gchristensen changed the topic of #nixos to: The Nix Ecosystem, https://nixos.org | NixOS: https://nixos.org/nixos/download.html | Nix: https://nixos.org/nix/download.html | Logs: https://logs.nix.samueldr.com/nixos/ | #nixos-dev, #nix-darwin, #nixos-aarch64, #nixos-chat
<{^_^}> [nixpkgs] @markuskowa opened pull request #66736 → glusterfs: 4.0 -> 6.5 → https://git.io/fjd9G
<{^_^}> [nixpkgs] @markuskowa merged pull request #65999 → mosquitto: 1.6.3 -> 1.6.4 → https://git.io/fjHMT
<{^_^}> [nixpkgs] @markuskowa pushed 2 commits to master: https://git.io/fjd9c
jgt has quit [Ping timeout: 264 seconds]
drakonis has joined #nixos
william123456 has quit [Remote host closed the connection]
Mateon1 has quit [Ping timeout: 248 seconds]
notgne3 has quit [Quit: ZNC 1.7.3 - https://znc.in]
notgne3 has joined #nixos
notgne3 is now known as notgne2
sheeldotme has quit [Ping timeout: 260 seconds]
dansho has joined #nixos
Mateon1 has joined #nixos
Ariakenom has quit [Quit: Leaving]
<{^_^}> [nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fjd9V
lambda-11235 has joined #nixos
iyzsong has joined #nixos
<{^_^}> [nixpkgs] @dtzWill opened pull request #66737 → azpainter: init at 2.1.4 → https://git.io/fjd9i
<{^_^}> [nixpkgs] @aanderse merged pull request #66014 → cre2: 0.3.0 -> 0.3.6 → https://git.io/fjHy8
<{^_^}> [nixpkgs] @aanderse pushed 2 commits to master: https://git.io/fjd9M
slyfox has quit [Ping timeout: 248 seconds]
slyfox has joined #nixos
Soo_Slow has quit [Remote host closed the connection]
thc202 has quit [Ping timeout: 268 seconds]
<{^_^}> [nixpkgs] @aanderse merged pull request #66692 → smartgithg: 18.2.9 -> 19.1.1 → https://git.io/fjdBQ
<{^_^}> [nixpkgs] @aanderse pushed 2 commits to master: https://git.io/fjd9y
<{^_^}> [nixpkgs] @l0b0 opened pull request #66738 → pythonPackages.vcard: init at 0.13.0 → https://git.io/fjd9H
count1 has joined #nixos
vika_nezrimaya has joined #nixos
<vika_nezrimaya> how to disable APXS2 in PHP? apr fails to cross-compile and I don't use apache anyway, this is an RPi, Apache is too bulky for it!
jgt has joined #nixos
<vika_nezrimaya> it'll probably just waste space on my SD card >.<
<count1> Hello, just started messing around with Nixos the other day and i'm very new to the workflow of things like nixpkgs. I was trying to download a package that is in the all-packages.nix file on github but nix-env keeps returning no matches. I get no issues downloading any other packages from that list at all. Have been looking online for a while now with no luck. Appreciate any help, if anything I could just but
<count1> I really wanted to go through and run the binary/set up dynamic linking myself but I wanted to do the conventional nix route
jgt has quit [Ping timeout: 252 seconds]
<jackdk> I don't have time to sit and help, I'm sorry, but I think if you mentioned the package and/or linked the all-packages.nix on github at the line for the package you're trying to install, it will help you get a better answer
sb0 has joined #nixos
chuenifule has quit [Quit: leaving]
<count1> yeah sure no problem it is wasabiwallet package
<drakonis> count1: which channel are you on?
<count1> I tried on the default and the unstable
<drakonis> its not on 19.03
<count1> shoot ok thanks my bad
<drakonis> it was included a few months ago
<count1> that makes sense
<drakonis> it is on nixos-unstable though
kleisli has quit [Ping timeout: 245 seconds]
<aanderse> vika_nezrimaya: looks like you need to set config.php.apxs2 = false
* aanderse digs a bit more
Guanin has quit [Remote host closed the connection]
<vika_nezrimaya> aanderse: I assume I need to override nixpkgs.php with that option?
<kaliumxyz> how should I set up the env for a script with writeScript?
<aanderse> vika_nezrimaya: yeah, that, or i believe it would be nixpkgs.config.php.apxs2 = false; appended to your configuration.nix (or equivalent)
<vika_nezrimaya> wow, so Nixpkgs has a system-wide config for packages? I never knew about that!
<vika_nezrimaya> Which packages support this, apart from PHP? I assume, to load something from config, you need to declare "config" in your function arguments?
<aanderse> vika_nezrimaya: it isn't great. see work from ... @oxiji or someone to improve it
<aanderse> gosh whats their nick...
<aanderse> anywho, some packages use nixpkgs.config, but not that many
<count1> Is there an extra step you have to do after you add unstable to your channels? I already did nix-channel --update. Do you have to specify to nix-env which channel you want to look under?
ris has quit [Ping timeout: 245 seconds]
<dansho> does anyone know how to compute the sha256 needed by requireFile?
<clever> dansho: i think its just the plain sha256 of the file
<vika_nezrimaya> oh well it seems like clucene fails to cross compile, making me unable to use dovecot without modifying its nix expression :3
selfsymmetric-mu has joined #nixos
<clever> dansho: nix-hash is recursive by default, you want to add --flat
<clever> or use plain old sha256sum
pandada8 has quit [Ping timeout: 248 seconds]
<dansho> clever: it was the --flat flag i was missing, ty!
pandada8 has joined #nixos
<vika_nezrimaya> um... remind me how to remove a known string from a list?
dl4mfo has joined #nixos
<vika_nezrimaya> and return that list without a string
<clever> vika_nezrimaya: builtins.filter
<vika_nezrimaya> clever: not lib.lists.remove?
drakonis has quit [Quit: WeeChat 2.5]
<clever> vika_nezrimaya: ah, that will do the == for you, if thats all you need
<{^_^}> [nixpkgs] @dtzWill opened pull request #66739 → readline80: 8.0p0 -> 8.0p1 → https://git.io/fjdHt
stigo has quit [Quit: WeeChat 2.4]
count1 has quit [Quit: WeeChat 2.4]
dl4mfo_ has quit [Ping timeout: 248 seconds]
stigo has joined #nixos
is_null has joined #nixos
Shouou has quit [Ping timeout: 252 seconds]
jgt has joined #nixos
jgt has quit [Ping timeout: 264 seconds]
slack1256 has joined #nixos
alex`` has quit [Ping timeout: 248 seconds]
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/7a712206a43 (from 84 minutes ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
<{^_^}> [nixpkgs] @aanderse merged pull request #65713 → macvim: 8.1.1517 -> 8.1.1722 → https://git.io/fj9sn
<{^_^}> [nixpkgs] @aanderse pushed 2 commits to master: https://git.io/fjdH0
justanotheruser has quit [Ping timeout: 272 seconds]
<slack1256> Apart from documentation on nixpkgs, is there a way to nix repl to know what arguments to pass to a function?
<slack1256> Analogous as :t on ghci.
justanotheruser has joined #nixos
<{^_^}> [nixpkgs] @aanderse merged pull request #66251 → cocoapods-beta: 1.7.5 -> 1.8.0.beta.1 → https://git.io/fjQVt
<{^_^}> [nixpkgs] @aanderse pushed 2 commits to master: https://git.io/fjdHV
selfsymmetric-mu has quit [Remote host closed the connection]
kleisli has joined #nixos
dbmikus has quit [Quit: WeeChat 2.5]
<{^_^}> [nixpkgs] @lopsided98 opened pull request #66740 → processing: 3.4 -> 3.5.3 → https://git.io/fjdHP
mexisme has quit [Ping timeout: 250 seconds]
<jackdk> > builtins.functionArgs (import <nixpkgs>)
<{^_^}> { config = true; crossOverlays = true; crossSystem = true; localSystem = true; overlays = true; platform = true; system = true; }
<jackdk> slack1256: ^
drakonis has joined #nixos
mexisme has joined #nixos
jgt has joined #nixos
englishm has quit [Excess Flood]
englishm has joined #nixos
jgt has quit [Ping timeout: 250 seconds]
ddellacosta has quit [Ping timeout: 245 seconds]
waleee-cl has quit [Quit: Connection closed for inactivity]
fragamus has joined #nixos
[Leary] has joined #nixos
<vika_nezrimaya> ohhhhh I think I finally have some great material for a Nixpkgs PR
<vika_nezrimaya> I have found a lot of packages that don't respect environment.noXlibs and I probably should make a patch to fix that :3
<slack1256> jackdk: Thanks jackdk
Lears has quit [Ping timeout: 245 seconds]
<slack1256> nix-repl> builtins.functionArgs pkgs.haskellPackages.override
<slack1256> { }
<slack1256> It seems it only work with default arguments specified.
is_null has quit [Ping timeout: 258 seconds]
kleisli has quit [Ping timeout: 272 seconds]
Supersonic has quit [Ping timeout: 250 seconds]
is_null has joined #nixos
mexisme has quit [Ping timeout: 264 seconds]
Supersonic has joined #nixos
<aanderse> vika_nezrimaya: search for some existing issues on github for that
<aanderse> might be @volth that has one open for some time
<aanderse> could be some discussion points worth reading before proceeding
<aanderse> but i can recall the details
is_null has quit [Ping timeout: 272 seconds]
jgt has joined #nixos
nisstyre has quit [Quit: WeeChat 2.5]
jgt has quit [Ping timeout: 250 seconds]
is_null has joined #nixos
<vika_nezrimaya> Wow, Lua can't cross-compile?! I'm kinda... disappointed
<vika_nezrimaya> because I'll have to get rid of one of my machine's services to get rid of Lua dependency...
<infinisil> Might be easy to fix
<vika_nezrimaya> It seems like it wants to have gcc
<vika_nezrimaya> but we're cross-compiling, so stdenv only includes armv7l-unknown-gnueabihf-linux-gcc :3
<vika_nezrimaya> I'm no pro at cross-compiling or Lua, but I seem to be good at writing Nix expressions, so I might need some hints...
<aanderse> vika_nezrimaya: reference i was thinking of https://github.com/NixOS/nixpkgs/pull/46013/files#discussion_r234411558 (cc infinisil)
ryantm has joined #nixos
<vika_nezrimaya> aanderse: that includes a lot of stuff I'm doing right now + some things I haven't done yet (because I don't use these packages)
<vika_nezrimaya> actually would be nice to `grep -e x11Support -e guiSupport -e withGnome -rl $nixpkgs` and add all of that
<vika_nezrimaya> once this thing is merged, I could probably do that...
<aanderse> yeah i think "is for hydra to build the X-less variants of the packages as well" is a pretty important step to make noXlibs more practically useful, or at least useful to a larger audience
<vika_nezrimaya> agreed
<vika_nezrimaya> infinisil: any ways to setup nix-shell for cross-compiling? so I can reproduce the environment
<vika_nezrimaya> alternatively - setting up nix-shell from a .drv
m0rphism has quit [Ping timeout: 268 seconds]
fragamus has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
vika_nezrimaya has quit [Ping timeout: 246 seconds]
lunik18 has quit [Quit: :x]
vika_nezrimaya has joined #nixos
<vika_nezrimaya> Wow I had a kernel panic rn
<vika_nezrimaya> I dunno what was that
<vika_nezrimaya> has anybody wrote something to me while I was offline?
lunik18 has joined #nixos
<infinisil> Nope
<vika_nezrimaya> whew
<vika_nezrimaya> I hope my /nix/store isn't corrupted :3
<infinisil> (There's logs btw)
<vika_nezrimaya> I'm in tty rn so no logs for me
<vika_nezrimaya> if they can't be read with Emacs of course
nisstyre has joined #nixos
nisstyre has quit [Client Quit]
aleph- is now known as Gilfoyle-
nisstyre has joined #nixos
<markasoftware> if it can't be read with emacs, it's not worth reading
zeta_0 has joined #nixos
tmaekawa has joined #nixos
kleisli has joined #nixos
<vika_nezrimaya> infinisil: Indeed it could be an easy fix! Except that I'll have to rewrite like, half of the nix expression? The toolchain defaults (native toolchain) is configurable via makeFlags but configurePhase already modifies makeFlags...
jgt has joined #nixos
<vika_nezrimaya> well I could modify makeFlags array but I don't guarantee it'll work on Darwin
<vika_nezrimaya> probably will
<vika_nezrimaya> wait. how to get the cross-compiler's name when you're actually cross-compiling?
fragamus has joined #nixos
jgt has quit [Ping timeout: 250 seconds]
<vika_nezrimaya> is ranlib prefixed when cross-compiling?
buckley310 has quit [Remote host closed the connection]
buckley310 has joined #nixos
<vika_nezrimaya> yes it is
zeta_0 has quit [Quit: rcirc on GNU Emacs 26.1]
fragamus has quit [Ping timeout: 245 seconds]
<vika_nezrimaya> wait it doesn't work?!
<vika_nezrimaya> I overrid the makeFlags with CC, AR and RANLIB, but I can't see them anywhere!
Rusty1 has quit [Quit: Konversation terminated!]
<vika_nezrimaya> oh well
<vika_nezrimaya> I guess I'm going to file an issue for that
endformationage has quit [Quit: WeeChat 2.5]
is_null has quit [Ping timeout: 258 seconds]
ottidmes has quit [Ping timeout: 245 seconds]
<vika_nezrimaya> Yay, I filed an issue! #66742
<{^_^}> https://github.com/NixOS/nixpkgs/issues/66742 (by kisik21, 4 minutes ago, open): lua fails to build when cross-compiling
<vika_nezrimaya> now I'll go to sleep since a lot of stuff depends on lua, making me unable to continue my efforts until a solution arises
<vika_nezrimaya> goodnight #nixos :3
vika_nezrimaya has quit [Quit: ERC (IRC client for Emacs 26.2)]
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
jgt has joined #nixos
palo1 has joined #nixos
jgt has quit [Ping timeout: 250 seconds]
palo has quit [Ping timeout: 268 seconds]
palo1 is now known as palo
kleisli has quit [Ping timeout: 245 seconds]
<Miyu-chan> How does NixOps deploy to DO?
<Miyu-chan> Err, dammit. I mean, how do you select droplet type?
<Miyu-chan> Doesn't give much info
orivej has joined #nixos
<Miyu-chan> Oh, you can find it in the source code. https://developers.digitalocean.com/documentation/v2/#list-all-sizes
k0ral has joined #nixos
<k0ral> Hello
<k0ral> many nix commands stall on my side when copying path '/nix/store/f74zj9535vgk9nlrskk1qgk3v6m7w6n4-stdenv-linux' from 'https://cache.nixos.org'
<k0ral> notably nix-shell and nixos-rebuild
slack1256 has quit [Ping timeout: 258 seconds]
<infinisil> k0ral: Throw a bunch of -vv's on
<k0ral> as a matter of fact, when I manually run: nix copy -v --from https://cache.nixos.org /nix/store/f74zj9535vgk9nlrskk1qgk3v6m7w6n4-stdenv-linux, it does not progress at all
<infinisil> k0ral: That command finishes successfully in a second for me
<k0ral> adding -vv to `nix copy` does not show much more, let me paste that
<infinisil> Add moooore
<infinisil> Oh not sure if -vvvvvv's work with nix commands, I'd try nix-shell
<k0ral> ok, but nix-shell seems just to be doing much more than nix copy so I fear I would be showing you a lot of unrelated stuff
<k0ral> I will paste nix-shell -vvvvvv output
<k0ral> btw, I am using nixos-unstable
<{^_^}> [nixpkgs] @c089 opened pull request #66743 → http-prompt: Allow installation on Mac OS → https://git.io/fjdQD
<jophish> Hi all
<jophish> Why doesn't: imports = [ (fetchFromGitHub{...} + "/mymodule.nix") ] work?
