<infinisil>
bourbon: Well that depends on how you defined the derivation in the file
drakonis has joined #nixos
<infinisil>
bourbon: E.g. if foo.nix is `{ foo = derivation ...; }` then you could install it with `nix-env -i -f foo.nix -A foo`
<infinisil>
v0id72[m]: You did specify `boot.supportedFilesystems = [ "exfat" ]`?
fusion809 has quit [Remote host closed the connection]
kaliumxyz has joined #nixos
<v0id72[m]>
no just added the boot extramodulepackages
MmeQuignon has quit [Ping timeout: 246 seconds]
<kaliumxyz>
is it possible to do an if statement without the else?
o1lo01ol1o has joined #nixos
<kaliumxyz>
in nix that is.
<gchristensen>
no
<gchristensen>
what do you want to do, which made you want that?
<infinisil>
v0id72[m]: Well I don't have any idea what the problem might be, but if it's exfat, then you should try with that option set
mkoenig has quit [Ping timeout: 252 seconds]
<v0id72[m]>
k thanks going to give it a go. I lost my old VM so rebuilding and running into the one issue.
MmeQuignon has joined #nixos
<kaliumxyz>
okay :D
<hexa->
Moredread[m]: more joy, they renamed it to PrusaSlic3r (release 2.0.0-rc is out)
<kaliumxyz>
I am asking because I want to not put a string in a document.
<kaliumxyz>
if its set to null I don't want the option in the settings, just put an empty string instead?
mkoenig has joined #nixos
<kaliumxyz>
actually yes, that makes the most sense.
<gchristensen>
cool, sounds good, so if ... then ... else ""
<kaliumxyz>
I'm already including the new line no matter what because thats where the statement in.
<kaliumxyz>
thanks c:
<gchristensen>
:)
<v0id72[m]>
@infinisil Looks that that was the fix or adding my user to the disk group :)
jasongrossman has quit [Remote host closed the connection]
<tdeo>
what's the right way to make something use the correct library extension per platform?
<tdeo>
if the build system doesn't do it
<tdeo>
ah, extensions.sharedLibrary
mexisme has joined #nixos
ddellacosta has joined #nixos
buffet_ has joined #nixos
<angerman>
Ericson2314, blackriversoftwa: the global base pointer seems to be a deficiency in GHCs internal linker. There are a few issues on the issuetracker.
buffet has quit [Ping timeout: 258 seconds]
lovesegfault has quit [Ping timeout: 252 seconds]
lovesegfault has joined #nixos
mexisme has quit [Ping timeout: 246 seconds]
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fjl36
Qubasa has quit [Read error: Connection reset by peer]
<angerman>
Ericson2314: yes. Maybe I should just implement _GLOBAL_OFFSET_TABLE_
<angerman>
Gcc seems to be more eager to put in GOT relocations than clang, which I believe is the underlying reason I did not see this with a pure clang toolchain.
<matthewbauer>
nh2: $outputDoc is just the name of the doc output var (either $out or $doc). ${!outputDoc} dereferences it giving you the var value in $outputDoc
adamantium has quit [Quit: Konversation terminated!]
i1nfusion has quit [Remote host closed the connection]
i1nfusion has joined #nixos
esmerelda has joined #nixos
<___laika>
im working on a haskell project, and so far ive been manually generating my project.nix file with cabal2nix, but i figure it probably makes more sense to just use the haskell nix lib functions to call cabal2nix on my project automatically. if im doing this, does it still make sense to check project.nix into source?
alfon has joined #nixos
drakonis_ has joined #nixos
alfon has quit [Ping timeout: 246 seconds]
iqubic has joined #nixos
drakonis has quit [Ping timeout: 252 seconds]
<iqubic>
Gosh darn it.
adamantium has joined #nixos
<iqubic>
Why don't I have horizantal scrolling?
Elon_Satoshi has joined #nixos
Copenhagen_Bram has quit [Ping timeout: 264 seconds]
<nh2>
___laika: you could also use `callCabal2nix`, which does such generation when nix runs, but that of course needs internet access. I guess checking the cabal2nix output into your code is something you might want to do if you want to make compiling "offline" from just a source checkout easier
<iqubic>
So I just switched from synaptics to libinput. When using libinput, horizontal scrolling should be enabled by default. However, I currently don't have horizontal scrolling.
<iqubic>
Do I need to reboot for the switch from one synaptics to libinput to start working?
<nh2>
matthewbauer: `--with-curl` expects both `lib` and `include` to be in the given prefix. Is there a function with which I can easily merge `curl.dev` and `curl.out` into a symlink dir?
<drakonis>
i don't understand something, i have allowunfree=true set on my local file
<drakonis>
but when i try to drop into a shell, it still yells about it not being enabled
<drakonis>
is allowunfree broken?
<drakonis>
oh.
azdle has joined #nixos
buckley310 has joined #nixos
selfsymmetric-mu has left #nixos ["gone to the land of dead hiccups and extinguished light bulbs"]
<iqubic>
so, why isn't libinput working for me?
<nh2>
I see, `symlinkJoin` does the trick. Still fails to compile though
<drakonis>
i'm having a hard time getting my hp printer to work
dycan has quit [Remote host closed the connection]
<iqubic>
I'd like my Brother Printer to work.
<___laika>
nh2: im using this function: https://github.com/NixOS/nixpkgs/blob/master/pkgs/development/haskell-modules/lib.nix#L47 which runs callCabal2nix, but yeah im basically just asking if distributing my code (which will be an executable at the end oof the day) with the generated project.nix is worth the upkeep of making sure i run `cabal2nix . > project.nix` when i do releases
<{^_^}>
[nixpkgs] @thoughtpolice opened pull request #61529 → pythonPackages.funcsigs: fix tests on pypy3 → https://git.io/fjlGe
<nh2>
___laika: I suspect it is not worth it unless some users ask, but do make sure to script it with `--hackage-snapshot` so that it is reasonably reproducible
<___laika>
:++:
<nh2>
without that outputs of cabal2nix will vary over time
<iqubic>
Also, why do my printer drivers work on Windows, but not on Linux?
<adamantium>
iqubic: stupid question but: have you tried rebooting
snajpa has quit [Ping timeout: 248 seconds]
<iqubic>
No.
<iqubic>
Should I?
<adamantium>
i would
<adamantium>
but just my opinion
aither has quit [Ping timeout: 246 seconds]
Glider_IRC__ has joined #nixos
<adamantium>
it'd be perfect to never reboot, and in linux we don't need to very often, but sometimes especially with drivers things get garbled up
iqubic_ has joined #nixos
<iqubic_>
Well, that worked...
<iqubic_>
But now vertical scrolling is way too fast.
aither has joined #nixos
<iqubic_>
How do I fix that?
<Ralith>
probably more effectively by googling, reading documentation, and trying things, than by repeating the same questions in an IRC channel over and over
ivan has quit [Read error: Connection reset by peer]
marusich has quit [Ping timeout: 268 seconds]
dustinm has joined #nixos
lovesegfault has quit [Quit: WeeChat 2.4]
sb0 has quit [Quit: Leaving]
ivan has joined #nixos
esmerelda has quit [Ping timeout: 246 seconds]
<lopsided98>
jeaye: 4.19 is an LTS release which is still supported, and the latest version (4.19.43) has patches for MDS. NixOS has 5.1.2 as well, but its not used by default.
o1lo01ol1o has quit [Remote host closed the connection]
<jeaye>
lopsided98: Ah, got it. Thanks.
o1lo01ol1o has joined #nixos
<tbenst>
@o1lo01ol1o, I think I was just reading your pull request on Subhask hah, was that you?
o1lo01ol1o has quit [Ping timeout: 245 seconds]
<tbenst>
:/
Supersonic has quit [Ping timeout: 252 seconds]
esmerelda has joined #nixos
drakonis has quit [Quit: WeeChat 2.4]
sellout- has joined #nixos
Supersonic has joined #nixos
Swus has quit [Read error: Connection reset by peer]
<kaliumxy1>
Yes, and its a great resource but does not cover my case (as far as I understand it).
<kaliumxy1>
I have the source locally and am not planning on putting it on cargo nor do I want to build using nix-shell (unless that can easily be automated?)
<kaliumxy1>
The main problem is cargo wants a .cargo folder in the home dir to cache things in.
<kaliumxy1>
It doesn't react friendly to being put in the homeless shelter (very understandable to be fair).
<kaliumxy1>
"buildRustPackage requires a cargoSha256 attribute which is computed over all crate sources of this package. Currently it is obtained by inserting a fake checksum into the expression and building the package once. The correct checksum can be then take from the failed build."
ivan has quit [Read error: Connection reset by peer]
dustinm has joined #nixos
knupfer has joined #nixos
o1lo01ol1o has joined #nixos
glasserc_ has quit [Quit: ZNC 1.6.6 - http://znc.in]
ivan has joined #nixos
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
mexisme has joined #nixos
<kaliumxy1>
:D things crash now like I expect them to crash
<kaliumxy1>
Progress :D
o1lo01ol1o has quit [Ping timeout: 258 seconds]
glasserc has joined #nixos
thePirateKing has joined #nixos
<li_matrix>
v0
<li_matrix>
v0id72: what are you calling a good deed
mexisme has quit [Ping timeout: 255 seconds]
<thePirateKing>
hey dont really know where to ask this, more of a nix related question than nixos, I am helping a friend set up a dev environment for class with nix on mac OS, and when we try to run nix-shell, we are getting this error:
<thePirateKing>
`ld: library not found for -liconv
<thePirateKing>
clang-5.0: error: linker command failed with exit code 1 (use -v to see invocation)
<thePirateKing>
does anybody know how to deal with build issues with nix-shell not on nixos
<thePirateKing>
or is there a channel for that
gentauro42 has joined #nixos
<kaliumxy1>
you can use locate or nix locate to locate the lib and use patchelf to fix it.
<kaliumxy1>
nvm
<kaliumxy1>
it appears you just don't have the lib in your env, try adding it after "nix-shell -p {lib}" or do you get this when you invoke nix-shell?
gentauro has quit [Ping timeout: 246 seconds]
<thePirateKing>
oh well it is a pretty big shell, trying to add libiconv to the shell rn
<thePirateKing>
will update
dansho has joined #nixos
lovesegfault has quit [Ping timeout: 252 seconds]
<thePirateKing>
this is invoking nix shell for a dumb python and R package that is apparently a pain to install to the point of switching package managers
<thePirateKing>
and other packages too but python37Packages.rpy2 is a nightmare
<thePirateKing>
tred adding libiconv to shell.nix as well as nix-env -iA nixpkgs.libiconv
<___laika>
i thought i was closer to a solution of providing a local stack2nix dependency into my nix-shell environment for IDE purposes, but im having issues again
<___laika>
at this point everything builds and i can run nix-shell, but none of the packages i have in my packages argument to shellFor are showing up in the shell
palo1 has joined #nixos
kaliumxyz has quit [Remote host closed the connection]
<thePirateKing>
ok i am having a slight breakthrough, how do i pin a package version or reset it? The problem may be that it requires R 3.5.3, while my friend is on R 3.6
<thePirateKing>
which may be causing issues
zupo has joined #nixos
<thePirateKing>
(I am on 3.5.3 from nixpkgs-unstable and no problem)
endformationage has quit [Ping timeout: 244 seconds]
sir_guy_carleton has quit [Quit: WeeChat 2.4]
gautier has joined #nixos
gautier has quit [Quit: Page closed]
hoijui has joined #nixos
zupo has quit [Read error: Connection reset by peer]
musteresel has joined #nixos
<musteresel>
Hi, does anyone have a working crossSystem definition for mips?
hyper_ch2 has joined #nixos
lovesegfault has quit [Quit: WeeChat 2.4]
thePirateKing has quit [Remote host closed the connection]
cfricke has joined #nixos
<sphalerite>
musteresel: I did, a while back. I'll see if I can dig it up
pukkamustard has quit [Quit: ZNC 1.7.3 - https://znc.in]
pukkamustard has joined #nixos
lovesegfault has quit [Quit: WeeChat 2.4]
another-user has joined #nixos
lovesegfault has joined #nixos
<tilpner>
Are texlive paths not supposed to be in store-paths.xz?
