<greenerworld>
so one quick question, does nix compile every software or does it just download it from the repo's or from a remote url and installs it?
<{^_^}>
[nixpkgs] @worldofpeace pushed commit from @thefloweringash to staging « jackaudio: restrict aften dependency to Darwin »: https://git.io/fj8wW
<{^_^}>
[nixpkgs] @worldofpeace closed pull request #61670 → jackaudio: restrict aften dependency to Darwin → https://git.io/fj8wl
jgt has joined #nixos
jgt has quit [Ping timeout: 248 seconds]
___laika has joined #nixos
endformationage has quit [Ping timeout: 268 seconds]
<gentauro>
anybody using `glances` for system monitoring?
<gentauro>
I noticed that it has the long `nixos` bin paths which are a bit challenging to read (it's the same with `htop`) but, if I start `glances -w` in a RESTful mode, it's only the binary name that is show
<gentauro>
my question is, have anybody in here actually managed to run the terminal version with only binary names?
jgt has joined #nixos
<slack1256>
gentauro: I don't know about glances, but on htop you can configure it to only show the basename of the program and not the whole path
phreedom has quit [Remote host closed the connection]
xelxebar has quit [Remote host closed the connection]
polman has quit [Ping timeout: 255 seconds]
polman has joined #nixos
rfold has joined #nixos
mexisme_ has quit [Ping timeout: 252 seconds]
mexisme_ has joined #nixos
jmeredith has quit [Quit: Connection closed for inactivity]
<jluttine>
I have some dbus-type systemd user-services. When I log in the first time after reboot, they don't work. No errors in logs or anything suspicious, they just don't work, that is, they cannot be reached via dbus. When I log out and log in again, then they work normally. What could be wrong?? (I'm using my own desktop environment creation, so I'm probably missing something there..)
<jluttine>
Those non-functioning services start together with graphical-session.target
<jluttine>
I noticed that some dbus activation environment stuff is run after graphical-session.target has already been started, so this is my first guess
<{^_^}>
[nixpkgs] @livnev opened pull request #61701 → tree-from-tags: init at 1.0 → https://git.io/fj8oj
<eyJhb>
thanks, will try tilpner :) - Btw. if you have any experience with testdisk/mac os x parition recovert etc. I would greatly appreciate it ;)
<tilpner>
I have never used a mac for anything
<tilpner>
If it's anything like recovery from ext4, get a large external drive, run the tool and come back a day later to find a billion files with names that make no sense
<eyJhb>
Haha, I have taken a complete copy using dd, so currently working on that. Didn't want to risk screwing the drive even more.. And currently It basically finds... Nothing :(
hamishmack has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<eyJhb>
And doesn't help, that even if it finds anything, it is encrypted (the whole partition)
<_d0t>
hi! How do I reinstall grub? `nixos-rebuild boot --install-bootloader` goes for a full system update, which isn't something I want right now.
hamishmack has joined #nixos
<rfold>
nix-prefetch-url keeps failing because I'm on a really crappy connection. So I downloaded the file using Firefox instead, and I now have it in my home folder. Can I cause fetchurl { url = <original url>; sha256 = "..."; } to copy this file to the Nix store?
<rfold>
Like, I have the file, and I want to put it in the Nix store /as if/ fetchurl put it there.
polman has quit [Excess Flood]
fendor has joined #nixos
esmerelda has joined #nixos
linarcx has quit [Quit: WeeChat 2.4]
<tilpner>
rfold: You can nix-store --add
<rfold>
tilpner: does the part after the hash in the Nix store path not matter?
<tilpner>
rfold: And then Nix will evaluate the fetchurl, look for the hash, and find it's already there
vid has quit [Read error: Connection reset by peer]
polman has joined #nixos
<rfold>
If I take the hash from the store path and put it in sha256 it says it's the wrong length.
<tilpner>
You're calculating the hash wrong
<rfold>
OK, I did nix-prefetch-url file:///home/r/Garbage/psalm.phar and this seems to work
<tilpner>
IINM use nix-hash --type sha256 --base32
<tilpner>
Maybe with --flat
<tilpner>
Yes, --flat seems right if it's a single file
<rfold>
Thanks, I didn't know that fetchurl only looks at the hash.
<tilpner>
I'm not entirely sure how the name is handled
<tilpner>
So I might have just told a bunch of nonsense
<tilpner>
But I think fetchurl derives the name from the URL
<tilpner>
If that doesn't work, you can at least use requireFile, which is meant to be used with manual downloading of files
<tilpner>
But that would require everyone else to also download it manually
troydm has quit [Quit: What is Hope? That all of your wishes and all of your dreams come true? To turn back time because things were not supposed to happen like that (C) Rau Le Creuset]
<lostman>
Hi folks. Would it be possible in NixOS to set my user's shell to be `nix-shell user-env.nix` so that I can always have some things in scope?
