<c74d>
no, if I listen closely I can hear the fan as well, and it's different
c0bw3b has left #nixos []
<c74d>
leaving it on longer did capture activity
<c74d>
all from systemd-journal, kworker/u2, and jbd2/vda-8
<clever>
normal stuff for a nixos machine
<clever>
and part of that is the logger saving dmesg, which creates more messages
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
<c74d>
hm, the noise pauses when the HDD makes actual read/write noises
<c74d>
at least sometimes
Fare has quit [(Ping timeout: 268 seconds)]
<c74d>
hm, switching away from the VM window to IRC doesn't affect the noise, but then switching back to the VM makes it pause briefly
bfrog_ has quit [(Ping timeout: 258 seconds)]
<clever>
this reminds me of an issue on one of my older laptops, the capacitors near the cpu would make an audible noise, that is synced to the cpu usage
<clever>
so i could hear the cpu load changing in real time
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
aneeshusa has joined #nixos
erasmas has quit [(Quit: leaving)]
estewei has quit [(Quit: Quit)]
redmq has joined #nixos
markus1189 has joined #nixos
eacameron has quit [(Ping timeout: 265 seconds)]
bfrog_ has quit [(Ping timeout: 246 seconds)]
markus1199 has quit [(Ping timeout: 258 seconds)]
eacameron has joined #nixos
nh2_ has quit [(Ping timeout: 260 seconds)]
herzmeister has quit [(Quit: Leaving)]
herzmeister has joined #nixos
eacameron has quit [(Ping timeout: 244 seconds)]
<gchristensen>
aszlig: ping
eacameron has joined #nixos
eacameron has quit [(Ping timeout: 246 seconds)]
eacameron has joined #nixos
thc202 has quit [(Ping timeout: 250 seconds)]
m0rphism has joined #nixos
[0x4A6F] has quit [(Ping timeout: 256 seconds)]
eacameron has quit [(Ping timeout: 246 seconds)]
aneeshusa has quit [(Ping timeout: 260 seconds)]
stepcut has joined #nixos
<clever>
gchristensen: do you know if hetzner uses dhcp or staticly configured IP's on each box?
angerman has joined #nixos
nh2_ has joined #nixos
<jmiven>
clever: dhcp, at least on the VPSs
<clever>
ah
eacameron has joined #nixos
stepcut has quit [(Remote host closed the connection)]
aneeshusa has joined #nixos
joelmo has joined #nixos
stepcut has joined #nixos
<nhooyr>
I'm trying to override the patches being used while building a derivation. right now, they are `patches = [ ./remove-tools-1.7.patch ./cacert-1.7.patch ];` but I want to make them `patches = [ ./remove-tools-1.7.patch ];` in my ~/.nixpkgs/config.nix but the problem is that ./remove-tools-1.7.patch is in the directory of the derivation. how can I make that path relative to that directory?
<nhooyr>
in the directory of the package*
Seichi has quit [(Quit: Konversation terminated!)]
<clever>
avn: what does this output: ls -lh /nix/store/003x4zwb9794iydhbclxl0h4cixq1vq0-cmake-2.8.12.2.drv.bz2
<avn>
Is not so hard to write script, which will purge files for non-exists stuff
<avn>
ls -la /nix/store/003x4zwb9794iydhbclxl0h4cixq1vq0-cmake-2.8.12.2.drv no such file
<clever>
not sure then, i would expect GC to do it
Rotaerk has joined #nixos
<avn>
clever: I expect it myself
derjohn_mobi has joined #nixos
<clever>
only thing i can think of is that you did some naughty things in your early nixos days and manualy deleted things
<clever>
ive done that, it deleted the active nixos generated, lol
ilja_kuklic has quit [(Ping timeout: 250 seconds)]
derjohn_mob has quit [(Ping timeout: 265 seconds)]
ThatDocsLady has joined #nixos
zraexy has quit [(Ping timeout: 258 seconds)]
rardiol has quit [(Ping timeout: 244 seconds)]
glines has joined #nixos
rardiol has joined #nixos
agjacome has quit [(Quit: leaving)]
<NixOS_GitHub>
[nixpkgs] garbas pushed 2 new commits to master: https://git.io/v1adt
<NixOS_GitHub>
nixpkgs/master 3a1d52f Rok Garbas: updateScript: initital work
<NixOS_GitHub>
nixpkgs/master d295d68 Rok Garbas: firefox-bin: updated the updateScript with suggestions from @Mic92...
aneeshusa has quit [(Ping timeout: 246 seconds)]
aneeshusa has joined #nixos
echo-area has joined #nixos
<gchristensen>
ugh nixops + hetzner is so annoying
<gchristensen>
we should seriously consider dropping everything but aws + hardware
<simpson>
gchristensen: We should keep GCE!
<clever>
gchristensen: i have thought about how to integrate my kexec idea with nixops
<clever>
gchristensen: in theory, you just need to ssh into the root account, inspect some state (ip addr mainly), build the kernel+initrd, scp, execute, and wait for ssh to come back, then it behaves like it booted from the install ISO
<clever>
gchristensen: in theory, it will work on any hoster that obeys the MBR loader
<gchristensen>
aye
<gchristensen>
simpson: I dunno, gce doesn't work as nicely as aws. I don't care what we have as long as it works well
<gchristensen>
and users don't have to fight with nixops to do it
<clever>
gchristensen: the kexec idea also opens up routes that arent currently available, like partition table control
<simpson>
gchristensen: GCE is cheaper than AWS and I'm not interested in leaving GCE.