<lc_>
Hi, I'm trying to package LinVst for nixos and getting the folliwing
<lc_>
/nix/store/6k95q6vvnmspf3c0nm8q55sixsvmn9nc-binutils-2.31.1/bin/ld: relocatable linking with relocations from format elf32-i386 (/nix/store/p234xmw3dnfn8xccpvx7s7qlim9m1yd4-wine-4.0/lib/wine/libwinecrt0.a(exe_entry.o)) to format elf64-x86-64 (lin-vst-server.aAqxyv.o) is not supported
the-kenny has quit [Ping timeout: 248 seconds]
the-kenny has joined #nixos
pukkamustard has quit [Quit: ZNC 1.7.3 - https://znc.in]
pukkamustard has joined #nixos
<gentauro>
I've been playing around with the setup up of my new laptop and grub has like a lot of different boots profiles (I would say far to many)
<gentauro>
is there a way to purge them down to at least 10 (ish) og rather 5 (ish)?
<angerman>
I've tried to do it without the release-lib logic as that introduces quite a bit of complexity I'd like to keep out for now. But I'm completely puzzled as to where the gmp issue comes from now...
<clever>
ive also had plans to replace the nixpkgs release-lib, since it doesnt really fit with our needs
<angerman>
My intention was to keep it as simple as possible, as my team isn't the most nix experienced, and i'd still like them to understand what that little code we have does.
jackdk_ has quit [Ping timeout: 244 seconds]
<angerman>
so I'm tying out stuff more verbose than it probably needs to be, but trying to keep it as easy as possible to follow. And I didn't expect `gmp` to come up as an issue. So now I'm wondering what the heck I did wrong.
linarcx has joined #nixos
linarcx has quit [Client Quit]
mexisme has joined #nixos
<angerman>
I though it should be as simple as setting `system` and `crossSystem`, but apparently I'm missing some vital ingredient.
<clever>
ive modified jobs= to expose that entire pkgs attr, so i can now view it in repl
<clever>
as a side-effect, hydra would try to (cross) build all of nixpkgs, so this kind of thing shouldnt be pushed
Ariakenom has joined #nixos
<angerman>
not? 🙄
adamantium has joined #nixos
<angerman>
interestingly enough if I just build `x86_64-mingw32.pkgs.gmp`, I get /nix/store/wndwgmxmnx53q6pdnld2l22jic9afn6g-gmp-6.1.2-x86_64-pc-mingw32
___laika has quit [Ping timeout: 256 seconds]
<angerman>
so (at least to me) it's getting more confusing by the minute.
<clever>
nix-repl> :b x86_64-mingw32.pkgs.gmp
<clever>
out -> /nix/store/wndwgmxmnx53q6pdnld2l22jic9afn6g-gmp-6.1.2-x86_64-pc-mingw32
<klaas_>
hi all; after an upgrade to 19.03 (from 18.03) it seems that virtualbox shares are broken. I must be doing something wrong, because this is hardly a unique configuration I'd say
<klaas_>
@averall, re extension pack -- I thought the line "virtualisation.virtualbox.guest.enable = true;" was enough... what else should I add or do?
<mpickering>
FRidh: Thanks. Apparently it should be `lib.version` instead but that worked
<klaas_>
"is the module loaded"... modprobe: ERROR: could not insert 'vboxsf': Unknown symbol in module, or unknown parameter (see dmesg)
<averell>
enableExtensionPack = true, just like normal enable. but pure mounts should probably work without, i don't remember.
<FRidh>
seems like the help in nix eval needs to be updated then
<averell>
that sounds like a kernel version incompatibility, maybe a reboot could help
<klaas_>
I've been rebooting many times recently so I doubt that
<klaas_>
"nixpkgs.config.virtualbox.enableExtensionPack = true;" added and rebooted once more, but no banana
<klaas_>
there's also this in my journalctl: vboxsf: loading out-of-tree module taints kernel.
jb55 has quit [Ping timeout: 246 seconds]
ayerhart_ has joined #nixos
jackdk_ has joined #nixos
ayerhart has quit [Ping timeout: 258 seconds]
ThatDocsLady has joined #nixos
<{^_^}>
[nix] @domenkozar pushed to mojave-$HOME « install-multi-user.sh: don't break on Mojave default $HOME »: https://git.io/fjlC1
<{^_^}>
[nix] @domenkozar opened pull request #2808 → install-multi-user.sh: don't break on Mojave default $HOME → https://git.io/fjlCM
<averell>
ah, i suppose it's a known issue, #58127
ee1943 has quit [Read error: Connection reset by peer]
___laika has joined #nixos
ee1943 has joined #nixos
orivej has joined #nixos
cfricke has quit [Quit: WeeChat 2.4]
<klaas_>
yeah that looks very much like my situation
<averell>
fix is in 19.03, so i guess the next time the channel updates it might work again :)
turion has joined #nixos
mexisme has joined #nixos
<turion>
What's the meaning of store paths like /nix/store/eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee-go-1.11.6/?
<averell>
is it an actual store path? otherwise it looks like "nuke-refs"
<averell>
(basically to prune dependencies)
Jetien_ has quit [Ping timeout: 258 seconds]
<turion>
averell: Good point, this path doesn't exist
<turion>
Ah, learned about a new thing, thanks :)
<turion>
averell: +1
<turion>
+1 averell
<turion>
How does this work again
Jetien_ has joined #nixos
<Synthetica>
averell: Is there a reason that's not mentioned in the manual?
<Synthetica>
(Or am I missing it?)
<klaas_>
averell, if I understand correctly, that also means that rebuilding my system against the 19.03 that's in git will fix the problem (at the expense of warming up my CPU to do a rebuild), right?
<{^_^}>
[nixpkgs] @endgame opened pull request #61536 → configuration-hackage2nix.yaml: lattices is not broken → https://git.io/fjlWl
jackdk_ has quit [Ping timeout: 246 seconds]
fendor_ has quit [Ping timeout: 246 seconds]
st4ll1 has quit [Ping timeout: 252 seconds]
jb55 has joined #nixos
vs^ has quit []
abaiste^ has joined #nixos
<averell>
yes, i'd say so
<averell>
it's probably just a kernel rebuild and some minor changes since yesterday anyway
<Izorkin>
etu: ok. Can replace buildInputs = [ pkgs.libcouchbase pkgs.zlib igbinary pcs ]; to [ libcouchbase zlib igbinary pcs ]; Need to normal work override package openssl to libressl
<{^_^}>
[nixos-weekly] @domenkozar pushed 3 commits to production: https://git.io/fjl8Q
<{^_^}>
[nixos-weekly] @domenkozar pushed to master « Call for Content: 2019/10 »: https://git.io/fjl8F
<{^_^}>
[nixos-weekly] @domenkozar opened pull request #94 → Call for Content: 2019/10 → https://git.io/fjl8b
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
<srhb>
alexarice[m]: Which one exactly? And yes, Hydra knows.
<srhb>
alexarice[m]: (But of course, it doesn't know how long it takes on *your* hardware :-P)
<alexarice[m]>
gcc-7.4.0
<alexarice[m]>
was just hoping for a rough bound, trying to git bisect nixpkgs
<srhb>
alexarice[m]: Right, check Hydra's time history then :)
<gchristensen>
alexarice[m]: you'll probably want to `git bisect skip` mass rebuilds
Mister_Magister has quit [Quit: bye]
<srhb>
(But yeah :-P)
<gchristensen>
alexarice[m]: you can determine if you have a mass rebuild or not by checking to see if this exits successfully (indicating it will not be a mass rebuild): nix-build . -A stdenv -j0
<alexarice[m]>
gchristensen: I'm not sure I have much choice at this point, I'm down to the last few commits
<gchristensen>
ah, well, ... yep :P
<alexarice[m]>
it is doing a mass rebuild, I'll try some other commits but I'm not hopeful
<srhb>
alexarice[m]: Looks like Hydra took about 30 minutes on ike the last few times. Not sure how parallel that is.
<alexarice[m]>
srhb: thanks, hopefully won't take days then
<alexarice[m]>
can deal with hours
<srhb>
As long as you don't need, like, a browser, a day is the upper bound for a full system, I find...
<gchristensen>
ike has 48 cores, and the `cores` parameter is set to 0
<srhb>
So quite parallel :D
<gchristensen>
it also runs 48 jobs, so ... expect contention I guess
<Izorkin>
etu: I want to use the following patch, which enables me to build plugins with libressl instead of openssl - https://pastebin.com/YiMrpM0h
<siers>
roberth, that comparison table is a bit lacking though :)
<etu>
Izorkin: I see
<siers>
roberth, I don't think icetan/Profpatsc+h have comprehensive tests
zupo has joined #nixos
kisik21 has joined #nixos
<kisik21>
So. Let's say I have a Python application and I don't wanna mock a Redis server in the check phase. Can I somehow launch a Redis server (in a container maybe?) for a test and tear it down when the tests finish?
<clever>
kisik21: if the nix sandbox is enabled, your already in a container
<siers>
oh, icetan is in this channel too hehe
<kisik21>
clever: does that container include TCP/UDP connections? Do I have my own port space at least for localhost?
zupo_ has quit [Ping timeout: 244 seconds]
<kisik21>
Interesting. If it's true that means that there are three, not two types of packages. Those that build in sandbox well, those that not build in it and those that can fail outside sandbox...
<clever>
kisik21: yep, thats how nix stops you from getting internet access
<kisik21>
Wow, cool!
jasongrossman has quit [Ping timeout: 246 seconds]
<siers>
roberth, but it's nice that there's some inovation in this regard now and that my code helped. A while ago there were no tools to do this.
Tucky has quit [Ping timeout: 245 seconds]
orivej has quit [Ping timeout: 245 seconds]
jasongrossman has joined #nixos
<roberth>
siers: very good point about the tests, I'll add it to the table
balsoft has joined #nixos
<balsoft>
Hello! Are there any hydra people here by any chance?
<andi->
balsoft: best to just ask your question
<balsoft>
I have a question: is it possible to run some commands after an evaluation of a jobset succeeded?
<Taneb>
balsoft: there's a "RunCommand" plugin exactly for that
<gchristensen>
just the evaluation? or after the builds?
<Taneb>
Oh, after evaluation
<Taneb>
No, I don't think there's currently a way to do that
npmccallum has joined #nixos
<balsoft>
Hmm, can you elaborate a bit on RunCommand plugin?
<balsoft>
Taneb
oleks_ has joined #nixos
<Taneb>
balsoft: it runs a command after a build succeeds, which I've realised isn't quite what you asked
<balsoft>
That's quite interesting as well, I might be able to use it. Can you please provide some links?
<lunik1>
but nixos-rebuild --upgrade won't try to swap to the new kernel by default?
