?
tmaekawa has quit [Ping timeout: 245 seconds]
<jethro`>
dminuoso: I simply run `nix-shell` in the root directory of my project, containing shell.nix
<jethro`>
dminuoso: I build my project using cmake: cd build/; cmake -DCMAKE_CXX_CLANG_TIDY=$(which clang-tidy) .. ; make -j4
rauno has joined #nixos
<jethro`>
dminuoso: for clarification, clang++ is can find the stdlib header files and the program compiles. However, clang-tidy is unable to find them, and produces diagnostic errors such as the following: https://pastebin.com/gEdeDgLH
<jethro`>
dminuoso: my current hypothesis is that clang++ is wrapper, while clang-tidy is an unwrapped binary, but I can't find a good workaround
<simpson>
jethro`: Does the problem still happen if you write an ordinary Nix derivation for building your project?
<dminuoso>
jethro`: I wonder, some googling reveals bits about using nixos.llvmPackages.stdenv rather than clangStdenv
<kyren>
jethro`: are you using the `-p` flag to clang-tidy to read the database written by cmake?
<jethro`>
simpson: I haven't tried that, need to write a corresponding default.nix
schjetne has quit [Ping timeout: 255 seconds]
arjen-jonathan has joined #nixos
Makaveli7 has joined #nixos
<jethro`>
kyren: no, but cmake is the one specifying to run clang-tidy, so it should be using the compile_commands.json by default. I tried running clang-tidy outside of cmake specifying the build path with the file, with the same effect.
<kyren>
jethro`: hmmm, okay so it *is* being called with the `-p` flag then, just by clang?
<kyren>
er, sorry just by cmake
<jethro`>
kyren: yes, that's what I think is happening
<jethro`>
dmniuoso: I tried changing clangStdenv to llvmPackages.stdenv, but doesn't change anything
<kyren>
jethro`: then I'm not sure, sorry.. I would *assume* that cmake would be passing in all of the relevant header directories, it must be skipping over the ones that are clang system headers?
<kyren>
does clang-tidy honor any environment variables for places to look for headers?
<jethro`>
kyren: this is why i'm wondering if clang-tidy needs to be wrapped as well, because a quick look at compile_commands.json shows that system includes are not part of the compile command, and encapsulated in the wrapped clang++. https://pastebin.com/gV7M8x31
<kyren>
okay, just to test, in your nix-shell do `echo $NIX_CFLAGS_COMPILE`
<kyren>
what happens if you set CFLAGS to NIX_CFLAGS_COMPILE, does that make the problem go away?
<kyren>
(I'm kind of guessing here)
<dminuoso>
It's times like these that I enjoy having a Haskell gig instead of a C++ gig.
<jethro`>
kyren: what should I be looking out for?
* dminuoso
chuckles
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<jethro`>
dminuoso: it's just a temporary PITA for me, since this is just a school project :P
<kyren>
jethro`: oh, I was using `nix-shell -p clang` to test, so that's probably why, what if instead in shell.nix you set CFLAGS to be `${clang}/include/`
<kyren>
you can do it with shellHook, but there might be an even easier way
<kyren>
but shellHook = ''export CFLAGS="${clang}/include"''; might do it
<{^_^}>
intel/gvt-linux#47 (by alex3kov, 26 weeks ago, closed): qemu process defunct immediately after starting VM
schjetne has joined #nixos
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
kvda has joined #nixos
PeterHK has quit [Ping timeout: 256 seconds]
<kyren>
added that one to the NixOS wiki
IRCsum has quit [Remote host closed the connection]
IRCsum has joined #nixos
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<elvishjerricco>
How are you actually supposed to use the `run-vm` script from `runInLinuxVM` with `--keep-failed`? It expects to have /build and to be able to write to $out
slack1256 has joined #nixos
johanot has joined #nixos
ibrahims has joined #nixos
ckauhaus has joined #nixos
ibrahims has quit [Read error: Connection reset by peer]
ibrahims has joined #nixos
Boomerang has joined #nixos
<sphalerite>
elvishjerricco: I'm not sure you're supposed to use it at all, but maybe you can set the NIX_BUILD_TOP and out env vars to make it work?
zupo has joined #nixos
<elvishjerricco>
sphalerite: Unfortunately the out path and /build are hard coded in a couple of places, one of which is the output path for the VM to write to, meaning replacing that also requires mounting the replacement in the VM
<{^_^}>
[nixpkgs] @dotlambda pushed 3 commits to staging-next: https://git.io/fhb6Q
<Philonous>
It's funny, this keeps happening with erc users. You'd think they'd implement some sort of protection against it
alex`` has joined #nixos
init_6 has joined #nixos
Melkor333 has joined #nixos
adamt has quit [Quit: WeeChat 2.2]
MinceR_ has joined #nixos
simukis has quit [Quit: simukis]
<sphalerite>
alex``: those two things seem unrelated
MinceR has quit [Ping timeout: 240 seconds]
<sphalerite>
alex``: did you put your own stuff in the directories mentioned by nixos-rebuild?
<alex``>
Nope
<sphalerite>
alex``: in that case, I'd try moving the directories mentioned out of the way and rerunning the activation script as root (/run/current-system/activate)
<sphalerite>
actually, `/run/current-system/bin/switch-to-configuration switch` is probably better
<alex``>
Moving all the directories mentioned to a tmp directory?
<sphalerite>
well preferably somewhere you can restore them from if everything goes wrong
schjetne has joined #nixos
<alex``>
sphalerite: it works the first time, but say the same after
<{^_^}>
attempt to call something which is not a function but a set, at (string):218:1
Melkor333 has quit [Quit: WeeChat 2.4]
<dminuoso>
Mmm, I use home-manager, and I get the following error: The option `programs.java' defined in `/home/dminuoso/.config/nixpkgs/home.nix' does not exist.
<dminuoso>
Setting that option system wide in /etc/nixos/configuration.nix works just fine.
schjetne has quit [Ping timeout: 246 seconds]
arjen-jonathan has quit [Ping timeout: 268 seconds]
<sphalerite>
dminuoso: home-manager and nixos have different config option sets.
<{^_^}>
[nixpkgs] @uvNikita opened pull request #56340 → sshd: fix startWhenNeeded and listenAddresses combination for ipv6 → https://git.io/fhbiE
<kyren>
is something like nixpkgs.sambaMaster overridable? I ask because that package is the result of calling samba4.override { ... } and what I want to do is enable printing support for sambaMaster
ibrahims has quit [Ping timeout: 246 seconds]
<{^_^}>
[nixpkgs] @andir opened pull request #56343 → nixos/gnunet: fix typo in `PrivateTmp` parameter → https://git.io/fhbXk
<ivegotasthma>
what's the difference between fetchgit and fetchfromgithub?
ramses_ has joined #nixos
agander has joined #nixos
<Yaniel>
fetchfromgithub knows about the zip archives github provides
<adamt>
ivegotasthma: There's actually two different versions of fetchgit, as well, and I can't remember which is which..