<tilpner>
lostman: Even if it is, why would you want that?
<tilpner>
lostman: You can use users.users.lostman.packages to make packages available to just one user
linarcx has joined #nixos
<lostman>
tilpner: for example if I just want to write some Rust that uses some C library then on NixOS it won't just work. I need to enter nix-shell
<tilpner>
See direnv and lorri
<lostman>
tilpner: even if I install it as system library or user package. It's going to be there but not accessible
troydm has quit [Quit: What is Hope? That all of your wishes and all of your dreams come true? To turn back time because things were not supposed to happen like that (C) Rau Le Creuset]
<lostman>
and I don't want to write little custom configs all the time
<lostman>
it doesn't make sense to me. For messing around with some things
<lostman>
So why not have some sort of "gobal" environment?
<lostman>
nix knows where it is and wrapped cargo knows where to look for it, but only in nix shell
<tilpner>
See direnv and lorri
Synthetica has joined #nixos
<lostman>
yeah, i know of them and I don't want to use them
alexherbo21 has joined #nixos
<tilpner>
Okay
<lostman>
might as well write shell.nix every time I want to do somethign
<lostman>
I'm just curious why is it so hard to have any sort of default?
<andi->
if you want your user shell to always inherit some enviroment just set `users.users.me.packages = [ somepkg ];`?
<lostman>
if I could somehow set up the right cflags and ldflags for the default user shell then most of the things I ever want to do would just work out of the box without piling on any additional tools
<lostman>
andi-: but if I add a library there, will it work with clang?
<symphorien>
lostman: adding something along the lines of `if [ -z $IN_NIX_SHELL ]; then exec nix-shell user-env.nix; fi` in .bashrc might work
<andi->
lostman: not sure why I'd like to have that... Why not just have that lib for the projects that need it? Isn't the benefit of nix being able to have a clean(er) environment?
<infinisil>
And with direnv it's very much automatic
<lostman>
andi-: because that's how every single linux system behaves? I can always write custom configs for my projects, and I do
<infinisil>
No need to manually nix-shell
<lostman>
just want a sane default, like on "normal" linux
<andi->
lostman: every single one besides NixOS (and Guix, …)
<lostman>
I don't see any downside of doing that
<lostman>
so why not?
<lostman>
it should cost nothing, do what I want and make me happy :) should be good enough!
<andi->
I obviously want my computer to just do something different from what seems to be your vision :D
jakkn has joined #nixos
<infinisil>
It will cost you some purity!
<lostman>
andi-: and that's why I'm asking how *I* can do it on my laptop and not telling you that *you* should do it on yours, haha
<infinisil>
lostman: i mean the nix-shell in bashrc solution by symphorien might work, try it
<lostman>
entering nix-shell is a bit slow though. I can probably live with that but is there no better way? can I put something in configuration.nix?
<andi->
lorri
<jakkn>
morning, I'm struggling to understand the section "Testing Configuration Changes in a VM" in this guide https://nixos.wiki/wiki/NixOS:extend_NixOS. What is the nixos-rebuild -I flag for?
<andi->
that should be faster if nix-shell entering is stlow for you
<symphorien>
jakkn: -I is using the clone of nixpkgs passed as argument instead of the default (upstream) one
Ryuul has joined #nixos
<lostman>
andi-: so just exec lorri?
<lostman>
from bash? could try that
<lostman>
but just so I know I covered all the options, is there no way of doing that cleanly from configuration.nix?
<tilpner>
Not an intended way. You can probably hack it together, but that's going to be a bunch of trial and error until it works
<jluttine>
How can I specify a directory in writeTextDir name? I wrote something like writeTextDir "foo/bar/john.doe" ''lorem ipsum'', and nix complains about invalid character '/'
<symphorien>
you would have to source the relevant part of stdenv/setup.sh in your bashrc
vid has quit [Ping timeout: 246 seconds]
__monty__ has joined #nixos
<andi->
jluttine: writeTextDir just means that it puts the file in a directory not a specific directory.
troydm has joined #nixos
<tilpner>
jluttine: You want writeTextFile
is_null has joined #nixos
Ryuul has quit [Client Quit]
<tilpner>
jluttine: It takes a destination argument. See trivial-builders.nix
<jluttine>
ok, thanks. then i don't understand writeTextDir.. i mean, how does it determine both the directory name and the filename if i'm not allowed to use '/' character..
