<simpson> Anyway, it sounds like you're trying to compile stuff. You probably want to write a Nix expression.
endformationage has quit [(Client Quit)]
stepcut has joined #nixos
estewei has joined #nixos
<sophiag> simpson: sorry, i don't understand any of that. i'm trying to locate a c header file on *NixOs*
stepcut has quit [(Remote host closed the connection)]
<simpson> sophiag: Sure, but to what end? Are you compiling stuff?
<mtetreault> I am trying to enable connman but, wpa_supplicant complains as soon as I try to rebuild-switch. Is there any magic required to have wpa_supplicant working properly?.
<Profpatsch> peti: What would speak against statically compiling all executables in haskellPackages by default?
<sophiag> simpson: yes, i'm compiling stuff :)
eacameron has quit [(Remote host closed the connection)]
<sophiag> the header file is likely under a different name in this distro
<Profpatsch> The immense amount of transitive .so Files slows down invocation of most Haskell executables considerably in most cases.
<Profpatsch> e.g. purscript statically linked is instant, dynamically linked is ~4 seconds on my machine.
<Profpatsch> And that’s with an i5 and SSD drive.
AllanEspinosa has quit [(Ping timeout: 260 seconds)]
<simpson> sophiag: In glibc-2.23-dev, I have include/sys/msg.h, but that might not be the right one.
<simpson> sophiag: NixOS doesn't do anything special with the header names. It's all straight from glibc, unless I missed something.
stepcut has joined #nixos
stepcut has quit [(Remote host closed the connection)]
[0x4A6F] has quit [(Ping timeout: 260 seconds)]
stepcut has joined #nixos
stepcut has quit [(Remote host closed the connection)]
<clever> sophiag: what package is msg.h usualy in?
[0x4A6F] has joined #nixos
<sophiag> clever: i think either msg or sys
<clever> sophiag: by default, locate will ignore /nix/store, so that would never have worked the way you wanted
eacameron has joined #nixos
<sophiag> simpson: i should maybe just see if gcc can find sys/msg.h. msg/msg.h sounded odd to me from the start
<simpson> sophiag: Best of luck.
cheecheeo has quit [(Ping timeout: 260 seconds)]
Supersonic112 has joined #nixos
elninja44 has joined #nixos
<clever> sophiag: its in glibc, sys/msg.h
<clever> sophiag: and glibc is already in the buildInputs
Itkovian has quit [(Ping timeout: 258 seconds)]
<sophiag> welp that worked. much ado about nothing. of course tons of other errors related to the library i'm going through a turoial on
<sophiag> thanks guys
<simpson> Peace.
eacameron has quit [(Ping timeout: 252 seconds)]
<Ralith> argh, nixos 17.03 still hasn't fixed -O2 being forced on in nix-shell? :/
<Ralith> I thought that was a gimme
mtetreault has quit [(Ping timeout: 240 seconds)]
byteflame has joined #nixos
mtetreault has joined #nixos
<mtetreault> ls
<clever> . .. passwords.txt
<mtetreault> ... wrong window...
<mtetreault> sorry about that!
<mtetreault> At least it wasn't a password :)
<clever> yeah
sigmundv_ has quit [(Ping timeout: 252 seconds)]
<mtetreault> I was wondering, is nixos supposed to detect all the hardware automagically?