<Yaniel>
so it can just grab those instead of doing a full git clone
<ivegotasthma>
Yaniel: ah, okay
<ivegotasthma>
I'm getting a vim plugin from github, but there is so much variance over the nix code everywhere, that I'm not sure about the best way to do things
emily has quit [Remote host closed the connection]
emily has joined #nixos
mallabori[m] has joined #nixos
orivej has joined #nixos
reinhardt has quit [Quit: Leaving]
<{^_^}>
[nixpkgs] @Ma27 opened pull request #56345 → nixos/sshd: implement support for `Match` groups → https://git.io/fhbX7
tdbgamer has joined #nixos
growpotkin has quit [Remote host closed the connection]
<jluttine>
has something changed recently in nixos-unstable regarding environment variables? inside FHSUserEnv, my postgre DB init has started failing complaining something about bad environment variables. initdb failed: b'initdb: invalid locale settings; check LANG and LC_* environment variables\n'
<jluttine>
i have had this setup for a few years and it has worked every day
<jluttine>
i haven't changed now anything except i just upgraded nixos-unstable branch from a few weeks old point to the recent one
schjetne has joined #nixos
<{^_^}>
[nixpkgs] @hedning opened pull request #56346 → gnome3.gnome-desktop: fix thumbnailing with missing fontconfig cache → https://git.io/fhb13
sam_d has quit [Read error: Connection reset by peer]
o1lo01ol1o has quit [Remote host closed the connection]
sam_d has joined #nixos
rauno has joined #nixos
<pie_>
jluttine, in the interim i guess you could probably use a pinned version of nixpkgs for that one thing?
o1lo01ol1o has joined #nixos
rauno has quit [Remote host closed the connection]
<pie_>
jluttine, alternatively you can downgrade/roll back, if you dont need something from unstable
<jluttine>
pie_: yep. just would like to understand is there something broken in nixpkgs now or has something changed so that i just need to do something differently
<duairc>
So on my NixOS system I have tonnes of fonts installed, but I'm doing some web development at the moment and I want to check how a page looks if a user doesn't have certain fonts installed. It should be possible to rig up some sort of nix-shell, which if I run Firefox inside of it, it won't have those fonts?
<symphorien>
I think even in an empty environment, firefox will still read /etc/fontconfig
<pie_>
i think its a reasonable thing to want to do though, hm. idk much about firefox (font) config.
<pie_>
my second thought is chroots.... symphorien fhsuserenv?
<symphorien>
I don't know how they work, sorry
<pie_>
yeah me neither
sphalerite has joined #nixos
o1lo01ol1o has joined #nixos
<pie_>
duairc, idk if this kind of thing is generally accepted for github issues, but maybe try opening a github issue if you get no answer after a while? xD (im somewhat curious about th)
sam_d has quit [Read error: Connection reset by peer]
sam_d has joined #nixos
<symphorien>
I think the easiest way is to remove the fonts from /etc/nixos/configuration.nix, rebuild, test, and rollback
<andi->
duairc: I run Firefox through bubblewrap. Just pass the X11 socket and your profile to the process. Can post a snippet later if needed.
<andi->
Mine doesn't have access to anything but /nix/store, the socket and my profile
o1lo01ol1o has quit [Ping timeout: 240 seconds]
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
slack1256 has quit [Ping timeout: 272 seconds]
<{^_^}>
[nixpkgs] @yrashk opened pull request #56351 → pugixml: always include pkgconfig file → https://git.io/fhbMa
IRCsum has quit [Remote host closed the connection]
knupfer has joined #nixos
dwdr has joined #nixos
IRCsum has joined #nixos
cfricke has quit [Quit: WeeChat 2.4]
sphalerite has joined #nixos
<andi->
emily: yeah, but it was more a "does it work and if so what does it take" thingy.. Not claiming it solves the X11 issues :)
magnetophon has joined #nixos
<gchristensen>
(you should see the insane stuff that is done to make Dragon work with Linux, the unsecured jsonRPC socket is terrifying. definitely on my to do list to fix that.)
<{^_^}>
[nixpkgs] @xeji pushed commit from @artemist to master « neovim: Update ruby provider to 0.8.0 (#56316) »: https://git.io/fhbDg
Makaveli7 has joined #nixos
<WhittlesJr>
Is there a way to tinker with packaging something with a long build process without triggering the long build each time? The only thing I'm changing at this point is postInstall.
reinhardt has joined #nixos
o1lo01ol1o has joined #nixos
simukis has joined #nixos
o1lo01ol1o has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @xeji pushed commit from @r-ryantm to master « gammu: 1.39.0 -> 1.40.0 (#55317) »: https://git.io/fhby2
slack1256 has joined #nixos
<{^_^}>
[nixpkgs] @dotlambda merged pull request #56312 → Fixes several broken python packages for python-unstable → https://git.io/fhb0u
<{^_^}>
[nixpkgs] @dotlambda pushed 8 commits to staging-next: https://git.io/fhbyy
ibrahims has joined #nixos
hyper_ch2 has quit [Ping timeout: 256 seconds]
rcshm has joined #nixos
ibrahims has quit [Ping timeout: 245 seconds]
<rcshm>
hi, i have got vscode corrupted on nixos. how do i force reinstall the package? - /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1/bin/.code-wrapped: line 35: /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1/lib/vscode/bin/../code: No such file or directory
lfish has joined #nixos
<rcshm>
thanks for your help
<lfish>
hello, I need to set a static ip. Is there some special way to do this on nixos or is it the same as other distros? Btw, I'm running nixos 18.09
shpx has joined #nixos
<srhb>
lfish: See the options under networking.interfaces :)
<srhb>
rcshm: It's unlikely that that's what happened if you're on NixOS (since the store is read-only) -- but nix-store --repair-path /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1
<rcshm>
srhb, i did run the repair-path you suggested but the error is still there though.
<lfish>
srnb: oh! this is something I've been wondering for some time... is there a general way to get the options for these things? Thanks!
<rcshm>
there is no way to force reinstall? srhb:?
<srhb>
rcshm: There _is_ (sort of) but it's unlikely to do anything. The thing about reproducibility is that if it works, it'll work even when you "reinstall". If it's broken, it'll be broken even if you "reinstall"
<srhb>
rcshm: I would guess your problem lies elsewhere, perhaps in local config or environment variables.
<srhb>
rcshm: If repair-path hasn't fixed anything, "reinstalling" won't either.
jb55 has joined #nixos
<rcshm>
it worked before and it just happened after restart.
<rcshm>
what should i do then?
<srhb>
rcshm: I would try launching it from a clean environment, with any local configuration in your homedir moved out of the way.
<rcshm>
which file is that? i am sorry i am just not used to this yet.
<srhb>
rcshm: I don't know, I don't know how VSCode works specifically :)
<srhb>
This is just general advice for "my store seems fine and consistent, but suddenly my application is failing"
<srhb>
That generally means the problem is outside Nix.
<srhb>
rcshm: Are you on some architecture other than x86_64-linux? Have you messed with your LD vars somehow?
<rcshm>
not really.
<rcshm>
i am just using it for code dev and not messing with anything at all.
<srhb>
rcshm: How exactly are you trying to launch code?
<rcshm>
command code .
<rcshm>
that launches it in the current dir.