<gchristensen>
that is fine, sure, whatever, we could support other providers via unofficial plugins
mguentner2 has quit [(Ping timeout: 250 seconds)]
mguentner3 has joined #nixos
<gchristensen>
the problem isn't GCE the problem is nixops + not enough time and effort behind anything than aws
<clever>
gchristensen: with a few more changes, my kexec thing could nixos-ify any host in one go, and leave it without a configuration.nix file, then you can target it with the none env in nixops
aneeshusa has quit [(Ping timeout: 248 seconds)]
<gchristensen>
yeah
<gchristensen>
that might be the thing
<clever>
only scary thing, is that you need to be real carefull where you aim the thing
<clever>
the something else can be in-lined with -E
<Rotaerk>
k, I see; thanks
data8 has joined #nixos
anelson has joined #nixos
rardiol has quit [(Ping timeout: 244 seconds)]
nh2_ has quit [(Quit: Leaving.)]
<Rotaerk>
think I'm going to start naming my cabal2nix-generated files <project>.cabal.nix, so that it's clear that it was generated
<Rotaerk>
and thus only to be "modified" through overriding
rardiol has joined #nixos
<stepcut>
I installed the latest hydra from github, but when I look at a build for a job I get this http://i.imgur.com/Sxkb0t5.png, is hydra broken, or did I screw things up ?
<clever>
stepcut: i'm guessing the db schema is out of sync with the code
<stepcut>
so, I thought that might be the case, so I started with a fresh database and ran hydra-init
<clever>
if the schema in hydra-init is out of sync with the code, it will happen every time
<stepcut>
ok, that hydra error I pasted later is because I upgrade to the new hydra, but my init scripts were broken so the old hydra server never stopped and the new one never started
<clever>
ah
<clever>
so i was half right, but didnt think about that possibility
<eacameron>
clever: I never thanked you for the solution to getting the git hash of my channel. Thank you! Your knowledge knows no bounds. (Whether no bounds is worth knowing, only you would know.)
<clever>
heh
<stepcut>
I spent waaay too long figuring that out :)
<stepcut>
it would help if, ps ax | grep hydra, had showed the server was running, but it shows up as starman
<stepcut>
this would likely work better if I was running hydra on NixOS instead of Ubuntu
<stepcut>
getting the init.d scripts to start *and* stop reliably is a black art
<clever>
ive got nixos and hydra on my router, and its simple to start/stop
<eacameron>
clever deploys hydra to his router O_o
<clever>
i call it a router, but its a dual-socket rack-mountable server, with 3gig of ecc ram, and 3 gigabit ethernet ports
<clever>
but it does dhcp and nat, so that makes it a router, right?
<simpson>
clever: Do you happen to have any nixos-container networking snippets? Every time I fight with nixos-container, I seem to have networking troubles.
<clever>
simpson: i just setup NAT when i was doing that, one sec
<simpson>
I want generally for my containers to just be free to access Internets.
<clever>
simpson: line 8 has to be the external ip of the host, on the LAN, and line 12 has to differ from the lan subnet
<clever>
18&19 need to be unique within that new subnet
<clever>
and 10 forwards things to the container
<simpson>
Mm. I hate reading networking configuration. But this looks really useful, thanks.
bfrog_ has joined #nixos
<eacameron>
simpson: I wouldn't mind reading it if I actually understand a word of it. :/
<eacameron>
Too bad my networking prof was the worst prof in the school installing WireShark was about as much as we learned that semester.
<clever>
having done the same thing in my nixos router, it was simple to do again
<simpson>
Ah, yes, code is always easier to write the third time around.
<clever>
eacameron: i have written wireshark decoding plugins, heh
<eacameron>
clever: Lol...you would
glines has quit [(Ping timeout: 246 seconds)]
<clever>
simpson: i have also previously ran a linuxfromscratch router before, and manualy turned my laptop into a router a few times with entirely manual iptables and ip commands
uralbash has joined #nixos
<simpson>
clever: Yeah, like, I remember doing all that stuff a decade ago, but then I forgot all of it.
bfrog_ has quit [(Ping timeout: 260 seconds)]
glines has joined #nixos
<Rotaerk>
there some way to use antiquoting in a path?
<Rotaerk>
./${project}.nix doesn't seem to work
mk-fg has quit [(Ping timeout: 256 seconds)]
<clever>
Rotaerk: ./. + "/${project}.nix" i think
<Rotaerk>
heh, that worked
<clever>
paths can have strings appended to them
<clever>
but after each append, nix will normalize the path
<clever>
so ./. + "/" == ./. (the trailing slash is dropped)
<clever>
so you cant do the obvious ./. + "/" + project + ".nix"
<clever>
( and ) can force it, but antiquoting it inside a string also works
<eacameron>
So hypothetical scenario. You have a nixops deployment and you lose the state file...how do you recover a situation where you can deploy again? (Presuming you have another way of accessing the machine to add a new key or something
stepcut has quit [(Remote host closed the connection)]
<clever>
none that i know of, somebody that i talked to a few months ago kept the nixops state in git (with some gpg crypto over it)
<clever>
which also allowed sharing the state with other people
<clever>
to sync it up
anelson has quit [(Remote host closed the connection)]
<Rotaerk>
I'm doing let ghcpkgs = nixpkgs.pkgs.haskell.packages.ghc801 in (ghcpkgs.callPackage ./myproject.nix {}), but some of the dependencies of my project, while they exist in ghcpkgs, aren't the right version
<Rotaerk>
I have access to a good version from git...
<Rotaerk>
should I basically do something like... let self = ghcpkgs // { myDependencyOverride = callPackage (mkDerivation (/* pull from git */)) }
<Rotaerk>
and then the expression above
anelson has joined #nixos
glines has quit [(Ping timeout: 260 seconds)]
Wizek_ has quit [(Ping timeout: 265 seconds)]
Wizek has quit [(Ping timeout: 260 seconds)]
glines has joined #nixos
Wizek has joined #nixos
Wizek_ has joined #nixos
<Rotaerk>
though really, if a local package is meant to have a reproducible build, it really shouldn't draw upon nixpkgs derivations at all...
<Rotaerk>
since updating nixpkgs could update one of the local package's dependencies
joelmo has quit [(Quit: Connection closed for inactivity)]
<eacameron>
Rotaerk: I'm not exactly sure I understand, but it seems like you want to override your package set with some patched version of a package.