<jakkn>
andi-: where is the default path to '/path/to/nixos/'? I'm on a fresh nixos install and wound up here as I'd like to test my config changes in a VM before applying them. I still struggle to understand what this option does..
<jluttine>
or the directory is just the nix store directory..
<jluttine>
yeah, i suppose so
<lostman>
tilpner: I tried constructing my own NIX_CFLAGS_COMPILE and NIX_LDFLAGS and that almost works. If I do nix-shell -p it picks those up. But from my user's bash it doesn't. Even though both use the same wrapped clang. Not sure what other magic nix-shell is doing so I thought I'll try using it
<jluttine>
ok, thanks, i'll use writeTextFile
<jakkn>
symphorien: gah, I tagged the wrong person above --^
<tilpner>
lostman: Right, I don't know all of what nix-shell does
<symphorien>
jakkn: the default is specified in $NIX_PATH
<Synthetica>
For some reason relative imports don't seem to work in the Python lib I'm trying to package, is this a known issue?
linarcx has quit [Ping timeout: 252 seconds]
<jakkn>
symphorien: ah thanks, do I understand it correctly then that the content of nixpkgs is what defines my system, and I need this option in order to clone the host to VM?
<symphorien>
without this option, the vm is not built with *your* changes to nixpkgs, but the original nixpkgs as defined in $NIX_PATH
<symphorien>
so you don't test *your* changes
<jakkn>
dammit, now I'm confused again
<symphorien>
with nixpkgs and a configuration.nix you can build a system
<symphorien>
nixos-rebuild build-vm does just that, and runs the system in a vm
<jakkn>
right, this I understand
<symphorien>
to test changes to nixpkgs, you can modify nixpkgs, and then build a vm from this copy of nixpkgs
<symphorien>
by default, nixos-rebuild uses the nixpkgs in $NIX_PATH
<symphorien>
this one is read-only
knupfer has joined #nixos
<symphorien>
so you after you modify your own copy of nixpkgs, you must point nixos-rebuild to this copy
<symphorien>
this is what -I nixpkgs=/path/to/clone does
acarrico has joined #nixos
<jakkn>
what is the exact definition of 'modifying my own copy of nixpkgs'? Is it simply changing configuration.nix, or something more?
<tilpner>
jakkn: If you only made changes to your configuration.nix, you don't need to pass a custom nixpkgs path
<jakkn>
I've made no changes to nixpkgs and only want to test changes to configuration.nix, so I should be fine to proceed with the default then. Thanks! :)
<tilpner>
If you're not able to log into the VM, you need to set a password with configuration.nix
<{^_^}>
[nixpkgs] @CarlOlson closed pull request #29478 → roswell: init at version 17.9.10.82 → https://git.io/v5dWz
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
Matthieu` has quit [Ping timeout: 258 seconds]
NinjaTrappeur has quit [Quit: WeeChat 2.3]
ixxie has joined #nixos
NinjaTrappeur has joined #nixos
<{^_^}>
[nixpkgs] @bignaux opened pull request #61709 → gbdk-n: init at unstable-2019-03-14 → https://git.io/fj8PV
<genesis>
:)
<aanderse>
Izorkin: hopefully easier for someone to merge then, thanks
<gentauro>
I just realized that I don't have "python" installed by default :o
<gentauro>
so I tried to add `python3` to my "configuation.nix"
<gentauro>
and then added "python37Packages.pip"
<gentauro>
to install pythong pkgs. The problem I'm getting is something about "read-only" ...
troydm has quit [Quit: What is Hope? That all of your wishes and all of your dreams come true? To turn back time because things were not supposed to happen like that (C) Rau Le Creuset]
<gentauro>
so I'm guessing I'm not doing "python the nixos way". Can anybody give a hint? Thx in advance
troydm has joined #nixos
<pie_>
is there a builtin for getting variables that are currently in scope?
<pie_>
also is there a proper enumeration of builtins/primops/whatever somewhere? even if they arent documented
<pie_>
or do i just have to read all the c sources
<jakkn>
I'm having a hard time getting past the display manager when running result/bin/run-vmhost-vm
_ris has quit [Ping timeout: 245 seconds]
<jakkn>
any pitfalls I should watch out for? I'm trying to run the same configuration.nix as host.
neeasade has quit [Ping timeout: 268 seconds]
<jakkn>
I log on, then screen goes dark and a terminal flashes, before I'm back at login prompt.