<ddellacosta>
hello, I'm having trouble with some QT5 apps picking up icons--dolphin and okular for example simply have no app icons, whereas keepassxc (which also uses QT5) is fine. When I do an strace I see that the only icon folder those apps are looking in is hicolor, even though breeze icons are present and otherwise they appear to be using that theme. Anyone know what this may be?
<samueldr>
ddellacosta: I can't help much more, but check either of XDG_CURRENT_DESKTOP or QT_QPA_PLATFORMTHEME
reinhardt has joined #nixos
<ddellacosta>
samueldr: thanks--I don't see either of those set, what values should those be, or where can I read up more on those?
<adamt>
rcshm / srhb: I just verified locally that it seems to work fine when launching it here, so whatever is meant to be in the nix store for vscode does work
<srhb>
adamt: Thanks.
<srhb>
I'm pretty sure it's some environment issue, but I can't thinkof anything in specific.
<samueldr>
ddellacosta: not sure exactly, but google around for what they mean, I think they're somehow relevant, but I'm not sure which is, and how it will be
knupfer has joined #nixos
<ddellacosta>
ah yeah, looks like the Arch linux wiki may have some related tips. Thanks samueldr !
<srhb>
rcshm: Sure, remove it from your systemPackages, rebuild, then use nix-store --delete /nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1, then add it back. But it is unlikely to change anything.
<srhb>
If it _does_ change anything, it's a bug
<sphalerite>
yep
<sphalerite>
oh right the question was if feature freeze is today
<rcshm>
let me try.
<samueldr>
>> This year, we’re aiming for a feature freeze and branch-off on the 25th of February in order to get a release* within March.
<rcshm>
it says cannot delete path '/nix/store/fpzpv94smsskgb11id4rwm79n51kr1s0-vscode-1.31.1' since it is still alive
<rcshm>
srhb:
<etu>
What's the best practice for migrating the /nix folder to a new partition? It's on my / partition today.
<srhb>
rcshm: That should tell you about all the roots keeping it alive
<srhb>
rcshm: Don't blindly delete them all though, some of them might be important generations.
ddellacosta has quit [Quit: WeeChat 2.2]
<clever>
srhb: nix-store --verify-path can test it for corruption
<clever>
if its not corrupt, then there is zero point in deleting and re-downloading
<clever>
--repair-path can aldo delete&redownload, while rooted
<srhb>
It's already verified and repaired, so this is supposed to be an exercise in futility. :-)
<sphalerite>
yeah I'll be doing the stuff in a few hours, I've been looking through what's in the milestone
<srhb>
The only thing I can think of is some weird loader shenanigans.. Or an impurity.
xkapastel has joined #nixos
<rcshm>
srhb: it comes up /nix/var/nix/profiles/per-user/va/profile-11-link
<srhb>
rcshm: Sounds like you also installed it as your user then
<rcshm>
and a few lines like that. how could i know which?
<srhb>
rcshm: (Using nix-env, probably)
<rcshm>
yes
<rcshm>
i tried it after it got corrupted.
mbrgm has joined #nixos
<rcshm>
i took it out of the configuration.nix and tried nixenv
<rcshm>
but took it out again.
ibrahims has quit [Ping timeout: 244 seconds]
<srhb>
I still don't believe it's really corrupted. But you'll have to get rid of the generations that reference it in order to be able to garbage collect it.
ym555 has joined #nixos
<srhb>
rcshm: See `man nix-env` and the section about --delete-generations for details.
ramses_ has quit [Ping timeout: 256 seconds]
<rcshm>
let me check
<averell>
aww man, tensorflow release will be like tomorrow. it's already committed but no binary release :/
<kreisys>
I just updated my nix channels and now there's no binary cache version of cachix; it therefore tries to build it and fails. this got me pondering about how nix channels actually work? I thought that the whole point is to not allow the channel to advance if packages are broken? what's the correct way to handle a situation where a channel update breaks packages I use? just revert the update and wait until it's fixed? use a
<kreisys>
pinned version of nixpkgs just for that package? am I missing something obvious?
<rcshm>
thanks clever, srhb and adamt:
<rcshm>
now it works!
<clever>
kreisys: cachix isnt part of the tested job, so the channel only requires that hydra has tried to build it
<kreisys>
oic
<kreisys>
so the only way to keep my system up-to-date is to have my own hydra infrastructure that tests my system build and produces its own channel?
<srhb>
rcshm:That's worrying. :/
drakonis has quit [Quit: WeeChat 2.3]
<clever>
kreisys: anything that can build to completion, will be in the cache, anything that fails to build, will also fail for you
Neo-- has joined #nixos
Kakky has joined #nixos
<kreisys>
yeah when anything starts building I assume it's going to fail
knupfer has quit [Ping timeout: 264 seconds]
<clever>
kreisys: it can also be unfree software, that hydra isnt allowed to cache
<wolfmantjg123>
I’m going to try again I’ll let you guys know where I get stuck, this is very frustrating
<mdash>
i've used gparted before, but only from a gparted rescue disk
<wolfmantjg123>
Is it normal to struggle installing this distro, so far I haven’t been able to install nixos.
nSeagull has joined #nixos
<emily>
if you're inexperienced with manually installing linux, then certainly
<JonReed>
wolfmantjg123: For me it's the easiest distro to install normally, because once you know what you're doing, everything is/should be reproducible from configuration.nix and it's easy to debug it.
NickSeagull has quit [Ping timeout: 246 seconds]
<mdash>
yeah, i would say nixos may be one of the hardest to install 1 time, but the easiest to install 10 times :)
<wolfmantjg123>
JonReed yes I see the benefits of this distro, will save me from nasty problems down the road, but installing it is painful, I guess I need to get better at the command line.
<mdash>
the problem is that you have to know (some of) what you want before you can do anything at all.
<goibhniu>
wolfmantjg123: you do need to be comfortable working on the command line, editing text files, and understanding that you'll get an error if you are e.g. missing a semi-colon.
<goibhniu>
^understanding^understand
<clever>
in my case, having come from gentoo and LFS, nixos was trivial to install
<wolfmantjg123>
I copied exactly what is in the documentation and it didn’t work, am I missing something.
<JonReed>
wolfmantjg123: What doesn't work specifically? What partitions did you create?
arjen-jonathan has joined #nixos
Kakky has quit [Quit: Leaving]
<betawaffle>
how does nixos know what package a missing command comes from?
<betawaffle>
the whole "You can install it by typing one of the following:" thing
<tilpner>
betawaffle: Hydra builds an index that ships with the channel
<betawaffle>
fancy
<betawaffle>
can i query something to find out what binaries are part of a package?
<wolfmantjg123>
I’m fixing my usb and putting nixos again, is it fine if I use etcher to make it bootable?
<clever>
betawaffle: you can install nix-index, and then use nix-locate to search it
<{^_^}>
[nixpkgs] @LnL7 opened pull request #56377 → nixos-rebuild: add changelog/docs for edit subcommand → https://git.io/fhb7e
<tilpner>
betawaffle: Do you want to find the package for a binary, or the other way around?
<betawaffle>
tilpner: other way around
<clever>
wolfmantjg123: the iso is already bootable on usb, just dd it directly to the device
<tilpner>
betawaffle: Just fetch it and list $out/bin?