<jophish> What I found on the wiki looks a little verbose for this: https://nixos.wiki/wiki/Module#Abstract_imports
<Miyu-chan> k0ral: Region?
<k0ral> Miyu-chan: fr
<Miyu-chan> Ah. Hm.
pie_ has joined #nixos
<Miyu-chan> Add more -vvv, and see the curl debug logs.
<k0ral> -vvvvvvvvv, is that enough ? :D
<k0ral> I checked DNS resolving and it looks fine, I restarted nix-daemon, I checked I could reach cache.nixos.org
<Miyu-chan> > closing daemon connection because of an exception
<{^_^}> undefined variable 'closing' at (string):262:1
<Miyu-chan> Well, that could have been because of C-c C-c?
<Miyu-chan> C-c*
<k0ral> Miyu-chan: this is me Ctrl+C-ing the process
<Miyu-chan> Hm.
<k0ral> otherwise it hangs forever
<Miyu-chan> That is indeed weird.
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/a2bd8da766c (from 4 hours ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
<k0ral> I found a somewhat similar github issue but it suggests nix-daemon consumes all memory and in my case memory usage is quite low
<{^_^}> nix#1988 (by chris-martin, 1 year ago, closed): Nix 2.0: nix copy stalls
<k0ral> also, I never get the `error: Nix daemon out of memory`, even after waiting for several hours
<k0ral> so a memory issue looks pretty unlikely in my case
<{^_^}> [nixpkgs] @Infinisil opened pull request #66744 → doc: Replace google groups reference with Discourse Patch category → https://git.io/fjd7e
jgt has joined #nixos
slack1256 has joined #nixos
<infinisil> k0ral: It seems like the cache in your region doesn't respond well
<infinisil> k0ral: You can probably build it with --fallback
<infinisil> Well, maybe not, not sure
jgt has quit [Ping timeout: 264 seconds]
<Miyu-chan> Stupid realization. Unfree is a Nixpkgs construct that doesn't exist in Nix at all.
<Miyu-chan> That means you can freely build unfree drvs without specifically allowing unfree.
<infinisil> Indeed :)
<infinisil> Same for cross compilation, which might be surprising
<kandinski> Miyu-chan: same for broken, right?
<Miyu-chan> kandinski: I think so, yeah.
<Miyu-chan> infinisil: Yeah. :)
<k0ral> infinisil: how do you check whether this is a region-only issue ? So that I do not bother you next time it happens :)
<infinisil> Well I guess you could check multiple regions, but that's hard unless you have multiple machines
<infinisil> Actually I do have a server in france, lemme see
kleisli has joined #nixos
<Miyu-chan> k0ral:You can try the fastly scripts.
<infinisil> Seems to be downloading stuff for the `nix copy -v --from https://cache.nixos.org /nix/store/f74zj9535vgk9nlrskk1qgk3v6m7w6n4-stdenv-linux` command
<Miyu-chan> OTOH, your sitch is very very different from what I've been experiencing.
<Miyu-chan> It fails for you even with very small files.
<k0ral> fastly script returns no error whatsoever
<k0ral> for f74zj9535vgk9nlrskk1qgk3v6m7w6n4.narinfo
<k0ral> infinisil: --fallback does not help, strangely
<k0ral> I guess it does not because error is not detected since it hangs forever and there is no timeout
fusion809 has quit [Remote host closed the connection]
<k0ral> I could simply disable binary caches in my configuration.nix, but this is just avoiding the issue, not fixing it
<thoughtpolice> err, first post
thedavidmeister has joined #nixos
<thedavidmeister> i'm getting timeout "error 200" on nixos-rebuild
<thedavidmeister> how can i increase the timeout?
<infinisil> ,xy thedavidmeister
<{^_^}> thedavidmeister: xyproblem is when you want to do X, and you think Y is how, so you ask about Y instead of X. See <http://www.perlmonks.org/index.pl?node_id=542341> or <http://mywiki.wooledge.org/XyProblem>
slack1256 has quit [Remote host closed the connection]
dasj19 has joined #nixos
<infinisil> Your problem is not that the timeout is too low (almost surely), but that the timeout is reached at all
<thedavidmeister> 200 isn't an error though
<thedavidmeister> HTTP error 200 (curl error: Timeout was reached); retrying in 352 ms
orivej has quit [Ping timeout: 244 seconds]
<thedavidmeister> i could discover whether timeout is the problem or not by attempting to raise it
<infinisil> Yeah so that most likely means that the server wasn't reachable, nothing that could be fixed by increasing the timeout
<infinisil> Just try again, if it still doesn't work, it's not a too low timeout
<infinisil> And I think it automatically increases the timeout after a bit anyways
<thedavidmeister> i just want to debug
<thedavidmeister> i'm not saying it is this or it is that
<thedavidmeister> i do have a shitty australian connection
<thedavidmeister> it's not unusual for me to hit timeouts on large downloads
<k0ral> thoughtpolice: `nix copy --from https://aseipp-nix-cache.global.ssl.fastly.net/` does not work either
<k0ral> hangs at the same step
<infinisil> thedavidmeister: Hm I see
<thedavidmeister> adsl is pretty normal here, even in major cities
<k0ral> [1/0/1 copied (0.0/0.0 MiB)] copying path '/nix/store/f74zj9535vgk9nlrskk1qgk3v6m7w6n4-stdenv-linux' from 'https://aseipp-nix-cache.global.ssl.fastly.net'
<thedavidmeister> if i crank the timeout way up and it dies then i'll try something else :shrug:
<infinisil> thedavidmeister: You're gonna have to provide some more info, like the full log output, because you'd have to override the specific derivation that produces the timeout
<thedavidmeister> oh ok, so there's no top level cli flag i could pass into the curl that fetches from the nix cache?
<thedavidmeister> i'll do it again and send you -vvv output
<infinisil> Nope there's not
trinityhex has quit [Ping timeout: 250 seconds]
<infinisil> Well I gotta go sleep, but what you'd have to do is probably something like `failingSrc.override { curlOpts = "<options to increase timeout>"; }`
<infinisil> Oh wait, there is a way
<infinisil> thedavidmeister: Probably just `NIX_CONNECT_TIMEOUT=<number> nixos-rebuild ...`
<infinisil> Apparently this is one of the few (intensional) impurities in Nix
<infinisil> s/Nix/nixpkgs
thedavidmeister has quit [Ping timeout: 246 seconds]
<infinisil> Although, that's probably the wrong timeout
thedavidmeister has joined #nixos
felixfoertsch has joined #nixos
felixfoertsch has quit [Client Quit]
thedavidmeister has quit [Ping timeout: 268 seconds]
thedavidmeister has joined #nixos
ng0 has joined #nixos
<k0ral> I guess I will open an issue with all the info I collected
<thedavidmeister> ininisil++
<thedavidmeister> infinisil++
<{^_^}> infinisil's karma got increased to 121
k0ral has quit [Quit: WeeChat 2.5]
jgt has joined #nixos
felixfoertsch has joined #nixos
dasj19 has quit [Quit: dasj19]
tmaekawa has quit [Quit: tmaekawa]
jgt has quit [Ping timeout: 250 seconds]
kleisli has quit [Ping timeout: 248 seconds]
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
felixfoertsch has quit [Client Quit]
felixfoertsch has joined #nixos
felixfoertsch2 has joined #nixos
felixfoertsch2 has quit [Client Quit]
<{^_^}> [nixpkgs] @FRidh pushed 214 commits to staging-next: https://git.io/fjd7z
FRidh has joined #nixos
<thedavidmeister> i'm seeing this
<thedavidmeister> copying path '/nix/store/npdlhmkx59rb0j8q6nc30nz1r1w2617i-chromium-76.0.3809.100.tar.xz' from 'https://cache.nixos.org'...
<thedavidmeister> warning: unable to download 'https://cache.nixos.org/nar/167mgpvinppr60f64zy768jcyhhk14al0zv4az7ch0397nhshrnr.nar.xz': HTTP error 200 (curl error: Timeout was reached); retrying in 259 ms
<thedavidmeister> error 9 while decompressing xz file
<thedavidmeister> y
<thedavidmeister> i try to download that file with wget outside nix building
<thedavidmeister> it says ~30 mins
<thedavidmeister> what am i supposed to do here??
<thedavidmeister> can i manually paste it into place?
<tilpner> thedavidmeister: Yes, fetch the file manually, then use nix-store to import it
<tilpner> Not sure if you need to --restore or --import, I sometimes confuse those
<tilpner> xz -dc 167mgpvinppr60f64zy768jcyhhk14al0zv4az7ch0397nhshrnr.nar.xz | nix-store --restore /nix/store/npdlhmkx59rb0j8q6nc30nz1r1w2617i-chromium-76.0.3809.100.tar.xz
<tilpner> Something like that should work
<{^_^}> [nixpkgs] @FRidh pushed 215 commits to staging: https://git.io/fjd7r
<thedavidmeister> interesting
<thedavidmeister> that would be awesome if it works
<thedavidmeister> i got similar issues with firefox
<thedavidmeister> can i get a second opinion on whether it is possible to have nixos-rebuild simply not timeout?
<tilpner> Not from me, mine usually behaves
<tilpner> But there is an experimental cache on Discourse, if you think that can help
felixfoertsch2 has joined #nixos
<thedavidmeister> downloading chromium from here is at least 30 mins, and nixos-rebuild seems to timeout after about 5 mins
lambda-11235 has quit [Quit: Bye]
<thedavidmeister> why would a cache help?
<tilpner> Well, you're fetching chromium from cache.nixos.org
<thedavidmeister> yeah but i'm on adsl
<thedavidmeister> this is just how fast it is
<tilpner> So if your issue is related to cache.nixos.org malfunctioning somehow, a different one might help
<thedavidmeister> it's not malfunctioning, i can download it from wget
<tilpner> But it it's just nixos-rebuild-side "this is taking too long"...
<thedavidmeister> it's nixos-rebuild timeout from curl
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
<thedavidmeister> yeah afaics it's just "taking too long" but what am i supposed to do about that?
felixfoertsch has joined #nixos
<tilpner> Have you tried telling curl to wait longer?
<thedavidmeister> i'm asking how to do that
<tilpner> NIX_CURL_FLAGS might help
<thedavidmeister> ah!
<tilpner> Don't know the curl flag you want, but you can try first inputting nonsense there to see if it errors
<{^_^}> [nixpkgs] @FRidh pushed 0 commits to python-unstable: https://git.io/fjd71
<thedavidmeister> https://nixos.org/nixos/manual/ not documented anywhere here?
<tilpner> That's not surprising :c
<tilpner> Seems like -m/--max-time <seconds> could work
Izorkin has quit [Ping timeout: 272 seconds]
<thedavidmeister> aight
<tilpner> Hmm
<tilpner> This might be useless actually
felixfoertsch2 has quit [Quit: Textual IRC Client: www.textualapp.com]
<tilpner> I found NIX_CURL_FLAGS in the context of fetchurl/fetchipfs
<tilpner> But if you're doing a store substitution from cache.nixos.org, those wouldn't be used
<thedavidmeister> oh
<thedavidmeister> yeah so my wget finished
<thedavidmeister> that's a 700MB download
<thedavidmeister> it's not reasonable to expect everyone to be able to dl that in a few mins
<tilpner> I agree that longer substitution times should be possible (and I don't know that they're not)
<{^_^}> [nixpkgs] @abbradar pushed to master « libbass: update and use real version numbers »: https://git.io/fjd75
<tilpner> thedavidmeister: Apparently this is a new issue, and (according to that comment) might be fixed by using newest Nix
<tilpner> (As in, new enough to not have a release yet)
<thedavidmeister> uh
<thedavidmeister> hmm
<thedavidmeister> well i'm trying to pull latest commit from nixpkgs
<thedavidmeister> so that's fine
<thedavidmeister> but i need the latest nix to get the latest nix?
<thedavidmeister> :thinking:
<tilpner> Downloading Nix should be smaller?
<tilpner> s/smaller/faster/
<tilpner> But let me check which version you need
<thedavidmeister> are you saying i should just comment everything out in my configuration.nix
<thedavidmeister> then put it back in?
<thedavidmeister> well, everything big...
<tilpner> thedavidmeister: Which channel are you on?
<thedavidmeister> i'm grabbing latest commit from nixpkgs
<thedavidmeister> a2bd8da766c7a7ea456afd099b86920dd902cec4
dansho has quit [Quit: Leaving]
<thedavidmeister> needs to be pretty new as i only got a fix for my scanner driver committed a few days ago
<thedavidmeister> i'm trying commenting out all large downloads, then putting them back in
<tilpner> master doesn't have a Nix version new enough for you
<tilpner> You'll need to overrideAttrs yourself
<thedavidmeister> ok i don't understand that
<thedavidmeister> how do i get the newest commit?
<thedavidmeister> that would have the new version of nix in it?
<tilpner> The new version of Nix is not in any version of nixpkgs
<thedavidmeister> oh
<thedavidmeister> how are those two things related?
aramiscd has quit [Ping timeout: 268 seconds]
<tilpner> Can you specify that some more?
<thedavidmeister> in the context of this conversation
<thedavidmeister> i expected the latest nixpkgs would be latest everything
<thedavidmeister> i dont know any other way to specify a "nix version" other than nix-channel --add xxx
<thedavidmeister> i dont know any other way to specify a "nix version" other than nix-channel --add xxx yyy
<tilpner> No, someone has to edit nixpkgs, send a PR, and then it has to be merged
<thedavidmeister> oh ok
<thedavidmeister> i see, so nix version *is* in nixpkgs but there's some kind of release cycle
<tilpner> So just because edolstra pushes to nixos/nix doesn't mean the package definition in nixos/nixpkgs is updated
<thedavidmeister> right i see
<tilpner> Well, the version you want is not released
<thedavidmeister> ok, that makes sense to me
<tilpner> So it makes for it to not be in nixpkgs
<tilpner> You can still get it, and try your luck
<thedavidmeister> so i can force a nix commit by overriding something in my configuration.nix?
<tilpner> Exactly
<thedavidmeister> ok i'd like to try that
<tilpner> Hold on
NightTrain has quit [Ping timeout: 268 seconds]
jgt has joined #nixos
<simpson> thedavidmeister: To demystify this for you, concretely you need a commit along the lines of https://github.com/NixOS/nixpkgs/commit/5399f34ad9481849720d14605ce87b81abe202e9
<simpson> Even after a Nix release, a commit like this still needs to happen.