stepcut has joined #nixos
byteflame has quit [(Ping timeout: 268 seconds)]
<mtetreault> (I am still trying to figure out a few things)
takle has quit [(Remote host closed the connection)]
stepcut has quit [(Remote host closed the connection)]
vaibhavsagar has joined #nixos
galen has joined #nixos
eacameron has joined #nixos
Sonarpulse has joined #nixos
ryantrinkle has quit [(Ping timeout: 252 seconds)]
stepcut has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
darlan has quit [(Quit: Communi 3.5.0 - http://communi.github.com)]
ebzzry has joined #nixos
eacameron has quit [(Ping timeout: 258 seconds)]
indi_ has joined #nixos
mekeor has quit [(Ping timeout: 258 seconds)]
stepcut has quit [(Remote host closed the connection)]
mtetreault has quit [(Quit: Leaving)]
endformationage has joined #nixos
indi_ has quit [(Ping timeout: 252 seconds)]
takle has joined #nixos
AllanEspinosa has joined #nixos
takle has quit [(Ping timeout: 260 seconds)]
matthewbauer has joined #nixos
ebzzry has quit [(Ping timeout: 240 seconds)]
takle has joined #nixos
mekeor has joined #nixos
takle has quit [(Ping timeout: 258 seconds)]
aneeshusa has joined #nixos
mekeor has quit [(Ping timeout: 258 seconds)]
byteflame has joined #nixos
takle has joined #nixos
<matthewbauer> hey does anybody know what's up with the hydra build logs?
<dtzWill> you scared them away, we're trying to lure them back with log treats shhh
<dtzWill> and appearing non-threatening
byteflame has quit [(Ping timeout: 240 seconds)]
<clever> 2017-03-20 12:45:56 < niksnut> if you urgently need a log, you can find them at URIs like: https://cache.nixos.org/log/l9qmwi2q0dk4ji8pcycc188gank0q5pb-pointedalternative-0.1.0.0.drv
<clever> matthewbauer: they are on the binary cache now, and hydra isnt aware of that
<copumpkin> I thought it was aware of it now
mbrgm has quit [(Ping timeout: 240 seconds)]
<clever> a month ago it wasnt, havent heard any update yet
<matthewbauer> yeah it was broken then it worked again and not it looks like it's broken
takle has quit [(Ping timeout: 260 seconds)]
mbrgm has joined #nixos
<matthewbauer> but maybe it hasn't been deployed yet?
<Profpatsch> Whoa
<Profpatsch> > strace -e trace=file result/bin/hoogle 2>&1 | grep "open.*so.*= 3.*" | wc -l
<Profpatsch> 128
<Profpatsch> strace -e trace=file result/bin/hoogle 2>&1 | grep "open.*so.*= -1.*" | wc -l
<Profpatsch> 8829
<Profpatsch> Something’s horribly broken with haskell executables.
ebzzry has joined #nixos
<Profpatsch> I was wondering why executing them takes so long if they are dynamically linked.
<Profpatsch> That would explain that.
takle has joined #nixos
<Profpatsch> That’s 8829 stats too many and probably the reason why hoogle needs 2 secs to start.
joneshf-laptop has joined #nixos
* dtzWill seems to remember a certain recent RFC/issue about this
<copumpkin> yeah there was an issue about using rpath vs. absolute paths
<Profpatsch> Same thing for purescript: > strace -e trace=file result/bin/purs 2>&1 | grep "open.*so.*= -1.*" | wc -l
<Profpatsch> 14615
<dtzWill> but yeah taht's really impressive re:counts lol
<Profpatsch> That’s 4+s upstart time.
<dtzWill> @.@
<Profpatsch> Try haskellPackages.purescript
<Profpatsch> And compare to just purescript (statically linked)
<dtzWill> call me crazy but ... doesn't it seem strange that 14k stat's takes 4s? I mean...
<Profpatsch> That’s disk IO.
<copumpkin> Profpatsch: maybe add data to https://github.com/NixOS/nixpkgs/issues/24844
<Profpatsch> I’m already on an SSD
<Profpatsch> It’s strange that they *only* take 4s
<Profpatsch> dtzWill: you also have to factor in that that’s a tree of .sos
<dtzWill> hahaha fair enough. And yeah my new laptop's SSD ruined me for everything, and of course even then it has limits.
<Profpatsch> So they form a dependency tree.
<Profpatsch> And are partially sequential.
<Profpatsch> And who knows how ld.so works.
