<Raito_Bezarius>
Is it expected that cptofs on QEMU ARM virtualization can take hours to complete?
<Raito_Bezarius>
(in the ext4 image creation process)
<Raito_Bezarius>
(I tried to gdb -p and it seems like it's running something, a very long syscall because strace shows no output for some time)
<eoli3n_>
qyliss as its simple to use it on nixos, i wanted to give it a try
<eoli3n_>
but is this really usefull under nixos as we can rollback
<colemickens>
rollback just means system config, not data
<eoli3n_>
snapshotting is not usefull then, i dont want to compress to not kill my cpu
<eoli3n_>
colemickens: right, but datas are just my dotfiles, already git managed, and pictures/movies/music borgbackuped
<eoli3n_>
so...
* colemickens
nods
<colemickens>
I have a similar setup. I still like it for libvirt images, docker layers, and honestly I make stupid mistakes in my home dir and I have the disk space to spare. But it sounds like you know what you're doing :)
<eoli3n_>
colemickens: yep, but on arch i use lvm for that. But i'm interested by docker layers on zfs, how do you manage that ?
<eoli3n_>
just created a pool for /var/lib/docker ?
o1lo01ol1o has quit [Remote host closed the connection]
<colemickens>
I think I have a pool for /var/lib and was lazy about it, but yeah. And then virtualisation.docker.storageDriver="zfs".
<colemickens>
and then docker natively can talk zfs to have zfs manage layers/deduping. afaik it just works and means I get to avoid dm/aufs/whatever else
<clever>
colemickens: from what ive read, instead of using unionfs to merge each layer together
<clever>
colemickens: it will use zfs to clone the entire layer, and then mutate the clone
<eoli3n_>
colemickens: i was using overlayfs2
<clever>
so, the layers are pre-merged, as a single zfs filesystem, leaving you with less computation overhead
<clever>
in theory, it also eliminates the limit of 128 layers (a problem in unionfs), but docker doesnt actually allow you to take advantage of that
mbrgm_ has joined #nixos
Rusty1 has quit [Remote host closed the connection]
* colemickens
TIL. thanks!
<clever>
so, when you import an image, it will unpack one layer
<clever>
clone it, and unpack a 2nd layer ontop of the clone
<clever>
clone it, and unpack a 3rd layer ontop of the clone...
<eoli3n_>
i need to check if podman support zfs
<eoli3n_>
i don't use docker on desktop
<qyliss>
aIt does.
<qyliss>
I should set that up
<eoli3n_>
sorry, what means "aIt" ?
<qyliss>
It was a typo
<qyliss>
I meant "It"
<eoli3n_>
oh ok
mbrgm has quit [Ping timeout: 268 seconds]
mbrgm_ is now known as mbrgm
philr has joined #nixos
<eoli3n_>
i added 1G memory, and i see it melt
leotaku has joined #nixos
<eoli3n_>
319M left
<eoli3n_>
why is a rebuild so greedy
o1lo01ol1o has joined #nixos
drakonis has quit [Ping timeout: 260 seconds]
* colemickens
has noticed it being cpu greedy lately, rebuild cripples video playback for a few seconds
drakonis has joined #nixos
Thra11 has joined #nixos
<eoli3n_>
then... unable to fork: cannot allocate memory
<eoli3n_>
rr
Rusty1 has joined #nixos
<{^_^}>
[nixpkgs] @andersk opened pull request #76757 → ruby: Revert rubygems to 3.0.6 for ruby 2.6 and earlier → https://git.io/JeAAr
<qyliss>
eoli3n_: how many (logical) CPU cores do you have
drakonis has quit [Ping timeout: 260 seconds]
drakonis has joined #nixos
<eoli3n_>
qyliss: 2
<eoli3n_>
"echo 1 > /proc/sys/vm/overcommit_memory" worked
<lovesegfault>
I dislike how ansible will fail halfway through a playbook and then:
<eoli3n_>
wow, reading text on that video fucked my brain
thc202 has quit [Ping timeout: 248 seconds]
<lovesegfault>
1. It's hard to debug what/how failed
<eoli3n_>
i think i need to sleep
<lovesegfault>
2. It will leave things halfway done (inconsistent state)
civodul has quit [Quit: ERC (IRC client for Emacs 26.3)]
<p01ar>
