<infinisil>
Yeah but that's impure (because network), so we need to pin it somehow
<infinisil>
Hmm, is it still impure when it's fixed-output?
<infinisil>
Not sure!
<infinisil>
Oh well
<noonien>
I'll probably give up, don't want to insist on it too much. But i currently have `fetchmultiurls`, which fetches downloads a list of files. In `fetchgithubfiles` i first get the list of files, then i download each one. It would be a shame to reimplement the fetching of urls
<noonien>
hmm, is lib.assertMsg not in nixos-18.09?
<noonien>
it seems not
thc202 has quit [Ping timeout: 252 seconds]
<{^_^}>
[nixpkgs] @Infinisil merged pull request #56625 → set phpOptions per phpfpm pool, instead of applying to every phpfpm pool → https://git.io/fhAVK
<{^_^}>
#56998 (by ivan, 1 day ago, closed): chromium-72.0.3626.121 is actually a vulnerable 72.0.3626.119
<infinisil>
Fare: Ignoring that problem, this would work to check the version of a package on a channel: nix eval nixpkgs.chromium.name -I nixpkgs=channel:nixos-19.03-small
Havvy has quit [Read error: Connection reset by peer]
Havvy has joined #nixos
fractal has joined #nixos
oldandwise has joined #nixos
<oldandwise>
guys, how do I make hostapd and dnsmasq work together? ie, to let dnsmasq gave out and ip address
<buckley310>
set up hostapd without dnsmasq first, test it with static IPs, and then when thats working, add dnsmasq
<buckley310>
after hostapd is set up, adding dnsmasq to the wireless interface will be the same as any other interface
fiddlerwoaroof has joined #nixos
<{^_^}>
[nixpkgs] @teto opened pull request #57060 → buildLuarocksPackage: adjust the manifest path → https://git.io/fhphQ
<fiddlerwoaroof>
I'm on mac, and I can't install darcs because it seems to be incompatible with ghc-8.6 (nix-env -i darcs fails with a complaint about not having the right version of base). Is there any way to downgrade ghc or force nix to use a lower version of ghc in a certain context?
oldandwise has quit [Quit: leaving]
Soo_Slow has joined #nixos
<jackdk>
fiddlerwoaroof something like `nix-env -iA haskell.packages.ghc844.darcs`?
<symphorien>
this only works when nix fills the disk, not another program
<{^_^}>
[nixpkgs] @danbst pushed commit from @oxij to master « pkgs/top-level/stage.nix: don't override `overlays` and `config` in `nixpkgsFun` »: https://git.io/fhhJk
lopsided98 has quit [Remote host closed the connection]
lopsided98 has joined #nixos
jasongrossman has quit [Ping timeout: 252 seconds]
reinhardt has joined #nixos
kvda has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<teto>
how to track the "instantiated" events logged in nix-build -vvvvv ? I have a bug in my code where I declare one lua environment but 2 are created (and one is wrong)
Mateon3 has joined #nixos
Mateon2 has quit [Ping timeout: 268 seconds]
Mateon3 is now known as Mateon1
init_6 has quit []
rprije has quit [Ping timeout: 245 seconds]
init_6 has joined #nixos
Makaveli7 has quit [Ping timeout: 245 seconds]
Makaveli7 has joined #nixos
smarti has quit [Remote host closed the connection]
<linarcx>
Hi. friends. i want to use i3. in my .config file there is a part that use i3bar(status_command). but i get an error:`Error:status_command not found or is missing a library dependency(exit 127)`
<{^_^}>
[nixpkgs] @Mic92 merged pull request #57023 → crystal: fix build on Darwin → https://git.io/fhpXj
beaky has quit [Read error: Connection reset by peer]
teehemkay has quit [Ping timeout: 257 seconds]
teehemkay has joined #nixos
beaky has joined #nixos
waleee has joined #nixos
duckfullstop has joined #nixos
simukis has joined #nixos
<duckfullstop>
howdy, dumb emergency question because everything will catch fire if I don't figure this out; how do you define a networking.bridge.(name) with no constituent interfaces (ie a blank bridge)?