<Profpatsch> Maybe they don’t parallelize anything.
<dtzWill> I'm just gonna say, part of my current phd work is showing you can make a useful system without this insanity lol
rardiol has joined #nixos
neeasade has joined #nixos
<Profpatsch> copumpkin: Wait, you are saying this isn’t just haskelPackages?
<dtzWill> how about we use search semantics few understand and symbol resolution basically no one understands, just to implement normal code deployment? what could possibly go wrong? :P
<Profpatsch> copumpkin: I was of the impression that putting static paths in our .so RPATGH
<Profpatsch> RPATHs would mean no search happens?
takle has quit [(Ping timeout: 260 seconds)]
<copumpkin> rpaths are why we have the searches
<Profpatsch> copumpkin: When resolving shared object dependencies, the dynamic linker first inspects each dependency string to see if it contains a slash (this can occur if a shared object
<copumpkin> and yes I think haskell packages just suffer the most from it because they link to tons of stuff
<Profpatsch> pathname containing slashes was specified at link time). If a slash is found, then the dependency string is interpreted as a (relative or absolute) pathname, and
<Profpatsch> the shared object is loaded using that pathname.
<Profpatsch> man 8 ld.so
<copumpkin> Profpatsch: FWIW on darwin we do almost entirely absolute linkage
<copumpkin> but the linker works a little differently there
<copumpkin> anyway, I'd urge you to chime in and tell them how much it hurts :)
<Profpatsch> copumpkin: Oh, I see.
<dtzWill> tbh I don't see why absolute linkage isn't the sane default
<Profpatsch> libHStime-1.6.0.1-ghc8.0.2.so => /nix/store/v7cp0hn5c912l5q10pgwfzhwyc1i70kg-ghc-8.0.2/lib/ghc-8.0.2/time-1.6.0.1/libHStime-1.6.0.1-ghc8.0.2.so (0x00007fd362976000)
<Profpatsch> That means the RPATH is set as libHStime-1.6.0.1-ghc8.0.2.so so searching happens.
<dtzWill> with good support for letting you change it if you'd like, but almost zero reason to ever discover at startup what code you're supposed to use
<Profpatsch> dtzWill: Of course, it’s the only sane default on nixpkgs.
<Profpatsch> I’m kind of hurt that this isn’t the default already. :(
nckx has quit [(Quit: restarting my ZNC)]
nckx has joined #nixos
<dtzWill> ya, in a way it almost is--I think it's mostly supposed to be the effect, in that we're careful about constructing rpath's that only contain the paths of the exact lib deps indicated
<dtzWill> but I guess per that issue we don't do it in a way that prevents lots of searching anyway
dridus has joined #nixos
roygbiv has joined #nixos
<dtzWill> on the subject, since y'all seem informed, is there a short/simple reason why we (nixpkgs) prefer rpath over runpath? Not saying it's a mistake but AFAICT it's unusual enough that it seems intentional.
takle has joined #nixos
mkoenig has quit [(Read error: Connection reset by peer)]
<Profpatsch> copumpkin: How do you do it on darwin, using absolute links?
<Profpatsch> Is the stdenv different?
derjohn_mob has quit [(Ping timeout: 258 seconds)]
takle has quit [(Ping timeout: 260 seconds)]
matthewbauer has quit [(Ping timeout: 255 seconds)]
MP2E has joined #nixos
Supersonic112 has quit [(Disconnected by services)]
Supersonic112_ has joined #nixos
Supersonic112_ is now known as Supersonic112
<dtzWill> everything's different, lol
spinus has quit [(Ping timeout: 240 seconds)]
imalsogreg has left #nixos []
takle has joined #nixos
mkoenig has joined #nixos
matthewbauer has joined #nixos
nh2 has quit [(Quit: Leaving.)]
takle has quit [(Ping timeout: 252 seconds)]
<dtzWill> lol, no one wants to answer my rpath/runpath question qq
<Profpatsch> dtzWill: rpath and runpath link to the same WP-article
derjohn_mob has joined #nixos
<Profpatsch> dtzWill: man 8 ld.so
<Profpatsch> The (colon-separated) paths in the DT_RPATH dynamic section attribute of the binary if present and the DT_RUNPATH attribute does not exist.