<eacameron>
Rotaerk: That is a tad tricky because if that package has downstream dependencies, they also need to rely on the patched version.
<eacameron>
Rotaerk: Let me show you what I have...
<eacameron>
It's cut out of one of my projects, but it basically shows how to come up with a new haskell package set and override some of the dependencies in a fully transitive way (no inconsistencies)...it also shows how to get a package directly from GitHub.
<eacameron>
I think reflex-platform already has a (likely better version) of my fetchCabal thingy
<eacameron>
It really should be called "callCabal"
<Rotaerk>
clever, interesting. don't think that approach would work here, though; looks like you're downloading an expression and then importing it, but in my case I want to get a different version of the source for this package from its github repo
sdothum has quit [(Quit: ZNC - 1.6.0 - http://znc.in)]
<clever>
Rotaerk: you can also aim that at your fork of nixpkgs, and include customizations
<clever>
Rotaerk: or put the customizations into the config argument, it supports overrides
<eacameron>
clever: Rotaerk: That's essentially what my example is doing, but only for haskell packages (i.e. not all of nixpkgs)
<Rotaerk>
ah, so it's explicitly pulling from cabal..
<eacameron>
Rotaerk: Yah sorry not tons of docs there. It downloads from GitHub, runs cabal2nix on it, and then runs callPackage on *that*
<eacameron>
:P
hotfuzz has quit [(Read error: Connection reset by peer)]
hotfuzz_ has joined #nixos
<eacameron>
Rotaerk: peti agrees this function should be included in nixpkgs...just haven't gotten around to it yet.
<eacameron>
Once you have an overridden set, you can say "dependencies = with myOverriddenSet; [ dep1 dep2 dep3];" and it will use your revised set
<Rotaerk>
yea I read that section; obviously didn't fully understand it though
<Rotaerk>
thanks, will re-read
<eacameron>
Rotaerk: "overrides = self: super: { ... }" really is confusing. "self" is the revised set that you're currently defining. "super" is the unrevised set that you're revising...
<eacameron>
so you'll notice "fetchCabal' = ... fetchCabal { haskellPackages = self; }" <- which is saying "when you build this package with cabal, build it with the new set of packages that I'm currently defining. Since packages are built in dependency order, that won't create a cycle of any sort.
angerman has quit [(Remote host closed the connection)]
<NixOS_GitHub>
[nixpkgs] rbvermaa pushed 1 new commit to release-16.09: https://git.io/v1Vtu
<NixOS_GitHub>
nixpkgs/release-16.09 7a28df4 Rob Vermaas: Fix location of broadcom sta driver.
<lewo>
hi. How could I write a file in a test VM (makeTest) before starting any services? Is there an exemple of such kind of "preprovionning"?
<lewo>
(this is besically to create a /var/run/my-file in order to simulate nixops deployment keys mecanism)
christine_ has joined #nixos
ckauhaus has joined #nixos
markus1189 has quit [(Remote host closed the connection)]
anelson has quit [(Ping timeout: 258 seconds)]
anelson has joined #nixos
hotfuzz_ has joined #nixos
hotfuzz has quit [(Ping timeout: 258 seconds)]
zagy has joined #nixos
markus1189 has joined #nixos
vandenoever has joined #nixos
vandenoever has quit [(Changing host)]
vandenoever has joined #nixos
ericsagn1 has quit [(Ping timeout: 246 seconds)]
<NixOS_GitHub>
[nixops] rbvermaa pushed 1 new commit to master: https://git.io/v1Vmi
<NixOS_GitHub>
nixops/master 50d6e15 Rob Vermaas: Add i2 and c5 as hvm instances.
eacameron has joined #nixos
<lewo>
ok, boot.postBootCommands doest the job
eacameron has quit [(Ping timeout: 245 seconds)]
<clever>
lewo: normaly, id use a prestart on the systemd unit, auto-generate random keys before starting up
<NixOS_GitHub>
[nixops] rbvermaa pushed 1 new commit to master: https://git.io/v1VYc
<NixOS_GitHub>
nixops/master 0740a03 Rob Vermaas: Remove the need to add new HVM instance types. HVM is the default nowadays for new instance types, so better to just check if it is not a PV GRUB instance type.
ktosiek is now known as ktosiekk
ktosiekk is now known as ktosiek
ktosiek is now known as ktosie
ktosie is now known as ktosiek
johann__ has quit [(Quit: Leaving.)]
tat_ has joined #nixos
<NixOS_GitHub>
[nixops] rbvermaa pushed 1 new commit to master: https://git.io/v1VYo
<NixOS_GitHub>
nixops/master b57820d Rob Vermaas: Regenerated ec2-properties.nix, which new instance types.
<tat_>
i just saw that the debian pkg's on hydra.nixos.org are failing to build, are you still actively supporting deb package installation ?
Wizek_ has quit [(Ping timeout: 265 seconds)]
<lewo>
clever, do you have an example?
<lewo>
clever, and are you overriding the prestart of an existing unit file, or do you create a new unit file that has to be run before all others?
<clever>
lewo: i usualy put the prestart on the unit that cares about the keys, generaly in the same module file
<clever>
lewo: the sshd service is an example of that already in nixos
latk has joined #nixos
ckauhaus has quit [(Quit: Leaving.)]
<lewo>
clever, not sure we are speaking about the same usecase. I have a module that can gets its secret from /var/run/some-file. In order to test this module, I want to write a test taht use this module and create this secret file (which would be normally created by nixops). So I don't want to put this part in the module itself since this is only used by the test.
<clever>
my thought, is to have it generate its own secret when nixops isnt at play
markus1189 has quit [(Ping timeout: 260 seconds)]
<clever>
and then the testcase can read the file out of the guest, to steal the secret and continue testing
<chattered>
Not sure whether I should be asking this here or in #emacs, but it seems my Emacs has lost all its dbus functions. It's compiled with dbus-support by default, right?