<yorick>
lunik1: it's impossible to swap linux kernels at runtime
<lunik1>
ok
<yorick>
it'll try to apply everything else (and fail if there are module changes)
oleks_ is now known as oleks
<gchristensen>
lunik1: the Linux kernel does support live/hot patching, but that is whant RedHat and Canonical charge six digits of support contract for.
abathur has joined #nixos
esmerelda has quit [Ping timeout: 252 seconds]
kisonecat has joined #nixos
agent13 has quit [Remote host closed the connection]
vid has joined #nixos
ilovealgebra has joined #nixos
srid has quit [Ping timeout: 252 seconds]
<siers>
roberth, I did say *comprehensive* test suite. :P But specifying that detail's ultimately up to you
<ilovealgebra>
Hey, so I need to delete the _newest_ generation of my system (I did a rollback and I want to delete the newest build). How do I do that?
acarrico has quit [Ping timeout: 268 seconds]
<roberth>
siers: I didn't say comprehensive comparison :P
esmerelda has joined #nixos
fendor has quit [Ping timeout: 244 seconds]
srid has joined #nixos
<ilovealgebra>
I don't think there is a command that does it so I probably need to remove some file? Can anyone help me understand how I can force the system to build the next generation from an older generation when I do a nixos-rebuild (i.e. fork the path and discard the version I currently have as most recent)
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<lucus16>
Does Nix enforce UTF-8 encoding or does it work with any encoding supported by the system?
<jasongrossman>
ilovealgebra: NixOS doesn't build a generation from a previous generation, it builds it from configuration.nix. So you have to find your previous version of that.
<ilovealgebra>
My configuration didn't change.
<jasongrossman>
ilovealgebra: Then I guess you don't have a problem.
<ilovealgebra>
I did an upgrade when I wasn't on any channel and it broke the system, now when I've added the correct channel and try to build from the next-to-newest generation it tries to use the work that went into that broken build
abathur has quit [Ping timeout: 245 seconds]
<jasongrossman>
ilovealgebra: NixOS doesn't build a generation from a previous generation, it builds it from configuration.nix.
<jasongrossman>
ilovealgebra: Have you done sudo nix-channel --update?
<ilovealgebra>
then why doesn't it overwrite the latest generation timestamp when I run this command sudo nix-env --profile /nix/var/nix/profiles/system --list-generations
<ilovealgebra>
yep I have now
__monty__ has quit [Quit: leaving]
<ilovealgebra>
but I hadn't when I built the broken system
jackdk_ has joined #nixos
linarcx has quit [Quit: WeeChat 2.4]
cfricke has joined #nixos
linarcx has joined #nixos
cfricke has quit [Client Quit]
<jasongrossman>
ilovealgebra: I don't know about generation naming - sorry. (I name my generations in a non-default way.)
rauno has quit [Ping timeout: 252 seconds]
<ilovealgebra>
okay I am going to try to boot into the new generation (I think the reason things broke was because I tried to switch into it and some services couldn't restart properly)
<ilovealgebra>
thank you
ilovealgebra has quit [Quit: Page closed]
<jasongrossman>
,tell ilovealgebra Please let us know whether rebooting solved your problem. (I hope it did.)
<{^_^}>
jasongrossman: I'll pass that on to ilovealgebra
vid has quit [Ping timeout: 245 seconds]
vid has joined #nixos
ilovealgebraicda has joined #nixos
another-user has quit [Ping timeout: 245 seconds]
<ilovealgebraicda>
hmm okay so this time it looks like the system services are sort of working (I have internet) but my wm is completely broken
alexherbo21 has quit [Ping timeout: 268 seconds]
alexherbo21 has joined #nixos
hoijui has joined #nixos
kisonecat has quit [Ping timeout: 245 seconds]
ilovealgebraicda has quit [Ping timeout: 256 seconds]
ilovealgebra has joined #nixos
kisonecat has joined #nixos
<ilovealgebra>
and also the new generation isn't even listed as 19.03 it still says 18.09 in it
<{^_^}>
ilovealgebra: 9 minutes, 9 seconds ago <jasongrossman> Please let us know whether rebooting solved your problem. (I hope it did.)
<ilovealgebra>
it did not
esmerelda has quit [Ping timeout: 255 seconds]
<hoijui>
this works on linux, but not on windows:
<ilovealgebra>
ah so the problem all along is that I was on windows
<hoijui>
how would it look on windows? I think it woudl need an extra parameter at the end "..., 0)"
<hoijui>
but also SYS_IOCTL and TIOCGWINSZ do not exist on windows
<hoijui>
I have no clue what this call does, btw. ;-)
<ilovealgebra>
I am back to square one, I want to delete the newest generation, how do I do it!
isHavvy has joined #nixos
knupfer has joined #nixos
orivej has joined #nixos
Havvy has quit [Ping timeout: 246 seconds]
<hoijui>
ah.. it returns the size of the terminal window. so I would need to know how to do that on windows
<jasongrossman>
ilovealgebra: Ah, that's why I mentioned NixOS. If you're on Windows then you're not using NixOS. Which doesn't solve your problem, but it's worth understanding anyway!
<jasongrossman>
ilovealgebra: Even on windows, it's still the case that Nix doesn't use a previous generation to build a new generation. It only uses your configuration files (and, optionally, channels).
talyz1 is now known as talyz
<jasongrossman>
ilovealgebra: So I think deleting generations is a red herring, if your aim is to get your new generation working.
<ilovealgebra>
I was joking
<ilovealgebra>
haha
<ilovealgebra>
I'm on nixos
<ilovealgebra>
was going to do a /s but it felt overblown
asheshambasta_m has quit [Quit: Leaving.]
<jasongrossman>
ilovealgebra: OK, good joke. So what about your new generation is not working?
<jasongrossman>
ilovealgebra: If the problem is your WM, then which WM? Etc.
<ilovealgebra>
I have a not fully set up xmonad so I use plasma
hedning_ has quit [Remote host closed the connection]
<ilovealgebra>
and when I logged into it the background picture was black and the bar didn't load
<ilovealgebra>
or it loaded and crashed
<ilovealgebra>
I was left with this firefox window
<ilovealgebra>
but couldn't really do anything
__monty__ has joined #nixos
<ilovealgebra>
if I rebuild switch into it then it's even worse because almost all systemd services are dead
esmerelda has joined #nixos
<ilovealgebra>
and I end up with a completely non functioning system
<ilovealgebra>
but then the wm works
<teto>
any idea how to get the nixos config from within a home-manager module ?
<teto>
infinisil: but home-manager can work without nixos ? so I should check it exists first right ?
<ilovealgebra>
but I strongly suspect this generation isn't rebuilding properly (for example the timestamp doesn't change and the version number doesn't change even after joining the channel and updating the configuration.nix)
<infinisil>
teto: Oh you want to do this for home-manager upstream?
<ilovealgebra>
I just did a sudo nix-store --verify --check-contents --repair and everything seems to be fine btw
<teto>
infinisil: kinda, I would like home-manager systemd services to inherit the nixos proxy settings. kinda systemd.user.sessionVariables = args.nixos.config.networking.proxy.envVars;
<teto>
I believe they should inherit everything from nixos sessionVariables but not sure
<infinisil>
Eh, you should probably rather use HM's NixOS integration and then propagate these options to the home-manager options from there
<Synthetica>
Is there a way to scaffold a VSCode extension on NixOS? They recommend `npm install -g yo generator-code`, but that doesn't work on NixOS
alexherbo21 is now known as alexherbo2
jackdk_ has quit [Remote host closed the connection]
cfricke has joined #nixos
fendor has quit [Ping timeout: 244 seconds]
<pbb>
does anyone know how to load a device tree overlay with NixOS on the Raspberry Pi 3B+?
asheshambasta_m has joined #nixos
silver has quit [Read error: Connection reset by peer]
knupfer has quit [Remote host closed the connection]
knupfer has joined #nixos
kisonecat has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @nixbitcoin opened pull request #61542 → rednose: disable test suite → https://git.io/fjlRh
silver has joined #nixos
<{^_^}>
[nixpkgs] @nixbitcoin closed pull request #61238 → [19.03] rednose: disable test suite → https://git.io/fjCaK
camsbury has joined #nixos
esmerelda has quit [Ping timeout: 245 seconds]
Makaveli7 has quit [Quit: WeeChat 2.4]
pie_ has joined #nixos
sigmundv has quit [Ping timeout: 268 seconds]
sigmundv has joined #nixos
jasongro` has joined #nixos
jasongrossman has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @vbgl pushed commit from @CohenCyril to master « coqPackages: refactor mathcomp packages »: https://git.io/fjl0R
Jackneill has quit [Read error: Connection reset by peer]
fendor has joined #nixos
Glider_IRC has joined #nixos
<Ericson2314>
pie_: yes i made the gazillion types of dependencies.
<Ericson2314>
clever: you actually can manually call stage.nix with build != host != target
<Ericson2314>
Need to make sure the right `pkgs{Build,Host,Target}^2` are passed in, but it "should" work otherwise, to the extent the individual packages can handle it
<matthewbauer>
clever: does clang need the same flag as well maybe?
jD91mZM2 has joined #nixos
<clever>
matthewbauer: my understanding is that clang just calls out to llvm to do 90% of the work, but it feels like clang has a whitelist of platforms, and is blocking what should just work
petrkr has quit [Quit: Leaving]
<clever>
although, i was also jamming a 2nd llvm into the buildInputs, rather then changing clangs llvm
<matthewbauer>
i mean that there are some libraries provided by clang, but also some binaries provided by llvm
jD91mZM2 has joined #nixos
<matthewbauer>
libclang, llvm-ar, etc.
<Ericson2314>
oh
<Ericson2314>
sur
<Ericson2314>
someday when cc-wrapper is a lot less tangled in everything
<Ericson2314>
this will be easier
Glider_IRC_ has joined #nixos
<jD91mZM2>
How does one update a nixops configuration? It seems nix-channel reveals it's using 17.09 (which is older than when I created my nixops deployment!), and it never seems specified anywhere
<jD91mZM2>
Or should one just update the channel using the regular nix way, even though the usual way to rebuild is through nixops deploy?
<aminechikhaoui>
jD91mZM2 you can have nixpkgs in the path of the deployment, e.g `nixops create -d <deployment> -I nixpkgs=channel:nixos-19.03`
<aminechikhaoui>
doesn't have to be like that, for example you could have a local checkout so that you manage the updates
Glider_IRC has quit [Ping timeout: 244 seconds]
<jD91mZM2>
Oh, that doesn't rebuild anything aminechikhaoui. Guessing my local channel overrides the remote's channel. So maybe I had the right channel after all!
<clever>
deploy also obeys -I foo=bar, but it wont remember it
<jD91mZM2>
What was that website to check the last released commit for each channel? I want to see if the latest kernel is released yet, heard about the ZombieLoad security issue thing
<clever>
jD91mZM2: and i found that last one, via the "spam" {^_^} just gave :P
<gchristensen>
is it too spammy?
<clever>
not really
<jD91mZM2>
Darn it, no kernel update released... Might have to override the channel with a git commit
<clever>
gchristensen: sometimes the reverse, ive had to search thru a dozen pages of irc logs to find that URL before
praduca has quit [Remote host closed the connection]
<qyliss^work>
jD91mZM2: make sure to upgrade your microcode too (patch is in master), or you won't actually have the mitigation. Just a kernel that knows it's vulnerable. AAUI
<lucus16>
How do I evaluate a nix file, without building or anything, just having it print the result of an expression?
