ng0 has quit [Quit: Alexa, when is the end of world?]
<Ralith>
rycee: got `error: Could not connect: Connection refused` from home-manager switch's dconf stuff just now
hke has joined #nixos
erictape1 has joined #nixos
<Ralith>
it seems to be preventing the switch from working
erictapen has quit [Ping timeout: 240 seconds]
acarrico has joined #nixos
dermetfan has quit [Ping timeout: 244 seconds]
Ralith_ has quit [Remote host closed the connection]
josephoenix has joined #nixos
<josephoenix>
Hi all
<Ralith>
rebooting made the problem go away once again
<gchristensen>
hello!
<Ralith>
this seems to be a recurring problem
<josephoenix>
All of a sudden, ACME isn't working with my nginx hosts. The service unit fails to start, and the server appears to be offering a self-signed cert... for example.com?
<josephoenix>
Has anyone else enecountered this? NixOS 17.09 (but upgraded everything to 18.09 and still see it), nginx with no fancy addons
<josephoenix>
The failure message from the Let's Encrypt / ACME service has been changing. One sec while I reboot and capture a recent output.
<Ralith>
it was working great, then I updated, now in some environments it's setting `TERM=dumb` and in others it's still okay
<Ralith>
okay, somehow nuking .direnv for the fourth time in a row worked...
<Mic92>
Ralith: see my direnv function in the nixos wiki
<Mic92>
Ralith: it sets TERM=dumb, when you are using direnv-mode
<Mic92>
in emacs
<Mic92>
and cache the return value of nix
<samueldr>
josephoenix: '2019-01-25 00:19:23,345:ERROR:simp_le:1446: CA marked some of the authorizations as invalid, which likely means it could not access http://example.com/.well-known/acme-chal'
<Ralith>
I'm already using a function like that
<Ralith>
for some reason it was setting TERM=dumb in a shell
<samueldr>
this makes it unclear whether their error message has example.com as an example or your setup has example.com somehow
<Ralith>
but after repeatedly and aggressively nuking the state it's not
<Mic92>
yes, my version saves TERM before calling nix
<averell>
the derivation has preliminary example.com self-signed certs too. so probably just validation failure
<Ralith>
it's clearly something to do with state getting screwy across an update
<Ralith>
my version also does that, but briefly stopped for unclear reasons
thc202 has quit [Ping timeout: 245 seconds]
<samueldr>
josephoenix: I see a chevron (>) is there more output to the right?
hke has quit [Ping timeout: 250 seconds]
<ToxicFrog>
Huh. Is lmsensors just broken on NixOS?
<ToxicFrog>
Like, it "works", but also it uses a compiled-in ETCPATH that points at the lmsensors package in the nix store, so it's impossible to add a sensors.conf
hellrazor has joined #nixos
<ToxicFrog>
And there's basically no system where lmsensors gives meaningful results without at least some fiddling with sensors.conf
<chrisaw>
How do I add an environment variable to a nix pkg?
<chrisaw>
As in - I need to build code with an environment variable already set - could just use export I guess but wondering if there's a more elegant way to do it. I'm sure I've seen one but can't remember it. :(
ddellacosta has joined #nixos
hellrazo1 has quit [Ping timeout: 250 seconds]
eadwu has joined #nixos
ddellaco_ has quit [Ping timeout: 244 seconds]
halfbit has joined #nixos
<infinisil>
chrisaw: attributes of mkDerivation get turned into env vars automatically
<infinisil>
So you can just `stdenv.mkDerivation { MY_ENV_VAR = "foo"; }`
<{^_^}>
[nixpkgs] @grahamc pushed commit from @mrVanDalo to release-18.09 « rustc: mark compileprocess as timeconsuming »: https://git.io/fhrF7
<chrisaw>
@infinisil Thanks for the advice - how would I achieve that in an overlay package though? I'm using airflow = super.pythonPackages.buildPythonPackage rec { }
<infinisil>
chrisaw: Just set the var in the attribute set you have there, buildPythonPackage should propagate it to mkDerivation
sigmundv__ has joined #nixos
<infinisil>
chrisaw: Also, you should use self.pythonPackages.buildPythonPackage
<disasm>
fpletz: I'm curious if a decision was made at all to switch to networkd as the default with the upcoming release.
<chrisaw>
infinisil: Heh, I had tried the same thing with single quotes - changed to normal quotes and problem fixed. Thanks a lot! :)
sigmundv_ has quit [Ping timeout: 240 seconds]
lawlesseel has quit [Ping timeout: 256 seconds]
hph^ has quit [Read error: Connection reset by peer]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
josephoenix has quit [Quit: Lost terminal]
hph^ has joined #nixos
hph^ has quit [Excess Flood]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
infinee_ has joined #nixos
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
hph^ has joined #nixos
hph^ has quit [Max SendQ exceeded]
doyougnu has quit [Ping timeout: 240 seconds]
silver has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @grahamc opened pull request #54564 → WIP: Bogus patch to efitools → https://git.io/fhrAr
erictape1 has quit [Ping timeout: 245 seconds]
hooo has joined #nixos
<{^_^}>
[nixpkgs] @ivan opened pull request #54565 → [18.09] buildFHSUserEnv: use runScript in env → https://git.io/fhrxU
infinee_ has joined #nixos
halfbit has quit [Ping timeout: 250 seconds]
o1lo01ol1o has joined #nixos
o1lo01ol1o has quit [Ping timeout: 240 seconds]
Synthetica has quit [Quit: Connection closed for inactivity]
Ralith_ has joined #nixos
slack1256 has joined #nixos
Supersonic has quit [Ping timeout: 268 seconds]
Supersonic has joined #nixos
tmaekawa has joined #nixos
tmaekawa has quit [Quit: tmaekawa]
iqubic has quit [Ping timeout: 245 seconds]
iqubic has joined #nixos
<iqubic>
Anyone know what might be causing my computer to be running so so slow.
gagbo has quit [Quit: I'm out !]
pie_ has joined #nixos
<iqubic>
Answer: A rouge GHC was the cause.
rprije has joined #nixos
halfbit has joined #nixos
tilpner has quit [Ping timeout: 246 seconds]
<jackdk>
those red GHCs always cause trouble
<iqubic>
I meant rouge. Alright...
<jackdk>
Sorry, I didn't see that someone else had made that joke in #haskell already. Now I'm stifling giggles at my desk
<iqubic>
It's fine.
<iqubic>
This is the first time that the dante plugin for emacs has actually screwed me over.
sigmundv__ has quit [Ping timeout: 268 seconds]
halfbit has quit [Ping timeout: 240 seconds]
tobiasBora has quit [Ping timeout: 272 seconds]
tobiasBora has joined #nixos
<slack1256>
On the haskell.section.md documentation say that to enable profiling one has to set the overrides. Is that still true after the overlay rearchitecture?
revtintin has joined #nixos
<ivan>
something keeps deleting all my zsh history
<clever>
phizzz: this causes a copy of whatever dir core.nix was in, to wind up at /run/current-system/nixcfg/
<clever>
and a new copy is made for every generation, so you now have history even without commits being made
<tdbgamer>
Have you guys had problems with Pulse audio speeding up the microphone input?
<tdbgamer>
I just spent a few days debugging it and finally just disabled pulse audio and it fixed it
<tdbgamer>
but now some of the KDE menus for audio fail cause they relied on Pulseaudio :/
<tdbgamer>
has anyone run into this before? I'm on NixOS 18.09
<phizzz>
ottidmes: thanks for the suggestion. it turns out i mangaged to boot from the installation media again, so i thought i lost my changes
tmaekawa has quit [Quit: tmaekawa]
<phizzz>
clever: cool!
<phizzz>
i have a headache from looking at /nix/store paths...
<phizzz>
i might have to read a tree book later
<einargs>
Is it okay if I complain about trying to get a go package to build with nix? No reason, just need to vent (and maybe someone will point something out I missed).
<jasongrossman>
phizzz: That's just an example of course!
<jasongrossman>
phizzz: It's important to realise that NixOS expects to manage things in /etc.
<phizzz>
jasongrossman: i was using visudo, isn't that the correct way to do it?
<iqubic>
Not in Nixos.
<phizzz>
iqubic: ok, i just read the top of the file # Don't edit this file. Set the NixOS options ‘security.sudo.configFile’ or ‘security.sudo.extraRules’ instead.
knupfer has quit [Remote host closed the connection]
<elvishjerricco>
einargs: so you're trying to use filterSource on a derivation's output? Yea unfortunately you can't do that. You'll have to either override the derivation to remove some output files, or you'll have to make a new derivation that copies / hard links / symlinks the files you want.
endformationage has quit [Quit: WeeChat 2.3]
<einargs>
elvishjerricco: I'm not sure how I would override the result of fetchFromGitHub.