<chattered>
(at least, I can't see --without-dbus in the configure flags)
ckauhaus has joined #nixos
<lewo>
clever, ok, I understand
derjohn_mobi has quit [(Ping timeout: 260 seconds)]
Judson has quit [(Ping timeout: 260 seconds)]
ThatDocsLady has quit [(Ping timeout: 260 seconds)]
<Cinnamon811rc>
Hello! Could you help me with installation of nixOS?
thc202 has joined #nixos
<goibhniu>
Hi Cinnamon811rc, what's up?
<NixOS_GitHub>
[nixpkgs] knedlsepp opened pull request #21016: openblas: Fix build issue when inside VM and CPU is undetected (master...fix-openblas-build) https://git.io/v1VnY
derjohn_mobi has joined #nixos
tat_ has quit [(Read error: Connection reset by peer)]
hyper_ch has quit [(Ping timeout: 260 seconds)]
<Cinnamon811rc>
Grub i386 failed to set on a separate partition, and I missed the step where the installer asked about root password
<Cinnamon811rc>
and now I cannot login into the system installed
<NixOS_GitHub>
nixpkgs/master 7833315 Joachim F: Merge pull request #21015 from nico202/non...
wkennington has quit [(Read error: Connection reset by peer)]
IITaudio has joined #nixos
<IITaudio>
Hi, I'm using pkgs.writeText to write some config file from configuration.nix. I'd like to create also a couple of scripts. Is there a way to set the permissions of files created by writeText?
<joachifm>
IITaudio: see build-support/trivial-builders.nix
christine_ has joined #nixos
<IITaudio>
joachifm: executable true, was that easy, thanks!
<kragniz>
domenkozar: on the front page now :)
<domenkozar>
so easy
<joachifm>
IITaudio: :)
johann__ has quit [(Quit: Leaving.)]
griff_ has quit [(Quit: griff_)]
AllanEspinosa has joined #nixos
<ben>
Why does nix-env -qasP hit the network? :(
AllanEspinosa has quit [(Ping timeout: 245 seconds)]
<gchristensen>
ben: "The third is S or -, indicating whether a substitute is available for the derivation."
<gchristensen>
ben: ie: it checks the substitution cache (cache.nixos.org) for the substitution
<ben>
Oh, so it tries to fetch every single substitute
<gchristensen>
mmm probably not actually fetching, probably a HEAD or something to see if it exists
Ivanych has quit [(Ping timeout: 250 seconds)]
<gchristensen>
ben: that is the `-s` option, btw
mudri has quit [(Ping timeout: 248 seconds)]
<gchristensen>
ben: `man nix-env` you can not pass `-s` and it won't check
<ben>
yeah
<ben>
Thanks; I guess I won't pass -s in the future because I appear to suffer from a network condition where simultaneously created connections cause most of them to silently be dropped
<ikwildrpepper>
that explains the xen error I had this morning, which I didn't want to look into :D
herzmeister has joined #nixos
<gchristensen>
which I'm fixing now. I had learned earlier this week that `nix-prefetch-url` can take an attribute, and it was so much nicer than build-and-wait-for-error
<LnL>
gchristensen: this is why I stopped using the nix-prefetch-tools
<gchristensen>
yeah ... so usually this problem would be caught by my build machine and I'd fix before merging... but that one isn't back up yet.
<LnL>
the annoying thing with fetchpatch is that it won't complain once it's in the store
<gchristensen>
eh?
<LnL>
so you have to run nix-build --check -E 'fetchpatch ...' or build on another machine
IITaudio has quit [(Quit: Leaving)]
<NixOS_GitHub>
[nixpkgs] grahamc pushed 1 new commit to master: https://git.io/v1Vz7
<NixOS_GitHub>
nixpkgs/master a2d6e8a Graham Christensen: xen: Fix patch hashes...
<LnL>
it would be nice if you could just give a derivation expression to nix-prefetch and it would detect the fetchurl/fetchpatch attributes and check those
<NixOS_GitHub>
[nixpkgs] grahamc pushed 1 new commit to release-16.09: https://git.io/v1Vzb
<NixOS_GitHub>
nixpkgs/release-16.09 c43a79e Graham Christensen: xen: Fix patch hashes...
Shoue has joined #nixos
<NixOS_GitHub>
[nix] edolstra pushed 1 new commit to master: https://git.io/v1Vg2
<NixOS_GitHub>
nix/master 47f5877 Eelco Dolstra: Probably fix a segfault in PathLocks
<sphalerite>
is there anything analogous to apt show for nixos? The metadata is all there, but what about querying it?
<sphalerite>
I guess I could grep through the actual expressions, but surely there's a way with a better UX?
<LnL>
nix-env -qaP is the closest thing
<sphalerite>
that's more like apt search
<manveru>
sphalerite: you'll like nox
<sphalerite>
that's also more like apt search
<manveru>
oh, my bad
<LnL>
ah, you can get the meta field from a drv
<manveru>
what format is the drv even? it's not nix or json...
<sphalerite>
although I guess nox does show most of the relevant metadata
<manveru>
yeah... it's not exactly a beauty, hopefully we'll get the new nix cli soon
<FRidh>
Ugh why am I using Nix for constructing a listening test.
<gchristensen>
b/c you're an addict
<FRidh>
Yes, I guess so.
<FRidh>
The first part that is working is kind of neat though. Now, I get to the point I need a PRNG and somehow I don't think that would be part of Nix.
edvorg has quit [(Remote host closed the connection)]
<gchristensen>
I've used /dev/random to make prngs in a runCommand derivation, and with an input parameter to use as a `seed` for when I want to regenerate it
<gchristensen>
so the randomness is stable until I want to new randomness
jgeerds has quit [(Quit: Leaving)]
<rly>
Do we have a jsonlinter packaged?
<FRidh>
gchristensen: and then import from derivation to get the result?
<gchristensen>
yeah
<FRidh>
I was thinking of doing something like that, although in Python. But I don't know, how many "layers" of IFD can nix handle?