<Profpatsch> DT_RPATH is deprecated …
<dtzWill> "Use of DT_RPATH is deprecated"
<Profpatsch> The distinction is that DT_RPATH takes precedence over the LD_LIBRARY_PATH environment variable, whereas DT_RUNPATH does not. The latter is more convenient and means that rpaths are less likely to get in your way, since you can override them.
<dtzWill> also LD_LIBRARY_PATH overrides one and not the otehr
<Profpatsch> There you have it.
<dtzWill> also maybe I'm remembering which we use and don't
<Profpatsch> You don’t want people to use dynamic overriding with nix.
Fare has quit [(Remote host closed the connection)]
<dtzWill> well but that also means setting LD_LIBRARY_PATH breaks all the things
<dtzWill> okay
<Profpatsch> You want it to be fixed and changed in the sources.
<dtzWill> I suppose I agree with that pretty strongly
<hodapp> huh. now that I have two slow NixOS machines on my network, I should really figure out how to do that remote-builds thingy
<Profpatsch> So, LD_LIBRARY_PATH should not take preference.
<dtzWill> only times I seen people have problems are reluctant converts to Nix that run on some Linux where they have all kidns of things in their LD_LIBRARY_PATH
<dtzWill> and they have to unset LD_LIBRARY_PATH to use anything nix related
<hodapp> does this facility take advantage of sending binary packages across the network too if another box already has them (as it's faster to get them from the LAN than from the Internet)?
<dtzWill> Profpatsch: nix sets runpath not rpath, my mistake. Anyway okay I'm satisfied with the answer, it's just unfortunate that it's ahrd to find someone's env that doens't have LD_LIBRARY_PATH set to somethign or other by the time I convince them Nix is worth their time xD
matthewbauer has quit [(Ping timeout: 260 seconds)]
<Profpatsch> Ah, I understand now!
pmahoney has joined #nixos
<Profpatsch> The problem with rpath.
<Profpatsch> So you have one rpath entry for each haskell Dependency.
eacameron has joined #nixos
<Profpatsch> And it tries to find foobar.so
pie__ has quit [(Ping timeout: 260 seconds)]
<dtzWill> and looks through all the things
<Profpatsch> It will linearly stat *every* rpath entry until it finds the so.
<Profpatsch> which means n^2
<Profpatsch> Or probably (n^2)/2 in the mean case.
<hodapp> Still O(n^2) :P
<Profpatsch> purs has 168 deps
<Profpatsch> -5 for the normal .so’s the Haskell runtime needs
<Profpatsch> 163^2 = 26.5k
<Profpatsch> /2 and we arrive at the 13k misses I measured before.
<Profpatsch> Science!
takle has joined #nixos
<dtzWill> ٩(^ᴗ^)۶
matthewbauer has joined #nixos
<hodapp> hah
eacamero_ has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
<pmahoney> i'm trying to package osquery (c++, cmake) on a non-nixos linux. there's a dependency chain of osquery -> libtsk -> libafflib -> libcurl. running 'ldd' on any of the intermediate shared libs shows libcurl.so.4 => /nix/store/... but running 'ldd' against my osquery binary shows libcurl.so.4 => /usr/lib/...
<pmahoney> and my binary fails to start with: error while loading shared libraries: libcurl.so.4: cannot open shared object file: No such file or directory
<pmahoney> anything in particular to watch out for?
takle has quit [(Ping timeout: 260 seconds)]
Itkovian has joined #nixos
eacamero_ has quit [(Ping timeout: 255 seconds)]
alunduil has joined #nixos
takle has joined #nixos
<alunduil> Does anyone know a way to generate UUIDs in nixops or nixos configurations?