<einargs>
And copying/hard-linking/sym-linking everything seems... icky.
dermetfan has joined #nixos
<elvishjerricco>
einargs: look into overrideAttrs in the nixpkgs manual. You can use it to add logic to postInstall or something
<einargs>
elvishjerricco: I didn't realize that fetchFromGitHub used mkDerivation. That's pretty cool!
<elvishjerricco>
einargs: Yea. Just be sure to change the sha256 every time you change the code. If nix sees a derivation with an output hash specified, it'll just reuse any existing store path with that output hash since there's no reason to build the same data twice. It doesn't know that the script has changed and would fail to produce that hash
kitl_ has joined #nixos
<einargs>
Ah, I see what you mean.
<einargs>
I didn't realize that the sha256 was actually used for the store path.
dermetfan has quit [Ping timeout: 250 seconds]
slack1256 has joined #nixos
kitl_ has quit [Remote host closed the connection]
markus1189 has quit [Ping timeout: 250 seconds]
markus1189 has joined #nixos
kitl_ has joined #nixos
tmaekawa has joined #nixos
tmaekawa has quit [Client Quit]
kitl_ has quit [Remote host closed the connection]
kitl_ has joined #nixos
jb55 has joined #nixos
palo1 has joined #nixos
ThatDocsLady has quit [Ping timeout: 272 seconds]
infinee_ has quit [Ping timeout: 244 seconds]
pjan has joined #nixos
palo has quit [Ping timeout: 272 seconds]
palo1 is now known as palo
tmaekawa has joined #nixos
tmaekawa has quit [Client Quit]
xok has joined #nixos
ottidmes has quit [Ping timeout: 240 seconds]
einargs has quit [Ping timeout: 256 seconds]
<{^_^}>
[nixpkgs] @Mic92 merged pull request #54559 → Add tests for Sonarr, Radarr, Lidarr and Jackett → https://git.io/fhr1w
<vonfry>
Do someone use krb5 on nixos? I find that the nixos.krb5 only has server, but not has the client(commands, such as kinit, klist and etc, cannot be found)
<etu>
,locate kinit
<{^_^}>
Found in packages: klibc, krb5.dev, heimdalFull, klibcShrunk
<{^_^}>
[nixpkgs] @dotlambda pushed commit from @stigtsp to release-18.09 « pythonPackages.python-gnupg: 0.4.3 -> 0.4.4 »: https://git.io/fhoIs
<vonfry>
Thanks. I have got it
<tdeo>
hi, what package is libstdc++.so.6 in? if i find it in /nix/store i get ...gcc-7.3.0-lib/lib/libstdc++.so.6 but i don't know how to reference that package
rcy has joined #nixos
<srhb>
tdeo: gcc-version-lib means it's the "lib" output
<srhb>
tdeo: In this case it's gcc-unwrapped.lib. nix-locate is useful for this (from this nix-index package)
Ariakenom has joined #nixos
<tdeo>
thank you! didn't know about nix-locate
<srhb>
tdeo: Very useful :)
jtojnar has joined #nixos
reinhardt has joined #nixos
vonfry has quit [Ping timeout: 268 seconds]
vonfry1 has joined #nixos
vonfry1 has quit [Client Quit]
o1lo01ol1o has joined #nixos
o1lo01ol1o has quit [Ping timeout: 246 seconds]
r5d has joined #nixos
pie__ has quit [Ping timeout: 240 seconds]
infinee_ has joined #nixos
<teto>
,locate pod2man
<{^_^}>
Found in packages: perl, perl526, perldevel, perl526Packages.podlators, perl528Packages.podlators, perldevelPackages.podlators
phizzz has left #nixos [#nixos]
orivej has joined #nixos
r5d has quit [Ping timeout: 240 seconds]
rcshm has joined #nixos
pie__ has joined #nixos
dermetfan has joined #nixos
hellrazo1 has quit [Ping timeout: 246 seconds]
hellrazo1 has joined #nixos
<iqubic>
Well, ACPI is screwing up again
<iqubic>
There is no way in hell that I have 272 hours left on this battery.
<iqubic>
Wait... That was just momentary glich. Re-running the command fixes this.
Tucky has joined #nixos
Guanin has joined #nixos
moredread[m] is now known as moredread[m]1
moredread[m]1 is now known as moredread[m]2
moredread[m]2 is now known as moredrea4
moredrea4 is now known as moredrea6
slack1256 has quit [Remote host closed the connection]
Guanin has quit [Client Quit]
moredrea6 is now known as moredrea8
rcshm has quit []
__Sander__ has joined #nixos
<{^_^}>
[nixpkgs] @buffet opened pull request #54568 → slurp: init at 1.0 → https://git.io/fhomq
jasongrossman has quit [Remote host closed the connection]
ng0 has quit [Client Quit]
rauno has quit [Remote host closed the connection]
agander has joined #nixos
dermetfan has quit [Ping timeout: 245 seconds]
jbaum98 has joined #nixos
<jbaum98>
is there a way to override linuxHeaders for a given package? I need to apply a patch to the system headers, so i think I need to override glibc in stdenv, but i'm not sure how
jasongro` has quit [Ping timeout: 245 seconds]
kai_w has joined #nixos
<kai_w>
I'm looking into pytorchWithCuda - I get an 'undefined symbol' error on importing the module. The file the symbol is in is on the RUNPATH, but not listed in the NEEDED fields so it is not being loaded
r5d has joined #nixos
<kai_w>
this is also the case on an ubuntu build of the same package, and it works there.
<kai_w>
I'm not very familiar with dynamic linking, any ideas?
rtjure has quit [Ping timeout: 246 seconds]
rtjure has joined #nixos
jasongrossman has joined #nixos
<edef>
kai_w: presumably it is loaded with dlopen(3)
<edef>
kai_w: so the search path is LD_LIBRARY_PATH, not the rpath
vidbina has joined #nixos
<edef>
kai_w: might want to strace/ltrace both versions and see if you can discern when/how loading happens, and where it looks
mounty has quit [Ping timeout: 250 seconds]
mounty has joined #nixos
xok has quit [Ping timeout: 245 seconds]
<kai_w>
edef: so if my symbol is in /lib/nvrtc.so.1, I'm looking for a call to dlopen with that file as the parameter?
<edef>
kai_w: yeah, though you can usually see the open(2) calls looking for it from an strace
<edef>
kai_w: the first bunch returning ENOENT, until it finally returns an fd
<phdoerfler>
Mic92, in the name of all the hours wasted on this planet by frustrated people needing the OracleJDK on NixOS, take my gratitude
<greatscottttt>
Mic92: thanks, I'm in a bit of an annoying situation, where we have a binary in our repo such that we're all using the same version of, in this case, clang-format
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
MerlinGttlinger[ has joined #nixos
nikola_i has joined #nixos
<Mic92>
greatscottttt: I'd say for clang-format you could write a shell.nix expression for nix-shell that takes the binary as the source and then uses the autoPatchelfHook.
<nikola_i>
I am trying to install nixos 18.09. Why is it compiling every package
<Mic92>
the only question is where you get your libllvm from?
<phdoerfler>
How long will there be security updates for past releases?
<greatscottttt>
Mic92: interesting, I saw that at the bottom of the that page...maybe I'll give that a quick shot
<phdoerfler>
Say… 17.09
copy` has joined #nixos
<Mic92>
phdoerfler: I was also annoyed everytime someone updated something jdk based in a pr.
<nikola_i>
is this normal installing nixos from 18.09 dvd is taking 7+ hrs
<Mic92>
phdoerfler: better not to use 17.09 anymore.
<phdoerfler>
Mic92: I got a server on 17.09 that is a minor headache to update to 18.x (because rspamd)
<{^_^}>
[nixpkgs] @voobscout opened pull request #54572 → zxing-cpp: init at git-20180320 → https://git.io/fho45
xkapastel has joined #nixos
silver has joined #nixos
nefix has joined #nixos
<nefix>
Has NixOS dropped support for 32bit CPUs?
<Mic92>
nefix: I think we are still maintaining some support but do not build all packages.
<Mic92>
There are still i686 installation isos.
erictapen has joined #nixos
<Taneb>
Setting unpackPhase to "" worked better than I feared it would
<Taneb>
(For not failing stdenv.mkDerivation if I don't have a sensible source to provide)
<nefix>
Mic92: well, I need a super simple netboot image
<nefix>
Also, could I customize a netboot (install some packages and change some run scripts)?
<Mic92>
nefix: because of multi-lib support the most important libraries should be still build.
fendor has joined #nixos
<srhb>
Taneb: phases is also relevant.
das_j_ is now known as das_j
<nefix>
Mic92: how am I supposed to build it then? I've seen that you are supposed to run the netboot function in the release.nix file, but I can't see how am I supposed to build it for 32 bits
<Taneb>
srhb: it's awkward when you want to do all of the default ones apart from one
<Taneb>
Or, at least, potentially all
<srhb>
Taneb: Sure :)
<Taneb>
Of course, it'd be nicer still if stdenv was fine with just not having any source
<Taneb>
But I guess most of the time that's a good sign something's been specified wrong
<Mic92>
nefix: I don't know. There is pkgsi686Linux for packages and then you can set --system 'i686-linux' otherwise.