<betawaffle>
what's the command for that? i'm still pretty unfamiliar with nix
<wolfmantjg123>
How long does it take to zero a 32 gb usb, it’s taking forever
<clever>
betawaffle: nix-build '<nixpkgs>' -A hello ; ls -l result/bin/
<goibhniu>
wolfmantjg123: it might also be a good idea to use the nix package manager on a distro you're familiar with for a while first
<clever>
wolfmantjg123: you dont have to zero the usb before writing the image
<tilpner>
betawaffle: find $(nix-build --no-out-link '<nixpkgs>' -A goxel)/bin -executable -type f
<clever>
looks like the same package is listed a dozen times
wolfmantjg123 has joined #nixos
ddellacosta has quit [Ping timeout: 245 seconds]
<tilpner>
Anyone familiar with buildPythonPackage? I get "Could not find a version that satisfies the requirement logbook (from matrix-nio==0.1) (from versions: )" even though python3Packages.Logbook is in propagatedBuildInputs
<wolfmantjg123>
Any tips on how to fix a messed up usb
<symphorien>
Maybe Logbook and logbook are distinct packages ?
<tilpner>
symphorien: They both are used with "import logbook"
<tilpner>
Well, "both". I don't think there are multiple
<goibhniu>
wolfmantjg123: what problem are you facing?
<samrose>
MoreTea: were you ever able to resolve that?
<wolfmantjg123>
Goibhnui well I put nixos on a bootable usb, after three hours of trying to install it yesterday I gave up and went to ubuntu, for some reason the usb is not recognized, I just finished zeroing it out, let’s see if I can reformat it
sam_d has quit [Read error: Connection reset by peer]
freeman42x has joined #nixos
<JonReed>
wolfmantjg123: I used to try things in a VM like VirtualBox. The feedback cycle is faster this way. For example you can create a disk there and then you can boot the installer `.iso`, you can also even create a disk for USB to test if you're creating a live USB correctly.
sam_d has joined #nixos
<JonReed>
Also, snapshots make it faster.
<wolfmantjg123>
JonReed: that sounds great, I think I’ll give that a try before trying it on hardware again
<LnL>
have you seen the installation section in the manual? that should get you started, but if you have any specific questions ask away :)
<wolfmantjg123>
LnL: I haven’t been able to install it using the manual, so I’ll post where I get stuck.
infandum has joined #nixos
<infandum>
I'm trying to run taffybar in stable. It installs fine, but running it gives the following error: libgtk-3.so.0: cannot open shared object file: No such file or directory
<srhb>
infandum: Hmm, that change shouldn't have made it into stable I thimk...
<srhb>
infandum: But it does sound like the gtk bug.
infandum has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @mtrsk opened pull request #56382 → Added jq to playonlinux deps → https://git.io/fhb52
infandum has joined #nixos
<infandum>
I added (taffybar.override { packages = with haskellPackages; (x: [pkgs.gtk3 aeson lens-aeson req]);}) to make it find gtk3, but that still did not work
<infandum>
<srhb>
infandum: How are you running it?
kisik21 has joined #nixos
tilpner has quit [Quit: WeeChat 2.4]
<joko>
Hey, if I use patchelf --shrink-rpath on a precompiled binary, the binary is not working anymore, is there a workaround?
<infandum>
srhb: taffybar
<infandum>
srhb: from /run/current-system/sw/bin/taffybar
<wolfmantjg123>
I’m about to try installing nixos again, do I type exactly what is on the manual?
<srhb>
wolfmantjg123: No, it's mostly a bunch of examples and you have to decide yourself on what to do on a number of points.
hyperfekt has joined #nixos
<srhb>
wolfmantjg123: If a point is unclear, do ask.
orivej has quit [Ping timeout: 240 seconds]
<wolfmantjg123>
srhb: ok, for the uefi(gpt) section of the manual do I copy the text exactly or make some changes
<wolfmantjg123>
?
<infandum>
Running ldd on what taffybar points to, grepping for libgtk, I get libgtk-3.so.0 => /nix/store/im3dqhs1mif1jdkpvf0lk7dgrfm2jl7b-gtk+3-3.22.30/lib/libgtk-3.so.0 (0x00007f322a58a000)
<infandum>
So it looks like it is present
<wolfmantjg123>
Section 2.2.1 uefi(gpt)
<aanderse>
wolfmantjg123: you're going to need to decide how you want to partition your disk(s) and have your commands based on that
<hyperfekt>
Does anyone know if there is any other way to set the active nixpkgs beyond nix.nixPath? I'm trying to extend the nixpkgs I get via the channel system without having to build my own nix-channel.
orivej has joined #nixos
<wolfmantjg123>
aanderse: I have no idea that’s why I came here, so the manual is just a reference?
<srhb>
wolfmantjg123: That section is workable as-is iff you agree with that partition layout.
<aanderse>
wolfmantjg123: have you ever installed a linux os before where you didn't use a graphical partition manager?
<{^_^}>
[nixpkgs] @FRidh pushed commit from @mtrsk to master « Added jq to playonlinux deps »: https://git.io/fhb5H
arjen-jonathan has joined #nixos
<srhb>
wolfmantjg123: (and if sda is the disk you want to nuke and use...)
<wolfmantjg123>
srhb: well isn’t sda the usb?
<srhb>
wolfmantjg123: I don't know, you can use lsblk or mount to check, for instance
<srhb>
wolfmantjg123: iirc it was sdb on my last install, but I don't think there's any guarantee of that, really.
ym555 has quit [Ping timeout: 268 seconds]
<srhb>
wolfmantjg123: (You could have other disks, too!)
<aanderse>
wolfmantjg123: the installer also has gparted right on the desktop as a shortcut
<aanderse>
so you could open up gparted and you will probably know exactly what is what pretty quick
<srhb>
That's good advice :)
kisik21 has quit [Remote host closed the connection]
<yl[m]>
I need help deciding something: At work, I'm currently installing Nix in multi-user mode for all machines. It works well, but there are a few issues: a) I hit the https://github.com/NixOS/nix/issues/2523 and had to workaround it b) I hit https://github.com/NixOS/nix/issues/1759 which lead me to believe that a large configuration will eventually break. What would be the cons of moving to single user installation?
<{^_^}>
nix#2523 (by periklis, 15 weeks ago, open): darwin: nix-daemon crashes due to OBJC_DISABLE_INITIALIZE_FORK_SAFETY
<{^_^}>
nix#1759 (by muji786, 1 year ago, closed): Nix on Mac OS 10.13.2
<wolfmantjg123>
Sda is the usb and nvme0n1 is my hardware, so I’m guessing nvme0n1 instead of sda
<wolfmantjg123>
?
<srhb>
wolfmantjg123: Sounds likely.
<aanderse>
wolfmantjg123: ohh you have one of those fun new hard drives! exciting!