<jakkn>
seeing various errors in journalctl about "unable to open display :0"
<gchristensen>
if your xserver's configuration is very specific to your host, like hardware, display configuration, etc.
dycan has quit [Remote host closed the connection]
NinjaTrappeur has joined #nixos
<ekleog>
anyone using docker on nixos?
srid has quit [Ping timeout: 248 seconds]
<ekleog>
I'm trying to `docker-compose up` a relatively trivial http://ekleog.xelpaste.net/xudUWq ; and it's failing with “disk full” while extracting it (while I'm pretty far from full disk)
toppler`` has joined #nixos
<clever>
ekleog: what does `df -i` report?
<clever>
ekleog: what about `watch -d "df -h ; df -i"` while docker-compose is running
<ekleog>
less than 1% everywhere, and `-` on `/` and `/nix` (which are 0 used / 0 free, they're btrfs)
<ekleog>
and the watch doesn't show any moving digits
ThatDocsLady has joined #nixos
<clever>
ekleog: what about the btrfs metadata stuff?
<ekleog>
(it's supposed to be <100MB according to docker-compose)
Synthetica has joined #nixos
<ekleog>
well, to be precise it moves from 31G free on / to 30G free and back to 31G free after having failed
<ekleog>
hmmm so it looks like from the names of the files, it's extracting a bit further (in alphabetical order) on each attempt…
<ekleog>
scratch that I just got contradicted by the last attempt
<{^_^}>
[nix] @edolstra pushed to run-in-pts « Set $TERM »: https://git.io/fj81u
<ekleog>
clever: so with `watch -d btrfs filesystem usage .` I can confirm that btrfs metadata don't go anywhere near to their may (still have 400MB free, and it moves by less than 20MB)
<ekleog>
anywhere near to their max*
<clever>
ekleog: all i can think of then is `strace -o logfile -ff docker-compose up`
<clever>
and then analyze the log to see what failed
<clever>
ekleog: perhaps starting with `grep ENOSPC logfile.*`
mexisme_ has joined #nixos
<ekleog>
clever: hmmm ISTR that doesn't work with docker because docker-compose will be delegating most operations to its daemon, but will try, thanks!
<clever>
ekleog: you can also do `strace -o logfiles -ff -p <something>` against the pid of the daemon
<ottidmes>
clever: I was wondering how do you handle secret files, such as SSH private keys, when e.g. using your kexec install. Do you copy those over manually once you enter the kexec'ed installer?
<clever>
ottidmes: for ssh private keys, i usually just let the server auto-generate random ones
srid has joined #nixos
mexisme_ has quit [Ping timeout: 246 seconds]
<clever>
ottidmes: but i can see how you cant verify the identify of the system you initially ssh into after a kexec, and it could be a malicious vm
<ottidmes>
clever: besides private keys there are other such files, generally what I believe you put in a secrets attrset
<clever>
ottidmes: most of the time, i would just scp them over, or use nixops deployment.keys
<ottidmes>
I originally planned to gpg encrypt them all based on .gitignore file listing the secret file patterns, but in practice its too annoying so I tend to just copy over the files via scp afterwards manually on per-need basis
<ottidmes>
guess I will combine the two, automatically scp the secret files based on the .gitignore listing trick
<clever>
ottidmes: i have plans to add the kexec stuff into nixops
<clever>
ottidmes: so you could just use deployment.keys directly
<ottidmes>
clever: thanks for your input, better to try to keep things simple
<ottidmes>
clever: that would be nice
<ekleog>
grmblwhat openat(AT_FDCWD, "/usr/lib64/python3.7/__pycache__/bdb.cpython-37.pyc", O_WRONLY|O_CREAT|O_CLOEXEC, 0644) = -1 ENOSPC (No space left on device)
wucke13 has joined #nixos
<ottidmes>
clever: right now I tend to just install NixOS directly, and only add nixops when this becomes convenient (so not really they way its mostly intended to be used)
<clever>
ekleog: this process may be inside a docker container, find the chain of ancestors (where clone() returned its pid), and then see if any of them did clone() with weird namespace related flags, or called mount
<clever>
ekleog: maybe just search for `mount(` in all logs, only one thing should be doing it
<clever>
ekleog: i think /var/lib/docker/btrfs/subvolumes/648592fdb43c4df55d71f9592c2d4202d9b83dc027c10d6b5158cda8d0d4aacf is a btrfs subvolume, that it is then using as the rootfs for the docker container
<clever>
ekleog: does that somehow have any special permissions like quotas?