Ariakenom has quit [Quit: Leaving]
<clever>
lucus16: nix-instantiate
<clever>
possibly with --eval
<jD91mZM2>
qyliss^work: Right, that should update along with the rest of the system. Apple Attachment Unit Interface?????
<qyliss^work>
As I understand it
<lucus16>
clever: I tried `nix eval -f file.nix` but it says more arguments are required
<jD91mZM2>
qyliss^work: Ah
<qyliss^work>
Yeah, it'll come along once it gets into a channel, but if you're going to cherry-pick the kernel upgrade, you probably want to cherry-pick the microcode update too
<Baughn>
jD91mZM2: I have nixops configured to also push the channel and set nixpkgs. Doesn't usually matter, but it makes nix-shell and such work as expected.
<lucus16>
Ah thanks, that works
<lucus16>
although the string it prints is not properly escaped
<{^_^}>
[nixpkgs] @cizra opened pull request #61546 → libfprint: added a fork for Lenovo ThinkPad → https://git.io/fjluR
jD91mZM2 has joined #nixos
<samueldr>
though `tested` failed in it, but just restarted it
<samueldr>
(a test timed out)
<jD91mZM2>
Sorry about that relog - patched now!
<samueldr>
qyliss^work: though this assumes the microcode update was in 6h ago
<qyliss^work>
microcode is the tip of nixos-19.03-small
<samueldr>
qyliss^work: bust since the next eval doesn't seem to have too many rebuilds it shouldn't be held back too much
MmeQuignon has joined #nixos
halfbit has joined #nixos
<qyliss^work>
I committed it 7 hours ago, and it made it into that channel's update 6 hours ago
srid has quit [Ping timeout: 244 seconds]
<samueldr>
(and yes, also verified by the Inputs tab of the eval, under Changes, 7cd2e4 to 570811 )
<Unode>
hi all, my laptop ran out of battery and I now can't boot due to systemd service errors. I tried older grub options but they all fail the same. Is there any option that provides a rescue boot?
<qyliss^work>
You could create an install disk and boot from that
<Unode>
systemd is just loop failing on starting name service cache daemon and login service
<Unode>
i dont have a stick at hand so i'm trying to salvage the current system
<bahamas>
hello. I'm trying to run `nix-shell release.nix` on the first file found on this paste https://bpaste.net/show/5a9745817ba5. I get an error "gradle: readCreateProcess: runInteractiveProcess: exec: does not exist (No such file or directory)" when building jvm-batching. can anyone tell me how to fix this?
<Unode>
fails that as well, complains that it cant open the password database
<{^_^}>
[nix] @edolstra pushed to print-build-logs « nix: Add --print-build-logs flag »: https://git.io/fjlu1
<Unode>
not sure what happened
chaperonemugdog has joined #nixos
<Unode>
:[
<samueldr>
is /etc there and populated as expected%?
<samueldr>
(should be able to see after debug1mounts)
srid has joined #nixos
<{^_^}>
[nix] @edolstra opened pull request #2810 → nix: Add --print-build-logs flag → https://git.io/fjluM
<Unode>
i'm seeing some errors durimg initrd but they go by too quickly to read
<chaperonemugdog>
Hi guys. I have just submitted PR #61545 and it seems that I will need to rebase this to staging. I was wondering which staging branch I should go for, as there are quite a few.
<chaperonemugdog>
Or is it sufficient to edit the base branch ?
<qyliss^work>
You can edit the base branch, but after that, if GitHub shows you lots of commits you don't recognise, you'll need to rebase on the command line and force push.
_e has joined #nixos
<qyliss^work>
The branch you want is the one just called "staging"
mexisme has joined #nixos
drakonis has quit [Quit: WeeChat 2.4]
sigmundv has joined #nixos
<chaperonemugdog>
Okay I did that. I indeed had to rebase on staging and force push
<{^_^}>
[nixops] @PsyanticY opened pull request #1149 → ec2 fleet and launch template resource. → https://git.io/fjlzv
esmerelda has quit [Ping timeout: 268 seconds]
<chaperonemugdog>
Thanks for the info :)
mexisme has quit [Ping timeout: 246 seconds]
domogled has joined #nixos
PhawaS0ze has joined #nixos
<pie_>
apparently path literals dont work if you reference something in the store?
chaperonemugdog has quit [Quit: Page closed]
srl295 has joined #nixos
<infinisil>
OHh I have the perfect solution for creating serial numbers for SOA DNS records
<infinisil>
The problem for those unfamiliar: Need to generate an increasing sequence number every time you change DNS entries. But the config should still be pure, so you can't just store the last sequence number in a file
<infinisil>
(well technically you can, but I'd like to avoid state)
<gchristensen>
oh?
<clever>
infinisil: impurely call `date` ?
<infinisil>
clever: Indeed, but there's a bit more to
<infinisil>
it
bahamas has quit [Quit: leaving]
<infinisil>
Hold on I'll paste the code
wfranzini has quit [Remote host closed the connection]
<gchristensen>
does your proposal handle rollbacks?
<clever>
yeah, rollbacks will randomly either go to an old seq number, or a new one (depending on how aggressive you GC)
<clever>
and then the dns wont go backwards
<gchristensen>
right, a rollback needs to generate a new, incremented serial
<clever>
one solution ive thought of, is to generate the seq number when bind starts
<infinisil>
Hmm okay I haven't thought about that, but let's ignore that right now
<{^_^}>
[nixops] @AmineChikhaoui pushed 2 commits to master: https://git.io/fjlzs
<gchristensen>
nice
<infinisil>
Important point is the `import (pkgs.runCommand ...)`, aka IFD
<gchristensen>
another decent option is just ignore serial numbers, use a dns service which just emits a current timestamp for serial numbers and a lowish ttl
<MmeQuignon>
Hi !
<infinisil>
Which takes as input all the records, without using them, and generates a date serial number
<infinisil>
Only if the records change, the derivation has to be rebuilt
<MmeQuignon>
An now I try to spawn the command : nix-shell --run "make html"
<MmeQuignon>
And in the makefile, the command sphinx-build is casted but it raises "command not found"
<MmeQuignon>
But if I spawn this command : nix-shell -p python3Packages.sphinx --run "make html", everything works
<MmeQuignon>
I do something wrong, but I can't figure out what
<aanderse>
so according to the nixpkgs manual: The version part of the name attribute must start with a digit (following a dash) — e.g., "hello-0.3.1rc2".
<aanderse>
but most packages which have "unstable" in the name put "unstable-YYYY-MM-DD" as the version
<aanderse>
hence the version string no longer starts with a digit
acarrico has joined #nixos
ubert has quit [Quit: Leaving]
<aanderse>
so documentation should be updated to clarify?
<MmeQuignon>
aanderse: I'm not sure, are you answering to me ?
<MmeQuignon>
^^'
<aanderse>
MmeQuignon: sorry, no, entirely unrelated
<aanderse>
ha sorry
<MmeQuignon>
no trouble ^^
<MmeQuignon>
that's what I thought ^^
<aanderse>
doing a quick search i saw both pname = "app-unstable" + version = "YYYY-MM-DD"; AND pname = "app" + version = "unstable-YYYY-MM-DD" in nixpkgs
<aanderse>
so i guess it doesn't matter *too* much, but consistency is preferred of course
<aanderse>
hmm
<dhess>
Ergh, looks like a garbage collection wiped boot config on 2 of my NixOS hosts: /nix/store/4ipwagwkdh4l6mw9sfw4spzaj6l8hax7-grub-2.04-rc1/sbin/grub-install: error: cannot find a GRUB drive for /dev/xvda. Check your device.map.
<dhess>
anyone know how to recover from that?
<infinisil>
gchristensen: clever: Although, if both master and slaves are controlled via NixOS (e.g. nixops), if you roll them back to the same point, serial number's don't matter, because they're only used to sync between master/slaves
<infinisil>
It would only be a problem if you roll back only master
waleee has quit [Quit: WeeChat 2.4]
<gchristensen>
true
<gchristensen>
so then who cares :)
<infinisil>
Well they still need to be increasing
<gchristensen>
what for?
<infinisil>
So slave knows it should sync the zone
<infinisil>
It's what serial numbers are for
<gchristensen>
why sync if you're using nixops
<infinisil>
Oh, good point lol
<infinisil>
Man now I feel stupid
<dhess>
ok any pointers on how to debug grub issues on NixOS then?
<gentauro>
I onle see -> `* nixpkgs.unoconv (unoconv-0.6)` :|
<gentauro>
I'm on 19.03 btw
vmandela has quit [Quit: Leaving]
<aanderse>
gentauro: try adding pkgs.libreoffice to configuration.nix and see if it installs
<ottidmes>
gchristensen: could you perhaps give me some more details on https://github.com/NixOS/nixpkgs/issues/42127#issuecomment-487027928 "although it has problems with running out of space on /boot". I just received my new laptop and want to make sure I organize my disks properly to make it have secure boot (at some point in the near future)
domogled has quit [Quit: domogled]
<gentauro>
aanderse: it did
<gentauro>
but as a "n00b" I find it difficult to search for packages ...
orivej has quit [Ping timeout: 258 seconds]
<aanderse>
gentauro: yeah, that doesn't sound right
<gentauro>
is it cos `nix` is out of date? I though it was updated everytime I run `nixos-rebuild switch`
<gchristensen>
ah
<gchristensen>
yes, ottidmes
Jetien_ has quit [Ping timeout: 245 seconds]
<ottidmes>
gchristensen: thanks :)
Jackneill has joined #nixos
<gentauro>
aanderse: I haven't done anything "crazy" I have installed everything (except Haskell stack) with `configuration.nix`
<gentauro>
so I have that going on for me :)
<gchristensen>
> Finally, you can have much fewer generations: each generation on mylaptop takes 20MB in /boot.
<{^_^}>
error: syntax error, unexpected ',', expecting ')', at (string):253:8
<gentauro>
any guides on how to install "custom" fonts which doesn't exist as a package?
Jetien_ has joined #nixos
<gentauro>
normally in other GNU/Linux distros, you just place fonts in `~/.fonts` and run `dpkg-reconfigure fontconfig-config`
<gentauro>
(debian / ubuntu)
<ottidmes>
gchristensen: ah thanks for the numbers :) I plan to keep room for an Ubuntu LiveCD and NixOS rescue on /boot so then 4GB should suffice even with 20MB generations
<samueldr>
gentauro: if you're fine keeping them at the user level, putting them in ~/.local/share/fonts (the new XDG compliant location) should be enough
<gchristensen>
cool
<samueldr>
and 99.9% of the time no fc-cache or fontconfig reconfiguration is necessary
<samueldr>
even on other distros
<ottidmes>
gchristensen: and I can do secure boot after the fact, right? (not that I put in the effort of installing and configuring my applications only find out this weekend that I need to redo it all)
<gchristensen>
yep
<ottidmes>
pfew :)
logzet has quit [Remote host closed the connection]
logzet has joined #nixos
<ottidmes>
gchristensen: ow BTW I remember you trying to go the most things in memory route, do you still do that, or did it not really work out? (got 32GB RAM in my new laptop so I now have room to try such things)
drakonis has joined #nixos
<gchristensen>
ottidmes: I have servers running all in RAM, but never for my laptop. my laptop has a / partition which is erased on every boot
[Leary] has quit [Read error: Connection reset by peer]
Lears has joined #nixos
<gchristensen>
(so, indeed, it doesn't use up precious RAM.)
drakonis_ has quit [Ping timeout: 244 seconds]
<ottidmes>
exactly, meant the hybrid approach (all memory does not really makes sense for a laptop)
<iddt>
ya, on the advice of some website that suggested i do it.