<adamt>
duckfullstop: networking.bridges.foo.interfaces = []; seems to be valid? At least it doesn't fail on build time
<{^_^}>
[nixpkgs] @Mic92 pushed commit from @dtzWill to release-19.03 « pythonPackages.m2crypto: swig2 as nativeBuildInput so it can be invoked »: https://git.io/fhhkE
ThatDocsLady_ has joined #nixos
orivej has quit [Ping timeout: 240 seconds]
reinhardt has quit [Ping timeout: 240 seconds]
duckfullstop has quit [Ping timeout: 268 seconds]
cfricke has joined #nixos
duckfullstop has joined #nixos
<Izorkin>
etu: and please run build in darwin php73-embed
tycho01 has joined #nixos
<tycho01>
hi. I accidentally overwrote my hardware-configuration.nix, and regenerated versions are missing the /boot section. how can I find which device was used for /boot in my present revision?
<jasongrossman>
tycho01: You could try running `lsblk`.
vk3wtf has quit [Quit: WeeChat 2.3]
Makaveli7 has quit [Ping timeout: 240 seconds]
reinhardt has joined #nixos
asymmetric has quit [Ping timeout: 255 seconds]
<arianvp>
is there a 19.03 channel already?
<gchristensen>
arianvp: 19.03 is in beta and has a channel, yeah
<arianvp>
I'm trying to boot a radeon vii
<arianvp>
alright, ill try switching to it, see if it causes a boot :)
<tycho01>
@jasongrossman that was helpful, thanks!
m1sosoup has joined #nixos
<etu>
arianvp: you might want to use a newer kernel
<etu>
arianvp: And you might want to activate kernel modules as well :/
<symphorien>
Denommus: nixpkgs has only nix 2.2 but the installer (which has nothing to do with nixpkgs) was from the latest version of nix
<symphorien>
nix upgrade-nix just uses the latest nix from nixpgks
<symphorien>
hence the downgrade
<clever>
> builtins.compareVersions "1.2" "2.2"
<{^_^}>
-1
<clever>
> builtins.compareVersions "2.2" "2.2.1"
<{^_^}>
-1
<clever>
it agrees that 2.2 < 2.2.1
<Denommus>
symphorien: so... how do I upgrade back to 2.2.1?
<clever>
Denommus: but, `nix-env -u` doesnt really "upgrade" it just guesses what attr a given package comes from, and installs whatever nixpkgs says it is on the new version
aswanson has quit [Quit: WeeChat 2.2]
asymmetric_ has joined #nixos
<symphorien>
the easiest is to wait for nixpkgs to have 2.2.1
<Denommus>
and what about nixopsUnstable? Why does it keep being replaced by nixops?
aswanson has joined #nixos
<symphorien>
> nixops.name
<{^_^}>
"nixops-1.6.1"
<symphorien>
> nixopsUnstable.name
<{^_^}>
"nixops-1.6.1pre2728_8ed39f9"
<clever>
Denommus: nix-env -u doesnt know what attr a package came from
asymmetric has quit [Remote host closed the connection]
<clever>
Denommus: so it just guesses nixpkgs.nixops, incorrectly
asymmetric_ is now known as asymmetric
<symphorien>
nixops unstable has a lower version number
<clever>
i dont think -u cares about version# at all
<symphorien>
it does
<clever>
it just finds a "random" derivation with a matching name?
<symphorien>
that's why --leq exists
<Denommus>
can I remove nixops with overlays?
<symphorien>
,declarative
<{^_^}>
There are multiple ways of managing declarative profiles. 1) Attrset, compatible with imperative use of nix-env https://git.io/fAQHW ; 2) buildEnv, providing more control over the paths that are linked into the profile https://git.io/fp0aU ; 3) home-manager, providing nixos-like config for your ~ https://github.com/rycee/home-manager
<clever>
Denommus: would be simpler to just set nixops = self.nixopsUnstable; with an overlay
<clever>
but fully declarative profiles are better
<Denommus>
symphorien: oooh, I've been using --leq
<Denommus>
that might be it
<symphorien>
Denommus: if nix-env frustrates you, use declarative methods as above
<symphorien>
nix is a great build system, but not so great a package manager
<Denommus>
I beg to differ, I really enjoy it as a package manager. But you need to really understand the nix language in order to make good use of it
<Denommus>
for instance, nixopsUnstable is currently broken, but I managed to fix it using only overlays
<symphorien>
once you have this understanding, the buildEnv solution mentioned in the factoid is easy to do and way better than raw nix-env
<betawaffle>
this is not really the right place to ask this, but there seem to be a fair number of zfs users here... if i don't enable dedup, can zfs still dedup all-zero blocks?
<srhb>
betawaffle: Are you talking about sparse zvols?
<betawaffle>
i guess?