<Profpatsch> alunduil: nixos configurations are declarative.
<Profpatsch> You can’t have randomness.
<Profpatsch> s/declarative/purely functional/
<Profpatsch> alunduil: What’s your use-case?
<NixOS_GitHub> [nixpkgs] Ralith opened pull request #24888: electron: 1.4.15 -> 1.6.2 (master...electron) https://git.io/vSHzB
calvinx has joined #nixos
<alunduil> Profpatsch, thinking of ways to generate uniform but distinct hostnames for nixops hosts in a cluster where the number of hosts is a nixops paramter.
takle has quit [(Ping timeout: 268 seconds)]
<Profpatsch> alunduil: host-1 host-2 host-3
byteflame has joined #nixos
<alunduil> Profpatsch, yeah, I prefer UUIDs when I can but that might have to do.
<Profpatsch> alunduil: why?
<alunduil> Profpatsch, do you know of a way to at least pad the numbers?
<Profpatsch> why?
<alunduil> Profpatsch, mostly an OCD aesthetic. No technical reason.
<Profpatsch> Sure there is a way, but why?
<alunduil> Profpatsch, mostly curiosity and to see what that solution would look like.
<Profpatsch> Somethingsomething list functions
<dtzWill> pmahoney: check rpath/runpath of the osquery binary? readelf helps
takle has joined #nixos
<Profpatsch> That might help.
<Profpatsch> take is in the library already.
<Profpatsch> Questionable still. :)
justin_ has quit [(Ping timeout: 255 seconds)]
acarrico has quit [(Ping timeout: 240 seconds)]
byteflame has quit [(Ping timeout: 258 seconds)]
<alunduil> Profpatsch, perhaps. Is take in builtins or lib?
<Profpatsch> lib/lists.nix
<pmahoney> dtzWill: thanks. not sure i'm checking for rpath correctly (seems to be absent?), but runpath is missing libcurl (using readelf -a result/bin/osqueryi | grep -iE 'RUNPATH|RPATH')
<alunduil> Profpatsch, thanks.
takle has quit [(Ping timeout: 258 seconds)]
zeus_ has joined #nixos
<pmahoney> dtzWill: i guess i'm not familiar with how transitive share lib deps are handled in nix, or interact with runpath, etc. reading up while i try a new build explictly adding '-lcurl'
indi_ has joined #nixos
takle has joined #nixos
Itkovian has quit [(Ping timeout: 258 seconds)]
matthewbauer has quit [(Remote host closed the connection)]
onethirtyfive has joined #nixos
takle has quit [(Ping timeout: 260 seconds)]
ryantrinkle has joined #nixos
indi_ has quit [(Ping timeout: 260 seconds)]
<onethirtyfive> Hello! My MacBook Pro running 17.0 won't wake up from suspend. I'm not even sure it finishes suspending. Does anyone know where I could go for troubleshooting info?
<onethirtyfive> *17.03
Guest60 has quit [(Quit: My Mac Pro has gone to sleep. ZZZzzz…)]
<onethirtyfive> I do have the 'resume' option set in modprobe configuration to my swap partition. I'm thinking it might be some system service or something. Any ideas?