<tilpner> thedavidmeister: Please keep in mind that you will be running Nix master, which may or may not eat your store. It is up to you to decide whether that risk is worth it
<tilpner> (But I just wanted to point that out again)
<thedavidmeister> eat my store?
<tilpner> thedavidmeister: https://tx0.co/fNB
<thedavidmeister> what are the consequense of that?
<thedavidmeister> a rebuild or a bricked laptop?
<tilpner> If you need your installation tomorrow for an important thing, you should delay this experiment
ardumont has quit [Ping timeout: 250 seconds]
<thedavidmeister> ok i will comment out everything then
<thedavidmeister> i don't *need* chrome and f
<thedavidmeister> f
<thedavidmeister> i have brave
<thedavidmeister> what is the frequency of nix releases?
jgt has quit [Ping timeout: 250 seconds]
<tilpner> thedavidmeister: The link I sent contains a snippet you can include in your config. That will change your system to use the currently most recent Nix master
<thedavidmeister> unfortunately i gotta gsd this weekend
<thedavidmeister> i'm already a bit frustrated from the delays so far
<thedavidmeister> i think i'll triage and try again later
ixxie has joined #nixos
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
asymptotically has joined #nixos
Guest59 has joined #nixos
Ariakenom has joined #nixos
LastBreath has joined #nixos
Yakulu has joined #nixos
ardumont has joined #nixos
Yakulu has left #nixos [#nixos]
Guest59 has quit [Client Quit]
<thedavidmeister> thanks for the help
chloekek has joined #nixos
thedavidmeister has quit [Ping timeout: 258 seconds]
thedavidmeister has joined #nixos
<thedavidmeister> exit
thedavidmeister has quit [Client Quit]
winem_ has joined #nixos
chagra has joined #nixos
<chagra> I login as my user from tty, is there an option to change the text displayed before the login prompt?
<symphorien> chagra: probably https://nixos.org/nixos/options.html#users.motd
Izorkin has joined #nixos
<tilpner> chagra: services.mingetty.greetingLine
gentauro has quit [Quit: /quit]
aswanson has quit [Quit: WeeChat 2.4]
chagra52 has joined #nixos
chagra52 has quit [Remote host closed the connection]
chagra has quit [Ping timeout: 260 seconds]
ixxie has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
rprije has quit [Ping timeout: 245 seconds]
<{^_^}> [nixpkgs] @FRidh pushed 74 commits to python-unstable: https://git.io/fjd5z
jgt has joined #nixos
lurpahi has joined #nixos
LastBreath has quit [Remote host closed the connection]
rprije has joined #nixos
LastBreath has joined #nixos
jgt has quit [Ping timeout: 264 seconds]
<{^_^}> [nixpkgs] @erikarvstedt opened pull request #66746 → <!-- Nixpkgs has a lot of new incoming Pull Requests, but not enough people to review this constant stream. Even if you … → https://git.io/fjd51
o1lo01ol1o has quit [Ping timeout: 272 seconds]
felixfoertsch23 has joined #nixos
felixfoertsch has quit [Ping timeout: 250 seconds]
freeman42x has quit [Ping timeout: 252 seconds]
<{^_^}> [nixpkgs] @erikarvstedt closed pull request #66746 → <!-- Nixpkgs has a lot of new incoming Pull Requests, but not enough people to review this constant stream. Even if you … → https://git.io/fjd51
<{^_^}> [nixpkgs] @erikarvstedt opened pull request #66747 → Androidenv fixes → https://git.io/fjd5H
alex`` has joined #nixos
mexisme has joined #nixos
<{^_^}> [nixpkgs] @lblasc opened pull request #66748 → libselinux, libsepol, selinux-sandbox: 2.7 -> 2.9 → https://git.io/fjdds
ilikeheaps has joined #nixos
<laas> Does anyone here use kakoune? I can't seem to figure out how to enable the autoload plugins
<laas> I only have a single file produced by the derivation called plugins.kak
<laas> it's empty
__monty__ has joined #nixos
<{^_^}> [nixpkgs] @timokau opened pull request #66749 → vimPlugins: update → https://git.io/fjddX
rprije has quit [Ping timeout: 272 seconds]
LouisDK has joined #nixos
nexgen has quit [Remote host closed the connection]
nexgen has joined #nixos
<LouisDK> Coming from Arch I've been playing around with NixOS unstable just wondering why some packages are way outdated eg. nano which is 9 months old.
<laas> nobody has bothered to update it probably
<laas> or it's being blocked by something
<laas> check PRs
jollyjester has joined #nixos
rprije has joined #nixos
<jollyjester> hello, i just installed nixos.fantasque-sans-mono but when doing fc-list | grep "Fantasque" nothing is showing up and also emacs gives me an error that the font is missing
<jollyjester> any idea on what i can do?
<tilpner> jollyjester: Use fonts.fonts, not nix-env or systemPackages
o1lo01ol1o has joined #nixos
tmaekawa has joined #nixos
<jollyjester> tilpner: how do i do that
<jollyjester> i'm new to nixos
<tilpner> You add fonts.fonts = [ pkgs.fantasque-sans-mono ]; to your configuration.nix
<jollyjester> makes sense
<jollyjester> thanks
<jollyjester> i wonder, what will happen when we add a few more thousand options to configuration.nix, will it get slower? how much more storage will it take/
<tilpner> Yes, it will get slower (assuming no improvements elsewhere)
<tilpner> Also see #57477
<jollyjester> hmmmm, nixos kind of tries to make you configure parts of your operating system easily with a global configuration file but that arrises problems such as these, i sitll believe vanilla systems somewhat like slackware will pass the test of time
<{^_^}> https://github.com/NixOS/nixpkgs/issues/57477 (by Infinisil, 22 weeks ago, open): `nixos-rebuild --switch` too slow, tracking issue
<tilpner> NixOS will certainly not replace all other distros soon
<tilpner> But the evaluation time is an acceptable cost to many people
LouisDK has quit [Remote host closed the connection]
<jollyjester> isn't the model of nixos kind of monolithic?
jgt has joined #nixos
<tilpner> I guess it is
<__monty__> I'd argue it's the opposite.
<jollyjester> well it uses a global configuration that configures everything, that's kind of monolithic
<jollyjester> instead of tools interacting with other tools
<simpson> jollyjester: If we imagine the filesystem as a store of state, then it's about mutable vs. immutable state.
<tilpner> Well, it still uses "tools interacting with other tools" to configure everything
<__monty__> *OR* it's the only tool that allows modular configuration, in the sense of doing one thing well.
<simpson> jollyjester: You are, to be clear, free to use non-NixOS. Nix should work on any Linux.
<jollyjester> yes i know
<jollyjester> i like nixos
<jollyjester> it's just that i'm concerned that adding more packages and options to work with will make the system slower
<__monty__> jollyjester: Slower than? Manually applying all the configuration and installing packages and setting up services?
<tilpner> __monty__: TBF having to wait 20s to add a package globally can be slower than apt install
o1lo01ol1o has quit [Remote host closed the connection]
<tilpner> I've been meaning to check how to separate my system into NixOS containers, so that I don't have to eval the entire thing at once
<simpson> But also, TBF, adding packages globally is something that drops off rapidly after the first week with NixOS.
jgt has quit [Ping timeout: 264 seconds]
<tilpner> simpson: Bad example then, use adding a single line to /etc/sudoers.conf
<__monty__> tilpner: Sure but they're not equivalent things. `apt-get install` is closer to imperative nix-env or nix-shell if we're ok with pretending imperative nix-env doesn't exist : )
<tilpner> */etc/sudoers
<__monty__> tilpner: Adding that to configuration.nix is still doing a lot more work than simply editing a text file though.
<simpson> tilpner: Sure. Totally agreed there, and this is the mutable vs. immutable tradeoff. Hard to improve this without falling into the potholes that tools like Puppet and Ansible have.
<tilpner> __monty__: I know, I'm just advocating for users that are no longer here because it seemed slow
domogled has joined #nixos
<{^_^}> [nixpkgs] @timokau merged pull request #66749 → vimPlugins: update → https://git.io/fjddX
<{^_^}> [nixpkgs] @timokau pushed to master « vimPlugins: update (#66749) »: https://git.io/fjdFa
<jollyjester> do you guys know anything about brave?
<jollyjester> it's version 61, latest stable is 64
<jollyjester> i emailed the maintainer, waiting for a response
<ivan> you can file issues on nixpkgs to request something be updated
<ivan> or submit a PR to do update it
<jollyjester> a PR?
<{^_^}> [nixpkgs] @WilliButz pushed to master « grafana-loki: fix typo in service config »: https://git.io/fjdFw
<ivan> it looks like it's 0.67.123 in master, so that could be backported
<ivan> jollyjester: pull request
<{^_^}> [rfcs] @FRidh opened pull request #50 → rfc 50: init → https://git.io/fjdFr
thc202 has joined #nixos
<jollyjester> oh yea
<jollyjester> makes sense
<jollyjester> i'll check out how to add backports
<jollyjester> backports are like the "bleeding edge" repo of nixos right, more up to date packages something like how debian has it
<{^_^}> [nixpkgs] @primeos pushed to master « signal-desktop: 1.26.1 -> 1.26.2 »: https://git.io/fjdFo
<jollyjester> how can i enable backports?
<jollyjester> can't find anything on it
<jollyjester> ivan:
<ivan> some things get backported to the stable release, there is nothing to enable
<{^_^}> [nixpkgs] @andir pushed 4 commits to release-19.03: https://git.io/fjdFK
<ivan> you could install brave from the unstable channel though
phreedom_ has joined #nixos
<ivan> jollyjester: nix-env -f https://nixos.org/channels/nixos-unstable/nixexprs.tar.xz -iA brave
<tilpner> -f channel:nixos-unstable will work too
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
m0rphism has joined #nixos
phreedom has quit [Ping timeout: 260 seconds]
chloekek has quit [Ping timeout: 268 seconds]
o1lo01ol1o has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
<jollyjester> thanks ivan
<jollyjester> works like a charm
<jollyjester> i tried guixsd, it was a mess
<jollyjester> barely any help and barely any documentation
<{^_^}> [nixpkgs] @primeos opened pull request #66750 → [19.03] signal-desktop: 1.26.1 -> 1.26.2 (backport) → https://git.io/fjdbI
o1lo01ol1o has joined #nixos
<jollyjester> ivan: what's the difference between nix-env -e and nix-env --uninstall
<jollyjester> also should you add packages to configuration.nix only when you want them to be "essential tools you can't live without"
<__monty__> jollyjester: One's shorthand.
<tilpner> jollyjester: -e is an alias for --uninstall
<jollyjester> tilpner: makes sense i guess
<jollyjester> how can i purge configuration files from installed packages
<jollyjester> and how can i clear cache, unused packages etc
<jollyjester> unused = orphaned i mean
<__monty__> Nix doesn't ordinarily touch $HOME
<tilpner> nix-collect-garbage
<tilpner> That will clean your store
<jollyjester> ok
<jollyjester> so i guess collecting garbage from home is done manually
<__monty__> jollyjester: Look into home-manager if you want more nix in your life.
<jollyjester> i just have done nix-collect-garbage and it deleted a lot like 157 MB worth of packages
<__monty__> And re what packages to include. I put the essentials in configuration.nix and most everything else in my home.nix.
<jollyjester> does nixos download ALL packages from the repos
<jollyjester> or something?
<tilpner> jollyjester: Look at its help page for flags to delete generations and free more space
Shouou has joined #nixos
<alexarice[m]> jollyjester: nixos doesn't delete packages when you uninstall them
<__monty__> 150MB is definitely not a lot in nix terms btw, not when talking disk space and not when talking RAM.
<jollyjester> oh wew
<jollyjester> i guess nixos might be the gentoo of binary packages then
<jollyjester> anyways, how can i switch to the unstable channel completely?
<jollyjester> like upgrade all my packages to it and use it
<alexarice[m]> jollyjester: change channel with `nix-channel` then rebuild
<jollyjester> ok
<tilpner> That will not affect your user profiles though
<alexarice[m]> tilpner: what do you mean
<{^_^}> [nixpkgs] @FRidh merged pull request #65262 → python3Packages.PyGithub: 1.36 -> 1.43.8 → https://git.io/fjM7z
<tilpner> alexarice[m]: IIRC nixos-rebuild and nix-env use different users channels by default, so changing just one will not affect their user profile
<tilpner> But it's been so long since I've used channels or seen a default config, so I might be wrong
<__monty__> root v. your user ordinarily.