happy new decade <3
* lovesegfault
screams about the decade starting in 2021 at p01ar
<p01ar>
*internally screams that for all purposes pointing out that there was never a year 0 is honestly like pointing out a tomato is a fruit
<eoli3n_>
lovesegfault: it's idempotent, so as you ignore_errors OR fix it, state should alway be the same
<lovesegfault>
eoli3n_: I'm not in the SRE team so I can't speak as to what our ansible setup does, but I've had issues where it will fail in provisioning and make the machine inaccessible
o1lo01ol1o has quit [Remote host closed the connection]
chloekek has quit [Ping timeout: 268 seconds]
<{^_^}>
[nixpkgs] @sorpaas opened pull request #76760 → Change types.attrs to recursively merge → https://git.io/JeAxY
iyzsong has joined #nixos
ebzzry has joined #nixos
lovesegfault has quit [Ping timeout: 248 seconds]
shibboleth has quit [Ping timeout: 240 seconds]
shibboleth has joined #nixos
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/JeAxz
drakonis has quit [Quit: WeeChat 2.7]
fendoer has quit [Read error: Connection reset by peer]
bvdw has quit [Read error: Connection reset by peer]
bvdw has joined #nixos
wildtrees has quit [Quit: Leaving]
<{^_^}>
[nixpkgs] @risicle opened pull request #76762 → pythonPackages.unicorn: redesign to become a wrapper package around unicorn-emu → https://git.io/JeAxM
Thra11 has quit [Quit: WeeChat 2.7]
lambda-11235 has joined #nixos
<{^_^}>
[nixpkgs] @Infinisil opened pull request #76763 → codeowners: Add infinisil to some more paths → https://git.io/JeAxd
justanotheruser has quit [Ping timeout: 246 seconds]
justanotheruser has joined #nixos
lambda-11235 has quit [Quit: Bye]
mexisme has quit [Ping timeout: 246 seconds]
drakonis has joined #nixos
felixfoertsch23 has joined #nixos
felixfoertsch has quit [Ping timeout: 268 seconds]
<shajra>
I'm really enjoying Haskell.nix... no question... just a statement.
jmeredith has joined #nixos
<{^_^}>
[nixpkgs] @andersk opened pull request #76765 → vagrant: Resolve crash by replacing gem symlinks with directories → https://git.io/JeAhP
<{^_^}>
[nixpkgs] @andersk closed pull request #76757 → ruby: Revert rubygems to 3.0.6 for ruby 2.6 and earlier → https://git.io/JeAAr
phi has joined #nixos
phi is now known as phi2020
Supersonic has quit [Disconnected by services]
Supersonic112 has joined #nixos
Supersonic112 is now known as Supersonic
zeta_0 has joined #nixos
ddellacosta has quit [Ping timeout: 260 seconds]
endformationage has quit [Quit: WeeChat 2.6]
fusion809 has joined #nixos
mexisme has joined #nixos
turq has joined #nixos
<turq>
I have installed nixos on my machine and I'm trying to use cachix, but when I run "cachix use <some-package>", this is the error that is emitted:
<turq>
MustBeRoot "Run command as root OR execute: $ echo \"trusted-users = root $USER\" | sudo tee -a /etc/nix/nix.conf && s
<turq>
udo pkill nix-daemon"
mexisme has quit [Ping timeout: 248 seconds]
<turq>
/etc/nix/nix.conf is read only, so the suggested modification isn't going to work
UndefinedIsNotAF has quit [Ping timeout: 252 seconds]
UndefinedIsNotAF has joined #nixos
GerdFlaig[m] has joined #nixos
<cvlad->
Anybody on `nixos` using `~/.XCompose`to be able to write non-standard characters? I'm not quite sure where I would begin setting it up.
Valodim[m] has joined #nixos
<simpson>
cvlad-: Your locale probably already has a handful of characters configured. Check out /share/X11/locale/$LOCALE/Compose for a listing.
<simpson>
For example, I have an Xlib with this compose file: /nix/store/5j7mqkqjrqmvm57vngfj358qzr2fv32y-libX11-1.6.7/share/X11/locale/en_US.UTF-8/Compose
civodul has joined #nixos
<cvlad->
@simpson
<cvlad->
simpson: I already have an Xconfig file form my ubuntu machine, and I had overridden the meta key as well. Just not sure how to replicate the results on nixos.