Rusty1 has joined #nixos
<srhb>
betawaffle: I'd rather not guess. :P
<infinisil>
betawaffle: There is #zfsonlinux
<betawaffle>
k
<gchristensen>
betawaffle: it'll compress them if they're compressable
<betawaffle>
oh, of course!
<betawaffle>
that's the answer i wanted
<infinisil>
Well only if you have compression enabled
<emily>
compression, the most local kind of deduplication~
<simpson>
Something something, "functions are compressed data"
johanot has quit [Quit: WeeChat 2.2]
work_ has joined #nixos
orivej has quit [Ping timeout: 252 seconds]
Lears has joined #nixos
<Denommus>
symphorien: the declarative stuff seems interesting, but is there a suggestion in which one I should choose?
[Leary] has quit [Ping timeout: 252 seconds]
<adamt>
Is there some kernel parameter that makes the init 1 stage more verbose? I have a bunch of new HP servers that spend ages in the "running udev" step, and looking at the code it seems like it's doing several things that might all be at fault
<adamt>
Bleh, I mean "stage 1" of the boot, not "init 1 stage" -- that's just gibberish
<adamt>
I haven't timed it, but it's several minutes, so I imagine they're reaching some sort of timeout before moving on
<{^_^}>
[nixpkgs] @callahad opened pull request #57077 → dsseries: init at 1.0.5-1 → https://git.io/fhhqN
<glasserc>
If I do nixos-rebuild switch -I /path/to/nixpkgs, should nix-info report the version of /path/to/nixpkgs ?
<srhb>
adamt: No location yet?
<glasserc>
Or more generally, how do I switch over to a local nixpkgs
<srhb>
glasserc: Try: -I nixpkgs=/path/to/nixpkgs
<adamt>
srhb: Nah, ktra just picked a date today, but he sounds very confident he can find a location. But if you have suggestions closer to Copenhagen than us.... ;-)
<srhb>
glasserc: Alternatively you can manipulate NIX_PATH directly.
<adamt>
It seems like it's udevadm settle that's slow. But that's probably going to be some exciting poking I can safe for monday.
<srhb>
adamt: :-P ok.
<srhb>
adamt: Well, thanks for the heads-up.
<glasserc>
srhb: Ah yes, I'm using -I nixpkgs=/path/to/nixpkgs, but nix-info still shows a different version than I expect
<srhb>
glasserc: After rebuilding?
<srhb>
glasserc: And switching, specifically.
<srhb>
glasserc: Ah, actually, nix-info does runtime things
<srhb>
glasserc: nixos-version is build-time.
<glasserc>
Hmm. I see that /nix/var/nix/profiles/per-user/root/channels/nixos is still the old version
<srhb>
glasserc: I'm not sure what you're saying exactly. -I shouldn't affect that anyway?
<glasserc>
Let me back up a bit.. I'm having trouble building a package in 19.03 that used to work in 18.09. So I'm trying to bisect what changed that made it start to fail
<glasserc>
So I was hoping to "switch back" to 18.09 by checking that out in a local nixpkgs, and then retry the build
<srhb>
glasserc: I'd suggest not doing rebuild for that, if you can avoid it.
<glasserc>
srhb: Yes, that's the immediate thing -- I think I know what that's about and I have a fix for it, but when I actually built the thing, it wouldn't run -- segv
<srhb>
glasserc: With xorgproto?
<glasserc>
Yeah
<srhb>
ok :)
<glasserc>
It's just annoying to track this down because rebuilding Gecko takes ~70 minutes on my computer
<srhb>
glasserc: Ouch. :/
<srhb>
glasserc: Maybe you can fiddle a bit with optimization flags. Might help a tiny bit.
<glasserc>
But actually in talking to you, I had an idea about what the segv is about -- I wonder if it's because I built against 19.03 although I haven't rebooted into 19.03 since switching a few days ago
<glasserc>
It might be a mismatch between Wayland versions or something
<srhb>
glasserc: I want to say no, that should never happen, but impurities like that have been known to happen especially in some gui toolkits.
asymmetric has quit [Ping timeout: 244 seconds]
<srhb>
glasserc: By the way, just making sure you know, 19.03 isn't released yet.