<alexarice[m]> tilpner: ah I see, I'm used to installing everything with `configuration.nix`
<{^_^}> [nixpkgs] @FRidh merged pull request #66242 → python3Packages.pylibgen: 2.0.1 -> 2.0.2 → https://git.io/fjQzb
<{^_^}> [nixpkgs] @FRidh pushed 2 commits to master: https://git.io/fjdbo
<alexarice[m]> I believe if you have no channels on your user then it defaults to the root channel
<tilpner> alexarice[m]: They were asking about nix-env flags before, so I figured they might have user profiles with nix-env around
<{^_^}> [nixpkgs] @FRidh pushed commit from @Pamplemousse to master « pythonPackages.capstone: 3.0.4 -> 3.0.5 »: https://git.io/fjdb6
<{^_^}> [nixpkgs] @FRidh merged pull request #64930 → pythonPackages.capstone: 3.0.4 -> 3.0.5 → https://git.io/fj1Td
<{^_^}> [nixpkgs] @FRidh closed pull request #63545 → pythonPackages.{hiredis,aioredis,channels_redis,django-crispy-forms,django-filter}: init; Added baddecisionsalex to main… → https://git.io/fjVV5
LastBreath has quit [Read error: Connection reset by peer]
domogled has quit [Quit: domogled]
domogled has joined #nixos
<{^_^}> [nixpkgs] @FRidh pushed 13 commits to python-unstable: https://git.io/fjdbS
mexisme has quit [Ping timeout: 264 seconds]
jollyjester has quit [Quit: leaving]
<{^_^}> [nixpkgs] @FRidh merged pull request #66118 → doh-proxy: 0.0.8 -> 0.0.9 → https://git.io/fjQJo
<{^_^}> [nixpkgs] @FRidh pushed 2 commits to master: https://git.io/fjdb7
<{^_^}> [nixpkgs] @FRidh pushed commit from @tadeokondrak to master « discord-canary: 0.0.91 -> 0.0.92 »: https://git.io/fjdbF
<{^_^}> [nixpkgs] @FRidh merged pull request #66579 → discord-canary: 0.0.91 -> 0.0.92 → https://git.io/fj561
<{^_^}> [nixpkgs] @FRidh merged pull request #66744 → doc: Replace google groups reference with Discourse Patch category → https://git.io/fjd7e
<{^_^}> [nixpkgs] @FRidh pushed commit from @Infinisil to master « doc: Replace google groups reference with Discourse Patch category »: https://git.io/fjdbx
rsoeldner has joined #nixos
ng0 has quit [Ping timeout: 260 seconds]
jgt has joined #nixos
jgt has quit [Ping timeout: 264 seconds]
rprije has quit [Ping timeout: 246 seconds]
<{^_^}> [nixpkgs] @volth opened pull request #66751 → treewide: name -> pname → https://git.io/fjdNq
ng0 has joined #nixos
<tilpner> FRidh++
<{^_^}> FRidh's karma got increased to 7
<{^_^}> [nixpkgs] @markuskowa merged pull request #66715 → mixxx: 2.2.1 -> 2.2.2 → https://git.io/fjdr2
<{^_^}> [nixpkgs] @markuskowa pushed 2 commits to master: https://git.io/fjdNs
<{^_^}> [nixpkgs] @markuskowa merged pull request #66656 → [19.03] borgbackup: 1.1.9 -> 1.1.10 → https://git.io/fjdTB
<{^_^}> [nixpkgs] @markuskowa pushed 3 commits to release-19.03: https://git.io/fjdNG
drakonis has quit [Quit: WeeChat 2.5]
rprije has joined #nixos
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/2d0a2044b3c (from 3 hours ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
rprije has quit [Ping timeout: 245 seconds]
winem_ has quit [Ping timeout: 250 seconds]
o1lo01ol1o has quit [Ping timeout: 246 seconds]
rprije has joined #nixos
felixfoertsch23 has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
[Leary] has quit [Ping timeout: 245 seconds]
dsiypl4 has joined #nixos
asymptotically has quit [Quit: Leaving]
LastBreath has joined #nixos
Ariakenom has quit [Quit: Leaving]
tmaekawa has quit [Ping timeout: 245 seconds]
is_null has joined #nixos
rprije has quit [Ping timeout: 245 seconds]
o1lo01ol1o has joined #nixos
LastBreath has quit [Read error: Connection reset by peer]
LastBreath has joined #nixos
b has joined #nixos
rprije has joined #nixos
slyfox has quit [Quit: \o/]
nexgen has quit [Quit: Leaving]
is_null has quit [Ping timeout: 244 seconds]
pingiun has quit [Read error: Connection reset by peer]
ottidmes has joined #nixos
is_null has joined #nixos
jgt has joined #nixos
<{^_^}> [nixpkgs] @bachp opened pull request #66752 → plex: 1.16.3.1402 -> 1.16.4.1469 → https://git.io/fjdN2
o1lo01ol1o has quit [Remote host closed the connection]
<{^_^}> Channel nixpkgs-19.03-darwin advanced to https://github.com/NixOS/nixpkgs/commit/a4179730db4 (from 40 minutes ago, history: https://channels.nix.gsc.io/nixpkgs-19.03-darwin)
jgt has quit [Ping timeout: 264 seconds]
slyfox has joined #nixos
o1lo01ol1o has joined #nixos
bakakuna has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
<{^_^}> [nixpkgs] @markuskowa opened pull request #66753 → openblas: 0.3.6 -> 0.3.7 → https://git.io/fjdN6
dasj19 has joined #nixos
bakakuna has quit [Ping timeout: 268 seconds]
Thra11 has joined #nixos
<{^_^}> [nixpkgs] @fadenb opened pull request #66754 → graylog: 3.0.2 -> 3.1.0 → https://git.io/fjdN7
<{^_^}> [nixpkgs] @primeos pushed to master « nvme-cli: 1.8.1 -> 1.9 »: https://git.io/fjdNA
juhe has joined #nixos
dm9 has joined #nixos
dasj19 has quit [Quit: dasj19]
dasj19 has joined #nixos
Lears has joined #nixos
<juhe> Hi, I'm trying to repair botched nix store after not smart usage of "nix-path --delete --ignore-liveness", so I booted from usb stick, mounted root to /mnt and used "nix-store --dump" on machine which contains required paths and "nix-store --restore" on target. This even though copies files, doesn't seem to work as I expected, because nixos-enter brings "bash-4.4" and not usual root promt. The nix-store --verify-path (from withing chroot)
<juhe> marks the copied paths invalid as well... Any idea what am I missing?
leotaku has quit [Ping timeout: 245 seconds]
rprije has quit [Ping timeout: 246 seconds]
<__monty__> If nothing better's suggested, as long as you have your configuration.nix I think nixos-install should get you set up right.
jb55 has quit [Read error: Connection reset by peer]
klntsky has quit [Read error: Connection reset by peer]
ng0 has quit [Read error: Connection reset by peer]
Palpares has quit [Remote host closed the connection]
oida has quit [Remote host closed the connection]
phreedom_ has quit [Remote host closed the connection]
klntsky has joined #nixos
jb55 has joined #nixos
<juhe> Yes, that would work, but I would like to know why copied path is not considered valid. Is there any extra metadata kept outside of /nix/store/ that was not updated by "nix-store --dump"?
ng0 has joined #nixos
<juhe> actually by "nix-store --restore"...
Palpares has joined #nixos
<__monty__> There's a database in /nix/var I believe but that should be updated by restore.
jgt has joined #nixos
phreedom has joined #nixos
leotaku has joined #nixos
tilpner has quit [Quit: WeeChat 2.4]
<juhe> Hm, maybe it is not updated with nix-store --restore, but nix-store --import. The difference between the two seems to be in --import/--export using some more metadata. But, I can't use that since it seems to use /nix/store and not /mnt/nix/store. I'll check man page if it is possible to change default /nix/store path with nix-store --import/export ones.
jgt has quit [Ping timeout: 250 seconds]
<__monty__> Shouldn't your store be in the right place after the chroot?
<gchristensen> --export and --import are better choices
<gchristensen> juhe: how did you come to decide to use --ignore-liveness?
o1lo01ol1o has quit [Remote host closed the connection]
anderslundstedt has quit [Remote host closed the connection]
<juhe> gchristensen: well, don't ask... :-|. I wanted to free up some space since the machine has only 32GiB storage. I know it is pretty stupid thing to do, but I didn't find the command that would remove some (unimportant) stuff from /nix/store even if it would render it somewhat broken. My idea was to nix-deploy to that machine and reboot and then collect garbage so the rest of the paths would be removed.
anderslundstedt has joined #nixos
<gchristensen> juhe: aye, but how did you learn that --remove-liveness exists?
<juhe> gchristensen: from manual page of nix-store
<gchristensen> okay
<gchristensen> it seems that while it is very silly to use the flag, the results seem more severe than I expect
<__monty__> Depends on what you remove though.
<juhe> Yes it's dangerous command :)
<gchristensen> the man page doesn't indicate that well
<adisbladis> bendlas: Ping?
<__monty__> adisbladis: How do you know I'm having trouble with ping?
<{^_^}> [nixpkgs] @Mic92 merged pull request #66751 → treewide: name -> pname → https://git.io/fjdNq
<{^_^}> [nixpkgs] @Mic92 pushed 2 commits to staging: https://git.io/fjdA6
<juhe> It could be inferred though. Maybe having explicit note that it removes also paths it is depending on (like system-path or some etc stuff), it would be fine though.
<Palpares> is nix intended to work fully on other distro such debian or is it normal 50% of packgs are broken there?
<{^_^}> [nixpkgs] @Mic92 merged pull request #66712 → pyprof2calltree: move to python-packages → https://git.io/fjdV1
<{^_^}> [nixpkgs] @Mic92 pushed 2 commits to master: https://git.io/fjdAP
<juhe> I'll try nix-store --export/--import with explicitly set NIX_{STORE,DATA,LOG,STATE,CONF}_DIR env variables, maybe it'll solve my problem.
oida has joined #nixos
<gchristensen> Palpares: Nix should work on other distros, yes
<gchristensen> there are some caveats with programs which use GL etc.
dm9 has quit [Quit: WeeChat 2.5]
<Palpares> tried to install qutebrowser, vim.youcompleteme without success on debian while this works fine on nixos. is it ok to post issues on github for debian ?
<bendlas> <adisbladis "bendlas: Ping?"> yep?
<gchristensen> Palpares: yep
LastBreath has quit [Read error: Connection reset by peer]
<Palpares> also, I do have package collision: eclipse and scala-ide both produce an eclipse executable and make both not working together. Is it a normal behavior, or should it be corrected in the nixpkg ?
Thra11 has quit [Ping timeout: 246 seconds]
Walker3R has joined #nixos
tilpner has joined #nixos
<adisbladis> bendlas: Why is the melpa update run as --fg-daemon instead of using --batch ?
<{^_^}> [nixpkgs] @Mic92 merged pull request #66564 → varnish5: fix build → https://git.io/fj52p
<{^_^}> [nixpkgs] @Mic92 pushed 3 commits to master: https://git.io/fjdxs
<bendlas> adisbladis: because of thread usage. with --batch, emacs immediately closes
<adisbladis> Right, that makes sense
<adisbladis> Thanks :)
<adisbladis> I'm debugging a case when the updater silently hangs
noonien has joined #nixos
<adisbladis> bendlas: Maybe I can pm you?
<bendlas> sure
__monty__ has quit [Ping timeout: 244 seconds]
__monty__ has joined #nixos
philr has joined #nixos
<{^_^}> [nixpkgs] @primeos pushed 4 commits to master: https://git.io/fjdxR
<juhe> Unfortunately using nix-store --import/--export and custom NIX_*_DIR env vars doesn't work, getting "error: path '...' is not in the Nix store.
eof has quit [Ping timeout: 246 seconds]
<Miyu-chan> nix-store --export does not export the closure btw.
<juhe> Miyu-chan: that's ok, I can loop through the referrers and get the missing paths.
<juhe> ideally, nix-copy-closure would have an option to specify alternate nix store path, that would be easiest.
jgt has joined #nixos
<nkaretnikov> i'm trying to define a custom set of packages. is there a way to automatically derive all inputs from the current scope when calling `import ./path/to/package { <here> };` instead of specifying all of them instead of `<here>`?
<nkaretnikov> i think this is somehow related to `inherit`, but i'm not sure. i recall it's possible, but maybe i'm just mistaken
<Miyu-chan> juhe: Just for reference, `nix-store -qR` gets the whole dependency list.
jgt has quit [Ping timeout: 250 seconds]
<nkaretnikov> here's the example, i'm trying to follow: http://sandervanderburg.blogspot.com/2014/07/managing-private-nix-packages-outside.html the file i'm referring to in my question is called custom-package.nix in the blog post.
<nkaretnikov> s/,//
Walker3R has quit [Quit: Quit]
Walker3R has joined #nixos
<nkaretnikov> the author does `gpm = import ./pkgs/gpm { inherit (pkgs) stdenv fetchurl flex bison ncurses; };`. can it be done implicitly, something like `gpm = import ./pkgs/gpm { ... };`
Walker3R has quit [Client Quit]
Walker3R has joined #nixos
<__monty__> nkaretnikov: with?
Walker3R has quit [Client Quit]
Walker3R has joined #nixos
<nkaretnikov> __monty__: not sure i understand. https://nixos.org/nix/manual/#idm140737317848160 looking here, it doesn't seem to do what i want. or can you provide an example?
Walker3R has quit [Client Quit]
Walker3R has joined #nixos
<{^_^}> [nixpkgs] @primeos pushed to master « gns3Packages.{server,gui}Preview: 2.2.0b4 -> 2.2.0rc3 »: https://git.io/fjdxp
<__monty__> No, you're right, that doesn't capture the names in the set.
hmpffff has joined #nixos
<nkaretnikov> ah, wait. i think it's shown at the end of the blog post
hmpffff has quit [Client Quit]
asymptotically has joined #nixos
<nkaretnikov> (with callPackage) let me try that
Walker3R has quit [Ping timeout: 252 seconds]
Walker3R has joined #nixos
chimay has joined #nixos
hmpffff has joined #nixos
LouisDK has joined #nixos
chimay has quit [Client Quit]
<{^_^}> [nixpkgs] @Mic92 merged pull request #66294 → firestarter: init at 1.7.3 → https://git.io/fjQ9E
<{^_^}> [nixpkgs] @Mic92 pushed 2 commits to master: https://git.io/fjdpU
Walker3R has quit [Remote host closed the connection]
<LouisDK> I tried to upgrade nano to 4.3 by editing nano/default.nix after cloning the github repo. I only changed the version af sha256sum (which I got from https://ftpmirror.gnu.org/nano/nano-4.3.tar.xz). However I get this error: https://pastebin.com/Qn4qGH7J - neither of the checksums in the output is in my nan/default.nix - do I need to edit another fil
<LouisDK> e too?
<{^_^}> [nixpkgs] @AluisioASG opened pull request #66756 → libreoffice-fresh: 6.2.4.2 -> 6.3.0.4 → https://git.io/fjdpI
<{^_^}> Channel nixos-19.03-small advanced to https://github.com/NixOS/nixpkgs/commit/a4179730db4 (from 3 hours ago, history: https://channels.nix.gsc.io/nixos-19.03-small)
<__monty__> LouisDK: How did you figure out the sha256?
<__monty__> Usually people use
<__monty__> ,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.
<AluisioASG> @LouisDK The hash I get is 108dzj6azqmini9bvgl26r0q1y59s3nnrw75hfzv91bs50davlq0. What is the hash you put in default.nix?
chloekek has joined #nixos
<{^_^}> [nixpkgs] @nh2 merged pull request #66490 → zlib: Properly clean up static/shared distinction → https://git.io/fj7hZ
<{^_^}> [nixpkgs] @nh2 pushed to master « zlib: Properly clean up static/shared distinction »: https://git.io/fjdpO
sb0 has quit [Quit: Leaving]
<__monty__> Is there a way to override something in a let? In particular, I'd like to override one of pass's extensions.
dsiypl4 has quit [Ping timeout: 244 seconds]
<tilpner> __monty__: Generally not, but it seems you're allowed to pass a function
<tilpner> pass.withExtensions (e: [ ... ])
Walker3R has joined #nixos
<tilpner> Just pass something that's not directly taken from e
Walker3R has quit [Remote host closed the connection]
Okinan has quit [Quit: Konversation terminated!]
bakakuna has joined #nixos
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
<__monty__> tilpner: Ah, clever. In this case I got around it by overlaying pass to be from the local checkout.
rauno has joined #nixos
rj1 has quit [Quit: bye]
rj1 has joined #nixos
rj1 has joined #nixos
rj1 has quit [Changing host]
fendor has joined #nixos
<LouisDK> @AluisioASG - How did you get that hash?
freeman42x has joined #nixos
hmpffff has quit [Quit: Bye…]
<{^_^}> [nixpkgs] @edolstra merged pull request #66686 → rustc: 1.36.0 -> 1.37.0 → https://git.io/fjdll
<{^_^}> [nixpkgs] @edolstra pushed 3 commits to staging: https://git.io/fjdhi
rauno has quit [Ping timeout: 250 seconds]
<siraben> How do I delete all generations except the last N ones?
<__monty__> siraben: Don't thing there's a way. Closest thing is going by age.
<{^_^}> [nixpkgs] @edolstra merged pull request #66739 → readline80: 8.0p0 -> 8.0p1 → https://git.io/fjdHt
<{^_^}> [nixpkgs] @edolstra pushed 2 commits to master: https://git.io/fjdjJ
<__monty__> ,tofu LouisDK
<{^_^}> LouisDK: 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.