alx741 has quit [(Quit: alx741)]
mguentner3 has quit [(Quit: WeeChat 1.7)]
aneeshusa has quit [(Ping timeout: 260 seconds)]
takle has joined #nixos
hexagoxel has quit [(Ping timeout: 260 seconds)]
sophiag has left #nixos ["ERC (IRC client for Emacs 24.5.2)"]
mguentner has joined #nixos
takle has quit [(Ping timeout: 252 seconds)]
galen has quit [(Ping timeout: 240 seconds)]
hexagoxel has joined #nixos
alunduil has quit [(Quit: leaving)]
takle has joined #nixos
takle has quit [(Ping timeout: 252 seconds)]
mguentner2 has joined #nixos
mguentner has quit [(Ping timeout: 240 seconds)]
darlan has joined #nixos
byteflame has joined #nixos
schoppenhauer has quit [(Ping timeout: 268 seconds)]
byteflame has quit [(Ping timeout: 260 seconds)]
siel has quit [(Ping timeout: 246 seconds)]
schoppenhauer has joined #nixos
siel has joined #nixos
<pmahoney> dtzWill: adding -lcurl did the trick. now to investigate why that didn't happen automatically
eacameron has joined #nixos
mguentner2 has quit [(Ping timeout: 260 seconds)]
<dtzWill> ha, yes indeed
hamishmack has quit [(Quit: hamishmack)]
eacameron has quit [(Ping timeout: 240 seconds)]
roygbiv has quit [(Quit: ™)]
roygbiv has joined #nixos
matthewbauer has joined #nixos
takle has joined #nixos
matthewbauer has quit [(Remote host closed the connection)]
roygbiv has quit [(Client Quit)]
eacameron has joined #nixos
siel has quit [(Ping timeout: 258 seconds)]
mizu_no_oto has joined #nixos
Itkovian has joined #nixos
takle has quit [(Ping timeout: 260 seconds)]
mguentner2 has joined #nixos
matthewbauer has joined #nixos
systemfault has joined #nixos
siel has joined #nixos
siel has quit [(Changing host)]
siel has joined #nixos
alx741 has joined #nixos
vaibhavsagar has quit [(Ping timeout: 260 seconds)]
AllanEspinosa has quit [(Ping timeout: 258 seconds)]
pmn has quit [(Remote host closed the connection)]
alx741 has quit [(Quit: alx741)]
pmn has joined #nixos
contrapumpkin has quit [(Quit: My MacBook Pro has gone to sleep. ZZZzzz…)]
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
Itkovian has quit [(Ping timeout: 252 seconds)]
takle has joined #nixos
contrapumpkin has joined #nixos
takle has quit [(Ping timeout: 260 seconds)]
byteflame has joined #nixos
<unlmtd[m]> what could be keeping my bluetooth headphones from showing up in pavucontrol? Iv switched to pulseaudioFull and they connected OK
justbeingglad has joined #nixos
justbeingglad has left #nixos []
<unlmtd[m]> I have this ``` [pulseaudio] bluez4-util.c: org.bluez.Manager.GetProperties() failed: org.freedesktop.DBus.Error.UnknownMethod: Method "GetProperties" with signature "" on interface "org.bluez.Manager" doesn't exist ```
takle has joined #nixos
byteflame has quit [(Ping timeout: 258 seconds)]
<unlmtd[m]> ahh got it
takle has quit [(Ping timeout: 260 seconds)]
systemfault has quit [(Quit: Bye!)]
edvorg has joined #nixos
stepcut has joined #nixos
takle has joined #nixos
stepcut has quit [(Remote host closed the connection)]
takle has quit [(Ping timeout: 240 seconds)]
stepcut has joined #nixos
hamishmack has joined #nixos
zeus__ has joined #nixos
indi_ has joined #nixos
zeus_ has quit [(Ping timeout: 258 seconds)]
<dash> unlmtd[m]: what was it :)
indi_ has quit [(Ping timeout: 252 seconds)]
takle has joined #nixos
seagreen has joined #nixos
<unlmtd[m]> dash: had to 'authorize service' in bluetoothctl. I also 'trusted' the device
matthewbauer has quit [(Read error: Connection reset by peer)]
<unlmtd[m]> some bluetooth voodoo
matthewbauer has joined #nixos
takle has quit [(Ping timeout: 268 seconds)]
rardiol has quit [(Remote host closed the connection)]
<dash> ah yeah, makes sense.
systemfault has joined #nixos
takle has joined #nixos
oahong has quit [(Quit: ZNC 1.6.4 - http://znc.in)]
takle has quit [(Ping timeout: 260 seconds)]
<hyphon81> Did somebody install NixOS in ZFS? I failed it...