Bodil has quit [()]
Bodil has joined #nixos
systemfault has joined #nixos
<gchristensen>
:)
johann__ has quit [(Read error: Connection reset by peer)]
johann__ has joined #nixos
Ivanych has quit [(Ping timeout: 260 seconds)]
amarsman has joined #nixos
agarzia has quit [()]
agarzia has joined #nixos
johann__ has quit [(Quit: Leaving.)]
Fare has quit [(Ping timeout: 260 seconds)]
<sphalerite>
I just saw that the closure for pavucontrol includes a whole lot of dev packages, including gtk+3 and all that it depends on. I'm guessing that's not really necessary? How can I debug it?
Wizek has quit [(Ping timeout: 260 seconds)]
anelson has quit [(Ping timeout: 245 seconds)]
anelson has joined #nixos
<joachifm>
sphalerite: one thing you can do is grep for store paths and work from there
rly has quit [(Ping timeout: 250 seconds)]
<joachifm>
sphalerite: then there's nix-store -q --tree &c
<sphalerite>
oooh yes that's nice
<sphalerite>
thanks!
emmanueloga has quit [()]
emmanueloga has joined #nixos
<NixOS_GitHub>
[nixpkgs] NeQuissimus pushed 2 new commits to release-16.09: https://git.io/v1Vw1
<NixOS_GitHub>
nixpkgs/release-16.09 a8e2b3e Tim Steinbach: linux: 4.4.36 -> 4.4.37...
<NixOS_GitHub>
nixpkgs/release-16.09 fbd303f Tim Steinbach: linux: 4.8.12 -> 4.8.13...
<NixOS_GitHub>
[nixpkgs] NeQuissimus pushed 2 new commits to master: https://git.io/v1VwM
<NixOS_GitHub>
nixpkgs/master e861a5f Tim Steinbach: linux: 4.4.36 -> 4.4.37
<NixOS_GitHub>
nixpkgs/master bfffbb5 Tim Steinbach: linux: 4.8.12 -> 4.8.13
<sphalerite>
hm, cairo-dev seems to be getting brought in by cairomm, but the only reference to cairo-dev I can find is in cairomm's propagated-native-build-inputs. Surely that shouldn't cause it to become a dependency?
<sphalerite>
(runtime dependency)
<joachifm>
it all depends on which store paths end up being written to the output
<joachifm>
there's all sorts of reasons why you might pick up unwanted refs
<Cinnamon811rc>
Hello! Could you, please, tell me how to add ModemManager in the config file? Is it services.modemamanger?
<Cinnamon811rc>
I am in countryside and can connect to the Internet via Modem only.
<Cinnamon811rc>
Maybe somebody knows?
Fare has joined #nixos
smola_ has quit [()]
smola_ has joined #nixos
<gchristensen>
Cinnamon811rc: we don't seem to have a service for ModemManager
<Cinnamon811rc>
really? only via wifi?
<gchristensen>
I don't know, I'm sorry :(
<Cinnamon811rc>
it is ok, thank you. When I tried to install guixSD there was ModemManager there. But they derived their OS from nixOS
<gchristensen>
one minute, Cinnamon811rc
pchiusano has quit [()]
<Cinnamon811rc>
aha thank you:)
pchiusano has joined #nixos
adulteratedjedi has quit [()]
adulteratedjedi has joined #nixos
johann__ has quit [(Quit: Leaving.)]
<gchristensen>
I don't know how modemmanager works
<gchristensen>
but you can get it from pkgs.modemmanager, `nix-shell -p modemmanager`
<Cinnamon811rc>
I mean it is Mobile Modem on the USB flash drive
cpennington has joined #nixos
<Cinnamon811rc>
what does it mean? single back quotes?
<Cinnamon811rc>
I am new and do not know what is nix-shell
<Cinnamon811rc>
should I write it in term?
<gchristensen>
I'm sorry, Cinnamon811rc, I've run out of time for right now. I'm hoping someone else can help you :) however: typing `nix-shell -p modemmanager` at the term will open a shell where modemmanager is in the PATH
<Cinnamon811rc>
tham you for your support and have a great day
<Cinnamon811rc>
thank you
<NixOS_GitHub>
[nixpkgs] joachifm pushed 3 new commits to master: https://git.io/v1VXw
<NixOS_GitHub>
nixpkgs/master ca7cc96 Joachim Fasting: grsecurity: enable PAX_INITIFY...
<NixOS_GitHub>
nixpkgs/master 9a63779 Joachim Fasting: grsecurity: use upstream url as the primary source
<NixOS_GitHub>
nixpkgs/master d1a5dc0 Joachim Fasting: grsecurity: 4.8.12-201612062306 -> 4.8.13-201612082118
yorick_ is now known as yorick
<NixOS_GitHub>
[nixpkgs] joachifm pushed 1 new commit to release-16.09: https://git.io/v1VXQ
<NixOS_GitHub>
nixpkgs/release-16.09 f8bea71 Joachim Fasting: grsecurity: 4.8.12-201612062306 -> 4.8.13-201612082118...
chaoflow has joined #nixos
uralbash has quit [(Quit: Leaving.)]
pi3r has joined #nixos
<pi3r>
is it possible to pass a resolver to buildStackProject ?
eacameron has joined #nixos
glines has joined #nixos
eacameron has quit [(Ping timeout: 244 seconds)]
exi has joined #nixos
Rotaerk has quit [(Quit: Leaving)]
<exi>
hi
<exi>
anyone here knows how to use nixos-rebuild switch --build-host <myhost>
<exi>
?
<exi>
i only get:
<exi>
error: imported archive of ‘/nix/store/9krlzvny65gdc8s7kpb6lkx8cd02c25b-default-builder.sh’ lacks a signature
<exi>
ah, one more line on top:
<exi>
copying 213 missing paths (0.48 MiB) to ‘builder@nixos.lan’...