<LouisDK> __monty__ I did "curl https://ftpmirror.gnu.org/nano/nano-4.3.tar.xz | md5sum | sha256sum". Does seem to be the proper way.
<siraben> __monty__: how do I do it by age then?
jgt has joined #nixos
<LouisDK> @{^_^}, thx will try that
<__monty__> siraben: nix-env --delete-generations 30d
<siraben> __monty__: Ah I see. Thanks!
<__monty__> siraben: Actually, seems like +N would do what you want.
<__monty__> Wait, no maybe that's specific gens.
<hyperfekt> LouisDK: lib.fakeSha256 is easier than remembering how many zeroes, by the way. You can also use the nix-prefetch-url command.
<__monty__> LouisDK: Or use the mnemonic, as many 0's as cards in a deck.
freeman42x has quit [Read error: Connection reset by peer]
<LouisDK> hyperfekt, __monty__, thx. Will look into that.
ddellacosta has joined #nixos
jgt has quit [Ping timeout: 252 seconds]
LouisDK has quit [Remote host closed the connection]
dsiypl4 has joined #nixos
freeman42x has joined #nixos
<infinisil> I mostly use 52i0 in vim
Ariakenom has joined #nixos
<clever> infinisil: ive found that inserts it at the wrong offset, and lands outside the quotes, its also abnormally slow
<clever> infinisil: ive been wanting to figure out and then "steal" this chunk of code, https://github.com/chessai/nixos-configs/blob/ac508a55caccc4d7e1dc8e6479e343c1b2af740f/pkgs/vim/config.nix#L13-L24
<infinisil> Neato
<__monty__> siraben: Yeah, if you want to keep the 3 most recent generations do `nix-env --delete-generations +3`, this is 3 *including* the current.
<siraben> __monty__: Ah that's even better. Thanks.
rj1 has quit [Quit: bye]
<{^_^}> [nix] @toonn opened pull request #3048 → Fix nix-env documentation for --delete-generations → https://git.io/fjdjo
<siraben> That sort of syntax is confusing. I wonder if nix-env --keep-generations would make more sense.
<__monty__> siraben: You could open a UX issue at NixOS/nix I guess. Tbh, I intuited the + to mean current+x.
<__monty__> I have a build that fails because the makefile is trying to change permissions on /etc/bash_completion.d, does nix have a way of patching/providing a good path for the bash_completions? (Better yet a way to turn bash_completions into fish completions?)
<{^_^}> [nixpkgs] @kim0 opened pull request #66757 → xmrig: 2.14.4 -> 3.0.0 → https://git.io/fjdjX
<__monty__> How do I set up ssh properly? When I ssh into a newly setup nixos box I end up in an environment that has almost no binaries on the path.
MmeQuignon has joined #nixos
<{^_^}> [nixpkgs] @primeos merged pull request #66750 → [19.03] signal-desktop: 1.26.1 -> 1.26.2 (backport) → https://git.io/fjdbI
<{^_^}> [nixpkgs] @primeos pushed 2 commits to release-19.03: https://git.io/fjFeq
<maralorn> __monty__: How does the PATH look? What binaries are you missing?
Walker3R has joined #nixos
<__monty__> maralorn: PATH=/usr/bin:/bin:/usr/sbin:/sbin:/nix/store/hash-openssh-7.9p1/bin
Walker3R has quit [Remote host closed the connection]
<__monty__> Almost nothing is available with this PATH.
<__monty__> Tmux for example, is not.
bakakuna has quit [Ping timeout: 245 seconds]
<{^_^}> [nixpkgs] @7c6f434c merged pull request #66756 → libreoffice-fresh: 6.2.4.2 -> 6.3.0.4 → https://git.io/fjdpI
<{^_^}> [nixpkgs] @7c6f434c pushed 2 commits to master: https://git.io/fjFeO
<andi-> __monty__: but you added it to the system environment / the users environment ?
<__monty__> andi-: Yes, locally everything's available.
<__monty__> I am *not* using a module yet btw. Just added openssh to my packages because I haven't translated my ssh config yet.
<andi-> You are also missing /run/current-system/sw/bin etc…
<andi-> how did you define the user?
<{^_^}> [nixpkgs] @Profpatsch opened pull request #66758 → tree-sitter: init at 0.15.7 → https://git.io/fjFe6
Profpatsch has joined #nixos
<{^_^}> [nixpkgs] @nh2 opened pull request #66759 → treewide: Remove unnecessary `--disable-static` → https://git.io/fjFei
<maralorn> __monty__: It looks like something is circumventing the nixos mechanism for setting PATH. This PATH comes from a non nixos system. Have you perhaps copied your non-nixos .bashrc or .zshrc to the ssh-host?
<__monty__> andi-: In configuration.nix, users.users.me, specified wheel group and my shell and isNormalUser.
<__monty__> maralorn: Nope.
<__monty__> I suspect I'm missing some ssh configuration that loads the right paths.
<{^_^}> [nixpkgs] @Ma27 merged pull request #66754 → graylog: 3.0.2 -> 3.1.0 → https://git.io/fjdN7
<{^_^}> [nixpkgs] @Ma27 pushed 2 commits to master: https://git.io/fjFeD
<maralorn> __monty__: But the default should definitely be that on nixos every login session is already setup with the right PATH. So I feel like something is overriding it.
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/56e5dad3d8b (from 2 hours ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
<clefru> Is there a way to talk to wpa_supplicant via wpa_cli when run regularly from NetworkManager? I can't seem to find the control socket for it, also I don't see it having any config file on the command line. It looks like it is spawned with "wpa_supplicant -u" for control via dbus, but how do I hook up wpa_cli with dbus?
orivej has joined #nixos
<__monty__> maralorn: Why would it override it only for ssh access though? And include just the ssh binary in the PATH? I suspect this is some sort of sandboxing done by nixos which is relaxed when using programs.ssh.
<maralorn> __monty__: Do I understand correctly that you installed openssh with nix on the client to connect via ssh?
<andi-> __monty__: can you post the configuration somewhere?
<maralorn> __monty__: Is that PATH in the ssh session?
slyfox has quit [Ping timeout: 248 seconds]
<__monty__> maralorn: Yes, and I have services.ssh.enable = true in configuration.nix.
<__monty__> andi-: Nix or ssh? The ssh config is just a bunch of keys and some hosts defined in .ssh/config.
<andi-> __monty__: probably the nix configuration unless you have a very fancy ssh config
<{^_^}> [nixpkgs] @LouisDK1 opened pull request #66760 → nano 3.2 -> 4.3 → https://git.io/fjFvc
<maralorn> __monty__: This seems a bit weird so maybe we are/I am having some misunderstanding here. But to me it looks like you are carying the client PATH over to the session in your host? So maybe you are using an ssh option which is transfering environment variables?
<__monty__> maralorn: I'm not specifying any ssh options, just the user@ip.
<__monty__> andi-: https://github.com/toonn/nix-config relevant parts are in nixos/ and home/, note that part of that home.nix is commented out here.
<{^_^}> [nixpkgs] @florianjacob opened pull request #66761 → nixos/utils: use removePrefix in escapeSystemdPath → https://git.io/fjFva
jgt has joined #nixos
<andi-> __monty__: have you tried with another shell? Not sure what fish does…
ryantrinkle has joined #nixos
<__monty__> andi-: I haven't, just ssh user@ip bash?
<andi-> i mean as the login shell of the user you are connecting as
<andi-> Is my best guess so far
<andi-> I'll have to run now. Good luck :-)
<clever> __monty__: youll also want a -t for ssh there
jgt has quit [Ping timeout: 264 seconds]
<ryantrinkle> is there a standard way to supply iana-etc stuff to builds?
<ryantrinkle> i.e. place something in the sandbox's /etc
<ryantrinkle> (without being impure)
<maralorn> And bash will probably not be in the PATH?
ddellacosta has quit [Ping timeout: 245 seconds]
<__monty__> Bash isn't, sh is but it indeed just inherits the path. Can I avoid rebuilding to just test a different login shell?
<maralorn> __monty__: You can probably find the path to a bash in the /nix/store manually?
<maralorn> find /nix/store -name bash
<maralorn> But I am not convinced that will help :-/
<__monty__> maralorn++ Stellar idea : ) The path is indeed correct if I `ssh user@ip -t /nix/store/blah-bash/bin/bash`.
<{^_^}> maralorn's karma got increased to 1
<maralorn> \o/
<maralorn> [x] My first karma.
<__monty__> So for some reason fish-over-ssh is not loading the same environment as fish-on-target, hmm.
<maralorn> __monty__: Is the home.nix already installed on the host your ssh-ing into?
juhe has quit [Quit: Leaving.]
<__monty__> Yes, modulo some commenting.
<maralorn> Well I can‘t tell. But maybe the culprit is in "~/src/dotfiles/fish/functions"?
<__monty__> maralorn: Those are just some aliases I use basically.
Shouou has quit [Ping timeout: 252 seconds]
wildtrees has joined #nixos
<maralorn> I still think that a vanilla fish should have a correct PATH so I wonder where it‘s geting overriden.
<maralorn> *being overriden?
<maralorn> Looks innocent, I admit.^^
wildtrees has quit [Remote host closed the connection]
wildtrees has joined #nixos
<__monty__> I am innocence incarnate!
<maralorn> __monty__: From https://nixos.org/nixos/options.html#users.users.%3Cname%3F%3E.shell "Don’t forget to enable your shell in programs if necessary, like programs.zsh.enable = true;"
<maralorn> __monty__: After reading nixos/modules/programs/fish.nix I am now nearly convinced that "programs.fish.enable = true" will solve your problem.
<__monty__> maralorn: Won't that clash with home-manager's programs.fish?
<maralorn> __monty__: Don‘t think so.
<hyperfekt> __monty__: iirc programs installed via home-manager appear in the path first, so it won't clash but it'll use home-managers package
<maralorn> The programs.fish in configuration.nix sets some environments files in /etc.
<{^_^}> [nixpkgs] @FRidh merged pull request #66596 → Update PyPy and FOSS EDA tools → https://git.io/fj5yl
<{^_^}> [nixpkgs] @FRidh pushed 8 commits to master: https://git.io/fjFfm
<maralorn> Exactly. home-managers programs.fish takes precedence, but it will also use the /etc config.
<{^_^}> [nixpkgs] @FRidh pushed 6 commits to python-unstable: https://git.io/fjFfs
Shouou has joined #nixos
<{^_^}> Channel nixpkgs-19.03-darwin advanced to https://github.com/NixOS/nixpkgs/commit/3864fd8f53b (from 63 minutes ago, history: https://channels.nix.gsc.io/nixpkgs-19.03-darwin)
orivej has quit [Ping timeout: 268 seconds]
hoijui has joined #nixos
freeman42x has quit [Quit: Leaving]
freeman42x has joined #nixos
Izorkin has quit [Quit: ZNC 1.7.4 - https://znc.in]
<{^_^}> [nixpkgs] @FRidh pushed to master « pythonpackages.gspread: fix build »: https://git.io/fjFf8
<ryantrinkle> if anyone has an interest in wireguard (particularly for mobile devices), could you help review this PR? https://github.com/NixOS/nixpkgs/pull/66689
<{^_^}> #66689 (by ryantrinkle, 1 day ago, open): wireguard: allow routes to overlap with other routes
Thra11 has joined #nixos
<ryantrinkle> this is necessary if your device is sometimes physically connected to the network it's VPNing into, and sometimes not, and you don't want to bother doing anything in between
<Thra11> Do I need zfs-kernel if I'm not planning to use any zfs filesystems?
<clever> Thra11: you can skip it
<Thra11> (It doesn't compile with the kernel sources I'm using, so I figure I'll try to disable it rather than fix it)
<Thra11> clever: Do you know what change to make to skip it when building an sd-image?
hoijui has quit [Quit: Leaving]
<clever> Thra11: first, you want to see what depends on it, check something like nix-store -q --tree
<{^_^}> [nixpkgs] @FRidh merged pull request #66631 → btrbk: 0.28.0 -> 0.28.3 → https://git.io/fjdeC
<{^_^}> [nixpkgs] @FRidh pushed commit from @jonringer to master « btrbk: 0.28.0 -> 0.28.3 »: https://git.io/fjFfg
<{^_^}> [nixpkgs] @adisbladis opened pull request #66762 → Python minimal → https://git.io/fjFf2
<{^_^}> [nixpkgs] @FRidh merged pull request #66668 → remove all instances of nix-env -i without -A in the NixOS manual → https://git.io/fjdY4
<{^_^}> [nixpkgs] @FRidh pushed commit from @symphorien to master « remove all instances of nix-env -i without -A in the NixOS manual »: https://git.io/fjFfa
<Thra11> It's linux-modules which depends on it
<clever> Thra11: what does `nix show-derivation` say about the linux-modules drv file?
freeman42x has quit [Quit: Leaving]
freeman42x has joined #nixos
freeman42x has quit [Remote host closed the connection]
kleisli has joined #nixos
freeman42x has joined #nixos
noudle has joined #nixos
<{^_^}> [nixpkgs] @nh2 opened pull request #66763 → freetype: Remove unnecessary `--disable-static`. → https://git.io/fjFfi
<Thra11> many things
<{^_^}> [nixpkgs] @tadeokondrak opened pull request #66764 → discord-canary: 0.0.92 -> 0.0.93 → https://git.io/fjFf1
<clever> Thra11: can you pastebin it?
jgt has joined #nixos
kleisli has quit [Ping timeout: 272 seconds]
kleisli has joined #nixos
orivej has joined #nixos
<clever> Thra11: it appears to be in the pkgs attribute, line 69
<clever> Thra11: and that drv is the result of pkgs.buildEnv, so the pkgs was generated here
jgt has quit [Ping timeout: 252 seconds]
iyzsong has quit [Quit: ZNC 1.7.1 - https://znc.in]
endformationage has joined #nixos
gentauro has joined #nixos
<clever> Thra11: which is being called in this area
<clever> i suspect zfs is in extraModulePackages
<clever> which happens if zfs is enabled
<clever> Thra11: which nixpkgs rev are you using?
<Thra11> clever: I suspect nixos/modules/profiles/base.nix line 52 is where it all starts
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66765 → ace: 6.5.5 -> 6.5.6 → https://git.io/fjFJO
<clever> that would do it
<Thra11> clever: I've removed it from the list and it no longer wants to build zfs-kernel :)
<clever> nice
orivej has quit [Ping timeout: 272 seconds]
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66766 → alfred: 2019.2 -> 2019.3 → https://git.io/fjFJW
<__monty__> maralorn++, hyperfekt++, andi-++, clever++ Thank you! This is much more comfortable.
<{^_^}> andi-'s karma got increased to 18, clever's karma got increased to 171, hyperfekt's karma got increased to 1, maralorn's karma got increased to 2
ixxie has joined #nixos
Izorkin has joined #nixos
<andi-> This bot surprises me every other day
philr has quit [Ping timeout: 248 seconds]
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
growpotkin has quit [Quit: ZNC 1.7.4 - https://znc.in]
growpotkin has joined #nixos
<nkaretnikov> is there a way to force rebuilds without messing with the store when building from a local directory (src = ./.)?