<srhb>
(Somehow we seem to be giving people the impression that releases are out as soon as the feature freeze hits)
<glasserc>
Yeah, I know, thanks. I guess I like to jump on unstable branches a little too early for my own good
<srhb>
Ok, just checking :)
<srhb>
This is how we get things fixed before release anyway, so kudos to you :-P
hlolli has joined #nixos
<glasserc>
Actually the reason I tried to switch to unstable was because I thought I wanted to contribute an updated version of some other package, and I didn't really have a clear idea about how to "run against master"
<srhb>
Same method as above, as you probably guessed now.
<glasserc>
Yeah. I still need practice
Ariakenom has quit [Ping timeout: 250 seconds]
<hlolli>
I'm readin the nixos-in-place readme.md, and it's recommended to do the installation in a tmux session to prevent ssh timeouts. Does it mean a tmux session from the client or tmux session inside the server?
<JonReed>
I miss a debugger with step-by-step execution for Nix. Right now debugging something like a stack overflow error in a large codebase with trace statements in lazy execution is like crawling through a Shawshank Redeption sewer that loops back into your prison cell.
<dminuoso>
JonReed: Have you tried using builtins.trace?
JonReed_ has joined #nixos
<JonReed_>
dminuoso: Yeah :(
<ToxicFrog>
hlolli: on the server, so that if your ssh session drops, it doesn't interrupt whatever command is currently running, and you can just ssh back in and reattach.
<ToxicFrog>
Running tmux on the client wouldn't help at all.
<infinisil>
goibhniu: I think JonReed_ meant nix evaluation
<ToxicFrog>
I wonder if anyone's installed NixOS on the NextThing CHIP
<goibhniu>
oh
<JonReed_>
infinisil: Yeah.
_kwstas has joined #nixos
<infinisil>
JonReed_: Yeah nix evaluation is *really* painful to debug
erasmas has joined #nixos
srl295 has joined #nixos
xkapastel has joined #nixos
_kwstas has quit [Quit: _kwstas]
Makaveli7 has joined #nixos
Dr8128 has joined #nixos
<samueldr>
ToxicFrog: it's a pretty standard allwinner setup IIRC, but armv7 so it includes all the general caveats: no upstream binary cache, incomplete user-provided, and slow to build for
<samueldr>
but depending on how cross-compilation improved, it might be good
<{^_^}>
[nixpkgs] @aanderse opened pull request #57078 → nixos/redmine: replace imagemagickBig with imagemagick → https://git.io/fhhY4
<ToxicFrog>
The main thing I'm nervous about is that it has no external SD card, just eMMC, so if it won't boot from USB (and the documentation doesn't say one way or another) I need to either figure out how to install NixOS from inside the sorta-debian install it ships with or figure out how to flash the eMMC storage over USB
<ToxicFrog>
Whereas with an rpi I can just slap a bootable image onto a uSD card and away it goes
* samueldr
missed the opportunity of getting a pocket chip before they folded
* vikingman
had the opportunity
<vikingman>
;)
<ToxicFrog>
Yeah, I'm hoping I can get away without running software where the install instructions are "run a shell script written for a five year old version of a distro you don't use" :P
adamt has quit [Ping timeout: 255 seconds]
<ToxicFrog>
I never even heard of the CHIP but a co-worker was giving away some random electronic bits and mixed in with all the AVRs that I was actually interested in were two CHIPs.
<samueldr>
ah, the main thing to understand is that it's possible to flash from outside the running operating system, and you probably need to familiarize yourself with it since chances are... you'll break it!
<ToxicFrog>
(and something that appears to be an ARM-based Arduino board, not sure what to do with that either)
Fare has quit [Ping timeout: 250 seconds]
<vikingman>
ToxicFrog: the wifi on chips are terrible and limited to around 10mbps
<vikingman>
and you cannot get bluetooth and wifi working at the same time
<ToxicFrog>
I am completely unsurprised by that, the rpi3 had the same issue
<ToxicFrog>
I don't actually have a use case for these things but they might be fun to fiddle around with
<ToxicFrog>
Maybe I can install octoprint on one and that frees up one of the rpis for other duties
<vikingman>
they made 2 revision of the chip
<vikingman>
one with toshiba memory with 8gb of storrage
<samueldr>
the pocket chip concept had some legs, with the integrated hardware + gpio; probably a neat thing to easily "get" into something to debug
<clever>
catern: that will be at cache.nixos.org/5blzqmi4735d90rdrw26byjnn2bdzm4b.narinfo
<catern>
clever: ok thanks
reilithion has joined #nixos
sondr3 has joined #nixos
<ardumont>
clever: gchristensen: i finally made the test work
<clever>
ardumont: nice
<ardumont>
i had issues with rights about kvm group though
<reilithion>
Ericson2314, Hey, are you around?