<aanderse>
wolfmantjg123: yes, definitely
orivej has quit [Ping timeout: 240 seconds]
<wolfmantjg123>
aanderse: System76
<srhb>
wolfmantjg123: And this is exactly the kind of thought process you will have to apply to fit the manual to your needs. :)
hamishmack has joined #nixos
<wolfmantjg123>
srhb: I hope all this effort is worth it, Ubuntu is much easier to install
<srhb>
wolfmantjg123: Depends on your temperament. Definitely worth it for me..
<simpson>
wolfmantjg123: There's no shame in preferring Ubuntu.
<srhb>
wolfmantjg123: example 2.3 is really great for getting a high level overview of what installation entails (minus partitioning)
<ryantm>
yl[m]: I believe people tried to make multi-user mode the default on MacOS but had to roll it back, probably because of issues like this. Cons of single user installation: harder to enforce sandboxed builds, can't have two users on the same computer using nix
<srhb>
wolfmantjg123: It is basically a birds-eye view of the whole process.
<srhb>
So, if you ignore that the `nano [...]` line contains a bunch of customization, it's really not a lot of steps :)
<yl[m]>
ryantm: sandbox is not enabled by default on Darwin anyway. I do not see a case where laptops will be shared by multiple users at the company. I think that I'll be fine then, right?
zenobit has joined #nixos
<ryantm>
yl[m]: I think so.
<yl[m]>
ryantm: is the sandbox enabled by default on linux non-nixos machines? Say Ubuntu?
<adamt>
wolfmantjg123: The period in my life were i learned the most about Linux, was when I installed and maintained a Gentoo setup. If you want to learn Linux, I'm sure NixOS can do something similar for you.
<ryantm>
yl[m]: You get sandboxed builds with multi-user Nix because the nix-daemon requires it (I think)
<MichaelRaskin>
Daemon can run without full sandboxing
<MichaelRaskin>
The notion of default is complicated because some ways of installation override source code defaults… by default
<yl[m]>
I think I'm just going to have to test my bootstrap script for the mono-repo on NixOS/Darwin/Ubuntu to make sure single user is feasable
<wolfmantjg123>
adamt: ok here it goes, hope I don’t jack up my laptop
<MichaelRaskin>
As for learning Linux — NixOS allows avoiding learning some core stuff by providing a different and consistent way to configure it via configuratoin.nix…
<yl[m]>
MichaelRaskin: most of my co-workers use nix-shell only (through direnv) they don't even change any of the installation
kisik21 has joined #nixos
<MichaelRaskin>
I think it sets sandbox=false
magnetophon has quit [Remote host closed the connection]
<yl[m]>
cool, I'll test it out
<yl[m]>
thank
ddellacosta has joined #nixos
<infandum>
Ah, the issue was it was getting a previous build from .cache
<infandum>
deleting taffybar from .cache fixed it
<wolfmantjg123>
My laptop has 1 terabyte, so will the commands from the manual work fine?
magnetophon has joined #nixos
<yl[m]>
ryantm: any way to convert multi-user to single-user or should I provide my co-workers with a script to fully uninstall the multi-user and install again?
<srhb>
wolfmantjg123: Which ones are you considering exactly?
infandum has quit [Remote host closed the connection]
<srhb>
wolfmantjg123: (Again, the eventual layout is up to you, within certain constraints. The manual gives some decentish minimums)
fusion809 has quit [Remote host closed the connection]
moet has joined #nixos
<LnL>
yl[m]: it needs a bit of configuration, but you can convert an existing installation
<yl[m]>
LnL: good to hear. Any guides I can follow?
<wolfmantjg123>
srhb: I’m using a System76 galago pro with 1 terabyte. I want to put nixos on all of my hardware. Just sure how to change the commands on the nixos manual for this?
<MichaelRaskin>
Maybe provide them with a script that enables sandbox in multi-user mode?
<srhb>
wolfmantjg123: Try reading point 2) of 2.2.1 carefully
<wolfmantjg123>
Any tips would be helpful, thanks guys
Melkor333 has quit [Quit: WeeChat 2.4]
<aanderse>
wolfmantjg123: have you ever used gparted before?
<yl[m]>
MichaelRaskin: sandbox won't help me in this case. Right now I have about 20 people with Nix installed, soon will have over a 100 and I want to make sure that I have least maintenance possible scenario
<wolfmantjg123>
aanderse: no will it be easier than the terminal?
<LnL>
yl[m]: it the daemon runs it will chown the store so make sure to remove/disable it
<betawaffle>
is it possible to prevent a particular kernel module from being loaded, even manually? what nixos option does that?
<ixxie>
how do I search proprietary packages in nix 2.0?
<aanderse>
wolfmantjg123: if you've never done hard drive partitioning from the terminal before then i believe it will be easier for you, yes
<samueldr>
ixxie: the same rules as before applies
<LnL>
yl[m]: what remains is removing build-users-group in nix.conf and chown -R /nix
<ixxie>
samueldr: wasn't there a search command added?
zenobit has quit [Ping timeout: 256 seconds]
<samueldr>
ixxie: yes
<yl[m]>
LnL: if I understand correctly: a) remove the daemon and delete the launchdaemon plist b) remove teh build-user-group from nix.conf c) chown the store to the current user d) test with nix-shell
<yl[m]>
LnL: I'll test it out, thanks!
<ixxie>
samueldr: hmmm doesn't seem to have an unfree flag
<wolfmantjg123>
How would I change the manuals commands for this?
<wolfmantjg123>
?
<LnL>
yl[m]: they will still have a separate default and per-user profile, but that won't interfere with anything
<betawaffle>
wolfmantjg123: is there something wrong with the commands in the manual?
<wolfmantjg123>
Typo 1 terabyte
<samueldr>
wolfmantjg123: the (example) partitioning commands use relative units so the main partition will fill the disk
<wolfmantjg123>
betawaffle: I just want to make sure I don’t mess anything up like last time
<betawaffle>
what happened last time?
<JonReed>
wolfmantjg123: As the manual says the command will fill create a partition which will fully fill the disk (all 1tb) expect for 512 mb in the front and 8gb in the back `parted /dev/sda -- mkpart primary 512MiB -8GiB`
<samueldr>
so, in 2.2.1 it creates a scheme looking like [3: 512MiB (ESP)][1: rest (root)][2: 8GiB (swap)]
<JonReed>
wolfmantjg123: You'll have to decide how much swap out of 1TB disk you want (if any).
<betawaffle>
how much ram do you have?
<wolfmantjg123>
betawaffle: i had to fix my usb after giving up yesterday, zeroing it out and reformatting it takes forever
<JonReed>
wolfmantjg123: 512mb is enough for boot, so you can leave that as it is
<clever>
wolfmantjg123: you dont need to zero or format the USB, writing the disk image is enough
<wolfmantjg123>
samueldr: JonReed ok guys I’m going to give it a go again, thanks for the tips.
<samueldr>
wolfmantjg123: I think I have a suspicion
<samueldr>
wolfmantjg123: which system76 laptop?
<samueldr>
does it have nvme?