<nkaretnikov> ideally, i'd rather define a package such that the rebuild is triggered automatically
<nkaretnikov> but it doesn't seem to be the case for such local builds
<nkaretnikov> (or maybe it's the case and my build script has a bug)
<balsoft> It should rebuild every time the source changes
<nkaretnikov> balsoft: but how would it detect it if there's no hash?
<balsoft> The problem usually is that it rebuilds way too much (even if you just mess with git or something)
<nkaretnikov> i literally just have src = ./. pointed to a local directory
<balsoft> It recalculates the hash every time for local directories I believe
<clever> nkaretnikov: every time you eval the expr, it will hash ./. recursively
<nkaretnikov> the directory in question is a symlink. could it be the cause of the issue?
<clever> yeah
<nkaretnikov> ugh :(
<balsoft> Yes
<clever> it will eval the symlink itself, not the destination
<nkaretnikov> okay, thx!
<clever> so it will just be a hash of the path its pointing to
<clever> if the sandbox is enabled, the destination wont exist, and it will fail
gentauro has quit [Ping timeout: 245 seconds]
slack1256 has joined #nixos
rsoeldner has quit [Read error: Connection reset by peer]
fusion809 has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66767 → btrfs-progs: 5.1.1 -> 5.2.1 → https://git.io/fjFJ2
rsoeldner has joined #nixos
fendor has quit [Ping timeout: 248 seconds]
<aveltras> does anyone have an example of a haskell project default.nix with an override of all-cabal-hashes ? it doesn't seem to work the way i think it works. Im trying to get the latest polysemy but i complains it's not found in the archive, here's my default.nix
lyvidev has quit [Ping timeout: 260 seconds]
LouisDK has joined #nixos
<clever> aveltras: you probably just want callHackageDirect
<clever> nix-repl> haskellPackages.callHackageDirect
<clever> «lambda @ /nix/store/xcn5jx4di3fbkfgdv3bjmj3ap3p65k9z-nixos-19.09pre179307.bc94dcf5002/nixos/pkgs/development/haskell-modules/make-package-set.nix:183:25»
<clever> 183 callHackageDirect = {pkg, ver, sha256}@args:
<clever> aveltras: you give it the name of a package, the version, and the sha256 of its tarball from hackage, and then it can operate without being present in all-cabal-hashes
<clever> ,tofu
lyvidev has joined #nixos
<{^_^}> 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.
<clever> the hash
<LouisDK> I'm searching for a way to use a git repo for specific packages in configuration.nix. The closest I've come is nixpkg pinning but I don't want to add a snapshot.
nexgen has joined #nixos
<nh2> How should I best test treewide changes PR'd to staging?
<{^_^}> #66759 (by nh2, 1 hour ago, open): treewide: Remove unnecessary `--disable-static`
<nh2> In general, what's the testing approach for staging? Do we let ofborg build it all, or how does it work?
chloekek has quit [Ping timeout: 268 seconds]
<clever> nh2: id let a hydra just build the entire release.nix
<nh2> clever: how do I start that?
<clever> nh2: and i think thats the main purpose of the staging branch, to pre-test such things before master gets broken
<nh2> clever: you mean just merge to staging and let hydra build it post-merge?
<clever> nh2: i'm thinking you can contact gchristensen and see about getting it tested better
<clever> nh2: either that, or have a dedicated jobset made for your branch
<nh2> staging stuff doesn't go to cache.nixos.org, right?
<clever> nh2: everything hydra.nixos.org does goes to cache.nixos.org
<aveltras> clever: is callHackageDirect supposed to be combined with callCabal2nix ?
<clever> aveltras: it does callCabal2nix for you
<clever> 183 callHackageDirect = {pkg, ver, sha256}@args:
<clever> 184 let pkgver = "${pkg}-${ver}";
<clever> 185 in self.callCabal2nix pkg (pkgs.fetchzip {
<aveltras> clever: it's complaining about missing argument type-errors-pretty which seems to be a really recent addition
<aveltras> and adding it with callHackageDirect in source-overrides doesnt fix it
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66768 → analog: 6.0.15 -> 6.0.16 → https://git.io/fjFJr
<clever> aveltras: youll need a 2nd callHackageDirect based override to introduce that to your package set
<clever> aveltras: try fetchzip in source-overrides then?, i think source-overrides handles the cabal2nix for you
<nh2> oh. OK.
<nh2> clever: if I want to do the full set of stuff that hydra does myself, how would I evaluate release.nix?
<clever> nh2: just eval it with nix repl, nix-build, or hydra-eval-jobs, or setup a full hydra
<nh2> clever: also which `release.nix` is it, `nixos/release.nix` or `pkgs/top-level/release.nix`?
<clever> nh2: pkgs/top-level/release.nix is just the pkgs tree, which is probably what you want
<clever> nixos/release.nix is all of the nixos tests
<aveltras> clever: aint this supposed to be enough ?
<clever> aveltras: oh, right, you need to use the set that this override applies to
<clever> aveltras: so the packages can inter-link
<nh2> clever: does `nixos/release.nix` include the full set of packages, and just does the automated tests in addition? I'd like to check both that my change passes all nixos tests, but also buidls all nixpkgs packages that are supposed to work that don't have NixOS tests
<clever> nh2: that would be release-combined.nix
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
<nh2> clever: so that would be `nix-build nixos/release-combined.nix` and I'm good?
<clever> nh2: note that the eval of the nix expr alone requires several gig of ram, and the resulting jobs need linux, darwin, and aarch64 build slaves, and at least 80gig of disk space
<nh2> clever: the disk and RAM by static-haskell-nix CI server can provide, but can I restrict it to do only Linux?
<clever> aveltras: https://gist.github.com/cleverca22/d386271d1c0e97dada9aab91ca219950 is what i prefer doing, the name on line 16 and 1 will have to match your cabal file
<clever> nh2: --arg supportedSystems '[ "x86_64-linux" ]'
<aveltras> clever: "haskellPackages.multi-ghc-travis has been renamedto haskell-ci, which is now on hackage" is this a known thing ?
<clever> aveltras: not to me
jgt has joined #nixos
kleisli has quit [Ping timeout: 258 seconds]
rsoeldner has quit [Read error: Connection reset by peer]
rsoeldner has joined #nixos
jgt has quit [Ping timeout: 252 seconds]
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
<{^_^}> [nixpkgs] @7c6f434c merged pull request #66767 → btrfs-progs: 5.1.1 -> 5.2.1 → https://git.io/fjFJ2
<{^_^}> [nixpkgs] @7c6f434c pushed 2 commits to master: https://git.io/fjFJ5
<aveltras> clever: can't seem to find naything on this, too bad for me, anyway thanks very much for the help
sudoforge has joined #nixos
phI||Ip has joined #nixos
rsoeldner has quit [Read error: Connection reset by peer]
rsoeldner has joined #nixos
<{^_^}> [nixpkgs] @primeos pushed 3 commits to master: https://git.io/fjFJA
ixxie has quit [Ping timeout: 268 seconds]
clever has quit [Ping timeout: 272 seconds]
phI||Ip has quit [Quit: ZNC 1.7.4 - https://znc.in]
obfusk__ has quit [Ping timeout: 252 seconds]
obfusk__ has joined #nixos
clever has joined #nixos
<Miyu-chan> The DO NixoOps is actually not too bad TBH.
phI||Ip has joined #nixos
<Miyu-chan> Wouldn't recommend it for production, but I'm using it for a Minecraft server.
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66769 → avidemux: 2.7.3 -> 2.7.4 → https://git.io/fjFUL
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66770 → bdf2psf: 1.192 -> 1.193 → https://git.io/fjFUm
<nh2> clever: hmm, it fails for me with `Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS`. Not sure if that means I ran out of RAM on the desktop, or that I have to increase something somewhere
<clever> nh2: the default heap size for the nix evaluator is fairly small
<nh2> how can hydra evaluate it? Does it set some options?
<clever> nh2: you need to increase the number the error says to increase
<__monty__> I'm trying to update the pass-otp extension but it uses this makefile: https://github.com/tadfisher/pass-otp/blob/v1.2.0/Makefile
<samueldr> and it can happen in relatively small amounts of ram, if there is a bunch of fragmentation in the heap
<__monty__> The build fails for me because /etc/bash_completions.d doesn't exist.
<__monty__> How can I patch that path/the makefile so it conforms to nix's expectations?
<clever> __monty__: create a patch (diff -u) and then add it to the patches list
<samueldr> setting BASHCOMPDIR to `${placeholder "out"}/share/bash-completion/completions"` should help
<nh2> clever: thanks, so I'm using now `GC_INITIAL_HEAP_SIZE=11000000000 nix-instantiate ...`, let's see how that goes
<samueldr> (as a makeflag)
<mightybyte> Is it possible to set extra-sandbox-paths inside a nix expression?
<mightybyte> I'm guessing no because that would break the sandbox
<samueldr> clever: in that case there's a makefile variable managing it, luckily :)
<clever> ah, that makes it easier
<{^_^}> Channel nixos-19.03 advanced to https://github.com/NixOS/nixpkgs/commit/9b3cb2d1f09 (from 19 hours ago, history: https://channels.nix.gsc.io/nixos-19.03)
fendor has joined #nixos
<__monty__> Is there a nix-y way to turn these bash completions into fish and zsh completions perchance?
chloekek has joined #nixos
rsoeldner has quit [Read error: Connection reset by peer]
rsoeldner has joined #nixos
phI||Ip has quit [Ping timeout: 248 seconds]
LouisDK has quit [Remote host closed the connection]
zeta_0 has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66771 → abcmidi: 2019.08.02 -> 2019.08.11 → https://git.io/fjFUV
<aveltras> clever: the multi-ghc-travis error is because i was trying to evaluate the whole haskellPackages, my bad :p
Izorkin has quit [Quit: ZNC 1.7.4 - https://znc.in]
<clever> aveltras: ah yeah, you need a -A to select what you want to build
Izorkin has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66772 → angband: 4.1.3 -> 4.2.0 → https://git.io/fjFU1
Izorkin has quit [Client Quit]
asymptotically has quit [Quit: Leaving]
Izorkin has joined #nixos
vika_nezrimaya has joined #nixos
shibboleth has joined #nixos
<__monty__> Ok, so I managed bumping pass-otp, *but* the build still fails because of some tests. I'm on a mac but these tests passed on another mac.
<{^_^}> Channel nixos-19.03-small advanced to https://github.com/NixOS/nixpkgs/commit/3864fd8f53b (from 3 hours ago, history: https://channels.nix.gsc.io/nixos-19.03-small)
orivej has joined #nixos
Havvy has quit [Ping timeout: 246 seconds]
<{^_^}> [nixpkgs] @Izorkin opened pull request #66773 → git: 2.22.0 -> 2.23.0 → https://git.io/fjFU5
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66774 → apacheHttpd: 2.4.39 -> 2.4.41 → https://git.io/fjFUA
cartwright has joined #nixos
Havvy has joined #nixos
<{^_^}> [nixpkgs] @Izorkin opened pull request #66775 → qemu: 4.0.0 -> 4.1.0 → https://git.io/fjFUh
phI||Ip has joined #nixos
Havvy has quit [Remote host closed the connection]
hmpffff has joined #nixos
waleee-cl has joined #nixos
Havvy has joined #nixos
rsoeldner has quit [Remote host closed the connection]
<{^_^}> [nixpkgs] @toonn opened pull request #66776 → pass-otp: 1.1.1 -> 1.2.0 → https://git.io/fjFTq
rsoeldner has joined #nixos
ddellacosta has joined #nixos
<__monty__> Urgh, so pass without extensions builds fine. But when I add the pass-otp extension, one of the pass tests fails. What could be going on?
<__monty__> On a different mac this problem didn't occur at all.
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66777 → bspwm: 0.9.8 -> 0.9.9 → https://git.io/fjFk2
ixxie has joined #nixos
<__monty__> Here's the specific buildfailure: http://ix.io/1Qni
cartwright has quit [Remote host closed the connection]
cantstanya has joined #nixos
<{^_^}> [nixpkgs] @mmahut merged pull request #66765 → ace: 6.5.5 -> 6.5.6 → https://git.io/fjFJO
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFkA
<{^_^}> [nixpkgs] @mmahut merged pull request #66766 → alfred: 2019.2 -> 2019.3 → https://git.io/fjFJW
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFkj
<{^_^}> [nixpkgs] @mmahut merged pull request #66768 → analog: 6.0.15 -> 6.0.16 → https://git.io/fjFJr
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFIJ
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66778 → barrier: 2.3.0 -> 2.3.1 → https://git.io/fjFIU
ddellacosta has quit [Ping timeout: 245 seconds]
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66779 → bird2: 2.0.4 -> 2.0.5 → https://git.io/fjFI3
npmccallum has joined #nixos
<{^_^}> [nixpkgs] @volth opened pull request #66780 → qemu: 4.0.0 -> 4.1.0 → https://git.io/fjFI8
Soo_Slow has joined #nixos
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/b770ded782c (from 88 minutes ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
<__monty__> Is there a way to see the difference between (pass.withExtensions (es: [])) and (pass.withExtensions (es: [ es.pass-otp ]))?
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66781 → avocode: 3.9.0 -> 3.9.2 → https://git.io/fjFIE
<symphorien> __monty__: nix-diff maybe ? it is pretty low-level though
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66782 → airsonic: 10.3.1 -> 10.4.0 → https://git.io/fjFIg
jgt has joined #nixos
<{^_^}> [nixpkgs] @mmahut merged pull request #66591 → nixos/zabbixProxy: fix database initialization logic → https://git.io/fj5M5
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFI2
<{^_^}> [nixpkgs] @volth closed pull request #66780 → qemu: 4.0.0 -> 4.1.0 → https://git.io/fjFI8
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
jgt has quit [Ping timeout: 250 seconds]
ris has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66783 → AgdaStdlib: 1.0.1 -> 1.1 → https://git.io/fjFIP
justanotheruser has quit [Ping timeout: 248 seconds]
Thra11 has quit [Ping timeout: 272 seconds]
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66784 → avfs: 1.1.0 -> 1.1.1 → https://git.io/fjFI1
slack1256 has quit [Read error: Connection reset by peer]
<{^_^}> [nixpkgs] @domenkozar pushed to haskell-updates « haskell: mark tomland as not broken »: https://git.io/fjFID
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66785 → acme-sh: 2.8.1 -> 2.8.2 → https://git.io/fjFIy
<{^_^}> [nixpkgs] @bdesham opened pull request #66786 → Weechat scripts → https://git.io/fjFIQ
Thra11 has joined #nixos
justanotheruser has joined #nixos
<freeman42x> when I try to install smartgit I am getting the following error: "hash mismatch in fixed-output derivation"
<freeman42x> how could I fix this?
__monty__ has quit [Ping timeout: 245 seconds]
<{^_^}> freeman42x: Please expand your question to include more information, this will help us help you :)
<freeman42x> freeman42x, I think I got it. This is fixed in nixpkgs but not in the channel yet
dsiypl4 has quit [Ping timeout: 245 seconds]
<freeman42x> so my question would be, how do I install something from a nixpkgs clone?