<goibhniu1>
gentauro: is the ownership correct for the font files?
<iddt>
too stupid to know better
alexherbo299 has joined #nixos
<gchristensen>
do you have a link to the website?
<gentauro>
samueldr: as a newcomer to NixOS I got terrified to hear the word "works fine here" :|
<ottidmes>
iddt: tip, never go with the manual deletion route of anything in /nix unless you absolutely know what you are doing, I did that when I started out with NixOS and made it buggy
alexherbo29 has quit [Ping timeout: 268 seconds]
<samueldr>
gentauro: :) thing is with installing fonts in ~/.local/share kinda sidesteps some of what NixOS does
<iddt>
ott, sounds like deep wisdom
<samueldr>
alternatively you could package the fonts you need within a (private I assume) overlay
<iddt>
gchristensen: i will try to dig it up
<gchristensen>
thanks
<emilsp>
hrm, last update to gnome-shell makes it not stable, especially when using virt-manager in fullscreen mode
<gentauro>
samueldr: all files have 644
<iddt>
gchristensen: is it safe to reboot now that i've done this awful thing?
<emilsp>
seems like alt-f4 in fullscreen mode of virt-manager will trigger a crash when keyboard shortcut interception is disabled.
Myhlamaeus has joined #nixos
<gchristensen>
hmm I hope so =) do you have backups?
<iddt>
it is not a precious system image to me. not a biggie to reinstall.
<gchristensen>
great to hear
ng0 has joined #nixos
<gentauro>
samueldr: folders 700 and files 644 ...
<gentauro>
I will delete everything under the folder and download the fonts once again ...
<gentauro>
it might help
<iddt>
yay, it rebooted. the booted-system symlink is back. didn't get any smaller tho. is there something i should now run to make it tighter?
<goibhniu1>
gentauro: maybe `fc-cache -r` helps?
<iddt>
it takes about 2.5gb w/ just about nothing installed on it, whereas it used to take about 1.7gb i believe, maybe less.
<iddt>
that seems like a big jump
<gentauro>
goibhniu1: and we have a winner -> `/home/johndoe/.local/share/fonts/proza-libre: caching, new cache contents: 12 fonts, 0 dirs` Thx !!!
<goibhniu1>
yay \o/
<gentauro>
goibhniu1: updating my script from `fc-cache -v` to add the `r` parameter :)
<ottidmes>
iddt: did you run garbage collect again?
<iddt>
ottidmes: yessir
<iddt>
ottidmes: nix-garbage-collect -d, right?
<ottidmes>
yep, as root, that ought to do it
fendor has quit [Ping timeout: 246 seconds]
<iddt>
ottidmes: i guess it just grew.
<gentauro>
iddt: how do you measure nixos size (just curious)?
dbmikus has joined #nixos
<iddt>
gentauro: df
<iddt>
gentauro: nothing exciting happening on the system before/after the upgrade
<iddt>
gentauro: basically a clean system
<ottidmes>
iddt: du --human-readable --total --summarize $(nix-store --query --requisites /run/current-system) > >(sort --human-numeric-sort), should give some insight
vid has joined #nixos
pie_ has quit [Read error: Connection reset by peer]
<gentauro>
iddt: fair enogh
<iddt>
ottidmes: that is illuminating. thank you. there are some decidedly bloated things on here.
<gentauro>
I went down from my previous laptop from a 512GB SSD to a 256GB cos I really don't have that much and was worried that nixos could go crazy and use all the space :)
<gentauro>
but it seems stable around 42 GB (where most is my data)
<hyperfekt>
What would be the most idiomatic way to wrap directories around something in Nix? Say I have a package that produces directory a and I want a package that has a directory b that contains directory a.
<gchristensen>
a symlink I suppose?
<hyperfekt>
Okay, so there's no readymade builder. Thanks!
<Unode>
wondering if you could give some hints about rebooting the system. I managed to get a USB stick with systemrescuecd and boot the system
<clever>
Unode: rebooting which system?
<Unode>
checked the filesystems, nothing obviously wrong
<clever>
oh right, the thing that isnt booting right
<Unode>
yeah sorry for lack of context
ilovealgebra has quit [Ping timeout: 256 seconds]
tdeo` has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
I now have a chroot into the main system but I'm clearly missing something since the bash that I'm running mostly complains that it doesn't know any commands
<clever>
Unode: you want nixos-enter
<Unode>
I remember that there was some script that could be sourced to setup most of this up but I can't find it on my notes and google isnt helping
<clever>
that will chroot, and set the env vars right
<Unode>
do I simply run that with a full path? /mnt/tmp/nix/store/.../nixos-enter ?
pshevtsov has left #nixos ["WeeChat 2.4"]
<clever>
that will likely fail
<clever>
it needs to be installed on the host end
<Unode>
so I should get a nixos live cd instead?
pereiraalex[m]1 has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
you can also try installing nix, and then just nix-env the right thing
<clever>
or just ignore nixos-enter, and manually setup the env vars
<clever>
that will get you 80% of the stuff you expect
grahamc[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
i'm not sure if its all documented in one spot, i mostly just memorized how everything works
<clever>
gchristensen: you got booted!
<gchristensen>
eh?
<clever>
gchristensen: from matrix
<clever>
2019-05-15 14:44:52 -!- grahamc[m] [gchristens@gateway/shell/matrix.org/x-tgmttfnyebwhgwkt] has quit [Quit: Idle kick: User has been idle for 30 days.]
<gchristensen>
oh hah
<gchristensen>
I'd turn it off but I don't know how :P
<clever>
it turned itself off :P
<gchristensen>
good enough for me
<Unode>
clever: thanks, I'll try to run with that
<Unode>
I think I also picked the wrong bash cause this one doesnt even give me completion
<clever>
use the bash from the path i gave above and you should be good
<clever>
though completion mostly comes from bash sourcing the bashrc stuff in /etc
torched[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle kick: User has been idle for 30 days."]
<Unode>
ok I see what happened
<Unode>
disk got full
<Unode>
and seems like nixos sql db got corrupted in the process
<clever>
what points towards it being corrupt?
<Unode>
when I run nixos-rebuild I get IO errors
<Unode>
but let me clear some space and see if that solves the issue
<clever>
that doesnt mean the DB is corrupt
<clever>
check dmesg
<clever>
what does dmesg say?
ststefa[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
___laika has joined #nixos
<gchristensen>
anyone have a link handy for docs on signing paths for distribution?
aaronduino[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
cleared some space and trying a reboot now...
<clever>
Unode: IO errors feels more like a failing drive
<Unode>
clever: yeah but I ran a couple of disk checks and everything passed
<Unode>
although the main partition was indeed full
<Unode>
so I wonder if that corrupted something
<clever>
sqlite is fairly hardened against a lot of things
<Unode>
boot is still unhappy after clearing space
<Unode>
giving it some time to see if it repairs
hyperfekt has quit [Ping timeout: 256 seconds]
johnw has joined #nixos
<Unode>
same errors as before on boot, I think I'll need to nixos-rebuild this one
dmitry-zen[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
is the sqlite db regenerated with it?
<clever>
Unode: the sqlite db keeps track of what is in /nix/store/
<clever>
and is modified every time that is modified
<clever>
if the db is lost, then so is all of /nix/store, and it will need to re-download/build everything
<Unode>
which is something I wasn't doing when the disk got full
<clever>
nix-env profiles will also be lost
kevincox[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
yeah, its unlikely to corrupt if it wasnt in use when you ran out
mkoenig has quit [Remote host closed the connection]
<Unode>
well, storage got full and then battery died
vid has joined #nixos
rfold has joined #nixos
<Unode>
after that it doesn't boot
<Unode>
I also see some root errors during initrd but too fast to read :(
<clever>
Unode: you can run dmesg to see those errors
<Unode>
I can't get a shell in that case
<Unode>
even single mode fails
<clever>
Unode: try from the usb stick you booted from, after mounting the rootfs
<clever>
no need to chroot
nokomprendo[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
hum, how does that work?
hoijui has quit [Ping timeout: 250 seconds]
<clever>
just mounting it may show some things
<Unode>
the usb stick has system rescue cd
<Unode>
oh
<Unode>
will try
melchips[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
mkoenig has joined #nixos
tuxubuntu[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<cizra>
Is there a way to exempt some generations from nix-collect-garbage -d?
peter2121[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<cizra>
ie. retain some historic known-good configurations
<clever>
cizra: thats why i just dont -d, and i manually delete generations
cocreature has quit [Quit: A monad is just a monoid in the category of endofunctors, what's the problem?]
<clever>
cizra: and instead of list, you can also ` --delete-generations 42`
<Unode>
rootfs is btrfs, it mounts fine. also ran btrfs check, no errors. dmesg shows 'disk space caching enabled', 'has skinny extents' 'enabling ssd optimizations' other than that is all fine
<Unode>
no additional errors
<clever>
Unode: try to chroot again then, using the bash from the system profile, and then `nixos-rebuild build` and see what happens
<Unode>
should I moutn bind /dev /proc /sys as well or not needed?
<dhess>
ugh, EC2 instance totally hosed, "error: symbol `grub_file_filters` not found."
<cizra>
clever: Perrrrfect! Delete everything from that list and only checkpointed generations will be left.
cocreature has joined #nixos
<dhess>
now is when I really regret using EC2 -- no console access
PhawaS0ze has quit [Ping timeout: 258 seconds]
<clever>
dhess: if you need to, you can disconnect the root EBS, reconnect it to a second instance, and then mount the device and poke around
<clever>
dhess: then reverse those steps to put it back where it belongs
<dhess>
clever: yeah I suspect that's what I'll have to do.
esmerelda has quit [Ping timeout: 258 seconds]
toppler`` has quit [Ping timeout: 258 seconds]
<clever>
dhess: ive also had plans to build ssh into the bootloader somehow, either directly into grub, or by booting into a thin distro, that presents a menu, and kexec's into the real distro
<dhess>
it's a mostly stateless machine so I could just recreate it.
<dhess>
what I'm really worried about is I have another EC2 instance where grub 2.04-rc failed and that one is not stateless.
<Unode>
clever: nixos-rebuild build shows a few warnings but no errors
<dhess>
I haven't rebooted it yet so I'm ok for now
<Unode>
nixos-rebuild switch fails with and error:
<clever>
Unode: switch will always fail under chroot
<clever>
yangm97: you want boot
<Izorkin>
etu: how to use overrideAttrs in nur-packages?
<clever>
oops, wrong tab-complete
<dhess>
so I should probably figure out how to repair this one first I suppose
<etu>
Izorkin: I think you do it the same as in your config?
<etu>
Izorkin: Why would it be different?
<Unode>
ok boot works as well
* etu
has never used NUR
<clever>
Unode: boot will update /boot for whatever it just build
<Unode>
clever: so all seems ok
<etu>
So I don't know, but my guess it's the same
<clever>
Unode: then try rebooting again
toppler`` has joined #nixos
<Unode>
clever: one more thing that might matter
<Unode>
I have the kernels in a separate boot partition
<Unode>
does nixos-rebuild take care of moutning it too_
<Unode>
?
<clever>
nope
<Unode>
or do I need manual action there as well?
<clever>
it assumes systemd took care of that already
<clever>
so you need to mount everything that matters where it belongs
<Unode>
ok
<dhess>
how do I manually re-install grub in NixOS?