<ekleog>
clever: I never set quotas for my btrfs, so I'd assume there's none
<ekleog>
(esp. none smaller than something that wouldn't fit the uncompressed data of a ~100MB compressed image)
<clever>
cant think of anything else to check
Deathmist[m] has joined #nixos
<ekleog>
`sudo btrfs qgroup show -pcreF .` -> `ERROR: can't list qgroups: quotas not enabled` (and the btrfs subvolume directory does not actually exist after completion of `docker-compose up`)
MmeQuignon has joined #nixos
goibhniu1 has joined #nixos
<xok>
how do I search for packages?..
<xok>
I am doing nix search PACKAGE
<ekleog>
… well I'm just going to run that docker from inside an ubuntu vm I guess
<xok>
but it doesn't alway return correct values...
fusion809 has quit [Remote host closed the connection]
esmerelda has quit [Ping timeout: 246 seconds]
nD5Xjz has quit [Ping timeout: 248 seconds]
MmeQuignon has quit [Read error: Connection reset by peer]
nD5Xjz has joined #nixos
MmeQuignon has joined #nixos
o1lo01ol1o has joined #nixos
detran has quit [Ping timeout: 258 seconds]
xelxebar has joined #nixos
freeman42x has joined #nixos
cybrian has quit [Quit: cya]
xelxebar has quit [Remote host closed the connection]
cybrian has joined #nixos
piratemj has joined #nixos
piratemj has left #nixos ["Leaving"]
xelxebar has joined #nixos
esmerelda has joined #nixos
gspia has quit [Quit: Leaving]
detran has joined #nixos
jmeredith has quit [Quit: Connection closed for inactivity]
goibhniu1 has quit [Ping timeout: 252 seconds]
mmlb2 has quit [Ping timeout: 246 seconds]
linarcx has joined #nixos
gchristensen has quit [Quit: Connection closed for inactivity]
eyJhb has quit [Quit: Bye]
eyJhb has joined #nixos
gspia has joined #nixos
<Ralith>
rycee: it's really frustrating how home-manager's dconf support breaks at the drop of a hat and prevents me from updating without restarting my session :/
<ekleog>
clever: hmmmmm
<ekleog>
when I try to run a VM I'm getting “The I/O cache encountered an error while updating data in medium "ahci-0-0" (rc=VERR_DISK_FULL).”
<Ralith>
starting a new login shell seems like a workaround
<ekleog>
and then I just resume the VM and it moves on a bit before doing it again
knupfer has joined #nixos
<ekleog>
I'm starting to wonder about some hardware failure…
<ekleog>
there's nothing frightening in smartctl --all though
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
orivej has joined #nixos
<ekleog>
… and zsh bails out on me with `zsh: can't rename /home/ekleog/.directory_history/home/ekleog/history.new to $HISTFILE` every second command
<ekleog>
nothing showing any kind of btrfs reaching full disk or metadata showing full disk, though…
<clever>
ekleog: it sounds like the host btrfs is "full" so it cant expand the disk image for the ubuntu vm
<ekleog>
yeah that's kind of what I'm feeling too
<clever>
ekleog: id say you should hit up some btrfs experts in maybe #btrfs, or switch to zfs :P
<ekleog>
but it's not metadata as I can just easily create files [a-f] in $HOME (exceptI have to run each command twice for zsh to let it through)
<clever>
dmesg?
<ekleog>
the day zfs will be supported by mainline I'll consider switching to it :p
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<ekleog>
nothing appearing
<ekleog>
but I do have write errors in journalctl -f too
<alexarice[m]>
Does anyone else loose connection to dns servers on any `nixos-rebuild switch` command?
<ekleog>
will try on #btrfs anyway, thanks :)
<alexarice[m]>
only seem to be able to fix it by restarting
<clever>
alexarice[m]: what are the contents of /etc/resolv.conf before&after?
<alexarice[m]>
will check now
dftxbs3e has joined #nixos
<wucke13>
zathura-pdf-poppler breaks constantly for me, as the download disappears once they have a new version upstream. What can we do about it? Why isn't it cached by Hydra?
MmeQuignon has quit [Ping timeout: 255 seconds]
<alexarice[m]>
wucke13: Cache will be reset whenever a dependency updates I expect
sb0 has quit [Quit: Leaving]
<wucke13>
alexarice[m], I think that's not the cause. AFAIK this is happening becaus of the poppler people changing something, not because nixpkgs changing?
gchristensen has joined #nixos
<alexarice[m]>
wucke13: I was trying to explain why hydra is not caching the build, and this is probably because it caches the old build, then a dependency updates and so it tries to rebuild but this fails and so it has no cached copy for the build you request
<alexarice[m]>
I could be misunderstanding though
<wucke13>
alexarice[m]: Ahh, I see
<wucke13>
alexarice[m]: guess you're right.