<tilpner> Wait, you can't do that, {^_^}!
zeta_0 has quit [Quit: rcirc on GNU Emacs 26.1]
<tilpner> freeman42x: Many Nix tools look up the path to nixpkgs in NIX_PATH
<tilpner> freeman42x: So setting NIX_PATH=nixpkgs=/path/to/your/clone nix-build ... will use your clone
rsoeldner has quit [Read error: No route to host]
<freeman42x> tilpner, ah, ok. I would rather just run the install command in the clone folder though. Do you know what the command to install from the clone is?
rsoeldner has joined #nixos
<tilpner> nix-env -f . -iA foo or something like that
<freeman42x> tilpner, cheers
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66787 → babeld: 1.8.5 -> 1.9.0 → https://git.io/fjFLL
<vika_nezrimaya> Continuing my cross-compiling adventure; spidermonkey seems to break with python2-related errors
<tilpner> ... and by "many Nix tools", I mean "all tools except nix-env"
<tilpner> Because nix-env is weird
<{^_^}> [nixpkgs] @mmahut merged pull request #66771 → abcmidi: 2019.08.02 -> 2019.08.11 → https://git.io/fjFUV
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFLt
<vika_nezrimaya> tilpner: nix-env is a weird thing; it uses Nix to emulate stateful and imperative package management
<vika_nezrimaya> which is... not exactly the thing Nix was designed for
<{^_^}> [nixpkgs] @mmahut merged pull request #66760 → nano 3.2 -> 4.3 → https://git.io/fjFvc
<{^_^}> [nixpkgs] @mmahut pushed commit from @LouisDK1 to master « nano 3.2 -> 4.3 (#66760) »: https://git.io/fjFLq
npmccallum has quit [Quit: npmccallum]
dsiypl4 has joined #nixos
freeman42y has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66788 → binaryen: 88 -> 89 → https://git.io/fjFLO
rsoeldner has quit [Remote host closed the connection]
<{^_^}> [nixpkgs] @mmahut merged pull request #66772 → angband: 4.1.3 -> 4.2.0 → https://git.io/fjFU1
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFL3
rsoeldner has joined #nixos
drakonis has joined #nixos
freeman42x has quit [Ping timeout: 252 seconds]
__monty__ has joined #nixos
FRidh has quit [Quit: Konversation terminated!]
<{^_^}> [nixpkgs] @grahamc merged pull request #66361 → maintainers: add github user IDs in service to NixOS/rfcs#39 → https://git.io/fj7mj
<{^_^}> [nixpkgs] @grahamc pushed 10 commits to master: https://git.io/fjFLW
<vika_nezrimaya> Why is polkit in my closure despite me disabling polkit?...
ottidmes has quit [Ping timeout: 272 seconds]
tomatze has joined #nixos
<tilpner> vika_nezrimaya: Use "nix why-depends" to answer that question
jgt has joined #nixos
<vika_nezrimaya> tilpner: is it normal that it wants to build 199 packages before giving me the answer?
<vika_nezrimaya> it's gonna fail
<tilpner> No, it doesn't need to build anything
slack1256 has joined #nixos
<vika_nezrimaya> weird.
<tilpner> You can pass store paths to it directly
<tomatze> hey, I'm new to nixos and I'm trying to make my first derivation of the following project: https://github.com/nealey/hdjd . The project says, that I should just "make" it and the just run it. My problem is, that mkDerivation fails, because it trys to "make install" it after making it. how can I tell mkDerivation to just copy the binary to the nix-s
<tomatze> tore instead of trying to do "make install" and failing with: "make: *** No rule to make target 'install'. Stop." ?
jgt has quit [Ping timeout: 250 seconds]
slack1256 has quit [Remote host closed the connection]
<symphorien> tomatze: installPhase = '' mkdir -p $out/bin/foo; cp foo $out/bin/foo '';
<symphorien> or the equivalent with `install` instead of cp
dsiypl4 has quit [Ping timeout: 245 seconds]
<tomatze> symphorien thx for the quick answer :) I'll try to do that.
slack1256 has joined #nixos
<vika_nezrimaya> tilpner: nix why-depends $systemClosure.drv $polkit.drv once again tries to build packages O___O
<vika_nezrimaya> I guess I'll just wait
lyvidev has quit [Remote host closed the connection]
<tilpner> vika_nezrimaya: Oh, by "store paths" I meant the built outputs, not drv paths or attribute paths
domogled has quit [Ping timeout: 268 seconds]
<tilpner> E.g. /nix/store/p0amfxmsmniapjy5rdrszkxm98cwx8p7-hello-2.10
kleisli has joined #nixos
<vika_nezrimaya> I have no build outputs because my builds are failing! :3
lyvidev has joined #nixos
<vika_nezrimaya> and I want to see what brings polkit in my system closure
<vika_nezrimaya> because I have security.polkit.enable = lib.mkForce false;
<vika_nezrimaya> and surprise, it still shows up in the build logs somehow!
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
<tilpner> vika_nezrimaya: You can use nix-store --query --tree on the derivation and search where polkit is mentioned
<vika_nezrimaya> that works
<vika_nezrimaya> it's udisks
<{^_^}> [nixpkgs] @mmahut opened pull request #66789 → transifex-client: fixing build → https://git.io/fjFLu
<tilpner> vika_nezrimaya: The cause of our misunderstanding was that there are two important closures. why-depends works on the runtime closure, while --query --tree will show you the build-time closure
slyfox has joined #nixos
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66790 → checksec: 2.0.1 -> 2.1.0 → https://git.io/fjFLg
<tilpner> And you can only determine the runtime closure of something if you have it realised already, so that you can scan it against the buildtime closure
<vika_nezrimaya> Ok. I seem to be in need of a quick git-blame on nixpkgs.freetalk
<tomatze> symphorien thx again, now it works like a charm, I'm still fascinated, how simple it is to build stuff with nixos in a reproducable way
<vika_nezrimaya> Because I don't know who put autoconf and automake in buildInputs instead of nativeBuildInputs :D
dsiypl4 has joined #nixos
<tilpner> Don't blame, just improve :)
<vika_nezrimaya> does texinfo need to be native or not?
<tilpner> I think that depends on how texinfo is used there
<vika_nezrimaya> ok I'm about to find out :3
<vika_nezrimaya> aaaand syntax error
<tilpner> Huh
Shouou has quit [Remote host closed the connection]
<tilpner> If it uses /bin during buildPhase/installPhase, I'd say nativeBuildInputs
Shouou has joined #nixos
wfranzini has quit [Remote host closed the connection]
<tilpner> If it uses /lib during runtime, buildInputs
<tilpner> And no, I don't know what to do if it does both
<vika_nezrimaya> well, it's an XMPP client
wfranzini has joined #nixos
<vika_nezrimaya> I doubt it would link to texinfo
<vika_nezrimaya> it's not Emacs
<tilpner> While you're fixing that package, please fix the indentation too
<tilpner> :)
<vika_nezrimaya> I'm overriding it with an overlay, not rewriting it
<tilpner> Aww. Maybe nixpkgs-format will do it eventually
<vika_nezrimaya> I'll upstream my stuff once I have this system built :3
<vika_nezrimaya> It contains a lot of nasty hacks I will not upstream, but rather file issues about them
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66791 → cni-plugins: 0.8.1 -> 0.8.2 → https://git.io/fjFL9
<vika_nezrimaya> Like the one with Lua
rsoeldner has quit [Ping timeout: 252 seconds]
tomatze has quit [Remote host closed the connection]
<vika_nezrimaya> It's a hard blocker for me, but I'm trying to power through as much as I can without it. Next thing I'm gonna complain about is probably Spidermonkey, which doesn't cross-compile too - but I an live without it...
<vika_nezrimaya> oh, also clucene-core fails to compile
<vika_nezrimaya> apr (Apache portable runtime) too
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66792 → ckbcomp: 1.192 -> 1.193 → https://git.io/fjFL5
c74d has joined #nixos
<c74d> What's the proper way to convert a Nix `int` value into a `float`? I'm using `builtins.fromJSON "${toString someInt}.0"`, but that feels... improvable.
rsoeldner has joined #nixos
slack1256 has quit [Read error: Connection reset by peer]
<vika_nezrimaya> Wow. Git doesn't want to link to libcurl
o1lo01ol1o has joined #nixos
<{^_^}> [nixpkgs] @worldofpeace merged pull request #63960 → gsctl: init at 0.15.4 → https://git.io/fjKJ1
<{^_^}> [nixpkgs] @worldofpeace pushed 3 commits to master: https://git.io/fjFLp
mildred has joined #nixos
<{^_^}> [nixpkgs] @mmahut opened pull request #66793 → nbstripout: moving buildInputs to checkInputs → https://git.io/fjFtJ
<vika_nezrimaya> What's the benchmark feature in the builders configuration for distributed builds? The wiki said it's a marker that builds on that machine take constant time, but how to utilize it? how does it get utilized?
<{^_^}> [nixpkgs] @mmahut opened pull request #66794 → mumble_overlay: fixing the library path → https://git.io/fjFtZ
<{^_^}> [nixpkgs] @worldofpeace opened pull request #66795 → nixos/pantheon: adjust to renamed gnome3 options → https://git.io/fjFtn
mexisme has joined #nixos
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
jgt has joined #nixos
<{^_^}> [nixpkgs] @worldofpeace merged pull request #66795 → nixos/pantheon: adjust to renamed gnome3 options → https://git.io/fjFtn
<{^_^}> [nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fjFtB
<{^_^}> [nixpkgs] @oxij opened pull request #66796 → anki: use wrapQtAppsHook to fix execution → https://git.io/fjFtu
<{^_^}> [nixpkgs] @mmahut merged pull request #66778 → barrier: 2.3.0 -> 2.3.1 → https://git.io/fjFIU
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFtz
jgt has quit [Ping timeout: 264 seconds]
<{^_^}> [nixpkgs] @mmahut merged pull request #66777 → bspwm: 0.9.8 -> 0.9.9 → https://git.io/fjFk2
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFt2
inkbottle has quit [Ping timeout: 245 seconds]
<{^_^}> [nixpkgs] @mmahut merged pull request #66782 → airsonic: 10.3.1 -> 10.4.0 → https://git.io/fjFIg
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFta
<gchristensen> ,-iA
<gchristensen> ,-A
<{^_^}> You'll usually want to use nix-env -i with -A. It's faster and more precise. See https://nixos.wiki/wiki/FAQ/nix-env_-iA for details.
<{^_^}> [nixpkgs] @mmahut merged pull request #66787 → babeld: 1.8.5 -> 1.9.0 → https://git.io/fjFLL
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFtr
<{^_^}> [nixpkgs] @mmahut merged pull request #66790 → checksec: 2.0.1 -> 2.1.0 → https://git.io/fjFLg
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFto
<{^_^}> [nixpkgs] @etu merged pull request #66789 → [19.03] transifex-client: fixing build → https://git.io/fjFLu
<{^_^}> [nixpkgs] @etu pushed 3 commits to release-19.03: https://git.io/fjFt6
<{^_^}> [nixpkgs] @mmahut merged pull request #66791 → cni-plugins: 0.8.1 -> 0.8.2 → https://git.io/fjFL9
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFti
selfsymmetric-mu has joined #nixos
gentauro has joined #nixos
Thra11 has quit [Ping timeout: 244 seconds]
iqubic has joined #nixos
<{^_^}> [nixpkgs] @mmahut merged pull request #66788 → binaryen: 88 -> 89 → https://git.io/fjFLO
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFty
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
<{^_^}> Channel nixos-19.03 advanced to https://github.com/NixOS/nixpkgs/commit/3864fd8f53b (from 6 hours ago, history: https://channels.nix.gsc.io/nixos-19.03)
<{^_^}> [nixpkgs] @mmahut merged pull request #66784 → avfs: 1.1.0 -> 1.1.1 → https://git.io/fjFI1
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFt9
<{^_^}> [nixpkgs] @mmahut merged pull request #66726 → httplz: 1.5.0 -> 1.5.1 → https://git.io/fjdPd
<{^_^}> [nixpkgs] @mmahut pushed 3 commits to master: https://git.io/fjFtQ
rsoeldner has quit [Read error: Connection reset by peer]
noudle has quit [Ping timeout: 268 seconds]
rsoeldner has joined #nixos
kleisli has quit [Quit: Leaving]
<{^_^}> [nixpkgs] @mmahut merged pull request #66792 → ckbcomp: 1.192 -> 1.193 → https://git.io/fjFL5
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFtN
___laika has joined #nixos
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
dasj19 has quit [Quit: dasj19]
<{^_^}> [nixpkgs] @Ma27 merged pull request #66769 → avidemux: 2.7.3 -> 2.7.4 → https://git.io/fjFUL
<{^_^}> [nixpkgs] @Ma27 pushed 2 commits to master: https://git.io/fjFth
jgt has joined #nixos
liberiga has joined #nixos
rsoeldner has quit [Remote host closed the connection]
mexisme has quit [Ping timeout: 252 seconds]
rsoeldner has joined #nixos
jgt has quit [Ping timeout: 250 seconds]
<freeman42y> getting this error on nixos-rebuild switch: https://gist.github.com/razvan-panda/fd8ed33922dd6219bcf0f2a038b6e1b5
<freeman42y> hmmm, it worked fine after a reboot
<{^_^}> Channel nixpkgs-19.03-darwin advanced to https://github.com/NixOS/nixpkgs/commit/67135fbcc5d (from 50 minutes ago, history: https://channels.nix.gsc.io/nixpkgs-19.03-darwin)
<infinisil> c74d: Oh wow that's a hacky workaround for int -> float
<infinisil> I wish I could tell you a better way..
<c74d> infinisil: I've now realised that I could use `int + 0.0` or `int * 1.0`
<{^_^}> [nixpkgs] @mmahut merged pull request #66764 → discord-canary: 0.0.92 -> 0.0.93 → https://git.io/fjFf1
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFqW
<infinisil> Oh that's much better :o
<infinisil> Nice
<joepie91> freeman42y: have you seen any other strange behaviour on your system, or just this? anything at all
<freeman42y> joepie91, is this strange? https://i.imgur.com/sZEJgO5.png
<c74d> infinisil: Although I'm not sure whether adding 0.0/multiplying by 1.0 might cause the value to change via rounding errors or something
<joepie91> freeman42y: no, not necessarily. I *think* that's a path in the nix store
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66797 → clickhouse: 19.13.1.11 -> 19.13.2.19 → https://git.io/fjFq8
<joepie91> freeman42y: the reason I ask is that I've only seen your problem twice, on my own system, and that was when I had memory corruption issues (hardware conflict and bad RAM stick respectively)
<joepie91> which had corrupted the Nix store
<joepie91> but that usually goes accompanied by other strange behaviour, like rendering glitches, applications suddenly not starting or looking weird, etc.
<{^_^}> [nixpkgs] @mmahut merged pull request #66752 → plex: 1.16.3.1402 -> 1.16.4.1469 → https://git.io/fjdN2
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFqB
<{^_^}> [nixpkgs] @worldofpeace merged pull request #66793 → nbstripout: moving buildInputs to checkInputs → https://git.io/fjFtJ
<{^_^}> [nixpkgs] @worldofpeace pushed 2 commits to master: https://git.io/fjFqR
<joepie91> freeman42y: if you want to exclude a bad RAM stick, it's worth running a memtest at some point
<joepie91> though if you've seen no other oddities, that's not very likely to be the case
<freeman42y> joepie91, "I *think* that's a path in the nix store" might be. I think I installed hie via nix-env. how can I remove it?