<Unode>
so this is a bit weird cause /boot isn't/wasn't empty
<clever>
Unode: thats why i like to `chmod 0 /boot` before mounting it
<clever>
so if i ever forget to mount, it fails hard
Purple-mx has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
rather then silently putting things in the /boot dir of /
<Izorkin>
etu: with this variant php73Packages-unit = pkgs.php73Packages.override { php = php73-unit; openssl = pkgs.libressl; libevent = pkgs.libevent.override ({ openssl = pkgs.libressl; }); libcouchbase = pkgs.libcouchbase.override ({ openssl = pkgs.libressl; }); }; - error - nixpkgs/pkgs/top-level/php-packages.nix:1:1 called with unexpected argument 'openssl', at /nixpkgs/lib/customisation.nix:69:12
___laika has quit [Quit: WeeChat 2.4]
<dhess>
clever: ok another q -- this host is managed by NixOps and I never nixos-rebuild on the actual host. Any trick to making sure nixos-rebuild uses the deployed config?
<clever>
dhess: one minute
<clever>
40 --install-bootloader)
<etu>
Izorkin: Override is not overrideAttrs
<Unode>
oh actually I might be mixing things, seems like this boot partition is from my old gentoo installation
<clever>
41 export NIXOS_INSTALL_BOOTLOADER=1
zupo has quit [Ping timeout: 246 seconds]
zupo_ has joined #nixos
offlinehacker has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
dhess: first, you cant really get nixos-rebuild to easily respect the nixops config
<etu>
Izorkin: With overrideAttrs you'd override the attributes, for example buildInputs = [ libressl ];
<clever>
dhess: but, you can just run `/run/current-system/bin/switch-to-configuration (switch|boot)` to get the same effect, using whatever is currently active (last deployed)
<Unode>
so I'll give it a reboot and cross fingers
<dhess>
clever: awesome, thanks
<clever>
dhess: and behind the scenes, `--install-bootloader` just does `export NIXOS_INSTALL_BOOTLOADER=1`, and switch-to-configuration then checks for that var
MmeQuignon has quit [Read error: Connection reset by peer]
MmeQuignon has joined #nixos
lycium[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
Thra11 has joined #nixos
bobvanderlinden_ has quit [Quit: Idle kick: User has been idle for 30 days.]
<dhess>
hmm not sure how I'll do that from a chrooted mounted EBS image. I guess I'll have to dig through the image's /boot/grub/grub.cfg and figure out where /run/current-system would come from
utdemir[m] has left #nixos ["Kicked by @appservice-irc:matrix.org : Idle kick: User has been idle for 30 days."]
<Unode>
still failing the same way :(
<clever>
dhess: one sec
<Unode>
is there any way to ask systemd to give me a shell while half booted?
<clever>
-r-xr-xr-x 1 root root 20377 Dec 31 1969 /nix/var/nix/profiles/system/bin/switch-to-configuration
Matthieu_ has joined #nixos
<clever>
dhess: /run is a tmpfs, so current-system wont exist on-disk
MmeQuignon has quit [Read error: Connection reset by peer]
<clever>
dhess: the above path is where the last active generation is kept
wolfman12345[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<dhess>
yeah that's the issue
<clever>
Unode: change the init= to init=/bin/sh
<dhess>
oh ok good to know!
<clever>
Unode: then systemd wont even be ran
hoijui has joined #nixos
orivej has joined #nixos
pie_ has joined #nixos
<Unode>
clever: that fails at stage 2 with /mnt-root/bin/sh not found
<clever>
make-package-set.nix: callHackage = name: version: callPackageKeepDeriver (self.hackage2nix name version);
<clever>
Cale: there ^^
o1lo01ol1o has quit [Read error: Connection reset by peer]
<Cale>
cool, thanks
o1lo01ol1o has joined #nixos
_ris has joined #nixos
<Unode>
clever: so something maybe useful, in journalctl, stage-2-init malformed JSON string,... in update-users-groups.pl line 11 followed by 'Activation script snippet users failed'
<Unode>
an from there a bunch of errors
tfc[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
invalid user root:root
<Unode>
Activation script snippet var failed
fusion809 has joined #nixos
<clever>
Unode: that sounds like your nix store is corrupt, so various things are failing on boot
theseriousadult[ has quit [Quit: Idle kick: User has been idle for 30 days.]
<clever>
Unode: the --verify --check-contents should confirm that
<Unode>
I'll let the verification finish
<Unode>
will it be able to fix itself?
<clever>
there is also a --repair command, but it will depend on what is corrupt, there are many choices
<Unode>
ok waiting for verify first
<Unode>
hopefully not much is damged
pleeplop has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
I currently don't have connection on this machine so repairing might be tricky
<Unode>
if it needs to download something
<clever>
wifi or wired?
fearlessKim[m] has quit [Quit: Idle kick: User has been idle for 30 days.]
<{^_^}>
overrideDerivation not yet supported for functors
<etu>
hmm, that last one worked locally
<cizra>
How do I teach nixpkgs to clone from git? Or should I just use GitHub's autogenerated tarballs?
<etu>
Izorkin: There should be no problem for you to override libcouchbase
<Unode>
clever: removed, now trying a reboot
<clever>
cizra: fetchFromGitHub
<cizra>
clever: ✓
<Unode>
clever: got a bunch of uninitialized errors but now system booted
WilliamHamilton[ has quit [Quit: Idle kick: User has been idle for 30 days.]
<Unode>
\o/
<clever>
Unode: should probably file an issue about uid-map corruption being non-obvious
<clever>
and maybe have it recover on its own, with a log to warn you
mexisme has joined #nixos
<Unode>
clever: so my user still doesn't login properly but I'll try to run nixos-rebuild now and see how far I get
<run500>
why is a sha256 hash required for git fetchers? isnt it redundant since the git rev hash is already hashing the content?
<clever>
run500: the nix itself doesnt understand git, so all it can do is hash the directory
<Yaniel>
and the rev can be a tag, which can be altered AFAIK
<clever>
run500: also, the rev is a hash over the commit, which contains the hash of the previous commit, so you need the entire chain of commits to validate the rev as being right
pie_ has joined #nixos
pie___ has joined #nixos
<run500>
i see, so it would take too long to actually verify the hash since you need to examine each commit?
<clever>
run500: and you would need to clone the entire history to do that
<Izorkin>
etu: php73Packages-unit = pkgs.php73Packages.override { php = php73-unit; }; - how to replace this variant? I can not understand
<clever>
run500: but fetchFromGitHub just downloads the auto-generated zip from github
<clever>
so you dont have any trace of the .git dir
Matthieu_ has quit [Quit: WeeChat 2.4]
vid has quit [Ping timeout: 264 seconds]
<run500>
in the case of fetchgit we have all the history
<clever>
run500: yep, but by default, it deletes the .git at the end, before nix validates the hash is correct
<clever>
run500: and there is still the issue of nix itself knowing how to validate that it matches a given rev
<clever>
run500: fixed-output derivations are just declaring the hash of $out, as a promise of purity, to be granted impure network access
<clever>
and the hash keeps them in line
<clever>
nix doesnt care how it creates $out, so you are free to use whatever you want (curl, torrent, git, svn, carrier pidgeon)
<etu>
Izorkin: Do you have your code with what you're trying to do?
<run500>
clever: is it feasable to teach nix about git's hashing to avoid having to double-specify the hashes?
<clever>
run500: it would need a new type of fixed-output derivation, which would require everybody to upgrade nix to be able to build it
<clever>
run500: and then every other VCS will want their hashing in nix, and it will get bloated
<{^_^}>
[nixpkgs] @herrwiese opened pull request #61548 → xbill: init at 2.1 → https://git.io/fjlVY
<run500>
clever: would that be such a bad thing for nix to know about different vcs's hashing systems? that would make the process of updating revs easier, which is something that happens a lot
<clever>
run500: nix-prefetch can already give you the hash of a rev
<run500>
clever: agreed, this would remove the need to run nix-prefetch
<pie_>
now if only the firefox history search wasnt garbage
<run500>
clever: i was wondering about this because anecdotally its a common source of complaints to have to double-specify the hash
<pie_>
(wrong chan)
<cizra>
Synthetica: Heya, it's my first foray into packaging stuff for NixOS :) Should we discuss it here, or in GitHub, or in private?
<Synthetica>
Wherever you want
<cizra>
Synthetica: $ nix-build --arg thinkpad true -A libfprint '<nixpkgs>'
<clever>
cizra: that wont pass thinkpad to libfprint
<cizra>
Synthetica: ↑ that's supposed to build my modified version, I think - yet it returns more or less immediately.
<clever>
cizra: that will import <nixpkgs>, check if that directly has a thinkpad arg (it doesnt), then discard the arg, and basically ignore it
<Izorkin>
etu: attribute 'php73Packages-unit.couchbase' at nur-packages/default.nix:35:79 already defined at /nur-packages/default.nix:33:3
<etu>
Izorkin: I don't know the context, so it's very hard for me to know what variables exists or not
<etu>
Izorkin: But just name it something else to try with?
Tucky has quit [Ping timeout: 258 seconds]
npmccallum has joined #nixos
MmeQuignon has joined #nixos
<Izorkin>
etu: https://github.com/Izorkin/nur-packages/blob/master/default.nix - me used this variant. This variant - php73Packages-unit = pkgs.php73Packages.override { php = php73-unit; }; - all plugins rebuild automatically, not need manual added. I wanted to automatically change openssl to libressl
domogled has quit [Ping timeout: 268 seconds]
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]
<{^_^}>
[nixpkgs] @c0bw3b closed pull request #29764 → msr-tools: nixos module to avoid CPU throttling → https://git.io/vdTN6
<MmeQuignon>
Hi (again) ! How to make a command installed from default.nix available in a nix-shell ? I would like to install "sphinx" and use "sphinx-build". Thanks!
<Unode>
clever: thanks again for all the help, I'm still fighting with X as I can't login but at least the system is working
<clever>
Unode: it could be more stray nulls in various files
<Unode>
would have taken forever to diagnose this. Thanks a log!
<Unode>
clever: yeah, now running a full system grep for those
<Unode>
just in case...
<Unode>
can't really pinpoint how that came to be but I'll be cautious about full disk from now on
zeta_wolf has joined #nixos
<dhess>
clever: hmm weird. I've got these nvme EC2 instances. They've worked fine for months. And now grub is failing to install because it's complaining that /dev/xvda doesn't exist, which it doesn't.
<clever>
Unode: something like that would turn me off an FS forever
<clever>
Unode: ive been using zfs for a few years, and havent seen any weird corruption like that
<clever>
dhess: sounds like a bug in nixops, the boot.loader.grub.device isnt set right
<Unode>
I can't really say if this is filesystem or something else causing it
<Unode>
but btrfs is a first for me as well
<dhess>
does NixOS only run grub-install when grub has changed?
<dhess>
I'm at a loss why this hasn't manifested earlier.
<slabity>
Aren't nvme devices under /dev/nvme*
wfranzini has quit [Ping timeout: 258 seconds]
<clever>
dhess: yeah
<dhess>
slabity: yes they are
<clever>
dhess: /boot/grub/state has the state of various grub flags
<dhess>
clever: yeah it only runs it when grub is updated?
<clever>
dhess: and if that file doesnt match, it will re-run grub-install
<clever>
one of the flags is the version of grub
<dhess>
clever: yeah and mine has /dev/xvda, not /dev/nvme...