<samueldr>
if it does, it won't be /dev/sdX
<wolfmantjg123>
clever: well fixed it somehow
<samueldr>
and if you used the (only) /dev/sdX, you might have clobbered your usb drive
<samueldr>
(I'm guessing in the dark here)
<aanderse>
samueldr: he did mention the nvme device
<wolfmantjg123>
samueldr: System76 galago pro, 1 terabyte, and its using nvme0n1
<samueldr>
oh, right, then sorry :) looks like that's not it
<samueldr>
(I assumed so because of what you said about the usb drive)
<JonReed>
wolfmantjg123: Next time, you can zero the beginning sectors and the end on usb using `dd`, like 1-10MB from each side. There is no reason usually to zero the whole drive and wait for it.
ikitat has quit [Ping timeout: 250 seconds]
dminuoso has joined #nixos
<wolfmantjg123>
samueldr: lsblsk just shows loop0 , sda, and nvme0n1 , so I think I need to use nvme0n1 for the manual?
<wolfmantjg123>
Right?
<betawaffle>
wolfmantjg123: yes
ikitat has joined #nixos
tmaekawa has joined #nixos
<wolfmantjg123>
betawaffle: ok thanks, I’m going to start typing commands, I’ll let you guys know when I get stuck
tmaekawa has quit [Client Quit]
wedens has quit [Quit: Connection closed for inactivity]
<wolfmantjg123>
JonReed: ok thanks for the tip, what would be the command to do that because I have to zero three more usbs that I messed up
<lejonet>
How do I use fetchgitPrivate? (using fetchgitPrivate says it doesn't exist, but I can find fetchgit/private.nix in the nixpkg pointed to by NIX_PATH
<lejonet>
is it just using fetchgit and not prefixing the url with like http:// or ssh://?
xkapastel has joined #nixos
<adamt>
lejonet: What about just regular fetchgit (either one of them)?
<lejonet>
adamt: I tried giving it a ssh:// url and it didn't like me, it tried using the repository owner as a port... (because : delimits stuff for SSH git)
<lejonet>
throwing ssh:// away seems to make it like me better
<sphalerite>
lejonet: yeah I'd say "don't"
<sphalerite>
just use builtins.fetchGit instead
<sphalerite>
it has nice caching and stuff and is much less weird
<clever>
wolfmantjg123: if you have a github account you can also `nix-env -iA nixos.gist` then `gist --login` and then `gist -p file.txt` to upload a file
aanderse_ has joined #nixos
magnetophon has quit [Ping timeout: 264 seconds]
aanderse has quit [Ping timeout: 244 seconds]
<betawaffle>
i was gonna ask if there's a pastebin service that can be used from curl
<bsima>
betawaffle: i like ix.io
<andi->
nc termbin.com 9999 < file
<betawaffle>
both good answers, thanks
nSeagull has quit [Ping timeout: 245 seconds]
johanot has joined #nixos
fendor_ has quit [Ping timeout: 245 seconds]
ibrahims has joined #nixos
<{^_^}>
[nixpkgs] @lheckemann merged pull request #52464 → nixos/fish: generate autocompletions from man pages → https://git.io/fpxdy
<wolfmantjg1234[m>
if someone could help me with that error, nixos manual 2.2.1
<wolfmantjg1234[m>
?
<barrucadu>
I've added my backup shell script to my gist
<clever>
wolfmantjg1234[m: you gave parted 2 devices, you should only give it one device
<averell>
i'm getting "warning: dumping very large path (> 256 MiB); this may run out of memory" on every nixos-rebuild, i'm not sure where the large file is being sourced from?
<ottidmes>
averell: in my experience its caused by an accidental src = ./.; containing more than you intended
moet has quit [Quit: leaving]
<tilpner>
barrucadu: I don't know either, that looks fine
<tilpner>
(Unless you have overlayed a non-nixpkgs duplicity)
<averell>
in my configuration.nix or where?
<barrucadu>
Nope
<tilpner>
There are also no recent changes to duplicity
<clever>
averell: when you run `nixos-rebuild dry-run`, it will give you a .drv file
<clever>
averell: hmmm, let me find a full cmd
<clever>
[root@system76:~]# du -hc --max=0 $(nix-store -qR $(nix-instantiate '<nixpkgs/nixos>' -A system)) | sort -h | tail
<clever>
averell: what does this show near the bottom?
adamt has quit [Ping timeout: 245 seconds]
<ottidmes>
barrucadu: instead of stdenv.lib. in meta everywhere, why not do meta = with stdenv.lib; as is commonly done, "${version}" is the same as just version, no period at the end of the description
<barrucadu>
tilpner: Oh hang on, it's not duplicity giving that error - it's `time $BACKUP_SCRIPT_DIR/host-scripts/$MY_HOST`. That's a Python script using the shebang "#!/usr/bin/env python3"
<averell>
hah. "1.6G /nix/store/cdygjx74xgpra6zbxw9pyqz78l9h7plq-nixpkgs" not sure how i managed that
<ottidmes>
whoops, wrong person
<clever>
averell: you probably have a ./. or /path/to/nixpkgs that contains a .git
<ottidmes>
monotux: some feedback: instead of stdenv.lib. in meta everywhere, why not do meta = with stdenv.lib; as is commonly done, "${version}" is the same as just version, no period at the end of the description
<barrucadu>
duplicity is a python2 thing, so I guess the environment variables nix-shell is setting are breaking my python3 script, hmm
<{^_^}>
[nixpkgs] @dtzWill pushed to master « llvm-polly,clang-polly: Add llvm/clang variants w/polly support (#55065) »: https://git.io/fhbFI
<averell>
well i'm building from a nixpkgs clone, but i'm doing that on other machines too without this problem
<monotux>
ottidmes, thanks! I'll update accordingly. still trying to understand how to build this, but just discovered the buildPythonApplication macro which seems fitting
<clever>
ottidmes: what ive noticed helps a lot more, is to open the chrome task manager, and then kill off processes
<clever>
ottidmes: for example, there is a proc for youtube, using 3.1gig of ram, and a 2nd youtube proc using 2.6gig
<clever>
just click "end process" on it, and you get the ram back, and can still refesh those tabs later to use them
<hyperfekt>
I think I'm gonna patch nix-channel to save commit hashes instead of downloading nixpkgs instead
<clever>
hyperfekt: just wait till i'm not online, then everything breaks :P
vidbina has joined #nixos
<betawaffle>
clever: 1000+ tabs? what's wrong with you?
<clever>
betawaffle: good question!
<betawaffle>
how do you navigate them?
<ottidmes>
clever: thats the better solution when you have as many tabs open as you do, for me restarting is hard to beat in how effective and fast/easy it is
fenedor has joined #nixos
<clever>
betawaffle: i have an extension called vimium, T lets me search all open tabs by name
<betawaffle>
ohhhh
<clever>
ottidmes: restarting results in it trying to actively use all 76gig of "ram" at once
nDuff has joined #nixos
matix has quit [Quit: WeeChat 2.4]
<hyperfekt>
clever: haha, it should probably also save the output has of the nixexprs so it can take them from store
<betawaffle>
that's fine if you have 96gigs of ram, clever
<ottidmes>
betawaffle: I used a tab tree plugin when I had < 150 open, but Firefox has better support for such plugins, where I use it extensively, it works really well when searching about a topic, you just do a search, open a bunch of pages and they get all organized under your search and then I just go through them one by one, and those pages in turn have links, so you get a search tree that can easily be understood
<ottidmes>
and closed when done
fendor has quit [Ping timeout: 246 seconds]
<clever>
hyperfekt: thats already what nix-channel does, it fetches the nixexprs.tar.xz, and then "installs" it into the channels profile with nix-env
<clever>
hyperfekt: and the tar's that are provided, include the git rev already
wfranzini has joined #nixos
<ottidmes>
clever: my motherboard can only handle 24G, so I already maxed it, but I dont run any VMs so its enough for me
<clever>
betawaffle: motherboard maxes out at 32g
<hyperfekt>
clever: problem is i wanna apply patches to that and it seems the only way to override the nixpkgs used for the nixos modules is to change the nixpath, which means in turn that i cant access the original version anymore
<clever>
hyperfekt: why do you want to patch the nixpkgs?