<sondr3>
anyone have issues with JetBrains IDEs? Every time I update them I get a SIGSEV for each and every one of them, Pycharm, CLion, Webstorm etc
<samueldr>
sondr3: how are you updating them? and "every time" does it mean you can make them work and next update it fails, or that you rollback and re-update later and it fails?
<catern>
is there a way to ignore my config.nix for one nix-build run?
<sondr3>
samueldr: just by nix-env -u, and then they stop working
<sondr3>
reinstalling them seems to fix it every time
<samueldr>
and sondr3, could you run `nix-shell -p nix-info --run nix-info` it may help figuring out the issue, if it is what I think it might be
<samueldr>
(speaking of, TIL nix-info will hang for a while if you're using remote builders and the builder isn't reachable)
<samueldr>
to confirm, you could run `strace -f -o strace.log $that_ide` and look for mixed gdk pixbuf plugins
<sondr3>
samueldr: I have already reinstalled them so they are working fine again, I'll keep it in mind for the next update though
<sondr3>
it seems fairly serious though
<samueldr>
it only happens when mixing a system using dependencies depending on pixbuf 2.36 and using software using 2.38 (e.g. from unstable)
gspia has quit [Quit: Leaving]
<sondr3>
samueldr: oh, that might be the case, I have some things I use from unstable
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
<samueldr>
when you install using -iA nixos.$something, it installs from the nixos channel, but (I assume) that nix-env -u doesn't really care about channels and finds the most up to date package *as named* in all the channels
<samueldr>
if you were to install the IDEs using systemPackages it shouldn't happen
<samueldr>
(though I'm using "doesn't really care" it's not a slight against nix-env nor channels, but more of an implementation detail)
gspia has joined #nixos
<delan>
anyone know how to make libvirt guests shutdown rather than suspend on host shutdown? libvirt-guests.sh seems to handle this and according to the source I could write ON_SHUTDOWN=shutdown to /var/lib/sysconfig/libvirt-guests but what’s the NixOS Way™ to do that?
<sondr3>
samueldr: might be a good push for me to finally check out home-manager
<sondr3>
thanks for the help, clears up a lot!
<samueldr>
sondr3: it probably would help, as long as it's using the right nixpkgs
<samueldr>
sondr3: and when 19.03 is released, and when you'll update, the issue should go away
<sondr3>
sweet
<catern>
there's something impure in my build, is there a way to ignore config.nix for one run? setting the environment or something?
<pie__>
garbas, ok here's what I dont get, pypi2nix -V "2.7" -e sflock -E libffi -E openssl -E "pkgconfig zlib libjpeg openjpeg libtiff freetype lcms2 libwebp tcl" fails on Pillow, but if i add -e Pillow it doesnt fail
drakonis has quit [Ping timeout: 240 seconds]
asymmetric has joined #nixos
<yl[m]>
I'm making a derivation that produces `outputHash` and `outputHashAlgo` so I can access the network with it, but it seems that I'm missing cacert in the sandbox as well
<yl[m]>
how to expose it?
ym555 has quit [Ping timeout: 252 seconds]
<clever>
yl[m]: have a look at the source for fetchurl
<Izorkin>
etu: in logs - sh: fork: Resource temporarily unavailable
<clever>
yl: ah, those env vars are better
<yl>
clever: definitely! It works, thank you!
rauno has joined #nixos
<noctux>
hello, I'm just starting to experiment with nixos (in the VM image), to see how it works. My current goal is to get a searx instance running with nginx and uwsgi. My current attempt is at https://paste.xinu.at/McsZ/, but somehow, uwsgi fails to find the python packages, even though they get specified in the uwsgi-config: https://paste.xinu.at/McsZ/#n66 and thus, I thought that python.withPackages
maximiliantagher has quit [Remote host closed the connection]
<clever>
,tofu
<{^_^}>
To get a sha256 hash of a new source, you can use the Trust On First Use model: use probably-wrong hash (for example: 0000000000000000000000000000000000000000000000000000) then replace it with the correct hash Nix expected.