<unlmtd[m]> ya you have to use legacy mounts
zeus__ has quit [(Remote host closed the connection)]
<hyphon81> I made zpool when install phase. However, NixOS says zpool not found when booting.
zeus_ has joined #nixos
Itkovian has joined #nixos
<unlmtd[m]> ohh ya theres the hostID
<unlmtd[m]> it mounts only on the same hostID, so you have to make sure its set to whatever it was during install
<unlmtd[m]> I tried to do a zfs install where the mounting was handled by zfs but I was told its not possible with nixos atm. so you need legacy mounts
<hyphon81> Yep. I set hostID and legacy mounts. I could nixos-install it success. The problem is booting.
<unlmtd[m]> hmm
zeus_ has quit [(Ping timeout: 258 seconds)]
<hyphon81> I could install on ZFS when NixOS 16.09 with uefi. However, 17.03 with MBR is failed because zpool is not found when it is booting process.
dridus has quit [(Quit: znc? wassat?)]
athan has joined #nixos
<hyphon81> Dose ZFS need UEFI boot? I don't think so. But I don't know detail about ZFS.
<c74d> it should work with MBR, but I've never managed it
takle has joined #nixos
<unlmtd[m]> maybe you need to specify zfs support for grub
<unlmtd[m]> I use uefi
<unlmtd[m]> not much to look at
<hyphon81> I refered it. https://nixos.org/wiki/ZFS_on_NixOS
<hyphon81> unlmtd[m]: Thanks!
byteflame has joined #nixos
<unlmtd[m]> maybe the boot.supportedFilesystems
takle has quit [(Ping timeout: 252 seconds)]
stepcut has quit [(Remote host closed the connection)]
byteflame has quit [(Ping timeout: 260 seconds)]
<unlmtd[m]> dash: did you enable boot.loader.grub.zfsSupport ?
galen has joined #nixos
<unlmtd[m]> that would be needed for mbr boot
stepcut has joined #nixos
<hyphon81> Oh, I didn't enable it. Thanks!
<unlmtd[m]> that should help
filterfish__ has joined #nixos
stepcut has quit [(Ping timeout: 260 seconds)]
filterfish_ has quit [(Remote host closed the connection)]
<NixOS_GitHub> [nixpkgs] dmjio opened pull request #24889: Remove hardening at HaLVM 2.4.0 (master...halvm-bump) https://git.io/vSHK2
marsel has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
indi_ has joined #nixos
indi_ has quit [(Ping timeout: 252 seconds)]
matthewbauer has quit [(Remote host closed the connection)]
ambro718 has joined #nixos
marsel has quit [(Ping timeout: 260 seconds)]
<hyphon81> Hmm, I have not done boot ZFS with MBR NixOS 17.03.
<hyphon81> It stop when it is booting phase "NixOS Stage 1".
systemfault has quit [(Quit: Bye!)]
fractal has quit [(Ping timeout: 264 seconds)]
freusque has joined #nixos
<hyphon81> I attempted to enable grub.zfsSupport and to set kernelParams "zfs_force=1".
<hyphon81> I saw "cannot import 'rpool': no such pool available" when booting. But, I can import it when I'm running installing ISO.
<clever> hyphon81: ive only ever done zfs booting with an ext4 /boot partition, and it just works for me
<calvertvl> hyphon81: you may need boot.initrd.availableKernelModules = [ "zfs" ] and boot.kernelModules = [ "zfs" ]
<calvertvl> I have grub boot from ZFS working with (IIRC) legacy BIOS
<clever> calvertvl: nixos will juat automaticaly include zfs in the initrd if you set a filesystem type to zfs
<c74d> clever: so that's one of the bits of obsolescence in the wiki?