<Mic92>
but adding 32-bit in the release.nix seems to have no effect.
init_6 has joined #nixos
nakkle has joined #nixos
rprije has quit [Ping timeout: 240 seconds]
erictapen has quit [Ping timeout: 240 seconds]
acarrico has joined #nixos
<nakkle>
I've noticed that if a nix expression for a job on a hydra server cannot be evaluated, the whole jobset is displayed as successful, and the one job that couldn't be evaluated is completely ignored. Can you somehow configure a hydra / a project / a jobset such that jobs that cannot be evaluated are still treated as mandatory for the jobset?
Lears has joined #nixos
[Leary] has quit [Read error: Connection reset by peer]
erictapen has joined #nixos
<{^_^}>
[nixpkgs] @alyssais opened pull request #54573 → libreoffice,libreoffice-fresh: fix build on Darwin → https://git.io/fhoEu
<ivegotasthma>
anyone here uses airvpn with nixos?
<ivegotasthma>
I downloaded the tar for the client, but there are problems starting the UI
<arianvp>
im trying to run a haskell repl in nixos
<arianvp>
I did nix-shell -p zlib ; cabal repl
<arianvp>
but I get " can't load .so for : libz
<arianvp>
and then cabal repl crashes
<arianvp>
cabal new-build works fine
<arianvp>
it's just cabal new-repl that fails
nakkle has quit [Ping timeout: 252 seconds]
nakkle has joined #nixos
noqqe has joined #nixos
<noqqe>
Hi - Need some help from you guys.
<noqqe>
when i update system config (using switch without --upgrade) it installs all kind of packages that i dont understand where they come from (wayland, postgres.. gtk, kde)
<noqqe>
im running testing(19.03) and dont know what to do
<noqqe>
in summary it tries to download around 30gb of packages
<arianvp>
oh wow
<noqqe>
any idea what i can do to find out what causes this?
<noqqe>
(its btw a tiny lamp stack on this server)
<nikola_i>
hello my system is not booting as nix is unable to find the new uuid of luks lvm setup. also the generated uuid in hardware Configuration.nix is different from the uuid output of blkid for the luks llvm partition.
<nikola_i>
is there a way around this
<nikola_i>
I am using live USB nixos 18.09
<arianvp>
nikola_i: I'm honestly not sure if the generator can deal with lvm-luks setups
<arianvp>
I have a btrf-luks setup with multiple subvolumes, and I had to write my hardware-config myself as the generator generated something that didn't work
<arianvp>
could you show me the otuput of lsblk and your current hardware-configuration.nix
<arianvp>
?
<arianvp>
or blkid*
ld34 has quit [Remote host closed the connection]
<nikola_i>
arianvp: it worked for me the first time I created luks llvm on my machine a year ago.
<arianvp>
but yeh I would just manually override the correct UUIDs in your hardware-configuration
<srhb>
I've mostly had to manually do the luks setup on most of my machines. I think it only gets generated correctly for very vanilla setups.
<jasongrossman>
ivegotasthma: I haven't used airpvn's client on NixOS, but of course you can use airvpn's servers via openvpn (or I think even via ssh?).
<ivegotasthma>
jasongrossman: yes, that's what I'm doing now, direct openvpn
Makaveli7 has quit [Quit: Leaving]
<nikola_i>
arianvp: here's the output of both hw config and blkid. I am using sda7 for luks+llvm
<{^_^}>
[nixpkgs] @grahamc pushed 0 commits to grahamc-patch-1: https://git.io/fhozY
ng0 has joined #nixos
<arianvp>
nikola_i: i am not near that laptop at the moment
<arianvp>
but the point is, you need to think well about how you want stuff to be mounted and in what order
jomik has joined #nixos
<arianvp>
and very exotic installs are not that well-supported by the standard generator
<arianvp>
so you'll have to write the config yourself
<arianvp>
I guess you have a LUKS Container and inside that an LVM container, right?
<srhb>
nikola_i: A brief look on your UUIDs suggest to me that they are correct and you just need to add the luks attributes, and possibly tweak the swap setup according to the encrypted options.
<nikola_i>
srhb: arianvp: i have lvm inside luks container. I am going to try to use the uuid from blkid /sda7 in configuration.nix to see if that works
<srhb>
nikola_i: Yes, you should use that for the boot.initrd.luks.devices.encrypted.device
<nikola_i>
Mic92: i am using ext4 not zfs
eadwu has joined #nixos
<Mic92>
ok. didn't catch the whole context.
<__monty__>
Mic92: Couple typos, "your as" Add all mounts to your as legacy mounts as described, and mount's, either mounts or `mount`s.
worldofpeace has joined #nixos
<gchristensen>
aanderse: did you click through to the full log? https://logs.nix.ci/?key=nixos/nixpkgs.54524&attempt_id=ae7278a1-7fde-4df9-a620-37ff10480fdf
<srhb>
aanderse: Maybe you have local overrides that fix darcs?
<Mic92>
__monty__: oh gosh. I was so distracted while writing this because I was looking for the right wiki template.
<__monty__>
Mic92: No worries, I always find typos when rereading. Especially when it's things that are embarrassingly permanent, like commit messages ; )
<nikola_i>
srhb: arianvp: finally it worked. used the uuid from blkid /sda7
<__monty__>
*rereading my own stuff
<gchristensen>
or like IRC messages
<srhb>
gchristensen: If you live in the EU you can always invoke GDPR :-P
<gchristensen>
what if it is on the blockchain!
Synthetica has joined #nixos
<srhb>
:O
<__monty__>
Yes, constantly, but they're not as embarrassing. People rarely look at the logs anyway, right?
<gchristensen>
good god, what happens if someone GDPR's their information out of nixpkgs
<gchristensen>
or the linux kernel
<srhb>
gchristensen: Indeed. Or just samueldr's logs.. :P
<srhb>
Maybe there's provisions for these cases, I don't know.
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
Guanin has quit [Quit: Lost terminal]
<Baughn>
My server has become unable to run NixOS tests.
<Baughn>
Except some of them work if run interactively.
<Baughn>
And the way in which they fail ranges from "systemd randomly receives SIGABRT" to "/dev/vda is somehow missing"
<Baughn>
And rolling back to a previous generation has no effect, nor does using older versions of the stable channel.
<Baughn>
Nor both in combination.
* Baughn
throws his hands up in the air. *Why?*
vidbina has quit [Ping timeout: 240 seconds]
anderslundstedt has quit [Quit: leaving]
<gchristensen>
anything in dmesg?
<Baughn>
Nope.
<Baughn>
The zfs tests fail because of a missing vda. If I run them interactively, they instead fail because the zfs module cannot be modprobed- "illegal argument", it says.
<Mic92>
Baughn: just to be sure, could it be that you have data corruption? Also this would solve when using sufficient old generations: nix-store --verify --check-contents --repair
<Baughn>
It's almost got to be that, right? Let's see..
<Baughn>
A sufficiently old generation does work. I'd guess a kernel update is at fault.
<Baughn>
I probably went through a few generations without rebooting.
patrl has quit [Ping timeout: 246 seconds]
o1lo01ol1o has quit [Ping timeout: 244 seconds]
<laas>
question: do overlays declared by extending the nix path in configuration.nix not apply to the packages installed through environment.systemPackages?
<symphorien>
if you change the nix path in configuration.nix, it will not apply the first time you rebuild
<symphorien>
is only take into account the value of the nix path in the shell which launches nixos-rebuild
<symphorien>
so you may need to rebuild once, create a new shell, and rebuild there
jabranham has joined #nixos
vidbina has joined #nixos
<samrose>
anyone have an example of packaging a nodejs app and running it as a service in nixos?
<laas>
ah
<laas>
thanks!
o1lo01ol1o has joined #nixos
<laas>
I just broke my system accidentally by doing `sudo nix-store --delete --ignore-liveness (which kak)`
<laas>
somehow this deleted more than just kakoune
<laas>
how do I fix it
<laas>
I don't have the basic nix packages anymore
<srhb>
laas: Ouch. If that which did not expand to a path, you probably need a live-usb
<laas>
well thankfully I have one
<symphorien>
ls /nix/store/*nix*/bin/nix
<laas>
can I just run nixos-install again/
<laas>
?
<laas>
I can't open a terminal, so I can't check that lol
<srhb>
laas: It'd then be equivalent to "nix-store --delete-everything #yolo"
<Mic92>
laas: yes.
<laas>
if I run nixos-install?