<exi>
error: imported archive of ‘/nix/store/9krlzvny65gdc8s7kpb6lkx8cd02c25b-default-builder.sh’ lacks a signature
<exi>
so it fails to copy paths over but there is no documentation to tell my why that is and how to fix it
Cinnamon811rc has left #nixos ["Leaving"]
philips has quit [()]
philips has joined #nixos
fritschy has quit [(Quit: Leaving.)]
alx741 has joined #nixos
glines has quit [(Ping timeout: 264 seconds)]
roconnor has joined #nixos
<exi>
anyone?
rly has joined #nixos
m0rphism has joined #nixos
glines has joined #nixos
johann__ has joined #nixos
Seichi_ has joined #nixos
Seichi has quit [(Ping timeout: 260 seconds)]
MercurialAlchemi has quit [(Ping timeout: 260 seconds)]
<rly>
How am I supposed to use bundix?
chaoflow has quit [(Quit: WeeChat 1.6)]
<manveru>
with a lot of enthusiasm
cpennington has quit [(Ping timeout: 264 seconds)]
<rly>
manveru: the documentation lists some commands, but it never bothers to say what it will achieve.
<Guest10411>
hey! how can one install node modules globally? Currently tried npm -g install 'module', but it prompts me for admin accesss
Seichi has quit [(Quit: Konversation terminated!)]
herzmeister has quit [(Quit: Leaving)]
spacekit1eh has joined #nixos
herzmeister has joined #nixos
kragniz has quit [(Read error: Connection reset by peer)]
kragniz has joined #nixos
<sphalerite>
fractal: neat, thanks for the pointer
spacekitteh has quit [(Ping timeout: 250 seconds)]
<sphalerite>
or rather investigating the issue
zagy has joined #nixos
cpennington has quit [(Ping timeout: 258 seconds)]
digitus has joined #nixos
lsix has quit [(Quit: Konversation terminated!)]
rly has quit [(Ping timeout: 248 seconds)]
adev has quit [(Remote host closed the connection)]
alx741_ has joined #nixos
hlavaty has left #nixos ["ERC (IRC client for Emacs 25.1.1)"]
alx741 has quit [(Ping timeout: 264 seconds)]
iyzsong has quit [(Ping timeout: 268 seconds)]
Guest350 has quit [(Ping timeout: 265 seconds)]
devhell has joined #nixos
<manveru>
zimbatm: i think we should add the bundled_gems to ruby...
amarsman has joined #nixos
L0g4nAd4ms has joined #nixos
<L0g4nAd4ms>
OK, i seem to be to stupid to change my x server keyboard layout to german. I have layout = "de", xkbModel = "pc105", but that does not seem to be enough.
<manveru>
zimbatm: also i'd love an option to exclude the rdocs for production
<gchristensen>
I wonder if we should be talking more loudly about these people using nixos (see the nix-dev thread)
c0bw3b has joined #nixos
goibhniu has quit [(Ping timeout: 244 seconds)]
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to staging: https://git.io/v1wmF
<NixOS_GitHub>
nixpkgs/staging ff62e8a Vladimír Čunát: glib: avoid runtime reference from gtk3...
ecraven has left #nixos []
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to staging: https://git.io/v1wmp
<NixOS_GitHub>
nixpkgs/staging b05b120 Vladimír Čunát: Merge branch 'master' into staging
<stepcut>
$ hydra-queue-runner --help
<stepcut>
error: unrecognised flag ‘--help’
<stepcut>
Try ‘hydra-queue-runner --help’ for more information.
<stepcut>
:(
<NixOS_GitHub>
[nixpkgs] vcunat closed pull request #20862: gnome-control-center: provides access to the tracker gsettings schemas (master...gnome-control-center-tracker) https://git.io/v1CIl
<gchristensen>
stepcut: :|
<stepcut>
I am getting this error, "possibly transient failure building ‘/nix/store/pd5arhhskwbkzmn21nzkh97wjh48lv23-appraisalscribe-paths-8564ecf.drv’ on ‘localhost’:" when hydra-queue-runner tries to run a job -- any idea if there is a way to get more information about why the step failed?
teknico has left #nixos []
<stepcut>
I think it is an issue with fetchgitPrivate :-/
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to master: https://git.io/v1wO2
<NixOS_GitHub>
nixpkgs/master e54650e Vladimír Čunát: gnome-control-center: port the parent commit to 3.22...
<pi3r>
It does compile the default.nix package but it is not available as an exec. I thought 'env' would do that for me ?
griff_ has quit [(Quit: griff_)]
zagy has quit [(Quit: Leaving.)]
<srhb>
pi3r: You probably want one of the ghcWithPackages ...
<srhb>
Oh, wait, you're expecting servant-client to produce an executable?
<srhb>
I'm confused. It's just a library, so there is no executable.
<NixOS_GitHub>
[nix] domenkozar pushed 1 new commit to master: https://git.io/v1wB8
<NixOS_GitHub>
nix/master 41d6523 Domen Kožar: Document builtins.match, fixes #1145
<pi3r>
srhb: no the default.nix is a supposedly producing the executable
brh has quit [(Quit: Lost terminal)]
spacekit1eh has quit [(Ping timeout: 258 seconds)]
spacekitteh has joined #nixos
<srhb>
pi3r: What executable?
pi3r has quit [(Ping timeout: 265 seconds)]
elurin has quit [(Quit: ERC (IRC client for Emacs 25.1.1))]
redmq has quit [(Ping timeout: 248 seconds)]
nh2_ has quit [(Ping timeout: 240 seconds)]
<Ralith>
srhb: he's invoking a package in the current directory
<NixOS_GitHub>
[nixpkgs] LnL7 pushed 1 new commit to master: https://git.io/v1wal
<NixOS_GitHub>
nixpkgs/master 2444baf Daiderd Jordan: include me in mention bot for darwin
josdotso has joined #nixos
zagy has joined #nixos
saintromuald has joined #nixos
<simpson>
garbas: (and also fridh who is missing) I just learned about https://wiki.python.org/moin/PyPIJSON information. Would this help us autogenerate Python packages at all?