maximiliantagher has joined #nixos
Fare has joined #nixos
maximiliantagher has quit [Read error: Connection reset by peer]
<{^_^}>
[nixpkgs] @peti pushed 4 commits to haskell-updates: https://git.io/fhhGb
zupo has quit [Ping timeout: 250 seconds]
zupo has joined #nixos
timor has quit [Ping timeout: 252 seconds]
biopandemic has quit [Ping timeout: 250 seconds]
<{^_^}>
[nixpkgs] @worldofpeace pushed to gnome-3.32 « fixup! gnome3.gnome-settings-daemon: bypass polkit by using an suid wrapper »: https://git.io/fhhZU
jabranham has quit [Quit: ERC (IRC client for Emacs 27.0.50)]
mariel__ has joined #nixos
ThatDocsLady_ has quit [Remote host closed the connection]
ddellacosta has joined #nixos
ErAsMuS has quit [Ping timeout: 245 seconds]
_kwstas has joined #nixos
random_yanek has quit [Ping timeout: 255 seconds]
erictapen has quit [Ping timeout: 246 seconds]
ddellacosta has quit [Ping timeout: 246 seconds]
pie__ has quit [Remote host closed the connection]
pie__ has joined #nixos
mariel__ has quit [Ping timeout: 255 seconds]
ddellacosta has joined #nixos
_kwstas has quit [Quit: _kwstas]
_kwstas has joined #nixos
_kwstas has quit [Remote host closed the connection]
jabranham has joined #nixos
zupo_ has joined #nixos
gagbo has joined #nixos
maximiliantagher has quit [Remote host closed the connection]
erictapen has joined #nixos
random_yanek has joined #nixos
<{^_^}>
[nixpkgs] @Infinisil merged pull request #56790 → nixos/nixos-install: tell the user what to do if setting a root password failed → https://git.io/fhxLW
<{^_^}>
[nixpkgs] @hedning pushed 100 commits to gnome-3.32: https://git.io/fhhZV
zupo has quit [Ping timeout: 252 seconds]
<{^_^}>
[nixpkgs] @Infinisil pushed commit from @bendlas to release-19.03 « nixos/virtualbox: add warning when for ineffective nixpkgs config »: https://git.io/fhhZr
<{^_^}>
[nixpkgs] @kalbasit pushed commit from @Izorkin to master « datadog-agent: 6.10.0 -> 6.10.1 (#57065) »: https://git.io/fhhnh
hamishmack has quit [Ping timeout: 252 seconds]
<yurb>
Is it normal that during a nix-env -u I see git-annex running its tests (or at least that's what I suppose it does)?
<ddellacosta>
hi folks, more haskell+nix newbie questions: if I want to try 8.6.3, it looks like that's only available in unstable. For my own repo, is The (A?) Right Way to modify the arguments to my default.nix and shell.nix in the repo like so? `{ nixpkgs ? import <nixpkgs-unstable> {}, compiler ? "ghc863" }:`
<{^_^}>
[nixpkgs] @xeji pushed commit from @dtzWill to master « gbenchmark: 1.3.0 -> 1.4.1 (#55646) »: https://git.io/fhhce
<ddellacosta>
infinisil: ...sorry, one question: reviewing the Nix 2.0 instructions, it's not clear to me where this file (chunk of code?) goes. Would I drop this somewhere in my project directory, then make sure whatever it returns is passed to default.nix somehow, is that right? And is the purpose to make sure that even though it's unstable it's fixed, vs. not knowing what the latest unstable may bring/break?
<ddellacosta>
er, sorry two questions...
<laalf>
how/where does linuxpackages_latest get created?
<{^_^}>
[nixpkgs] @xeji pushed commit from @dtzWill to master « ntp: 4.2.8p12 -> 4.2.8p13 (#57059) »: https://git.io/fhhcN
zupo_ has quit [Ping timeout: 245 seconds]
<gchristensen>
tilpner: it'd be pretty wild to require a gcc patch update to ensure 40,000 builds suceeded
<gchristensen>
not that it is out of the question
drakonis has quit [Ping timeout: 258 seconds]
<gchristensen>
(how'd we get to 40k? macos, aarch64)
<gchristensen>
(...more if we count cross)
monokrome has joined #nixos
<monokrome>
yo
rprije has quit [Remote host closed the connection]
<{^_^}>
[nixpkgs] @xeji pushed commit from @dtzWill to release-19.03 « ntp: 4.2.8p12 -> 4.2.8p13 (#57059) »: https://git.io/fhhcA
<monokrome>
I'm setting up a server and the ethernet address is `eno1` within the install image. If I install w/ `usePredictableInterfaceNames = true;`, will that change? Once I install/reboot I don't have access unless booting and networking works lol
<{^_^}>
[nixpkgs] @xeji pushed commit from @dtzWill to release-18.09 « ntp: 4.2.8p12 -> 4.2.8p13 (#57059) »: https://git.io/fhhCv
<gchristensen>
monokrome: let me get you an updated example.