<Mic92>
it is idempotent
<Mic92>
(nixos-install)
<laas>
nice, thanks for the help, I feel stupid for bricking my system that easily lopl
<symphorien>
laas: you can probably also reboot to an old generation if it does not depend on the version of kak you removed
<laas>
good thing I kept my live usb
Soo_Slow has joined #nixos
<srhb>
laas: Don't use ignore-liveness ever :-)
<srhb>
laas: (Instead, make the path dead first next time, before using delete)
ingenieroariel has joined #nixos
<clever>
,tofu
<{^_^}>
To get a sha256 hash of a new source, you can use the Trust On First Use model: use probably-wrong hash (for example: 0000000000000000000000000000000000000000000000000000) then replace it with the correct hash Nix expected.
<laas>
well, time to reboot and see what I can do, thanks again
<ingenieroariel>
hello nixos - I am trying to enable content_by_lua on nginx and do not know what to change on my /etc/nixos/configuration so nginx comes with luajit
<ingenieroariel>
I moved away from using too much from nixos nginx service in order to have more control over the config (replaced the acme stuff for a more manual http server set up)
<Baughn>
gchristensen: nix-build successfully builds the zfs test. nixops doesn't. They're both trying to build the *same derivation*, which means it already exists. (So why does nixops try to build it at all?)
<Baughn>
...well, let's take a look at the command line.
<Baughn>
Increasingly suspecting hardware failure. I suppose I should run memtest86 on it for a while.
<gchristensen>
must not be the same
<Baughn>
Same hash.
<gchristensen>
the the .drv matches?
<Baughn>
It's the same .drv.
<Baughn>
Same path.
<gchristensen>
mysterious
<Baughn>
I really should have gotten ECC memory. >_>
endformationage has joined #nixos
shabius has quit [Read error: Connection reset by peer]
shabius has joined #nixos
clefru has joined #nixos
<ingenieroariel>
I am trying now `let stable = import ./pkgs/servers/http/nginx/stable.nix; in callPackage stable { modules = [ nginxModules.lua ]; }` as a derivation - to see if my config runs on a nix shell with that one (gcc is compiling luajit-2.0 right now)
greatscottttt has quit [Quit: WeeChat 1.9]
<clever>
ingenieroariel: callPackage will import automatically if the 1st arg is a path
iqubic has left #nixos ["ERC (IRC client for Emacs 26.1)"]
<nefix>
Hey there! I'm trying to build a netboot image for booting it through iPXE, but when I run `nix-build -A netboot nixos/release.nix` I get this error: error: a 'aarch64-linux' with features {} is required to build '/nix/store/14hg3piib4gaiqmkmr3z8lmfgajgj1ln-append-initrd-secrets.drv', but I am a 'x86_64-linux' with features {benchmark, big-parallel, kvm, nixos-test}
<clever>
nefix: this is a nixos module, that does everything you want and more
<clever>
it sets up dhcpd for netbooting, it sets up bind for dns relaying, it sets up NAT so your netboot-slaves have internet, yet are still isolated from the LAN you dont control fully, and it creates the netboot image, with justdoit configured
<clever>
and it hosts the netboot files on nginx
<Soo_Slow>
looks like somebody has made a sandbox tool, based on nix package manager https://github.com/jollheef/appvm . Did anybody there try that?
<nefix>
Woah
<pbb>
How can I configure networking for VMs created by nixos-build-vms?
PyroLagus has quit [Ping timeout: 252 seconds]
<nefix>
clever: not sure how to use it though
<clever>
nefix: if you have a machine with 2 network cards, you can just add /path/to/nixos-configs/netboot_server.nix to your imports field, and then set netboot_server.network = { wan = "eth0"; lan = "eth1"; };
<clever>
nefix: wan is the side of the machine facing the internet, lan is the side facing the machines you want to netboot
<clever>
it is acceptable for wan to be facing another router
<laas>
So I'm trying to repair my broken system with nixos-install, but when I run it I get a FileNotFoundError in some python script. It says /nix/var/nix/profiles/system-1-link/kernel doesn't exist. Any ideas on what is causing this?
<clever>
i originally designed this for use on a laptop, where wan is my wifi, and lan is the ethernet
<nefix>
clever: well, I just want to build a customized version of NixOS, that when gets booted (through iPXE) automatically runs a shell script that opens an X window
PyroLagus has joined #nixos
<Baughn>
That line about switch_root looks... I don't know. It's a bash script. It's marked executable, sure enough. *How* can it get 'operation not permitted'?
<clever>
nefix: then your interested in just lines 7-27
<clever>
Baughn: is the fs mounted with noexec?
<Baughn>
I should hope not.
<Baughn>
This is from `nix-build ~/dev/nix/system/nixos/tests/zfs.nix -A stable |& ix`
<clever>
nefix: line 11 pulls in the netboot profile, line 12 is optional, line 13 refers to the module on 20, and 20-27 is just the contents of a configuration.nix file
<clever>
Baughn: oh, the 9plan stuff in qemu for mounting the host fs to the guest, sometimes messed up with zfs
knupfer has joined #nixos
<clever>
Baughn: the inodes get cross-wired, and a directory contains another directories files
<Baughn>
This used to work. I swear it did.
<Baughn>
Also, what.
ingenieroariel has joined #nixos
<nefix>
clever: I see, this is extremely helpful
<Baughn>
clever: Reference?
Ariakenom has quit [Ping timeout: 240 seconds]
<clever>
Baughn: let me find that screenshot...
Tucky has quit [Quit: WeeChat 2.2]
<Baughn>
Oh, there's no bug filed?
<symphorien>
laas: is /boot mounted correctly ?
<Taneb>
Is there any library functions that's like "if foo ? bar then [ foo.bar ] else []"?
dbmikus has joined #nixos
<laas>
<freenode_sym "Las: is /boot mounted correctly "> yes boot is mounted
<nefix>
clever: is the init kernel boot parameter required to boot?
<nefix>
clever: because I need to send other boot parameters too. And also I have to support different achitectures (that use the same menu) but they aren't going to have the same boot image
<Baughn>
zfnmxt: If it's an *interactive* shell, it may not like not having a TTY.
<Baughn>
What does `journalctl --user -u desktop-bridge` say?
<clever>
nefix: each arch should have its own netboot.ipxe, at a different URL
<clever>
nefix: though if you want anything more, you may want to just make a custom netboot.ipxe, copying the code from nixpkgs
<nefix>
clever: but that's the issue. This code isn't the same always. And my custom iPXE menu in the binary isn't going to mutate
<clever>
nefix: what if the menu just leaves the extra kernel params in a certain variable, and then the different selection of premade netboot.ipxe's all append that to the kernel params?
<nefix>
clever: I think I can manage to make it work. Since the images are going to be downloaded automatically, I can download there a custom netboot.ipxe too
<samrose>
clever: double aha! thanks
<laas>
symphorien: Yeah it is mounted and the system was actually "fixed", although I get the same error after trying to rebuild inside the install
<symphorien>
what do you mean by "rebuild inside the install" ?
<laas>
oh deleting the symlink fixed it
asymmetric has joined #nixos
<laas>
well I did nixos-install, rebooted into the new installation, and did `nixos-rebuild`
<clever>
Baughn: still looking...
<Baughn>
zfnmxt: Starts and stops again in about a second? :)
<Baughn>
zfnmxt: The last line is suggestive. It's getting EOF on stdin -- of course it is, systemd doesn't provide any input. You need to run it in non-interactive mode.
<clever>
Baughn: minor problem, its in this dir of 1208 screenshots, lol
jomik has quit [Quit: WeeChat 2.2]
<zfnmxt>
Baughn: You can't :(
<Taneb>
Goal for this weekend: package the dasher on-screen keyboard
<zfnmxt>
Baughn: Could I start it inside of tmux or something? Think that might work?
<Baughn>
zfnmxt: It doesn't have a non-interactive mode?
<zfnmxt>
Baughn: Nope. Jank beta software.
jomik has joined #nixos
<ottidmes>
I thought `nix eval` and `nix-instantiate` were basically the same except of the UI, but `nix eval` automatically instantiates derivation expressions, while nix-instantiate does this only for the returned derivation
__Sander__ has quit [Read error: No route to host]
<Baughn>
zfnmxt: If not, then some use of `expect`.
vidbina has quit [Ping timeout: 244 seconds]
fooinabar has joined #nixos
<zfnmxt>
Outright failed with `script`, let me check out `expect`. Thanks for the help!
vidbina has joined #nixos
<fooinabar>
I'm trying to build a package that includes some python bindings that are in a subdirectory of the source code. Is there a way I can use buildPythonPackage to build them?