<alexarice[m]>
I guess the only option for fixing the package is having a mirror of the file?
<alexarice[m]>
wucke13: if you need a temporary work around you could pin a nixpkgs version and hydra might have a cached build available
<alexarice[m]>
clever: /etc/resolv.conf is wiped out by the switch. It starts with `# Generated by NetworkManager` before and starts with `# Generated by resolvconf` after. Is there a way to fix this
esmerelda has quit [Remote host closed the connection]
<wucke13>
Hmmh
<rycee>
Ralith: Yeah, the dconf modules requires dbus which doesn't work very well under nixos.
<alexarice[m]>
clever: actually reloading sway somehow fixes this for the moment
<Ralith>
rycee: `sudo -u $(whoami) -i` to get the latest env seems to be an effective workaround, would it make sense to wire that in?
<Ralith>
or is this a nixos bug?
<rycee>
Ralith: Not certain how specifically it would help but if it works then I would be happy to merge a PR.
<Ralith>
I have no idea how any of this works but I'm guessing it's something to do with nixos having started a new dbus while my session still has environment variables pertaining to an old one
sm has left #nixos ["ERC (IRC client for Emacs 27.0.50)"]
<{^_^}>
[nixpkgs] @c0bw3b opened pull request #61714 → Package some DNS security tools → https://git.io/fj8DU
<rycee>
On the other hand we try to work around this in HM by starting a new dbus user session just to read the dconf settings.
<rycee>
I'm not sure how to improve things. Generally it works well for me so I haven't had much opportunity for debugging. So if a sudo works then I'd be happy to use it. Not sure how to guarantee that sudo is set up, though.
<Ralith>
do you rarely/never update NixOS without restarting your session?
ddellacosta has joined #nixos
<ekleog>
clever: so if you're interested in #btrfs' verdict: I had 400MiB metadata free but that's not enough, one needs at least ~1G, and I did not have any space left (all the free space was allocated to data), so I needed to balance the btrfs to get some free data space reallocated to metadata :)
<astal>
i changing value in pkgs.derivation (i.e confluence) through the packageOverrides. how can i bind this with service derivation?
esmerelda has joined #nixos
<astal>
enough clues as i need to look
<clever>
ekleog: thats what i initially thought it was
ddellacosta has quit [Ping timeout: 246 seconds]
detran has quit [Ping timeout: 246 seconds]
<ekleog>
indeed, I had been tricked by the fact 400MiB was like 20% of the metadata space, but apparently it's because I created this btrfs a few years ago with an allocator that was wrong for SSDs
<ekleog>
(which resulted in a too small metadata range)
<gentauro>
I'm assuming that by using `cron` this way -> https://nixos.wiki/wiki/Cron it would be a system "crontab" right? How do I make a user specific cron with NixOS?
<clever>
ekleog: ive found that its always too small for nixos, due to how /nix/store/ behaves
<ekleog>
clever: it's not even on my /nix partition :D
<clever>
ah
<clever>
i just zfs for every filesystem on the box :P
detran has joined #nixos
<ekleog>
the /nix partition has like metadata that is 4 times what the metadata for / was, for a smaller fs, so I guess btrfs correctly noticed metadata needed to be big
MmeQuignon has joined #nixos
jboy has quit [Remote host closed the connection]
Glider_IRC_ has joined #nixos
jboy has joined #nixos
mmlb2 has joined #nixos
Glider_IRC has quit [Ping timeout: 244 seconds]
o1lo01ol1o has joined #nixos
domogled has joined #nixos
drakonis has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @symphorien opened pull request #61716 → nixos/all-firmware.nix: only include firware for the host platform → https://git.io/fj8Db
xok has quit [Quit: Leaving.]
Notkea has quit [Ping timeout: 246 seconds]
Notkea has joined #nixos
<genesis>
bonjour Notkea
<Notkea>
hi :o
<rycee>
Ralith: Hmm, I mostly do nixos-rebuild switch and/or home-manager switch while logged into an X session. Typically I don't bother relogging afterwards.
<Ralith>
doing the latter after the former seems to be very dicey for me
xok has joined #nixos
endformationage has joined #nixos
<rycee>
Ralith: On systems where I use home manager as a nixos module then I might occasionally relog because HM is bad at detecting whether a user systemd is already running. Which seems to be related to the dbus user session issue again, user systemd does not work properly unless the dbus is socket activated.