<FRidh>
garbas: i use both this and the xmlrpc API to get the meta data
<FRidh>
but it doesn't include dependencies
<FRidh>
oh, simpson: ^
<simpson>
FRidh: Aha.
<garbas>
simpson: FRidh: yup also it doesnt check if specifications are possible to resolve
<simpson>
FRidh, garbas: That's too bad. But I guess it makes a bit of sense; we're the only ones AFAIK that do dependency resolution so early in the build process.
Guest37568 has quit [(Changing host)]
Guest37568 has joined #nixos
Guest37568 has quit [(Changing host)]
Guest37568 has joined #nixos
<NixOS_GitHub>
[nixops] domenkozar created deployment.hasFastConnection (+1 new commit): https://git.io/v1wru
<NixOS_GitHub>
nixops/deployment.hasFastConnection cde351b Domen Kožar: Move has_really_fast_connection() to Nix...
Guest37568 is now known as scott
ldng has joined #nixos
nh2_ has quit [(Ping timeout: 258 seconds)]
Guest350 has quit [(Ping timeout: 258 seconds)]
<NixOS_GitHub>
[nixops] domenkozar force-pushed deployment.hasFastConnection from cde351b to 8042933: https://git.io/v1wrS
<NixOS_GitHub>
nixops/deployment.hasFastConnection 8042933 Domen Kožar: Move has_really_fast_connection() to Nix...
dstufft has joined #nixos
<dstufft>
Hi!
<FRidh>
dstufft: hi
<dstufft>
I'm told by simpson that y'all are looking for some sort of API on PyPI?
* simpson
communicating
<FRidh>
simpson: simpson asked whether the pypi api could be of use. I've looked into it in the past, and we can use it to auto update our packages. It doesn't include dependencies, but that's a known problem :)
<dstufft>
FRidh: so it's kind of a mess right now :/ The "info" box there is populated by the first file that was uploaded to PyPI or the last time setup.py register was ran. If the first file someone uploaded was (A) a wheel and (B) was uploaded using twine, then PyPI will know about the dependencies
<dstufft>
this isn't a 100% solution, because different wheels could have different dependencies (e.g. python specific ones) if someone isn't using environment markers for their conditional dependencies
ryanartecona has quit [(Quit: ryanartecona)]
rardiol has quit [(Read error: Connection reset by peer)]
<dstufft>
If someone uses twine and does like ``twine upload dist/*``, then twine will ensure that it uploads a wheel first (it obviously can't if someone explicitly calls twine multiple times with different files though)
<dstufft>
gchristensen: heya
<FRidh>
dstufft: okay, interesting. i guess this could be parsed with the same tools used for requirements.txt?
<dstufft>
FRidh: dunno what you're using for requirements.txt, but yes, or you can use packaging
<dstufft>
pip doesn't have a public API, but packaging does
<simpson>
FRidh: I'm trying to remember all of the problems that our current system has. Aside from deps, wasn't there something about versions that we needed too?
<FRidh>
simpson: with the api we can get urls and hashes of all versions, and also determine the latest stable version, so that part is solved.
<FRidh>
one issue that we have is that we typically want packages that are all compatible with each other, because we don't resolve versions like pip or conda
<dstufft>
FRidh: things are kind of tricky with sdists. Even beyond the current format. There are folks who really do not want to be locked into a static dependency list for sdists evne as we move away from setup.py, projects with, complex dependency requirements
<simpson>
The feature system is also a thing. And it complicates things more.
<dstufft>
so it's possibly we might never get complete dependency coverage even for new projects
<LnL>
FRidh: I created some expressions that generate a fixed output derivation using pip download
<dstufft>
My _goal_ is to make it so that porjects that need dynamic data for dependencies explicitly opt into that though
<LnL>
FRidh: I did that because we have an internal pypi server at work
<FRidh>
dstufft: could you explain further?
<dstufft>
FRidh: So right now, projects can apply logic in their setup.py to determine what their install_requires is going to be yes?
<FRidh>
yes
glines has joined #nixos
<dstufft>
a lot of projects do this, the typical use cases are things like platform differences, python version differences, etc. Those are all handled fine by environment markers now which allows them to be static again
<dstufft>
HOWEVER
<dstufft>
some projects, like things that depend on Numpy, are more complicated
<FRidh>
exactly
<dstufft>
if something depends on numpy, it wants to be able to dynamically set install_requires based upont he version of numpy it was built against
<dstufft>
for built packages like wheels etc
<dstufft>
so we're probably always going to need some sort of escape hatch that lets people have dynamic install_requires
<dstufft>
which those projects won't be able to have a sdist that says "these are my deps" on PyPI
<dstufft>
hopefully we'll get to a place those are the exception rather than the rule though.
<domenkozar>
dstufft: is there no way to avoid dynamic install_requires for numpy?
<simpson>
Definitely we want those to also be exceptions to the rule in nixpkgs.
proteusguy has joined #nixos
<dstufft>
domenkozar: You're probably asking the wrong person :) I'm a proponent of static install_requires, but the science folks are fairly adament they need it
<domenkozar>
I'd really like to hear those arguments :)
<domenkozar>
nevertheless, this could be solved with configuration syntax again
<dstufft>
domenkozar: It'll likely happen on distutils-sig when we look at finally implementing pluggable build systems again
<dstufft>
Nathaniel Smith is the one who primarily argued for it
user__ has joined #nixos
<gchristensen>
dstufft: the science folks do are fairly adamant about a lot of things :P
* gchristensen
has been jaded
<dstufft>
gchristensen: all in all, the relationship between the science folks and the rest of the packaging ecosystem has gotten a lot better lately, particularly with njs and co acting as laisons
<user__>
My system is not working properly (there was something wrong with udev during boot, and X is all garbled). There seems to be some erros on dmesg. Could someone take a look? And also tell me how to use a pastebin from the CLI?
<domenkozar>
as long as a single boolean control these 0.1% overrides, I think that's fine for packaging.