<Baughn>
zfnmxt: The correct fix is of course to file a PR adding a non-interactive mode. ;-)
<zfnmxt>
Baughn: I don't think it's actually open source atm =/
<zfnmxt>
Baughn: But, in searching for it, seems like I stumbled upon an open source alternative that may fix all my problems :D
eadwu has quit [Ping timeout: 240 seconds]
clr_ has joined #nixos
clr__ has joined #nixos
eadwu has joined #nixos
jomik has quit [Quit: WeeChat 2.2]
<nefix>
clever: what should be between line 20 and 27?
tg has quit [Ping timeout: 246 seconds]
puck has quit [Ping timeout: 246 seconds]
jomik has joined #nixos
hark has quit [Ping timeout: 246 seconds]
<nefix>
clever: don't understand what's justdoit. It's like an installation script, but I don't see why I need it. I just need to open a X graphic session and open a program
eadwu has quit [Ping timeout: 240 seconds]
eadwu has joined #nixos
<clever>
nefix: you can just delete the justdoit lines
<zfnmxt>
Baughn: Running it in a detached screen session worked! :D
<clever>
nefix: 20-27 is your nixos config, so you would put services.xserver.enable = true; inside there
<clever>
media-center.nix runs the plex frontend on bootup
<clever>
other then loosing the login creds, i could just jam that into a netboot image, and have entirely disk-less plex frontends
<nefix>
clever: yeah, I'm trying to do something like that
erasmas has joined #nixos
<clever>
nefix: media-center.nix may also interest you, as an example of how to run a gui app on bootup
<rycee>
Ralith: Hmm, I haven't seen that issue before. The dconf module starts up a short lived dbus session if DBUS_SESSION_BUS_ADDRESS is not set so perhaps you ran into some issue with that logic.
<nefix>
clever: I use iPXE built-in login menu, authenticate with a server that generates a token and boot with that token as boot argument
<nefix>
clever: then I read the token from /proc/cmdline and open the GUI program using this token
<clever>
it has a password login (though the pw is in plaintext in the script) which locks shell access
<nefix>
clever: well I have a Go server that generates the iPXE menus dinamically so, in the end, the menus are really simple
<clever>
nefix: ah, i was going a different route, with server-side scripting to change what the script was entirely
<clever>
nefix: in my case, it was just a dumb switch-case on the mac, to give a different fixed boot option
<nefix>
clever: you could do that in the DHCP, so you wouldn't need that server
<clever>
nefix: yeah, i can see how i would now, but it opens up the option of swapping it for a DB, and its safer to edit a php script then dhcp config
<clever>
nefix: get the dhcp wrong, and the whole network dies
<nefix>
that's true
<clever>
my dhcp server is also nixops managed, so i have to do a nixops deploy to change the cfg
<nefix>
that's another level! :) After really understanding how Nix and NixOS works, I'm planning on mirgrating my servers to NixOS and manage them with NixOps
<nefix>
But that's in the far future
<clever>
nefix: i initially managed the router with normal nixos-rebuild, but later migrated it to nixops
<nefix>
clever: are sessioncommands run in a shell? (Can I set variables and this stuff?)
<nefix>
Also, how could I "pack" a custom script and add it to the PATH so I can run it when starting the session?
<clever>
nefix: the session commands are inserted into a bash script that is responsible for starting the chosen desktop env (kde, gnome, xfce)
<clever>
nefix: because i put an exec on line 22 of media-center.nix, that script dies mid way thru, and never launches one
<clever>
instead, only running plexmediaplayer
tmaekawa has joined #nixos
<clever>
nefix: combined with the autoLogin flag for slim, it will then run sessionCommands on bootup of X
<nefix>
clever: Ok, so I could avoid "packing" the shell script and just run the commands in there
<clever>
yep
tmaekawa has quit [Client Quit]
<clever>
of note, if the main gui exits for any reason, X will shutdown, then slim restarts, auto-logs again, and it re-runs the sessionCommands
<clever>
i opted to disable that in mine, because its running on my NAS
<nefix>
clever: and could I just shutdown / reboot the whole computer?
nikivi has quit [Read error: Connection reset by peer]
<clever>
yep
<clever>
if you are non-root, i believe dbus to logind is the simplest way
<clever>
but you could also mess with sudoers to allow `sudo shutdown -h now`
<clever>
without a pw
nikivi has joined #nixos
<clever>
or just throw all sanity out the window, and run your frontend as root!
<nefix>
clever: xD
<nefix>
Let's see first if this works
Ariakenom has joined #nixos
<clever>
also of use, if you put the frontend stuff into a dedicated file like i did, you can do imports = [ ./frontend.nix ];
<clever>
but, you can also add it to the imports of a non-netboot machine, for testing
<nefix>
clever: error: value is a function while a set was expected <- nix-build is throwing this :/
asymmetric_ has joined #nixos
<nefix>
(I'm trying the version you sent [with modifications in the configuration)
<clever>
nefix: what does --show-trace give, and can you put both the nix file and backtrace into a gist ?
<clever>
nefix: wait, i see the problem
<clever>
nefix: line 1, you want import <nixpkgs> {}
<nefix>
clever: yep! that's building now
<clever>
`import <nixpkgs>` is a function, that takes the nixpkgs cfg
rcshm has joined #nixos
sigmundv__ has quit [Ping timeout: 244 seconds]
asymmetric has quit [Ping timeout: 250 seconds]
<ingenieroariel>
Ok, I got nginx compiled with lua support via /etc/nixos/configuration, I and verified with ldd that libluajit-5.1 is linked to the nginx the system is using - but nginx still does not load the configuration with lua blocks :/
<ddellacosta>
so I'm trying to install NixOS, and as part of that I'm trying to use ZFS with encryption. I've followed the instructions here: https://nixos.wiki/wiki/NixOS_on_ZFS, which say to set `boot.zfs.enableUnstable` to true, but when I try to use the example line to encrypt the whole pool on creation, it complains about options: `property 'ashift' is not a valid pool property`. Any ideas what I'm missing?
<srhb>
sotrhraven: The NixOS manual has a step by step guide that should Just Work, unless you're doing something exotic.
<ddellacosta>
ashift seems to be a pretty standard property, and unrelated to encryption, so I'm not sure what is going on here regardless
<Baughn>
Is ashift still required? Most disks should have correct sector-size properties.
<clever>
ingenieroariel: and if you read the nginx.conf in the error, what block is it in?
<gchristensen>
Baughn: most disks lie about it
<ddellacosta>
Baughn: yeah I have to admit I'm cargo culting these instructions just to get this up and running, I don't know what is critical and what is not
<clever>
Baughn: every single disk ive checked (even nvme) claim 512 byte sectors
<Baughn>
gchristensen: They export a 'logcal size' of 512, and 'physical size' of 4096.
vidbina has quit [Ping timeout: 250 seconds]
<Baughn>
ZFS should be looking at the physical sector size parameter.
<clever>
Namespace 1 Formatted LBA Size: 512
<gchristensen>
ZFS has a (long) list of disks which lie about it
<gchristensen>
Baughn: and the list got long enough that they gave up maintaining it
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
<clever>
Baughn: HDIO_DRIVE_CMD(identify) failed: Inappropriate ioctl for device
<clever>
nvme, not sata
<Baughn>
Ah.
<Baughn>
NVMe is different. The correct size really might be 512 bytes, though?
<Baughn>
I can't imagine any reason why you'd lie about it for such new hardware.
sotrhraven has left #nixos ["Leaving"]
<clever>
i cant see why they would make the erase blocks so small
<gchristensen>
Legacy™
hedning has quit [Remote host closed the connection]
<ddellacosta>
so, that's all well and good, but with every option I remove it just barfs on the next: `property 'altroot' is not a valid pool property`, then `property 'encryption' is not a valid filesystem property`, etc.
<__monty__>
ingenieroariel: Are you an underwater welder or something? Wonder what a mermaid engineer is : )
<clever>
nefix: oh, that reminds me...
<ingenieroariel>
__monty__: I wish it was so colorful - Ariel is just a given name for me - was biblical before Disney - but oh well
<clever>
nefix: a few months ago, somebody in here was making a usb stick, with every distro on it, and had trouble getting the nixos iso to work there
<{^_^}>
[nixpkgs] @worldofpeace opened pull request #54584 → gnome3.gnome-settings-daemon: bypass polkit by using an suid wrappe → https://git.io/fho55
ld34 has quit []
ld34 has joined #nixos
jomik has quit [Quit: WeeChat 2.2]
ingenieroariel has left #nixos [#nixos]
jomik has joined #nixos
ubert has quit [Quit: Leaving]
<ddellacosta>
clever: so as long as I turn off the `zfs-zed.service` which is running for some reason, I can do `rmmod zfs`, but then running `modprobe zfs` after that gives me `modprobe: ERROR: could not insert 'zfs': Unknown symbol in module, or unknown parameter (see dmesg)` (dmesg doesn't really tell me much other than listing the symbols it couldn't find). Any chance you've seen this before?