_kwstas has joined #nixos
<clever>
betawaffle: zramSwap.enable can automate that
_kwstas has quit [Remote host closed the connection]
<pingveno>
Once upon a time, I was using security.pki.certificateFiles with absolute paths (/etc/nixos/somecertificate.cert). I just tried it again and I'm getting an error about the files not existing. What do?
<clever>
betawaffle: but that adds even more cpu costs to things
<pingveno>
Is there a way to include the files from /etc/nixos?
<hyperfekt>
clever: so i can apply more than one change to modules that live in different branches
<lejonet>
sphalerite: erhm, it seems like it is caching a little bit too much lol, I've pushed new commits since the first deploy of the package, and bumped the version nr on the package, but its still fetching the old version
<clever>
pingveno: any unquoted path can be used in a nix file, and refer to any file that exists
<clever>
hyperfekt: you can override modules without patching nixpkgs
<lejonet>
sphalerite: but the path is properly called the new version nr :/
<hyperfekt>
clever: replacing a whole module outright means that a) i have to download the whole nixpkgs if it depends on any other files and b) i have to keep pulling upstream changes, when i really i just want to apply diffs
<pingveno>
Ah, I didn't have them unquoted.
<clever>
hyperfekt: for most changes, you can do them without even changing the module file
<hyperfekt>
yeah i did that for my last thing but it's really not pretty, extracting a submodules attrset, patching the options and patching the config and passing to the original module
<{^_^}>
[nixpkgs] @jules2689 opened pull request #56384 → Use Protobuf 3.4 for MySQL 5.7 to fix build in mac → https://git.io/fhbF9
<clever>
hyperfekt: what if you apply a diff to "${pkgs.path}/nixos/services/databases/postgresql.nix" while also doing disabledModules = [ "services/databases/postgresql.nix" ];
<clever>
hyperfekt: and then imports the patched version?
knupfer has joined #nixos
<hyperfekt>
clever: oh i didn't know about .path!
<spacekookie>
So, if I add a new nixos module, what should be the commit message be? Because there's not really a version to put into "init"
<spacekookie>
Or is it just `<nixos/module>: init` then?
<clever>
hyperfekt: basically, there is a path = ./../..; somewhere in nixpkgs
<{^_^}>
[nixpkgs] @matthewbauer pushed to master « vagrant: only use libvirt on linux »: https://git.io/fhbFF
<NemesisD>
would there be some way to evaluate an attribute of a nix expression? i need to get the cabalFlags out of a package to pass to cabal new-configure
<spacekookie>
Hey hyperfekt btw o/
<sphalerite>
dmj`: oops, sorry, didn't see that. I'm about to push release-19.03, could you lock it?
eadwu has quit [Quit: WeeChat 2.4]
<dmj`>
@gchristensen ^
<NemesisD>
i think i may be able to do this with a shellHook...
sepi` has joined #nixos
<sphalerite>
dmj`: nah he's not here atm, I tried a private message
eadwu has joined #nixos
<clever>
NemesisD: double-check `env`, it may already be an env var?
<dmj`>
@clever @sphalerite
<sphalerite>
clever isn't a github org admin is he?
<clever>
sphalerite: only normal push
sepi has quit [Read error: Connection reset by peer]
fenedor has quit [Quit: Leaving]
<sphalerite>
yeah I have that too, I need an admin :p
<NemesisD>
clever: to clarify, my shell.nix uses shellFor with the project's package. from what i understand, this ensures all deps are installed but doesn't actually build the package i'm working on. it would also seem like it doesn't configure the package
<hyperfekt>
spacekookie: oh hi! nice to see you here! :D
ddellacosta has quit [Ping timeout: 245 seconds]
<betawaffle>
it's not march yet
<NemesisD>
clever: so my plan is to add a shellHook "cabal new-configure ${pkgs.lib.concatStrings mypkg.configurFlags}"
<clever>
NemesisD: yeah, that could work
<hyperfekt>
clever: that's definitely a step forward, but still means any changed files have to be specified manually :think:
<clever>
NemesisD: but it hink you want concatStringsSep
<sphalerite>
ikwildrpepper: hey, are you around just now?
ibrahims has joined #nixos
<NemesisD>
clever: good catch!
<dmj`>
anyone here who has used RDS w/ nixops, how do I get the host name from the resources.rdsInstance.<name>.<??> field ?
<{^_^}>
[nixpkgs] @lheckemann pushed 0 commits to refs/tags/19.03-beta: https://git.io/fhbbT
<{^_^}>
[nixpkgs] @lheckemann pushed to release-19.03 « Set channel and version to 19.03 »: https://git.io/fhbbk
ddellacosta has joined #nixos
<hyperfekt>
hell yeah, my PR (#52464) made it in in time for 19.03 thanks to sphalerite :). too bad i didn't get video acceleration into chromium in time but getting my backups running had priority
<rodarmor>
Is it possible to install packages when booted into the live cd? Install commands, i.e. `nix-env -i vim`, appear hanging, but are maybe churning in the background with no output and I just need to wait longer
<clever>
rodarmor: yeah, it should install, but -iA will be faster
<{^_^}>
[nixpkgs] @dtzWill opened pull request #56386 → lua-modules/generated-packages.nix: fix 'lua' to propagate → https://git.io/fhbbi
<{^_^}>
[nixpkgs] @lheckemann pushed to master « 19.09 is Loris. »: https://git.io/fhbbP
immae has joined #nixos
dgarzon has joined #nixos
rprije has joined #nixos
shpx has quit [Quit: shpx]
<dgarzon>
Hi, I am trying to use: pkgs/development/tools/build-managers/bazel/bazel-deps/default.nix on Linux, but it fails to build. I think the issue is that `bazel-deps` needs to download things from the internet, and the sandbox prevents it from doing so. Is there any way to disable the sandbox for this particular derivation?
<hedning>
dgarzon: --no-sandbox should do the trick (you need to be a trusted user for it to work on NixOS IIRC)
vk3wtf has joined #nixos
ddellacosta has joined #nixos
simukis has quit [Quit: simukis]
dcol has quit [Ping timeout: 258 seconds]
PLPD-Bot has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @acyuta108 opened pull request #56387 → cypress: init at 3.1.5 → https://git.io/fhbbp
ikitat has joined #nixos
stepcut_ has quit [Remote host closed the connection]
stepcut has joined #nixos
johanot has quit [Quit: WeeChat 2.2]
__monty__ has quit [Quit: leaving]
knupfer has quit [Ping timeout: 252 seconds]
shpx has joined #nixos
stepcut has quit [Ping timeout: 255 seconds]
<JonReed>
wolfmantjg1234[m: A bit of a late response, but to quickly wipe the device I would use a helper function like: https://pastebin.com/e8pQ10b4
<JonReed>
wolfmantjg1234[m: Otherwise, you have to calculate where to seek in `dd` command or accidentally do something wrong.