<Ralith>
I'm not using the nixos module at all; been using HM since before that was a thing and never changed
<rycee>
If you're having issues with robustness when running HM standalone I think you'd have even more issues running it as a nixos module :-/
<Ralith>
that is a bit counterintuitive
zupo has joined #nixos
Synthetica has quit [Quit: Connection closed for inactivity]
shibboleth has quit [Quit: shibboleth]
domogled has quit [Ping timeout: 258 seconds]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<symphorien>
You can see it with man configuration.nix
<steve-chavez>
Recently I learn that I can see https://nixos.org/nixos/packages.html through `nix repl`. Was wondering if it was possible to do the same with options.
<symphorien>
I don't know about nix repl
o1lo01ol1o has joined #nixos
tilpner has quit [Quit: WeeChat 2.4]
<steve-chavez>
symphorien: Thank you! Didn't know about man configuration.nix.
xok has left #nixos [#nixos]
knupfer has quit [Remote host closed the connection]
mertdhjkl has joined #nixos
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<v0id72[m]>
hi all
<v0id72[m]>
just wondering if someone could help me out with making changes to the i3status.conf file. As its in a store folder which I cannot edit
<clever>
v0id72[m]: what part do you want to change?
<v0id72[m]>
most of the file :) Like wifi etc - add the weather just really customize it more.
<clever>
it looks like its currently in the same package as the binary, so you need a override against the i3status package
<symphorien>
The file in the store is just a default
<v0id72[m]>
cool thanks a bunch!
<symphorien>
If nothing in your home directory or /etc is found
<clever>
Unable to find the configuration file (looked at ~/.i3status.conf, $XDG_CONFIG_HOME/i3status/config, /etc/i3status.conf and $XDG_CONFIG_DIRS/i3status/config)
<v0id72[m]>
I have been trying on my own to find the answer for far too long. "_
<clever>
the error mentions it looking in 4 places
<v0id72[m]>
yep
<v0id72[m]>
I used locate and it was only in the store folders
toppler`` has joined #nixos
Tiez has quit [Quit: WeeChat 2.4]
knupfer has joined #nixos
tilpner has joined #nixos
<v0id72[m]>
@clever could you point me to a doc about doing the overide? Think I did it incorrectly
<alexoundos[m]>
Is there a way to check integrity/authentity of `narinfo` file? I see there is a signature field, but it seems to target the `nar` file (not narinfo).
<talqu>
hi, inside nix shell i run `cabal new-repl` and the project compiles with no errors. However, outside nix shell, executing `nix-build release.nix` i get errors from the module in which i use lucid library: Couldn't match expected type ‘T.Text’ with actual type ‘[Char]’ In the first argument of ‘value_’, namely..
<talqu>
I was missing {-# LANGUAGE OverloadedStrings #-} in that module. Hmm, cabal new-repl didn't complain about that. Is there any flag I can pass to make cabal new-repl complain the same way nix-build does
<clever>
talqu: it may be differences between your .cabl file, and whatever new-repl reads
<clever>
talqu: nix ignores the new thing
alexherbo213 has joined #nixos
alexherbo21 has quit [Ping timeout: 244 seconds]
hhes has quit [Quit: hhes]
mexisme_ has joined #nixos
djahandarie has joined #nixos
djahandarie has quit [Changing host]
<__monty__>
talqu: Check if you maybe have a .ghc.environment file.
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
oida has quit [Quit: byez]
mexisme has joined #nixos
oida has joined #nixos
mexisme_ has quit [Ping timeout: 246 seconds]
gagbo has joined #nixos
drakonis_ has quit [Ping timeout: 246 seconds]
alexherbo2137 has joined #nixos
alexherbo213 has quit [Ping timeout: 258 seconds]
ng0 has joined #nixos
drakonis_ has joined #nixos
cheriimoya_ has joined #nixos
alp has joined #nixos
mertdhjkl has quit [Quit: Page closed]
oida has quit [Quit: byez]
ondrejs has joined #nixos
oida has joined #nixos
steve-chavez has quit [Quit: Page closed]
<ondrejs>
hello. I want to compile (a haskell project) on my local nixos and then move the binary to a server. When I try that, the server does not understand how to run. Can you please advise? I tried static linking but that turned out horrible (eventually needed a hack with replacing files which I don't know how to do as /nix is read-only https://www.fpcomplete.com/blog/2016/10/static-compilation-with-stack)
alexherbo2137 has quit [Ping timeout: 268 seconds]
<nh2>
ondrejs: what OS is your server?