<dhess>
clever: ahhhhhhh ok
<dhess>
that is what is goign on
<clever>
another is the grub device
<dhess>
because grub just bumped to 2.04
<clever>
so if it was nvme, then incorrectly got changed to xvda, it will correctly want to re-run grub-install
<dhess>
so somehow the EC2 AMI I'm using on these things was built for grub on /dev/xvda and when I run the AMI on an nvme EC2 instance, it boots fine
<dhess>
until you run grub-install
<dhess>
and I guess that hasn't happened until now because NixOS grub has been on 2.02 for ages.
<dhess>
*and* it appears my NixOps config does not set boot.loader.grub.device
<dhess>
for these instances
<clever>
dhess: nixops should set it for you, based on the instance type
<clever>
dhess: you can also try just setting it yourself, to the correct value
<dhess>
clever: it probably would. But I'm provisioning everything with Terraform these days.
<dhess>
and just using NixOps for deployment
<clever>
dhess: oh, that would explain it
zupo_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<clever>
nixops doesnt know its aws, and doesnt know what defaults to use
<dhess>
clever: yeah I will, now that I know this is a problem
wfranzini has joined #nixos
<dhess>
clever: Nixpkgs knows it's AWS because I'm loading the amazon-image.nix and setting ec2.* config stuff. But yeah I didn't realize that the config was split between Nixpkgs and NixOps
<dhess>
I wonder whether I can just get away with setting boot.loader.grub.device to /dev/nvmeX and that's good enough
<zeta_wolf>
hello guys, how do i setup sasl for rcirc in emacs? i cannot find any instructions and i cannot get into any of the emacs channels until this is set up?
<dhess>
hmm does NixOps use pvgrub for EC2 instances?
<clever>
dhess: for extra fun, on real hardware, you CANT boot legacy nvme, because it just doesnt appear over the normal legacy api's, so uefi is the only way to boot it
mexisme has quit [Ping timeout: 264 seconds]
lovesegfault has joined #nixos
fendor_ has joined #nixos
<dhess>
clever: yeah I gradually switched most of my physical hardware to nvme and was using EFI boot
fendor_ has quit [Remote host closed the connection]
Ralith_ has joined #nixos
<zeta_wolf>
anyone? if not i will try to find another channel to ask this question?
<clever>
dhess: and anoyingly, despite grub "supporting" nvme, grub DOESNT support nvme!
<clever>
dhess: grub assumes the EFI firmware provides nvme drivers
<infinisil>
zeta_wolf: This channel is about nixos, not rcirc, not emacs, so yes, find another channel
<clever>
so if your motherboard lacks drivers in the rom, grub cant read the nvme, and you cant cheat by putting grub on a legacy disk
<zeta_wolf>
this sasl has given me so much problems
<dhess>
clever: so any idea how NixOps handles the bootloader for NVMe EC2 instances?
<dhess>
this is weird because the NixOps table you pointed me to says t3* doesn't support nvme but both of my instances are t3 and they're running nvme
<clever>
dhess: /home/clever/apps/nixpkgs/nixos/modules/virtualisation/amazon-image.nix: boot.loader.grub.device = if cfg.hvm then "/dev/xvda" else "nodev";
<infinisil>
Y'all should start using uuid's instead!
<clever>
dhess: i suspect that the nvme aws machines might also be uefi, you should look into what the state of a freshly booted ami is, `nixops deploy --create-only` to stop it from messing with it
<dhess>
clever: any way to tell whether it's UEFI?
<dhess>
I've always wondered how to tell from a running kernel
<zeta_wolf>
ok i'll keep searching
<dhess>
infinisil: I don't think that works with grub does it?
<clever>
[clever@system76:~]$ mount | grep efi
<clever>
efivarfs on /sys/firmware/efi/efivars type efivarfs (rw,nosuid,nodev,noexec,relatime)
<clever>
dhess: that FS can only mount when booting via EFI
<infinisil>
dhess: Ah darn, probably you're right, these /dev/disk/by-* things are set up by udev..
<dhess>
clever: no efi here then
<clever>
dhess: also, the old /boot/grub/state from before you deploy anything should give details
<dhess>
clever: yeah that's when I noticed this was weird. I just booted an AMI that I created from an official NixOS AMI. So its /boot/grub/state uses /dev/xvda as you would expect
<dhess>
but I'm booted on an nvme instance and it works fine
<___laika>
im installing weechat from nixpkgs, but i seem to be having trouble getting it to build with plugin support. does it query my nix store/env for python/perl when installing? is there any way to tell it to build those automatically?
<tobiasBora>
Couldn't execute ld HASCOMPILERhT0p/TESTHrUY.o -o HASCOMPILERhT0p/TESTHrUY.bundle -mmacosx-version-min=10.10 -bundle -undefined dynamic_lookup -L/no-such-path/lib -fstack-protector-strong : Inappropriate ioctl for device at Makefile.PL line 8.
<tobiasBora>
Sorry! YAML::Syck requires a compiler in order to be built.
<clever>
dhess: and right at the edge on line 50, you can see a fragment of stage 1.5
<clever>
dhess: definitely looks like a legacy grub install
<tobiasBora>
So I have two questions: 1) how can I test if my code builds on darwin as I don't have darwin 2) I never used mac, any idea what could cause this error?
<clever>
dhess: i suspect aws did some weird stuff to make legacy boot on nvme without any trouble, so you just need to set boot.loader.device right and it will keep working
<{^_^}>
[nixpkgs] @nh2 opened pull request #61552 → git: Don't depend `curl-config` when cross-compiling. → https://git.io/fjlwa
<dhess>
clever: yeah I'll try that
<{^_^}>
[nixpkgs] @nh2 closed pull request #61527 → git: Depend on curl.dev to provide `curl-config`. → https://git.io/fjl3G
<dhess>
there isn't really any other explanation for how it booted in the first place :)
<clever>
dhess: oh, one thing i can think of, if nixops is setting device = "nodev"; on nvme machines, then it will just never update stage 1.5 and the MBR
<dhess>
clever: ahh interesting. That's probably what I'll do for new instances
<dhess>
thanks so much for the help as usual
wfranzini has quit [Remote host closed the connection]
<clever>
dhess: but it will still update the modules in /boot/ i think, and after a while, they could become incompatible with the stage 1.5
<clever>
dhess: oh, and that might be why the other machine bricked itself!
<dhess>
clever: yeah it probably is. It's complaining about missing flags
<dhess>
at the grub menu
___laika has quit [Quit: WeeChat 2.4]
<dhess>
I managed to get grub 2.02 back on that machine before rebooting. Any way you know of to use grub 2.02 to update /boot with the versions it wants?
ThatDocsLady has quit [Ping timeout: 252 seconds]
___laika has joined #nixos
<clever>
dhess: just letting switch-to-configuration run grub-install should sync everything to the same version
<clever>
if grub.device is set right
<dhess>
yeah that didn't seem to work
<dhess>
but OK I have it mounted in a rescue instance so I can mess around here
<___laika>
looks like using the new style `nix build nixpkgs.weechat -c weechat` worked to give me plugin support like i expected
zupo has joined #nixos
<___laika>
are there any helpful writeups outside of the nix manual on the differences between nix-* commands and the new porcelain nix commands?
<cizra>
Synthetica: Would you suggest making a variant package of fprintd as well, or leave that as an option?
selfsymmetric-mu has quit [Remote host closed the connection]
zeta_wolf has quit [Remote host closed the connection]
<Synthetica>
cizra: Probably? Would that just override libfprint to be libfprint-thinkpad?
<cizra>
yes
<cizra>
Synthetica: This trick from manual should work, alternatively: environment.systemPackages = [ (pkgs.emacs.override { gtk = pkgs.gtk3; }) ];
<{^_^}>
[nixpkgs] @matthewbauer merged pull request #61526 → haskell: servant: Don't write to $doc if disabled. → https://git.io/fjl3s
<{^_^}>
[nixpkgs] @matthewbauer pushed 2 commits to master: https://git.io/fjlwF
<Synthetica>
cizra: Yes, but it would be difficult for anyone to discover that that is even possible
<cizra>
Synthetica: OK, variant package it is, then!
<{^_^}>
[nixpkgs] @c0bw3b pushed 2 commits to release-19.03: https://git.io/fjlwA
___laika has joined #nixos
acarrico has quit [Ping timeout: 246 seconds]
zeta_wolf has joined #nixos
mexisme has joined #nixos
avn has quit [Ping timeout: 258 seconds]
zeta_wolf has quit [Remote host closed the connection]
PhawaS0ze has quit [Quit: Leaving]
klntsky has quit [Remote host closed the connection]
klntsky has joined #nixos
<cizra>
Synthetica: So the usual way of getting fprintd is to say service.fprintd.enable=true; and the ThinkPad way would be to mention fprintd-thinkpad in systempackages first, then enable 'fprintd' (not fprintd-thinkpad) service, did I get it right? Or should I make the service name also configurable?
<Synthetica>
cizra: You should look at how i3-gaps does it
<Synthetica>
e.g. you make the package configurable
<cizra>
Synthetica: Oooh no - the fprintd module mentions it by name, so I'll have to change the module as well.
mexisme has quit [Ping timeout: 248 seconds]
<tilpner>
cizra: You can use overlays to make the module use a different package without altering the module
mexisme has joined #nixos
<dhess>
hmm dare I reboot this other ec2 instance now that I've set the grub boot device properly :)
<cizra>
tilpner: The i3-gaps method is straightforward enough. Except I don't understand the imports=... in the bottom of it.
mkoenig has quit [Remote host closed the connection]
mkoenig has joined #nixos
mexisme has quit [Read error: Connection reset by peer]
troydm has joined #nixos
___laika has quit [Quit: WeeChat 2.4]
___laika has joined #nixos
avn has joined #nixos
samrose_ has joined #nixos
samrose has quit [Ping timeout: 255 seconds]
Glider_IRC_ has quit [Ping timeout: 246 seconds]
balsoft has quit [Ping timeout: 250 seconds]
troydm has quit [Client Quit]
<avn>
Folks, anyone have grub issues after recent master update?
<cizra>
Synthetica: OK, it's good to go, tested and working. Also tested packaging the non-thinkpad vanilla version -- it builds and runs, and it doesn't detect my reader so I suppose it's in working condition for other users.
<clever>
avn: dhess and i just finished a long convo about that
<clever>
avn: what exactly is the error you got?
<avn>
clever: missed symbol filter_something.... (was need "screenshot" with mobile, but forgot -- have other issues)
<avn>
Also found that boot mirror not works as intended, both grubs written on same partition :/
<avn>
(but it other, not-so-important problem)
<clever>
avn: 2019-05-15 15:06:11< dhess> ugh, EC2 instance totally hosed, "error: symbol `grub_file_filters` not found."
<dhess>
avn: yep I had that problem.
<avn>
clever: exactly this!
<avn>
efi+zfs, if you ask about configuration
<clever>
avn: i suspect its due to the stage 1.5 (from boot.loader.devices) not being updated correctly
<avn>
my configs in github if need
<clever>
but efi doesnt use loader.devices...
<dhess>
avn: mine is pretty specific to EC2 though. In the end I believe I had a bad boot.loader.grub.device setting
<dhess>
I'm about to reboot to test that
<avn>
I have issue on my desktop :/
<clever>
avn: in your case, i'm guessing you have 2 efi system partitions? and one of them has the "old" grub
<avn>
(combined with dying cooler, why I miss to make shot)
<avn>
clever: two have both new, and one have very old
<avn>
clever: also, can we have basic conf with bootpath living also on efi?