<ddellacosta>
so I'm having very basic problems getting a repo going with ghc 8.6.3, it's not finding base--I pushed it up here in case anyone is willing to check it out and tell me what obviously wrong thing I'm doing: https://github.com/ddellacosta/servant-pt1
gilligan_ has joined #nixos
<monokrome>
gchristensen: Thanks <3
<gilligan_>
evening
<gchristensen>
hi gilligan_
<gilligan_>
o/
<gchristensen>
monokrome: I think you just need to change ip4 to ipv4
<monokrome>
oooooooooooh makes sense
<monokrome>
thnx :)
erasmas has joined #nixos
<monokrome>
Are the docs in GitHub? I'd love to start contributing if possible :)
<gilligan_>
hm, how do I end up in a scenario where the result of `nix-store --query --deriver <some-store-path` yields a .drv path that doesn't exist? (I'm fiddling around with retrieving all derivations that end up in a docker image and running vulnix on them)
<{^_^}>
[nixpkgs] @Infinisil pushed commit from @bendlas to release-18.09 « nixos/virtualbox: add warning when for ineffective nixpkgs config »: https://git.io/fhhCk
<{^_^}>
[nixpkgs] @xeji pushed commit from @apeyroux to master « chirp: 20190206 -> 20190304 (#56869) »: https://git.io/fhhC4
reilithion has quit [Quit: Leaving]
<gilligan_>
ls $(nix-store --query --deriver $(nix-build -A hello)) -- i'm confused - Running this on nixpkgs root shouldn't the file always be present? It's not though
<gchristensen>
ah, no, actually
<{^_^}>
[nixpkgs] @roberth pushed 0 commits to nixpkgs-pkgs-doc: https://git.io/fhhCz
<gchristensen>
the drv can be garbage collected separately
<gchristensen>
there is a nix.conf option which keeps them
<gilligan_>
i see
<LnL>
in this case the drv would be there since nix-build evaluated it
<gilligan_>
well it isn't though x__X
disasm has joined #nixos
<LnL>
however the registered deriver is the one that was originally built, not your local one which isn't necessarily the same
<noonien>
what does nix-hash use to compute the hash?
<LnL>
depends what hash you ask for
<noonien>
i mean, what does it use as input to the hashing function?
<noonien>
ofcourse it uses the file contents, but, what else?
<gchristensen>
I think just the contents?
<dminuoso>
Okay wow. The Xfce4 package is *frustrating*.
<LnL>
nothing else, unless you're hashing a directory instead of a flat file
<noonien>
yes, should have said that, i'm hashing a directory
<gilligan_>
LnL, what i'm *actually* trying to do is get the all store paths that go into the result of a dockerTools.buildLayeredImage and run vulnix an all of them - but vulnix fails to parse the .drv file which is always missing. And now i am pretty confused because if vulnix relies on the drv and there is no guarantee for it to be present how does it work in the first place?! :)
<gilligan_>
LnL, well I guess so.. it will explode with output like `Command '['nix-store', '-qR', '/nix/store/r2x1jikav339n7z7spqv0s9l46dacc5m-bash-4.4-p23.drv']' returned non-zero exit status 1.`
<gchristensen>
noonien: nix store paths don't have mtimes, they're all set to 1 second past the epoch
<noonien>
yes, i know, however, i'm hashing a path outside the nix store
<LnL>
gilligan_: oh so it's one of the dependencies that are missing?
<gchristensen>
ah, I think it isconverted to a NAR which loses that property
<{^_^}>
flyingcircusio/vulnix#21 (by plumps, 2 years ago, open): RuntimeError: ('Cannot determine deriver. Is this really a path into the nix store?', '/nix/var/nix/gcroots/booted-system')
<LnL>
given that you have the same expressions you can generate "equivalent" derivations, but they could be slightly different from the original ones depending on the nix version, etc.