<simpson>
cvlad-: You can do all the same stuff. The xmodmap binary is in the xorg.xmodmap package.
Chiliparrot has joined #nixos
jmeredith has joined #nixos
<{^_^}>
[nixpkgs] @timjb opened pull request #76783 → networkmanager: fix dispatcherScripts example → https://git.io/Jexko
zupo has joined #nixos
opthomasprime has joined #nixos
opthomasprime has left #nixos [#nixos]
cjpbirkbeck has quit [Quit: Quitting now.]
ng0_ has joined #nixos
Lears has joined #nixos
[Leary] has quit [Ping timeout: 265 seconds]
ng0 has quit [Ping timeout: 268 seconds]
ng0_ is now known as ng0
<cvlad->
simpson: Okay, it ended up working. I needed this in my configuration.nix:
<kahiru>
hey, what's the best way to configure and run a service which isn't packaged for nixos and there are no confiuration modules for it? Apart from packaging it and writing it ofc
<{^_^}>
[nixpkgs] @misuzu opened pull request #76786 → freeswitch: fix gsmopen build and enable by default for linux → https://git.io/JexIx
<eoli3n_>
kahiru: you could use flatpak
<eoli3n_>
check if package exists
<eoli3n_>
or running a container
<eoli3n_>
with podman
thelg has quit [Read error: Connection reset by peer]
<kahiru>
yeah, that doesn't sound too bad
philr has quit [Ping timeout: 258 seconds]
ddellacosta has quit [Ping timeout: 260 seconds]
<eoli3n_>
as podman runs docker images and can fire rootless containers, it should match any of usecases
UndefinedIsNotAF has quit [Ping timeout: 264 seconds]
UndefinedIsNotAF has joined #nixos
ddellacosta has joined #nixos
<kahiru>
I'm aware of that, I was just wondering if there isn't some other, more pure solution
<Unode>
Is there any way to control the order of entries in /etc/sudoers?
<Unode>
I have one user for which several passwordless rules exist. This user is also part of the wheel group. Since the wheel directive appears last in the sudoers file and defaults to asking a password, sudo always asks a password for this user despite having passwordless entries.
<Unode>
if I move the wheel directive to the top of the file, everything works as expected.
thelg has joined #nixos
<Unode>
what I'm trying to do is to move the wheel directive to the top but this is not something controlled by me. Additionally this worked (no password prompt) in 19.03 but stopped working (always asks password) in 19.09
<Unode>
I'm tempted to open an issue unless someone has an idea and I'm the one doing something wrong...
fenedor is now known as fendor
thelg has quit [Ping timeout: 268 seconds]
wildtrees has joined #nixos
thelg has joined #nixos
thelg has quit [Remote host closed the connection]
<sphalerite>
Unode: I think it should work if you use lib.mkBefore on your extraRules
<{^_^}>
[nixos-org-configurations] @grahamc pushed to master « status.nixos.org: add a noscript indicator, and a loading indicator »: https://git.io/Jexqq
<{^_^}>
[nixpkgs] @freepotion opened pull request #76792 → ivan: add icon and desktop item → https://git.io/JexmW
<{^_^}>
[nixpkgs] @tomberek opened pull request #76793 → bash-my-aws: init at 20191231 → https://git.io/Jexm8
obadz- has joined #nixos
<pie_>
ok quasselclient seems to work on unstable
obadz has quit [Ping timeout: 252 seconds]
obadz- is now known as obadz
shiver has joined #nixos
zarel_ has quit [Ping timeout: 260 seconds]
danbst has joined #nixos
<danbst>
hello o/
zarel has joined #nixos
<danbst>
I have a problem - persistent environment variables __ETC_PROFILE_DONE and __NIXOS_SET_ENVIRONMENT_DONE
<danbst>
when I start a clean bash (`env -i bash --noprofile --norc`), those variables are absent, as expected
alex has joined #nixos
<danbst>
however, when I launch it through Gnome (gnome-shell run), they are set every time
<danbst>
I can't find anything in internet about "persistent x session environment varibales" - I never defined those vars in my files
alexherbo2 has joined #nixos
<danbst>
I believe reboot will help here, but I wonder if anybody have seen such persistent variables
<mehlon>
I have both of these vars as well
<mehlon>
env | grep __
<danbst>