<clever>
avn: and the bios loaded the old one i'm guessing
<avn>
clever: I picked very old one by hand, andit allow me booting
<dhess>
avn: so in my case I think that the new grub (2.04-rc) updated the files in /boot but did not install properly in the MBR
<dhess>
and that left me with a broken boot
<clever>
avn: i think the problem can happen if the efi binary is out of sync with the modules in the /boot/ dir
balsoft has joined #nixos
jb55 has quit [Ping timeout: 244 seconds]
troydm has joined #nixos
Anton-Latukha has quit [Quit: Leaving.]
<avn>
any ways to debug/resolve? (and/or prevent it happens in future?)
<clever>
avn: ensure that grub-install updates every ESP and /boot partition at once
___laika has quit [Quit: WeeChat 2.4]
zeta_wolf has joined #nixos
<clever>
avn: `nixos-rebuild --install-bootloader` will force it to re-run grub-install
petersjt014 has joined #nixos
wfranzini has joined #nixos
<avn>
clever: `sudo nixos-rebuild boot --install-bootloader` will paste result and reboot (If I have a luck)
<avn>
installing the GRUB 2 EFI boot loader into /boot/sda/efi...
<avn>
Installing for x86_64-efi platform.
<avn>
and same for /boot/sdb/efi
<avn>
trying my luck
avn has quit [Quit: Lost terminal]
avn has joined #nixos
civodul has quit [Quit: ERC (IRC client for Emacs 26.2)]
knupfer has joined #nixos
knupfer has quit [Client Quit]
<zeta_wolf>
hello guys,i have xmonad-brightness setup in xmonad.hs, but the only thing i have not been able to setup is the permissions(udev rules), i have done some research and here is what i have so far:
<avn>
clever: well, I booted from 5th attempt. Coorect grub tell his version as 2.04-rc1, so /boot is correct, but efi vars probably point to wrong bootfile
knupfer has joined #nixos
<zeta_wolf>
other than the missing kernel is it correct?
<clever>
avn: efibootmgr -v
<avn>
I feel I peek legacy boot next time :/
aveltras has quit [Quit: Connection closed for inactivity]
<clever>
zeta_wolf: `udevadm monitor` will tell you what is being added/removed from the system, then you need to mess with `udevadm info` to query a device
<clever>
judson_: the generation# is only assigned after the nix level builds are finished, so you cant access it then
<clever>
judson_: rollbacks also become an issue, if you ever do a nixos rollback, the serial# goes backwards, and dns wont update
<clever>
judson_: i think the best option is to have a pre-start script for the dns server, that compares the state in a runtime state dir, to the state from /nix/store, and if they differ, increment a serial# also in the runtime state dir, and sed that into the zone file
hoijui has quit [Remote host closed the connection]
fenedor has joined #nixos
mexisme has quit [Ping timeout: 250 seconds]
<{^_^}>
[nixpkgs] @dtzWill opened pull request #61561 → lua-modules/generated: regen for formatting, luv -> 1.28 → https://git.io/fjloZ
<clever>
avn: i suspect you can do some funny bussiness with mdadm or device mapper, to mirror 2 parititions together (without a header) and then mount the result to /boot/EFI/
<clever>
avn: but if anything tried to modify one of the disks on its own, the mirror will be "corrupt" and i have no idea how it will decide what the "right" state is
<jeaye>
Trying to use gpg in the Nix docker image on Win10, but pinentry isn't found. I have it installed and I've set my pinentry-program to /nix/var/nix/profiles/default/bin/pinentry-curses, but gpg can't find it. I figure this has to be a Nix-related issue; ideas?
<avn>
clever: I ok, if bios allow me pick anything EFI from other disk :/ it for case if missed disk missed
<clever>
zeta_wolf: udev cant manipulate the permissions of things under /sys
<avn>
but I need to restore my boot order
<clever>
zeta_wolf: so a udev rule will never fix that
<clever>
zeta_wolf: which is why the readme file didnt mention udev
gentauro has joined #nixos
<avn>
otherwise I can not survive powerloss
<zeta_wolf>
clever: then what is an alternative?
<clever>
zeta_wolf: exactly what they said, chmod the file, probably with a systemd service
<judson_>
Does anyone do that automatic DNS serial number management?
<judson_>
Or just manually like the stone ages?
<clever>
avn: decide which partition should be your ESP, and fix up the config, then confirm that efibootmgr is refering to its uuid correctly, and with the right path
<clever>
judson_: infinisil is impurely using the date at build time as a serial#
<clever>
both normal rollbacks, will undo the serial# and go backwards
<infinisil>
judson_: I might end up not using it though, because if I use all my machines with nixops, master/slave sync isn't really needed
<clever>
and also, if you perfectly undo a change in the nix code, and the derivation matches an older one, nix will reuse the old build, and its old serial#
<clever>
infinisil: yeah, thats far simpler, just make every server a master, and let nixops manage the sync
<judson_>
My NixOps machine is a stealth master, so I really want sync to work.
freeman42x has quit [Ping timeout: 252 seconds]
<infinisil>
clever: Args
<infinisil>
Argh!
<infinisil>
clever: Well they still need to take over the zone if one of them fails so I can't do that
<infinisil>
clever: And anyways, can't have more than one authorative name server
<tobiasBora>
I'm trying to find the good way to deploy secrets. For now I have two methods. First, I have in my configuration a folder secret/, with all the files that belongs to root (I deploy this by scp the configuration from my laptop, and I set the permission to 600 for everything). Then I have two methods for putting a password for an application (like openldap). First method is to put in this secret/ folder a txt
<tobiasBora>
file, and use something like "rootpwFile = '/etc/nixos/secret/ldap.txt'" to setup the password. But the problem of this method is that ldap cannot access this file anymore because it does not have the good permission. So I was thinking instead to use my second method: put in the secrets/ a .nix file, that contains {password_ldap = "my password"}, and then import it from "rootpw" directly. Is this second method
<tobiasBora>
secure, and the good way to proceed? If not what is the recommended way to proceed to share secret to a non-root process?
ottidmes has quit [Quit: WeeChat 2.4]
<clever>
configure them all as master, but name one as authorative in the SOA record
<clever>
if bind is told that its master, it wont try to constact the authorative one, since it believes thats itself
adamantium has joined #nixos
<avn>
clever: where i can see which ESP have which UUID?
<clever>
avn: run blkid on the block device
<clever>
avn: also check what fdisk -l says about it
<clever>
with gpt, every fs now has 2 uuids, the partuuid, and the fsuuid
<clever>
Boot0004* UEFI OS HD(1,GPT,27c99b08-455d-4dfe-a44f-6150cbc09ef8,0x800,0x100000)/File(\EFI\BOOT\BOOTX64.EFI)..BO
<clever>
and its the partuuid that is used by efi
<clever>
which makes sense, the firmware wants to find the partition directly, without having to open (and understand) every single fs, and parse the fs uuid's
<avn>
clever: both have same partuuid ;)
<clever>
then your GPT tables may have been cloned, and the bios has no way to know which disk to use
<clever>
u change partition UUID
<clever>
if i switch fdisk to advanced mode (via x) i have the option to change the part uuid
<clever>
zeta_wolf: that will be undone when you reboot
camsbury has quit [Ping timeout: 246 seconds]
<clever>
zeta_wolf: so you need a systemd service that runs that again next time you boot
<infinisil>
clever: I don't think it would be a good idea to have multiple authorative servers, it just sounds bad
<clever>
brb
balsoft has quit [Ping timeout: 252 seconds]
MmeQuignon has quit [Ping timeout: 252 seconds]
linarcx has quit [Quit: WeeChat 2.4]
rfold has quit [Quit: WeeChat 2.4]
<avn>
clever: ty, it now point to file on fresh formatted sdb1, which written after with --install-bootloader, I hope it match "platform files"
<avn>
But I probably postpone testing :/
erasmas has quit [Quit: leaving]
<avn>
I have less than 8 hours to next workday
<avn>
btw, does with mdraid/devicemapper mirroring partitions have _different_ uuids?
<zeta_wolf>
clever: ok, i'll setup the permissions in systemd in home-manager(home.nix), do you know of any good instructions on how to do this, i don't know much about systemd? i'll go chech the home-manager manual to see if there is any instructions on how to do this. if you know of any good instructions please pastebin, thanks
BlessJah has quit [Remote host closed the connection]
cyraxjoe has quit [Quit: I'm out!]
<avn>
clever: thank you for help. I will PM you after next reboot if I success or not ;)
BlessJah has joined #nixos
<zeta_wolf>
uhh, this is frustrating, i wish they paste instructions on how to do this
pie_ has quit [Ping timeout: 252 seconds]
cyraxjoe has joined #nixos
<Matthieu_>
As everything I need in my python projects is defined in default.nix, is there a way to specify that I do not need a setup.py file ?
fenedor has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @Infinisil opened pull request #61570 → [Partial] Move renames to their appropriate places → https://git.io/fjl62
<zeta_wolf>
i can't seem to find systemd in the home-manager manual?
ppum has quit [Ping timeout: 256 seconds]
alp has quit [Remote host closed the connection]
<clever>
zeta_wolf: home manager cant do anything with /sys, since it doesnt get root
<clever>
zeta_wolf: you need to read the nixos manual
<{^_^}>
[nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fjl6i
<zeta_wolf>
clever: are you sure udev rules won't work, i see udev instructions for the `light` package(scroll down to `optional` section): https://haikarainen.github.io/light/
<clever>
zeta_wolf: ah, thats not properly telling udev to fix the permissions as it makes things in /dev/
<clever>
zeta_wolf: thats just telling udev "run this" any time a backlight thing is loaded
<clever>
zeta_wolf: but you could use those exact rules as an example to just chmod o+w the file your using
Guanin has quit [Quit: Leaving]
<{^_^}>
[nixpkgs] @Infinisil merged pull request #61143 → allow overriding perl528 and perldevel in overlays → https://git.io/fjcS4
<avn>
clever: btw, if I remember right -- you use weechat for irc with weeslack for slack (figuring from your nixos-configs ;). Can you share weechat configs as well?
<clever>
avn: irssi for irc, weechat for slack
<clever>
avn: the entire weechat config is in the nixos-configs
mbrgm_ has joined #nixos
<zeta_wolf>
you said earlier to re-do the the chmod, to i still have to do that?
<clever>
zeta_wolf: udev will redo the chmod every time the backlight driver is loaded
mbrgm has quit [Ping timeout: 258 seconds]
mbrgm_ is now known as mbrgm
<hexa->
I wonder, shouldn't the microcode update appear in my grub.cfg as the first element of the initrd command?
<avn>
clever: I see only plugin installing, so default configs -- right?
<clever>
avn: yep
<clever>
avn: the only config that isnt default, is the auth token to access slack
<avn>
I'd like to migrate all chats to it, but config looks like designed by martians
<avn>
will try, ty
<clever>
avn: i dont really use weechat, i just have it logging things
<zeta_wolf>
why did you say ealier that the `udev rules` will not work and that i would have to use `systemd` instead?
<avn>
clever: I see. But I annoying how 4 slack chats (tabs in chromium) consume 1gb memory each, I feel I find better applicaton for RAM ;)
jackdk has joined #nixos
endformationage has quit [Ping timeout: 255 seconds]
m0rphism has quit [Ping timeout: 244 seconds]
<infinisil>
avn: Damn that's a lot..
<clever>
avn: twitter on its own is using 2.5gig of ram