<infinisil> c74d: The problem with floats is only that they're not associative, adding 0 or multiplying by 1 will still be identities, so this should be fine
<joepie91> freeman42y: there's a remove flag for nix-env
<joepie91> I forgot what it is, I never use nix-env
<joepie91> should be in the manpage
<joepie91> `man nix-env`
mexisme has joined #nixos
<joepie91> freeman42y: out of interesting, are you running on a system with an AMD CPU/GPU by any chance?
<joepie91> interest*
<freeman42y> joepie91, no, It's an intel i7 with an nvidia deicated gpu
<joepie91> ah, seems `--uninstall` is the flag
<joepie91> freeman42y: ah, then at least it won't be the hardware conflict I had :D
fendor_ has joined #nixos
<joepie91> but yeah, I have no idea what the problem /is/ then
<vika_nezrimaya> how can I add a certain derivation to environment of a builder? I need curl-config to appear, but it doesn't. The derivation is... git
<freeman42y> joepie91, well, the issue is gone after reboot, so it's fine
<freeman42y> joepie91, this uninstall feedback though... https://i.imgur.com/qJRAj5Q.png
<{^_^}> [nixpkgs] @mmahut merged pull request #66743 → http-prompt: Allow installation on Mac OS → https://git.io/fjdQD
<{^_^}> [nixpkgs] @mmahut pushed 2 commits to master: https://git.io/fjFqE
<joepie91> freeman42y: alright. make sure to watch for any other odd behaviour. if you see other odd behaviour (especially behaviour that cannot be easily explained by a simple developer mistake), you should immediately shut down the system and run a memtest -- continuing to run with a bad RAM stick can end up corrupting data on disk, so it's not a good idea to do that if you have a suspicion of RAM issues
<joepie91> heh
<joepie91> yeah, that's not very informative...
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66798 → calibre: 3.45.2 -> 3.46.0 → https://git.io/fjFqu
ixxie has quit [Remote host closed the connection]
<joepie91> (also, in the unlikely case that it really /is/ a RAM issue, you can repair your nix store using the repair flag in the nix-store command, but you should only do that AFTER removing the bad RAM, otherwise you run the risk of creating a ton of corruption in your store)
<{^_^}> [nixpkgs] @Ma27 opened pull request #66799 → curlie: init at 1.2.0 → https://git.io/fjFqg
mexisme has quit [Ping timeout: 264 seconds]
notgne2 has quit [Quit: ZNC 1.7.3 - https://znc.in]
notgne2 has joined #nixos
<c74d> infinisil: thanks
<freeman42y> neither nix-env --uninstall nor nix-env -e worked . How do I actually uninstall a package?
<drakonis> is it installed with nix-env or part of your config?
fendor has quit [Quit: Leaving]
fendor_ has quit [Ping timeout: 248 seconds]
<infinisil> ,fancyUninstall freeman42y
<{^_^}> infinisil: Did you mean fancy-uninstall?
<{^_^}> freeman42y: Fancy way to uninstall packages, needs fzf installed: nix-env -q | fzf | xargs -I{} nix-env -e {}
<infinisil> nix-env -e is supposed to work
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66800 → cutemaze: 1.2.4 -> 1.2.5 → https://git.io/fjFq1
<infinisil> I often use above command to not mistype the package name
<freeman42y> the issue I was having is that I was trying to do nix-env -e as root, it worked when I called it as an user
<pie_> infinisil: 0_0 owo whats this
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66801 → ocamlPackages.cppo: 1.6.5 -> 1.6.6 → https://git.io/fjFqD
<pie_> nice'
<pie_> infinisil: is the bot database accessible somewhere? :P
<pie_> neato
<rycee> infinisil: You're still using Home Manager as a NixOS module?
<infinisil> rycee: Yup
<rycee> If you have NUR set up, could you try changing the import to `nur-no-pkgs.repos.rycee.modules.home-manager`?
<rycee> I've been trying to set up my nur-expressions repo to automatically update when HM builds happen and to define a NixOS module there.
domogled has joined #nixos
ottidmes has joined #nixos
<{^_^}> Channel nixos-unstable-small advanced to https://github.com/NixOS/nixpkgs/commit/f217c3b6fe2 (from 52 minutes ago, history: https://channels.nix.gsc.io/nixos-unstable-small)
domogled has quit [Ping timeout: 268 seconds]
<infinisil> rycee: I see. I don't have NUR set up currently, but I could try that out
<rycee> No worries. I've been trying it out and it seems to work pretty well. Just wanted some extra confidence.
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66802 → conky: 1.11.3 -> 1.11.5 → https://git.io/fjFqA
<rycee> My hope is to have this be the "official" way to use the HM module.
<rycee> Then people won't have to set up a channel or anything just for HM, only for NUR.
Mateon2 has joined #nixos
<rycee> Also, with the setup I have now the NUR expression will only update if the tests pass.
<infinisil> Neat
<rycee> So if some bad commit were to end up in master then there is another line of defence :-)
Mateon1 has quit [Ping timeout: 248 seconds]
Mateon2 is now known as Mateon1
trfl has quit [Quit: ZNC - http://znc.in]
<infinisil> rycee: Yup, can confirm it works through NUR :)
<rycee> Sweet :-)
<rycee> I'm pretty pleased with it, especially with the automatic version picking.
<infinisil> Automatic version picking?
* infinisil takes a closer look
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66803 → cimg: 2.6.7 -> 2.7.0 → https://git.io/fjFqh
<infinisil> Oh neat
<infinisil> Yeah that's pretty cool
jgt has joined #nixos
domogled has joined #nixos
<{^_^}> [nix] @goodwillcoding closed pull request #292 → Add named id to all chapters, sections, and appendices so they are easier to link to. → https://git.io/Y4uECw
<rycee> It was a bit of a struggle to get all the puzzle pieces to line up. And gitlab was having problems at the same time so I was tearing my hair out when nothing worked and I couldn't figure out what I did wrong.
<rycee> And it was gitlabs fault all along :-)
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66804 → clipgrab: 3.8.3 -> 3.8.4 → https://git.io/fjFmv
___laika has quit [Ping timeout: 245 seconds]
trfl has joined #nixos
<jasom> How do I get scsi generic devices (/dev/sg*) to exist on nixos?
jgt has quit [Ping timeout: 264 seconds]
o1lo01ol1o has quit [Remote host closed the connection]
<jasom> Ah, well "modprobe sg" gets it to load manually; I'll look into how to get modules to load on boot.
dsiypl4 has quit [Ping timeout: 268 seconds]
<infinisil> jasom: Probably with boot.kernelModules = [ "sg" ]
<jasom> which looks like boot.kernelModules
<jasom> thanks infinisil!
<worldofpeace> ,locate libattr.so.1
<{^_^}> Found in packages: attr
<infinisil> Hehe
___laika has joined #nixos
liberiga has quit [Ping timeout: 260 seconds]
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
o1lo01ol1o has joined #nixos
__monty__ has quit [Quit: leaving]
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66805 → ccid: 1.4.30 -> 1.4.31 → https://git.io/fjFmL
___laika has quit [Ping timeout: 245 seconds]
dsiypl4 has joined #nixos
o1lo01ol1o has quit [Ping timeout: 245 seconds]
dsiypl4 has quit [Remote host closed the connection]
domogled has quit [Ping timeout: 248 seconds]
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
realrokka has quit [Ping timeout: 272 seconds]
___laika has joined #nixos
rsoeldner has quit [Ping timeout: 252 seconds]
MmeQuignon has quit [Read error: Connection reset by peer]
<infinisil> There's couple similarities to NixOS here: https://lobste.rs/s/yiumwr/distri_linux_distribution_research_fast
MmeQuignon has joined #nixos
rsoeldner has joined #nixos
<infinisil> I sure wish NixOS was a bit faster.. rebuild times are too high
___laika has quit [Ping timeout: 246 seconds]
<infinisil> Huh indeed
___laika has joined #nixos
<infinisil> adisbladis: Maybe he meant distri being written in Go?
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66806 → chafa: 1.0.1 -> 1.2.1 → https://git.io/fjFm4
<adisbladis> Maybe?
<adisbladis> I have to admin I didn't look at the code
<infinisil> adisbladis: Huh, I just checked, I can't find the source code
<adisbladis> infinisil: https://github.com/distr1/distri
<infinisil> Ah
<adisbladis> (it was _not_ easy to find in that blog post)
<infinisil> Ah yeah I see it linked in the blog now
hmpffff has quit [Quit: nchrrrr…]
<infinisil> Yeah I tried hard to find it too
<infinisil> It's not even on the website
<{^_^}> [nixpkgs] @r-ryantm opened pull request #66807 → cgdb: 0.7.0 -> 0.7.1 → https://git.io/fjFmR
<infinisil> Though it looks like the website is just a github readme mirror
Mateon1 has quit [Ping timeout: 246 seconds]
<infinisil> I've been thinking this for a while:
<infinisil> A way to speed up Nix is to have the daemon keep evaluations in memory
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
<infinisil> E.g. you could say it should keep the evaluation for `nixos-rebuild` in memory, and anytime any file that influences it changes, it gets reevaluated
<infinisil> And only those parts get reevaluated that need to be
<infinisil> Then a `nixos-rebuild` command can just fetch the result from memory instantly
<infinisil> The only problem is that this will use up a lot of memory all the time..
<infinisil> I guess this could be solved by writing the result to disk when a while has passed, to get it out of memory, reload it on demand
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
<infinisil> With this, the whole evaluation step could be hella fast
<infinisil> Then we'd have to look into what making building so slow..
___laika has quit [Ping timeout: 272 seconds]
<pie_> obviously tons more caching could be done than what is done
<infinisil> pie_: Caching what?
<pie_> intermediate eval results
<pie_> except maybe not so "obviously"
orivej has quit [Ping timeout: 245 seconds]
jgt has joined #nixos
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
Mateon1 has joined #nixos
sigmundv_ has joined #nixos
jgt has quit [Ping timeout: 252 seconds]
<infinisil> Okay so after a quick benchmark, for a trivial derivation (the simplest you can have), it takes Nix about 2 seconds to build it
<infinisil> But most of that time is spent querying substituters
<infinisil> With --option substitute false it takes 180ms!
rsoeldner has quit [Remote host closed the connection]
rsoeldner has joined #nixos
Soo_Slow has quit [Remote host closed the connection]
<andi-> and with --option sandbox false?
tertl3 has joined #nixos
<tertl3> nixos ftw
ng0 has quit [Quit: Alexa, when is the end of world?]
shibboleth has quit [Quit: shibboleth]
rsoeldner has quit [Ping timeout: 252 seconds]
<infinisil> andi-: Another 50ms shaved off, about
<infinisil> And for the sandbox thing you can control by setting allowSubstitutes = false
<emily> hm, nix should race substitutions and builds :p
<infinisil> emily: Yeah am thinking the same
<emily> would make for annoying CPU spikes I guess
<infinisil> emily: Though it's a bit problematic, because if it can substitute it doesn't need to download dependencies for the build
<emily> why can't you just cache all the hashes the caches have?
<emily> (tongue-twister)
<emily> like apt and so on do
<emily> I guess you need an explicit update step then is the problem
<infinisil> Not sure what you mean
<emily> as in, rather than querying over the internet, just look up whether the substituters have the hash available locally.
<emily> ofc this would require architectural changes.
<infinisil> Ohh I see, neat idea
<infinisil> That might work really well
<infinisil> To just download a differential update of the hash list once when the build starts, then base all the lookups on that
<adisbladis> The entire cache manifest in a bloom filter would be <3
<adisbladis> Ze speeds
<infinisil> Hehe that's also a nice idea
<infinisil> Bloom filter has false positives or negatives?
<adisbladis> False positives
<adisbladis> infinisil: Which would be perfect for the cache
<infinisil> It would sometimes end up thinking it's in the cache when it's really not?
<adisbladis> Yeah
<emily> oh a bloom filter would be great :)
<emily> I'm guessing "every hash hydra has" would be pretty big to download initially so that might actually be a good idea
chloekek has quit [Ping timeout: 245 seconds]
<infinisil> Ping gchristensen for maybe getting a count on the number of paths in hydra's store
evanjs has joined #nixos
<adisbladis> infinisil: You'd have to check the s3 bucket directly
<evanjs> Okay, going to finally just ask here. Any idea how I might idiomatically get the newest version of a package (and by extension, a list of packages) from a given set of channels (e.g. config.nixpkgs.config.packageOverrides)?
<evanjs> Currently trying to do so with JetBrains packages, and while it works, it's sorta messy and I can't figure out how to use it with a package list (I'm calling the function on each package individually as shown here - https://github.com/evanjs/nixos_cfg/blob/master/modules/jetbrains.nix#L11)
<evanjs> Before I forget, here's the function that's referenced in the my JetBrains "module" https://github.com/evanjs/nixos_cfg/blob/ad002a4ec63b775c9785d83fea54292ee67e2476/overlays/libs/versions.nix#L4
<davidtwco> I've just accidentally removed "users.users.<name>.shell" for my user and rebuilt, now when I ssh to my remote machine I get "/run/current-system/sw/bin/zsh: No such file or directory" - any ideas on how I can get back in?
<{^_^}> Channel nixos-19.03-small advanced to https://github.com/NixOS/nixpkgs/commit/67135fbcc5d (from 3 hours ago, history: https://channels.nix.gsc.io/nixos-19.03-small)
<markasoftware> If you have root SSH access, you can specify a shell when switching users with su
<markasoftware> If you don't, you're sorta screwed. You probably still have sftp access, though
<infinisil> evanjs: So you want to get it from e.g. pkgs if that happens to be ahead of unstable-small?
<davidtwco> markasoftware: root ssh is disabled unfortunately.
<evanjs> infinisil: right so like, given the channels [ unstable-small unstable stable ], pick whatever is newest
<infinisil> evanjs: I mean, that solution looks fine to me, there's no more idiomatic way to do this
<selfsymmetric-mu> infinisil: Your daemon idea...would that be like lorri for NixOS, or does that do something different?
<infinisil> selfsymmetric-mu: Yeah that would be very similar, it would be an improved version of lorri
<evanjs> infinisil: Right, so here's the big problem. Trying to mimic your config and I hit a wall with how to apply it to the packages options for the new JetBrains module I made - https://github.com/evanjs/nixos_cfg/blob/277cf2fa0141ce9aad416fec1b59ba9add1d30ca/config/new-modules/jetbrains.nix#L23 - WIP - https://pastebin.com/K9MbXsM4
<infinisil> evanjs: Ah I see, how about changing the option to `listOf str`
mbrgm has quit [Ping timeout: 250 seconds]
mbrgm_ has joined #nixos
<infinisil> So you specify the package name instead of the package itself when assigning the option, which allows you to freely choose the version in that module
mbrgm_ is now known as mbrgm