<ddellacosta>
poking at the zfsonlinux list now to see if I can find any further clues.
<clever>
ddellacosta: you also have to reload spl, and the other zfs related modules
<ddellacosta>
clever: okay, thanks. Any pointers to docs where I can find all of this better explained? I hate asking these kinds of questions here if there is some doc that can guide me, but the ZFS on NixOS wiki page clearly ain't it
tg has joined #nixos
<clever>
ddellacosta: check `modinfo zfs` to see what it depends on, and unload those as welll
<ddellacosta>
clever: okay, thank you
vidbina has joined #nixos
<ddellacosta>
the thing that is confusing me is that I don't remember doing _any_ of this the last time I got zfs up and running...wish I'd taken better notes.
<ddellacosta>
any of this module fiddling I mean
<clever>
ddellacosta: if zfs isnt available in the installer image, then the modules wont be loaded initially, so the first load gets the version you just installed
<clever>
ddellacosta: also, if you dont use the unstable zfs, its much simpler
<clever>
ddellacosta: i also consider the zfs crypto to be unstable still, and prefer zfs on luks
patrl has joined #nixos
<ddellacosta>
clever: yeah, what I'm saying is that I installed zfs once before, with encryption enabled, and it wasn't nearly this challenging...I'm confused as to what I missed this time around
<ld34>
hi, i'm trying to install nix with fish shell on macos and it is failing. is there any documented workaround?
<bsima>
ld34: what is the error message?
<elvishjerricco>
ld34: I don't know the first thing about fish, but I think the Nix installer only modifies the necessary files for bash and zsh. There's like one line you have to add to your shell init so that you have the stuff installed by Nix on your PATH
<wedens>
is there something like netctl available in nixos?
<wedens>
(not NM)
<simukis>
"error: attribute 'linuxPackages_4_18' missing" can I find out why?
<ld34>
bsima: it isn't erroing, it at least doesn't modify necessary files
<simukis>
okay, so 4_19 and 4_20 are all that exists. I guess it is time to update and reboot.
* simukis
feels their 117 day uptime cry out a painful scream of agony.
<ld34>
elvishjerricco: it, indeed, is modifies bash and zsh, but i don't know what exactly should i do to make it work
<ld34>
elvishjerricco: what is that line
<simukis>
Thanks, clever, that’s the best way to search for packages I’ve seen so far.
<elvishjerricco>
ld34: Just check out what it did to the bash / zsh files, and see if you can figure out how to replicate that for fish. I don't know anything about fish so I can't provide much help
<simukis>
clever, does that perhaps also support search by filename somehow?
nefix has quit [Ping timeout: 256 seconds]
<clever>
simukis: nix-locate
<simukis>
great! Thanks.
asymmetric_ has quit [Remote host closed the connection]
asymmetric_ has joined #nixos
<ld34>
elvishjerricco: thanks, i'll look into, but, could you provide link on such files?
<WilliamHamilton[>
I have `riot-web` among the packages in my nixos configuration, but after the last nixos-rebuild switch I'm no longer able to find a folder matching *riot-web* in the /nix/store (I find the .drv files though). Why could this be happening?
<elvishjerricco>
ld34: At my NixOS machine currently, so can't quite look... But I think it's either `/etc/bashrc` and `/etc/zsh`, or `~/.bashrc` and `~/.zshrc` depending on whether you did a multi-user or single-user install
agander has quit [Quit: Leaving]
<Myrl-saki>
Hm.
Makaveli7 has joined #nixos
<Myrl-saki>
Gonn need some comments on this.
<Myrl-saki>
Adding a quick option to mksquashfs.nix.
<Myrl-saki>
For rapid testing.
<ld34>
elvishjerricco: thank's, gonna try to find them
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
mudrii has quit [Ping timeout: 256 seconds]
Makaveli7 has quit [Client Quit]
<ddellacosta>
clever: okay, I've finally got a pool created and I can start actually installing now. Thanks for your help, it was invaluable! I may add some notes to that wiki page to try to make the process a bit more clear
jomik has quit [Quit: WeeChat 2.2]
<clever>
ddellacosta: your welcome
erictapen has quit [Ping timeout: 268 seconds]
<Myrl-saki>
Hm. What's the best way to have debugging options in a NixOS module without making it an option?
<wedens>
Mic92: do you still use iwd? (I've noticed you in maintainers)
<kreisys>
hey I have a question... what does it mean when nix fails with 'resource deadlock avoided'? when this happens then simply retrying always fixes it... is there a way to avoid it?
<srhb>
kreisys: I think it originally didn't exit on that, so it was just informative, but I too have experienced it recently, I think.
<srhb>
kreisys: It's caused by multiple concurrent builds for some reason. I'm not sure what resource is the issue, never looked into it.
Makaveli7 has joined #nixos
<elvishjerricco>
Myrl-saki: What do you mean? How can it be an option without being an option? :P
<Myrl-saki>
elvishjerricco: Not a module option, but an option such as a flag.
rcshm has quit [Remote host closed the connection]
<elvishjerricco>
Myrl-saki: How would a user set this flag?
rcshm has joined #nixos
<kreisys>
I found the relevant issue on github. at least I'm not alone in this. 😅
<Myrl-saki>
elvishjerricco: Yeah, that's exactly what I'm trying to figure out with my question.
<elvishjerricco>
Myrl-saki: Oh. So what's wrong with making it a regular option? I can't really imagine any more appropriate way to do this
<Myrl-saki>
elvishjerricco: IMO, the only reason someone would want to use this is for debugging, and having it as a CLI flag(,if possible) allows one to turn it on and off quickly.
<elvishjerricco>
Myrl-saki: You could use `builtins.getEnv` to test an environment variable at eval time
<Myrl-saki>
elvishjerricco: Oh, hm, I guess that is an option.
<elvishjerricco>
Then users would be able to do `DEBUG_MYRL_SAKI=1 nixos-rebuild switch`
<elvishjerricco>
Myrl-saki: But I'd still probably stick to a NixOS option. That's pretty lightweight to me
<Myrl-saki>
elvishjerricco: Thanks.
<Myrl-saki>
elvishjerricco: Technically, I could do `{ config.imports = []; config.debug = false; }` too
<elvishjerricco>
Myrl-saki: It'd be interesting to add support for command line flags to influence the NixOS config
<Myrl-saki>
Hm.
<Myrl-saki>
--arg might work.
<Mic92>
wedens: never really because it does not work with the eduroam at my university.
<Myrl-saki>
error: attribute 'isoImage' in selection path 'config.system.build.isoImage' not found
<Myrl-saki>
Doesn't seem to work tho. :C
fendor has quit [Ping timeout: 246 seconds]
<Myrl-saki>
Seems like `--arg` cannot traverse down. I'm not sure how I feel about that.
<Myrl-saki>
Especially since -A allows traversing.
<elvishjerricco>
Myrl-saki: Yea I think adding configs on the CLI would require a few changes to `<nixpkgs/nixos/default.nix>`
<elvishjerricco>
Myrl-saki: But it doesn't sound like a very difficult change
<johnw>
Myrl-saki: how would --arg traverse down?
And has joined #nixos
<johnw>
I understand with -A, it's attribute selection, but --arg passes in arguments to a function
<Myrl-saki>
johnw: deep merge on the inputs, I guess.
<Myrl-saki>
--arg x.y foo --arg x.z foo
<Myrl-saki>
Should be the same as --arg x '{y = foo; z = foo;}'
<johnw>
hmm
<Myrl-saki>
Anyways, you make a good point.
<Myrl-saki>
-A has easier sematnics than --arg for nested sets.
<johnw>
the logic to desugar dotted attrset names could be reused for this I bet
<And>
Hi, is it possible to add nixos live install image on multiboot usb?
<And>
now
<johnw>
i.e., { a.b.c = "foo"; } is a similar situation
dermetfan has quit [Ping timeout: 250 seconds]
<And>
It stucks on stage 1 during load, cannot find mnt-root
asymmetric_ has quit [Ping timeout: 240 seconds]
<hyper_ch>
hmmmm, a bit offtopic, but do you of you have recommendations for mesh networking?
<gchristensen>
I hear zerotier is nice
<Mic92>
hyper_ch: like vpn or like wifi?
<hyper_ch>
wifi
<hyper_ch>
make wifi home coverage great again :)
<hyper_ch>
gchristensen: ah, that's vpn :)
<Mic92>
hyper_ch: stay away from wifi repeater.