<domenkozar>
controls*
Sonarpulse has quit [(Ping timeout: 250 seconds)]
<gchristensen>
that is really great to hear, dstufft, I definitely appreciate that work -- rising tides raise all boats
k0001 has joined #nixos
<dstufft>
domenkozar: yea, I don't think I really understand fully why they can't, but I think if there's some sort of explicit opt-in to it, then it'll be fine.
<dstufft>
most projects are unlikely to do dynamic install_requires just for kicks
<dstufft>
particularly since the penality will be future pip will end up being much slower resolving deps for your project
<k0001>
I'm getting this error with Hydra, and I'm not sure how that `store_uri` is supposed to be formated: `hydra.conf: binary_cache_secret_key_file is deprecated and ignored. use store_uri=...?secret-key= instead`. Anyone can point me in the right direction? I'm not sure what to put before the `?` in that URI.
<NixOS_GitHub>
[nixops] domenkozar force-pushed deployment.hasFastConnection from 8042933 to b356e9a: https://git.io/v1wrS
<NixOS_GitHub>
nixops/deployment.hasFastConnection b356e9a Domen Kožar: Move has_really_fast_connection() to Nix...
<FRidh>
(heh science person here.) I can understand somewhat why it is an issue. Packaging pure Python is straightforward, especially now with the environment markers. But as soon as you consider extension types you likely want other libraries/compilers to be there, and then it gets messy.
alx741_ has joined #nixos
<k0001>
domenkozar: Oh my! That is fantastic! It doesn't "run" yet, though, right? Or does it?
<domenkozar>
No :)
alx741 has quit [(Ping timeout: 244 seconds)]
<domenkozar>
k0001: but I'm writing down plans how to get there
<k0001>
domenkozar: I'll take a look at it, maybe I can help with some bits here and there.
<gchristensen>
NeQuissimus: ping
<domenkozar>
k0001: that would rock
<domenkozar>
k0001: #nixos-hercules
<domenkozar>
k0001: I'm catching up learning about product-profunctors, but currently we'are at the phase where queries needs to be written
<gchristensen>
hrm. a kernel update broke the build
<k0001>
domenkozar: let me guess, you are using Opaleye?
<domenkozar>
yes, jophish coworker is the author
<FRidh>
dstufft: what kind of dependency resolving will pip get? SAT similar to conda?
<domenkozar>
k0001: but I'm interested to hear about alternatives
ckauhaus has joined #nixos
<domenkozar>
let's move the discussion to hercules channel
<k0001>
domenkozar: there aren't many :( I've been working on my own based on Opaleye here and there for a while https://github.com/k0001/tisch
<domenkozar>
oliver charles is also working on it
<domenkozar>
it's a really grey area in Haskell
<k0001>
Yeah, we're often frustrated with the same things ocharles_ and I :)
<ale-batt>
i just installed nixos on my thinkpad, but my touchpad is making me crazy, each touch on it move the cursor at the position ... any idea how to solve that ?!
<gchristensen>
I think you need to enable palm rejection
RchrdB has quit [(Quit: Leaving)]
redmq has joined #nixos
<ale-batt>
i'll check that, thank you. it's a config from nix or kde ?
<pi3r>
$nix-build default.nix -I nixpkgs=https://github.com/NixOS/nixpkgs-channels/archive/nixos-unstable.tar.gz works fine and build the project with the exec.
<pi3r>
I thought I could then have a shell.nix with just:
<pi3r>
with (import <nixpkgs> {}).pkgs;
<pi3r>
(import ./.).env
<pi3r>
but that does not bring the exec in the shell. Do I really need to create a new "env" derivation in the shell.nix ?
hiratara has quit [(Remote host closed the connection)]
oida has quit [(Quit: WeeChat 1.0.1)]
hiratara has joined #nixos
Seichi has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to release-16.09: https://git.io/v1w7d
<NixOS_GitHub>
nixpkgs/release-16.09 64e2d4d Vladimír Čunát: xorg-server: maintenance 1.18.3 -> 1.18.4...
mooho has quit [(Ping timeout: 258 seconds)]
<ale-batt>
FRidh: i can't nixos-rebuild services.xserver.synaptics.palmDetect; i have to put it in /etc/nixos/configuration.nix right ?
<ale-batt>
i found why, sorry
cpennington has joined #nixos
FRidh has quit [(Remote host closed the connection)]
<ale-batt>
gchristensen: I didn't find anything with palm rejection and synaptics, any other idea ?
<ale-batt>
with the nixos live usb i didn't had the issue
virtualmaster has joined #nixos
<virtualmaster>
hi
c0bw3b has joined #nixos
nh2_ has joined #nixos
ldng has quit [(Quit: Leaving)]
glines has quit [(Ping timeout: 250 seconds)]
virtualmaster is now known as shravan
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
shravan is now known as shra1
shra1 has left #nixos []
virtualmaster has joined #nixos
virtualmaster has left #nixos []
ale-batt_ has joined #nixos
ale-batt has quit [(Ping timeout: 260 seconds)]
ryanartecona has quit [(Quit: ryanartecona)]
jarlg has joined #nixos
ale-batt_ has quit [(Ping timeout: 260 seconds)]
cpennington has quit [(Remote host closed the connection)]
glines has joined #nixos
redmq has quit [(Ping timeout: 258 seconds)]
vandenoever has quit [(Ping timeout: 250 seconds)]
nh2_ has quit [(Ping timeout: 244 seconds)]
k0001 has quit [(Ping timeout: 246 seconds)]
anelson has quit [(Remote host closed the connection)]
anelson has joined #nixos
redmq has joined #nixos
m0rphism has joined #nixos
k0001 has joined #nixos
roconnor has quit [(Ping timeout: 268 seconds)]
oida has joined #nixos
bencevans has quit [(Quit: No Ping reply in 180 seconds.)]
Tsutsukakushi has quit [(Ping timeout: 240 seconds)]