<clever>
ondrejs: if you are dealing with nix on both ends, then you can just use nix-copy-closure, to copy the binary, and everything it depends on
<ondrejs>
nh2: nixos as well
<ondrejs>
clever: the links seems to be for when you're using nix instead of stack, right? I am not there yet
<ondrejs>
clever: nix-copy-closure looks very promising! Going to try that, thank you
<clever>
ondrejs: any binary you build with bare stack will break when copied to another machine, and can even break if you simply run a garbage collection
<ondrejs>
all right, I see it'd good to convert my project to nix then (ideally preserving stack as it's nice for iterative builds)
<ondrejs>
*it'd be
talqu has quit [Remote host closed the connection]
<ondrejs>
or maybe using stack's nix integration is enough? e.g. stack build --nix
<clever>
ondrejs: that still doesnt put the binary into the nix store, so it has the same problems
MmeQuignon has quit [Ping timeout: 244 seconds]
<clever>
only something you can build with nix-build will be portable to other nix machines
MmeQuignon has joined #nixos
ym555 has joined #nixos
<ondrejs>
clever: I see, nix build it is. Thank you. It's kind of sad I have to maintain two ... dependency managers or what are they
<clever>
ondrejs: you can auto-generate the nix code from the stack.yaml file
<clever>
ondrejs: and you can still get (single-cabal level) incremental building with nix
<v0id72[m]>
clever: still lost as I added it to my configuration.nix file and still no go
vid has quit [Ping timeout: 268 seconds]
<clever>
v0id72[m]: environment.etc."i3status.conf".text = ''config goes here''; i believe
ayerhart has quit [Quit: ayerhart]
<v0id72[m]>
lol im still running in circles it appears :)
hedning_ has joined #nixos
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
lunik1 has quit [Ping timeout: 258 seconds]
lunik1 has joined #nixos
__monty__ has quit [Quit: leaving]
<nh2>
ondrejs: use stack2nix. That way you can continue to incrementally build with cabal, and build stuff with nix easily using your normal lts resolver
<nh2>
ondrejs: sorry, I meant "incrementally build with stack"
<ondrejs>
nh2: thanks, that's what I found. I've been compiling it the past 15 minutes :-)
<{^_^}>
[nixpkgs] @Izorkin opened pull request #61722 → nginxModules.pinba: init at 13.05.2019 → https://git.io/fj89D
<cizra>
If a package produces 2 executables, one of which is only useful during building the packaging (it downloads non-code game assets), how should I manage it? I wouldn't want to install it in system.
<cizra>
I suppose I could compile it in a preBuild hook, but it looks very inelegant.
<nh2>
cizra: couldn't you just skip copying it to `$out`?
<nh2>
so that the built package doesn't have it?
<cizra>
nh2: Well, my main package gets built by executing scons on project root. The helper/downloader gets built by running scons in a subdir. Should I make separate nix expressions for building them?
<nh2>
cizra: I'm not super sure, but you can also delete stuff from `$out` in e.g. `postInstall` if you just want to get rid of the executable at the end
esmerelda has joined #nixos
<cizra>
Yeah, but how should I make it get built in my nixpkgs?
<v0id72[m]>
clever: seems like nothing is going well. Just wondering could I not just make it a the super package its own .nix file and add that into the import?
<clever>
v0id72[m]: environment.etc should be the simplest way, that will just create a file in /etc/
<clever>
v0id72[m]: what happens when you try that method?
<v0id72[m]>
after rebuild check and the file was not present
<clever>
v0id72[m]: can you pastebin your configuration.nix?
<v0id72[m]>
sure give me a min to fix it :)
lambda-11235 has joined #nixos
gchristensen has quit [Quit: Connection closed for inactivity]
<samueldr>
it's bound to make a diff, with the added stuff; maybe you can then just include that bit into the list in your config, depending on how the emacs infra works
ondrejs has quit [Quit: Page closed]
<samueldr>
I did something similar for vim stuff, updated the generated set, added what was missing, used the diff to know how to import it
fenedor has joined #nixos
MmeQuignon has joined #nixos
<zeta_wolf>
samueldr: ok, sounds good, i setup nixos to do automatic upgrades, so that is fine, the only thing i haven't done yet is set emacs to do automatic upgrades, i'll try that to see if the package will be recognized afterwards, thanks for the help
<zeta_wolf>
samueldr++
<{^_^}>
samueldr's karma got increased to 87
fendor has quit [Ping timeout: 244 seconds]
esmerelda has joined #nixos
rfold has quit [Quit: WeeChat 2.4]
<marek>
any way to generate a configuration nix file from a drv?
gchristensen has joined #nixos
<nh2>
marek: what exactly does that mean? A configuration.nix file is a text file like any other, so you sure can write a derivation that generates one
<nh2>
but maybe that's not what you mean
zeta_wolf has quit [Remote host closed the connection]