<Mic92>
They half your airtime
<hyper_ch>
I know :(
<hyper_ch>
that's why I asked for mesh :)
<Mic92>
The Munic Freifunk uses NixOS with batman-adv I think
alex```` has quit [Read error: Connection reset by peer]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<WilliamHamilton[>
is there a way to add some libraries to steam-run?
dermetfan has joined #nixos
<ottidmes>
gchristensen: you had some experience with rnix, no? do you perhaps know how to produce nodes e.g. a string literal node based on a Rust string? so that I replace a bit of the source code with something new
<das_j>
WilliamHamilton[: I use environment.systemPackages = with pkgs; [ (steam.override { extraPkgs = pkgs: [ pkgs.corefonts ]; }) ];
<das_j>
If that is waht you need
<WilliamHamilton[>
das_j: let me see if I can put that to use, thanks!
<WilliamHamilton[>
das_j: wait, do you open a nix-shell with that?
<das_j>
I put that into my system config
And has quit [Remote host closed the connection]
knupfer has quit [Ping timeout: 240 seconds]
alex```` has joined #nixos
<gchristensen>
no idea, ottidmes
<gchristensen>
never actually used it
<ottidmes>
gchristensen: ah ok, well I got plenty of workarounds, guess I will use one of them for now
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
erictapen has joined #nixos
eadwu has quit [Quit: WeeChat 2.3]
<WilliamHamilton[>
das_j: is that configuration picked up by the `steam-run` command?
<das_j>
WilliamHamilton[: Oh. I'm not sure about that :/
<das_j>
sorry
<WilliamHamilton[>
because I added gnome3.gtk to the steam override and I still get the same error
<WilliamHamilton[>
np, thanks!
dermetfan has quit [Ping timeout: 240 seconds]
freeman42x has joined #nixos
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
goibhniu has quit [Ping timeout: 240 seconds]
halfbit has quit [Ping timeout: 240 seconds]
Judson has joined #nixos
Judson is now known as Guest26119
halfbit has joined #nixos
rcshm has quit [Remote host closed the connection]
chessai has quit [Remote host closed the connection]
alex```` has quit [Read error: Connection reset by peer]
chessai has joined #nixos
And has joined #nixos
<gchristensen>
disasm: here?
alex```` has joined #nixos
And has quit [Remote host closed the connection]
_deepfire has quit [Ping timeout: 240 seconds]
_deepfire has joined #nixos
Alling has joined #nixos
<Alling>
"NVidia drivers don't support wayland"
<Alling>
Is it really that simple?
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<{^_^}>
[nixpkgs] @yegortimoshenko opened pull request #54587 → toggldesktop: bump qt-oauth-lib (see #53676) → https://git.io/fhojR
<Alling>
I mean, is it simply not possible to use Wayland if you have an Nvidia GPU?
<samueldr>
AFAIUI yes, since wayland builds upon EGL (IIRC) and nvidia doesn't implement it
<samueldr>
(if it isn't EGL, it's another specific thing that the nvidia drivers don't implement)
<samueldr>
ah, GBM, EGLStreams is the nvidia-specific one
<Alling>
I see. So I guess I have three options:
<Alling>
1) Only using one out of three monitors (because the other two are "disconnected" in xrandr and I guess I need Nvidia drivers to fix that).
<Alling>
2) Three monitors with _massive_ tearing.
<Alling>
3) Windows.
<simpson>
Alling: Is nouveau not an option?
<simpson>
Also is Xorg not an option?
<Alling>
simpson: It most certainly is! I haven't thought about that. I just read in the manual about Nvidia GPUs.
<Alling>
simpson: I don't understand your second question.
<simpson>
Also, FWIW, you can ask xrandr to force an output to be enabled, which sometimes is needed with pre-HDMI connectors.
<simpson>
Alling: Oh, IOW, couldn't you just not use Wayland? Plenty of folks are getting by just fine without it.
<Alling>
In this case it's actually two PG278Qs connected via DisplayPort that are not detected by xrandr.
halfbit has quit [Ping timeout: 240 seconds]
<Alling>
I don't know. I've been having MASSIVE tearing on my laptop ever since I installed NixOS, and I heard Wayland is the only way to get rid of tearing, so I thought that would be the way to go now that I want to try NixOS on my main PC.
patrl has quit [Ping timeout: 252 seconds]
<simpson>
There are several ways to get rid of tearing. One technique that works reliably on several of my machines is to use a compositing manager; I use `compton`, which is a fork of `xcompmgr`.
<simpson>
It's very bare-bones; while it's not running, video tears, and while it's running, video doesn't tear, but otherwise it's not noticeable.
halfbit has joined #nixos
eadwu has joined #nixos
<Alling>
simpson: Can I use Gnome 3 and compton together?
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
eadwu has quit [Client Quit]
<Alling>
I must admit that I don't know too much about window managers, desktop managers and whatnot, so it's not obvious to me how to enable compton.
<simpson>
Alling: I'm not sure; I don't know much about GNOME 3. But maybe, or maybe GNOME has its own compositor that can eliminate tearing.
<simpson>
Oh, I don't have a service or anything like that; it's just something that starts as part of a script.
<Alling>
simpson: I found an option "services.compton.enable".
<simpson>
Interesting, TIL.
<Alling>
I can'
jmeredith has joined #nixos
<Alling>
I can't tell whether it should work for me though. I believe I have Xserver, GDM, Wayland and Gnome 3, although I don't know what any of those are.
<symphorien>
Alling: compton is Xorg only according to the arch wiki
<Alling>
OK, and I don't have Xorg because I have Xserver?
ld34 has quit [Remote host closed the connection]
<symphorien>
Xorg == Xserver
<symphorien>
wayland replaces xorg
<Alling>
So I _have_ Xorg?
<Alling>
I see.
<Alling>
So I should be able to use compton but not Wayland at the same time.
<symphorien>
(well not exactly, but this is a good approximation)
<simpson>
I am not sure why you *want* Wayland, is the thing. It's new and shiny but that's about it.
<simpson>
It sounds more like you want to eliminate tearing.
<Alling>
simpson: I don't actually want Wayland. I just want to be able to use all my monitors and without tearing, and preferably with "how the windows look like and behave" which I guess is called Gnome 3.
<Alling>
Yeah; I thought I needed Wayland for that.
<Alling>
I'm gonna have to switch to my laptop so I can reboot my main, brb.
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
Alling has quit [Quit: Page closed]
Alling has joined #nixos
<_deepfire>
Anyone else seen libHSbase-orphans linking issues when using ghc822Binary ?
<_deepfire>
I've seen it to carry over to ghc863Binary..
jabranham has quit [Quit: ERC (IRC client for Emacs 27.0.50)]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fhKfI
<Alling>
Hm. Interesting. I disabled Wayland, enabled compton and Nvidia and I got all my screens working. I get the Gnome login screen as usual and it works, but then I have nothing but a mouse cursor on a plain gray background. :D
vk3wtf has quit [Ping timeout: 252 seconds]
<simpson>
Log out, double-check that `GNOME` is selected as your session of choice, and log back in; sounds like you're on one of the fallback/default `none` sessions.
nefix has joined #nixos
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<Alling>
simpson: I checked. It was set to xterm. I selected GNOME instead, but same thing. Everything is just gray except for the mouse cursor.
<simpson>
Curious. I'm not sure what's wrong or what to check next.
<Alling>
I mean, I barely even know what I'm trying to do. :P
<WilliamHamilton[>
anyone encountered `GLib-GIO-ERROR **: 20:51:27.817: No GSettings schemas are installed on the system`?
<Alling>
"Should" it be enough to just set compton.enable = true and it should work with "the windows and such I usually have (Gnome 3?)"?
_kwstas has joined #nixos
<Alling>
Is it strange that I get a fully working login screen and then when I login the whole system is unusable? Or is that normal/expected?
eadwu has quit [Ping timeout: 245 seconds]
q6AA4FD has quit [Ping timeout: 245 seconds]
<gchristensen>
I would not expect that, but I do tend to expect my computers to work
<__monty__>
That'd define snafu.
<Alling>
gchristensen: OK, I don't really do that with NixOS.
<__monty__>
But no, nixos shouldn't mess up your system.
<Alling>
You should have seen me and an expert NixOS friend trying to make my AirPods work with my laptop. :D
<Alling>
He gave up after several hours.
<Alling>
Anyway, if I switch tty from the "gray one" (F2) and then back, it's not gray anymore but black instead.
<Alling>
And the journal is being continuously spammed with g_variant_get_va: assertion 'value != NULL' failed.
<Alling>
gchristensen: Do you know by any chance how (in general) to make NixOS work on a system with an Nvidia GPU and three monitors? Preferably with something similar to Gnome 3. It may not be very fancy, but i'm used to its keyboard shortcuts and such.
ixxie has joined #nixos
<Alling>
It's not strictly necessary to have Gnome 3 though.
<gchristensen>
I just use the nouveau driver and arandr to configure the three displays and it works for me
<gchristensen>
but I haven't done that in some time, I typically stick to one monitor
<Alling>
OK! How do I select nouveau?
<Alling>
Also do you have tearing?