mehlon yeah those are set in all NixOS sessions
<danbst>
however in my case, those are set even when I start bash without /etc/profile and without /etc/bashrc. Smbdy provides initial environment
<danbst>
which shouldn't happen
pie_ has quit [Ping timeout: 258 seconds]
<danbst>
if we talk about XY problem - I did `nixos-rebuild switch` after adding an environment variable, but it didn't show up in `env`. I backtracked it to this issue - because `__NIXOS_SET_ENVIRONMENT_DONE=1` the file with nixos-defined env vars is no longer sourced
<{^_^}>
[nixpkgs] @dudebout opened pull request #76794 → document nix-env bug relating to multiple output installation → https://git.io/Jexmp
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
<{^_^}>
[nixpkgs] @fpletz pushed to master « proxmark3: refactor & init unstable at 2019-12-28 »: https://git.io/JexYm
eoli3n__ has quit [Remote host closed the connection]
eoli3n___ has joined #nixos
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<mananamenos>
hi, why does latest nixo (loris in this case) use kernel 4.19 instead of the latest one by default?
<ivan>
linux release cycle is scary and the LTS kernels are less scary, new 5.4 LTS is still in its infancy (but has been unproblematic for me)
<ivan>
(not an official reason)
<ivan>
you can set the kernel to whatever you need
<sphalerite>
mananamenos: nixos's stable release don't update major versions generally. You can "opt in" to the latest linux kernel by setting boot.kernelPackages = pkgs.linuxPackages_latest; in your config though.
<mananamenos>
yeah, actually i did, and this solved the issue of gdm + gnome3 not working with 4.19 kernel
<sphalerite>
mananamenos: more specifically — nixos always uses longterm kernels, not stable ones
<mehlon>
and manually doing systemctl stop plex; systemctl start plex?
<systo>
same result in systemctl status plex with updated timestamps
pie_ has quit [Client Quit]
<mehlon>
I guess I can't help any further
<systo>
hmm, doing on separate lines only indicates one fail, doing stop/start on same line indicates too fails, some monitor is trying to keep it active at all costs?
<systo>
TY for the help thus far
cosimone has joined #nixos
kaliumxyz has quit [Remote host closed the connection]
leotaku has quit [Ping timeout: 258 seconds]
pie_[bnc] has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
fusion809 has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @Lassulus pushed commit from @bfortz to master « bitwig studio 3.0.3 -> 3.1.1 »: https://git.io/Jexn3
<eoli3n___>
lovesegfault i'm decided to switch, but i'm switching from archlinux/xorg-i3 to nixos/sway so i need to change everything single things
<eoli3n___>
every single thing*
eoli3n___ has quit [Remote host closed the connection]
<eoli3n_>
still here :)
<genesis>
bonjour eoli3n_ :)
<eoli3n_>
o/
<genesis>
\o
gustavderdrache has joined #nixos
lord| has quit [Read error: Connection reset by peer]
lord| has joined #nixos
spinus has joined #nixos
civodul has quit [Quit: ERC (IRC client for Emacs 26.3)]
<spinus>
hello, I cannot find it, is there any way to filter input of derivation? I tried builtins.filterSource but that seem not working as expected, nix still says it grabbing more than 256M where most of it is node_modules for local dev
<spinus>
but when I manually remove `node_modules` it works fine without any warning
<ryantm>
spinus: Sounds like the node_modules files are getting in some other way then. Do you have a minimal nix expression that has trouble?
<spinus>
ryantm: nah, I would need to extract it to make it any useful to look at. Thanks for the tip, I'll look for other places when this can be fetched and than start preparing example so maybe this will show me the way, cheers.
growpotkin has joined #nixos
LeshaInc has joined #nixos
philr has joined #nixos
<ajs124>
I'm trying to fix the build of pythonPackages.pysvn right now, but I'm getting lots of "locale.Error: unsupported locale setting" errors with their test suite and I'm not quite sure what those are supposed to mean