<JonReed>
I usually have a file which I can source `source file.sh` which contains a bunch of helpful functions and aliases.
<JonReed>
This function, for example, will echo first the dd commands with all parameters calculated and ask if you want to execute them, not to accidentally make a mistake or wipe something you don't want
<clever>
JonReed: there is also the wipefs command
<hyperfekt>
oh geez im dumb. i guess i can just disable the top-level module? cc clever
<clever>
hyperfekt: the top-level module?
<hyperfekt>
well the module equivalent of all-packages.nix
<clever>
hyperfekt: you would need to import eval-modules.nix yourself
<matix>
Can someone close #26111 ? It's obsolete after #56277 was merged
<{^_^}>
https://github.com/NixOS/nixpkgs/pull/26111 (by therealpxc, 1 year ago, open): vim-utils: include man pages in the output of vim_customizable.customize, add option to include gvim wrapper
<{^_^}>
[nixpkgs] @xeji closed pull request #26111 → vim-utils: include man pages in the output of vim_customizable.customize, add option to include gvim wrapper → https://git.io/vHGWG
<matix>
thanks
erasmas_ has quit [Quit: leaving]
xeji has quit [Quit: WeeChat 2.2]
amosbird has quit [Ping timeout: 240 seconds]
IRCsum has quit [Remote host closed the connection]
IRCsum has joined #nixos
IRCsum has quit [Remote host closed the connection]
clefru has quit [Ping timeout: 246 seconds]
clefru has joined #nixos
IRCsum has joined #nixos
Dagger has quit [Excess Flood]
Dagger has joined #nixos
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
WhittlesJr has quit [Quit: Page closed]
camsbury has quit [Remote host closed the connection]
camsbury has joined #nixos
eadwu has quit [Quit: WeeChat 2.4]
hchchchchchchchc has joined #nixos
camsbury has quit [Ping timeout: 246 seconds]
ikitat has quit [Ping timeout: 245 seconds]
dgarzon has joined #nixos
<dgarzon>
Hi, is there a way to know in which channel this derivation is cached? https://hydra.nixos.org/job/nixpkgs/trunk/bazel.x86_64-linux. I am currently using the latest `nixos-unstable`, but it still build bazel locally instead of pulling it from the cache!
<edef>
the nixpkgs channel advances without NixOS tests passing
<edef>
it should generally be ahead of nixos
zemm has quit [Ping timeout: 246 seconds]
zemm has joined #nixos
<samueldr>
in past weeks it was the other way around due to the darwin infra being overloaded :)
<samueldr>
though yeah, looks like nixpkgs-unstable is more fresh right now
* samueldr
checks if anything in nixos-unstable is having issues
<dgarzon>
According to http://howoldis.herokuapp.com/, they are all pretty recent and the bazel derivation succeeded 3 days ago!
<dgarzon>
I am trying to avoid building it as it takes a really long time :(
<samueldr>
dgarzon: click on "Help" in the "Actions" column :)
<samueldr>
and use nix-store --realise /nix/store/that-path
<samueldr>
depending on what you want to do with it
* samueldr
looks to see what's up
<dgarzon>
Thanks samueldr :) I am trying to set it up using nix-shell so I don’t think installing it that way will work for me!
ng0 has quit [Ping timeout: 256 seconds]
<dgarzon>
Any suggestions?
<samueldr>
dgarzon: can you paste the output of nix-info? (just in case)
<clever>
dgarzon: if you check the inputs tab on a hydra build, you can see what rev of nixpkgs it used
vk3wtf has quit [Ping timeout: 264 seconds]
<ToxicFrog>
I'm writing a systemd service script (systemd.services.<name>.script). It needs to run a python script that depends on libusb.
ibrahims has joined #nixos
<clever>
ToxicFrog: one sec
<ToxicFrog>
Is there a way to tell nixOS that the script depends on python36Packages.pyusb, or do I just need to put `nix-shell -p python36Packages.pyusb --run path/to/python/script ...` in the systemd unit script?
<ToxicFrog>
elvishjerricco: yes, I use it constantly and I know about .path, I just didn't know if there was anything I could put there that would do the right thing.
<dgarzon>
Thanks clever and samueldr, here is the output of nix-info:
<dgarzon>
I pointed to the right revision, and when I load my shell it still builds bazel from source :(
<samueldr>
(this could have been kept as a single lien)
<clever>
dgarzon: the default mode (with no params) is better for pasting to irc
o1lo01ol1o has quit [Remote host closed the connection]
<clever>
samueldr: he did `nix-info -m`, which uses markdown and many lines
ibrahims has quit [Ping timeout: 245 seconds]
o1lo01ol1o has joined #nixos
<samueldr>
line*
<dgarzon>
Actually, it worked!
<dgarzon>
It is building something else!
shibboleth has quit [Remote host closed the connection]
<dgarzon>
Thank you very much clever and samueldr :)
<ar1a>
what are some more secure ways of sending files over lan? i've just been ad-hocing it with netcat but it feels so wrong
<ar1a>
not ssh based preferably
masterdonx has quit [Ping timeout: 246 seconds]
<samueldr>
both your user's channels (nixos, nixpkgs) seemingly are about 14 days old
<samueldr>
how is your shell.nix (or default.nix) written? do you use `import <nixpkgs>`?
<clever>
ar1a: either scp or nfs is what i use mostly
<ar1a>
clever: nfs isnt really easy to set up on the fly though
<ar1a>
i basically want encrypted netcat i think
o1lo01ol1o has quit [Read error: Connection reset by peer]
<clever>
ar1a: scp for just copying things then
<slack1256>
how often is nixos-unstable updated? every 30 minutes?
johnnyfive has quit [Remote host closed the connection]
<clever>
slack1256: every time it passes testing on hydra
johnnyfive has joined #nixos
<ar1a>
clever: that requires turning password auth back on though
<slack1256>
got it
<samueldr>
slack1256: every ~12 hours it's evaluated by the build farm, and then when everything is finished building AND if tests passes, it will update*
<clever>
ar1a: keypairs
<samueldr>
* (except the small channels, which will only wait for tests to pass and a small subset)
<ar1a>
clever: yeah maybe a scp-only user and a chroot jail for that user
<ar1a>
requires forethought though
<ar1a>
although much easier with nix 😅
<samueldr>
ar1a: what's the use case, looks like you want to ad-hoc send files across computers you don't necessarily own, right?
<clever>
ar1a: you can set an ssh key to only allow scp
<ar1a>
samueldr: yeah pretty much, say at work and you need to send a coworker a file
wavewave has joined #nixos
<samueldr>
yeah, nfs/scp are annoying in that case :/