<gchristensen>
I don't think so
<gchristensen>
services.xserver.videoDrivers = [ "nouveau" ]; I think
vidbina has quit [Ping timeout: 272 seconds]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<Alling>
It's quite interesting that my main monitor, currently the only one showing anything at all, IS in portrait mode in NixOS, even though the entire screen is gray. So there is _something_ running that remembers my portrait mode setting.
<Alling>
Wait a sec ... I removed compton and it seems to work now!
<simpson>
Excellent.
<Alling>
It may very well have been the nvidia drivers that did the trick.
rcshm has quit [Remote host closed the connection]
<Alling>
I must say again that you guys on this channel are the best! I'm amazed at how helpful you've always been! (Y)
rcshm has joined #nixos
<gchristensen>
=) I'm really glad to hear that
<simpson>
I'm amazed too; I'm not normally helpful. It's good that things are working for you.
<__monty__>
Then I'll say it again, #nixos you're one of the most helpful channels on freenode, thank you.
MichaelRaskin has joined #nixos
_kwstas has quit [Quit: _kwstas]
IRCsum has quit [Ping timeout: 250 seconds]
IRCsum has joined #nixos
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
q6AA4FD has joined #nixos
<Alling>
This is actually really nice! Now I can use NixOS on my main PC with three monitors. Much nicer than having to use Git Bash and all that.
drakonis has joined #nixos
palo has quit [Quit: WeeChat 2.2]
palo has joined #nixos
rcshm has quit [Remote host closed the connection]
<gchristensen>
I just fetched the NixOS 18.09 graphical live CD, put it on a disk (cp the.iso /dev/sda) and rebooted. I selected the USB disk on the boot screen... and then I just get a "_" in the top left of the screen, and the rest is black. any ideas?
ng0 has quit [Quit: Alexa, when is the end of world?]
<nefix>
sorry for asking again, but is there any way to have multiple functions in the default.nix file and specify which do you want to execute when running nix-build and if there's no specification run both?
<symphorien>
if default.nix is { name ? "default" }: throw "called with name"
<clever>
nefix: you can return a set, like { foo = ...; bar = ...; }
<clever>
nefix: and then use -A foo or -A bar to access one or the other
<clever>
if you dont specify -A, then nix will create result result-2 result-3, one for each attribute
<symphorien>
then running nix-build . --argstr name foo will throw with name=foo
<clever>
wether you use -A or --argstr, depends on what exactly you want to do
<nefix>
i want to have like "ipxe" and "imaeg
<clever>
i think an attrset is better for that
<nefix>
" and build ipxe or the nixos image depending
<nefix>
ok, thanks both
lawlesseel has quit [Remote host closed the connection]
lawlesseel has joined #nixos
crmlt has joined #nixos
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
dbmikus has quit [Ping timeout: 240 seconds]
jasongro` has joined #nixos
knupfer has joined #nixos
<{^_^}>
[nixpkgs] @symphorien opened pull request #54591 → neovim: remove spurious references to compilation flags → https://git.io/fhKtS
eadwu has joined #nixos
jasongrossman has quit [Ping timeout: 250 seconds]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
halfbit has joined #nixos
jasongro` has quit [Ping timeout: 240 seconds]
ryantrinkle has joined #nixos
lord| has quit [Ping timeout: 250 seconds]
lord| has joined #nixos
<{^_^}>
[nixpkgs] @pacien opened pull request #54592 → imagemagick: add djvu support → https://git.io/fhKOn
shpx has joined #nixos
<{^_^}>
[nixpkgs] @pmahoney opened pull request #54593 → freeswitch: expose compilation option to enable postgres support → https://git.io/fhKOz
ixxie has quit [Ping timeout: 244 seconds]
pie__ has joined #nixos
pie_ has quit [Remote host closed the connection]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
Guest26119 has quit [Ping timeout: 250 seconds]
eadwu has quit [Quit: WeeChat 2.3]
eadwu has joined #nixos
<nefix>
clever: it's not working :( It hangs at "booting the kernel"
eadwu has quit [Client Quit]
<clever>
nefix: how much ram on the machine, how big of an initrd? how fast of a cpu?
eadwu has joined #nixos
<nefix>
600MB of initrd, (that's a lot :( ) and the machine is a VM, running in a hypervisor (fast). As for the ram, it has 1GB
<nefix>
clever: ^
<nefix>
Should I try with 2GB?
eadwu has quit [Client Quit]
<clever>
nefix: yeah, try more ram, it needs to be able to uncompress the gzip on the initrd
rcshm has quit [Remote host closed the connection]
<nefix>
Makes sense
rcshm has joined #nixos
<clever>
nefix: `gzip --list` on the initrd can tell you how much
johnw has quit [Ping timeout: 240 seconds]
_kwstas has quit [Quit: _kwstas]
johnw has joined #nixos
tdbgamer has joined #nixos
<gchristensen>
is there a way to tell nixos to not start X, from the bootloader? seems my normal tricks to escape the login manager aren't working (Alt-F<N>, Ctrl-Alt-Backspace, etc)
goibhniu has joined #nixos
<clever>
gchristensen: i think there is a systemd trick to disable services, from the kernel cmdline
<clever>
so you can just tell systemd, never start display-manager.service
<clever>
gchristensen: also, do you have ssh? chvt lets you change vty's
<gchristensen>
I don't
<gchristensen>
systemd.mask= mayb
<clever>
gchristensen: another nasty trick, change init= to init=/bin/sh
<clever>
then have fun getting the system happy enough to `nixos-rebuild boot` a change into effect
<pbb>
there's also this systemd recovery mode, but is there a kernel arg to boot into it?
<symphorien>
gchristensen: add 1 to the kernel cmdline (1 means single user)
<clever>
pbb: all kernel args are visible in /proc/cmdline so systemd can peek at it, once stage2 hands systemd control
<pbb>
systemd.unit=emergency.target ?
<nefix>
clever: yep, that worked. Thanks again
<clever>
so nixos doesnt have to do much to support those things
<clever>
nefix: oh, something else, one min
<clever>
nefix: do you need `man` and `info` to work on this machine? probably not
<clever>
[ { list = [ "d" ]; otherAttr = "someOtherValue"; } { list = [ "a" "b" "c" ]; otherAttr = "someValue"; } ]
<_d0t>
ok, so the problem I'm trying to solve is that somewhere between two system updates the media keys on my laptop stopped working in X
<clever>
pbb: and this will take an attrset, run a function on every key/value pair, and return the list of return values
<clever>
_d0t: was it a change in your config, or a change in nixpkgs?
<clever>
gchristensen: that will do it!
luigy has left #nixos [#nixos]
<_d0t>
clever: I was thinking the former, but it doesn't seem so anymore
<clever>
gchristensen: you want to either disable mutable users, or set the initialHashedPassword
<_d0t>
the only thing I changed was bumblebee
<gchristensen>
indeed :)
<gchristensen>
thanks1
<clever>
_d0t: i'm guessing nixpkgs then, you would need to look over the entire nix-diff, and/or bisect nixpkgs between the 2 revs
<_d0t>
ok
<ddellacosta>
sometimes `nixos-rebuild switch` seems to hang on "setting up tmpfiles," what is that about? How can I debug it? I've got --show-trace added but it's not telling me anything at that stage
cantstanya has quit [Remote host closed the connection]
<Guest33062>
_d0t, I know the feeling :)
Soo_Slow has quit [Remote host closed the connection]
Guest33062 is now known as judson
<judson>
That is, I know the feeling
cantstanya has joined #nixos
tilpner has quit [Quit: WeeChat 2.3]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
<ddellacosta>
now I've got it hanging on "starting the following units: accounts-daemon.service," which also happens intermittently. This is really frustrating--the only thing I can do seemingly is a hard reboot. How can I debug `nixos-rebuild` hanging?
<ddellacosta>
and furthermore I don't know if it's successfully completed or not, if I'm screwing up the system state by performing a hard reboot, etc.
<clever>
ddellacosta: i would check `ps -eH x` and `systemctl list-jobs` to confirm what its actually blocking on
<ddellacosta>
the problem is that I'm in single-user (rescue) mode, so I don't know how to do anything other than wait
<ddellacosta>
or, hard reboot
<ddellacosta>
kind of a catch-22 situation here
acarrico has quit [Ping timeout: 250 seconds]
<clever>
ddellacosta: run `screen` from the single-user shell
<clever>
then you can use normal screen commands to get multiple shells
<ddellacosta>
oh man didn't think of that. Thanks clever
jasongrossman has joined #nixos
<ddellacosta>
now I just have to install screen with... nixos-rebuild
<clever>
nix-env -iA nixos.screen
<ddellacosta>
okay, I'll have to go read up on what that is
<ddellacosta>
I feel like the NixOS install instructions expect you to already be very comfortable using NixOS
rcshm has quit [Remote host closed the connection]