disasm has quit [Quit: WeeChat 2.0]
Shouou has quit [Ping timeout: 268 seconds]
justanotheruser has joined #nixos
seanparsons has joined #nixos
disasm has joined #nixos
<gilligan_>
LnL, well i'm talking about spitting out store paths directly from a dockerTools.buildLayeredImage call - or even wrapping the call in something which then calls vulnix - having a different Nix version is not a concern of mine
<gilligan_>
but to me it seems like vulnix really shouldn't be depending on the drv's in the first place
<LnL>
if this is to avoid scanning only outputs instead of the entire build closure I think it should be done the other way around
<LnL>
start with the entire drv tree and skip everything that's not in the outputs
<LnL>
out->drv can change but not drv->out
<LnL>
ldlework: try doInstallCheck
<ldlework>
LnL my hero!
kp__ has quit [Quit: WeeChat 2.4]
<gilligan_>
LnL, actually it does scan the entire build closure.. you end up getting CVEs for subversion scanning some ndoejs app
kp__ has joined #nixos
seanparsons has quit [Remote host closed the connection]
<LnL>
yeah I thought so too, then use nix-instantiate and pass the drv directly instead of the output
pie____ has joined #nixos
seanparsons has joined #nixos
<pie____>
im trying to build a newer, but still old, version of the capstone bindings and i run into this: capstone-3.0.5rc2-py2-none-manylinux1_x86_64.whl is not a supported wheel on this platform.
<pie____>
i have no idea what to doabout it...
<pie____>
bennofs[m]: youre listed as maintainer, d oyou maybe have an idea?
<{^_^}>
#57080 (by kalbasit, 5 hours ago, open): buildGoModule: building Go binaries in two phases
<LnL>
gilligan_: depends on the expression but something like vulnix $(nix-instantiate -A contents) might work
<gchristensen>
yikes, go stuff is way outside my wheelhouse, yl
<yl>
gchristensen: irrelevant of Go stuff :) I'm redoing the infra to work more or less like bazel's infra
<yl>
gchristensen: so fetchAttrs and buildAttrs
<gilligan_>
LnL, ah D'OH that easy. ok thanks
<yl>
gchristensen: also is it frowned upon to move code around? go-modules provides buildGoPackage, I moved it to go-packages to make room for buildGoModule
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<{^_^}>
[nixos-org-configurations] @matthewbauer opened pull request #78 → Add nixpkgs-19.03-darwin to update channels → https://git.io/fhhWJ
<{^_^}>
[nixpkgs] @Infinisil opened pull request #57095 → gtk3: Fix for missing symbols (again) → https://git.io/fhhWU
<gilligan_>
LnL, ah uhrm, well i just realized that in my case my input will be limited to a file with stores paths - can't run nix-instantiate on those
<gchristensen>
you have a drv!
<gchristensen>
nix-instantiate ./test.nix -A mycontainer.closure :)
<LnL>
not following
<gilligan_>
gchristensen, ooooooh. yes of course i do indeed
<gilligan_>
LnL, sorry.. graham was helping me on that earlier on. It's not exactly complicated but missing some context i guess
ng0 has quit [Quit: Alexa, when is the end of world?]
gilligan_ has quit [Ping timeout: 258 seconds]
rcshm has joined #nixos
<selfsymmetric-mu>
Hi! I've disabled a patch by overriding with `patch = []`. Is there a way I can get a notification if the patches upstream are updated? In general, is there a way I can tell if my modifications might have the moral equivalent of merge conflicts?
<selfsymmetric-mu>
The exact situation I'm in is that `pipenv` is broken in NixOS because it has a bad patchfile that got out of date. I've killed the patch but that could be a nasty bug later on if I don't keep track of how it's affected.
<selfsymmetric-mu>
In other package managers (say, el-patch), conflicts coming down the pipe notify the user. Is something like that available here?
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
<{^_^}>
#57008 (by matthew-piziak, 1 day ago, open): Bad virtualenv patch breaking pipenv on Python 3.7.2
illegalprime has quit [Ping timeout: 255 seconds]
rcshm has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
talqu has quit [Ping timeout: 268 seconds]
o1lo01ol1o has joined #nixos
Cale has quit [Ping timeout: 252 seconds]
<{^_^}>
[nixpkgs] @FlorianFranzen opened pull request #57097 → qtbase: Fix paths returned by qmake -query → https://git.io/fhhW1
<disasm>
I'm trying to setup grub (using efi) in place of systemd-boot. I've tried with canTouchEfiVariables but it doesn't seem to be working. efibootmgr doesn't show the path to grub. I also tried with efiInstallAsRemovable but it didn't create the /boot/EFI/BOOT directory (like systemd-boot does).