<calvertvl> clever: I'd had some trouble with it at some point, and adding those fixed it; I agree it shouldn't be needed
<clever> fileSystems = {
<clever> "/" = { device = "amd/root"; fsType = "zfs"; };
<clever> this is enough for nixos to include zfs in the initrd, and find the pool on bootup
<unlmtd[m]> I concur, my setup work that way
<clever> the above even works if the zfs is on lvm, and the lvm is on luks
<clever> as long as the luks is also configured correctly
<unlmtd[m]> zfs on lvm?
<unlmtd[m]> lol
<clever> i wanted swap on the luks
byteflame has joined #nixos
<clever> but i know swap on zfs is a problem
<unlmtd[m]> I see
<clever> so i used lvm to split the luks between zfs and swap
<unlmtd[m]> I think it only cant hibernate
<unlmtd[m]> which I never use
<clever> thats a problem when the swap has a new key every time it boots
<clever> i'm using the same key to cover both swap and root, so in theory, it should be able to hibernate
<clever> but it had an error last time i tried, havent debugged it fully yet
<unlmtd[m]> have you tried zfs encryption yet?
<clever> not yet
<unlmtd[m]> I want to give it a try, just too busy with other things
snikkers has quit [(Ping timeout: 268 seconds)]
<c74d> ZFS encryption was reimplemented in OpenZFS?
<clever> calvertvl: do you know if the import error is coming from nixos or from grub?
<unlmtd[m]> id just encrypt home and swap
<unlmtd[m]> hyphon81 is having the troubles
Itkovian has joined #nixos
<unlmtd[m]> he said he got to stage 1
<clever> calvertvl: can you gist the configuration.nix and hardware-configuration.nix ?
<calvertvl> sounds like it's from NixOS or the initrd
<clever> calvertvl: and also the output of "zpool history"
<hyphon81> OK. I will paste my configs. Please wait.
byteflame has quit [(Ping timeout: 258 seconds)]
<Mic92> zfs encryption is not merged yet
<Mic92> calvertvl: do you have a dedicated not-zfs boot partitition?
<calvertvl> Mic92: no, boot is on zfs for me
<calvertvl> it's been working just fine since install back in December, unless I do something silly
<calvertvl> (like nix-collect-garbage -d without first testing boot on the latest generation)
eacameron has quit [(Remote host closed the connection)]
<Mic92> calvertvl: ok, you have already passed stage-1, which means your bootloader is fine booting.
<Mic92> zpool
davidak has quit [(Quit: Leaving.)]
<Mic92> calvertvl: can you enter the rescue shell from stage-1?
<calvertvl> Mic92: I've had mine fine - I'm currently logged into IRC from that system :P
<calvertvl> hyphon81 is having problems getting past stage-1
<clever> ah
<clever> got the names mixed up
galen has quit [(Quit: WeeChat 1.7)]
<clever> hyphon81: can you gist the configuration.nix and hardware-configuration.nix and also the output of "zpool history"
<Mic92> Looks like we have a lot of zfs users on nixos
ambro718 has quit [(Ping timeout: 255 seconds)]
<hyphon81> Thanks for many replying. It's my configs and zpool history. https://gist.github.com/hyphon81/0346a5e170722825795fc5ba5d6d0e64
pmn has quit [(Remote host closed the connection)]
<clever> hyphon81: at a glance it should work, but i suspect the extra / in rpool/root/nixos may confuse some scripts
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<clever> boot.loader.grub.zfsSupport isnt required, because your /boot is ext4
<clever> and boot.supportedFilesystems = [ "zfs" ]; isnt required because nixos just figures that out via fileSystems."/" on its own
eacameron has joined #nixos
<clever> hyphon81: and the docs say that the zfs_force=1 should only be done once
<hyphon81> Yep. I set "zfs_force=1" to attempt. I wish this effects that NixOS can't find rpool.
<hyphon81> But It didn't effect.
<clever> are you able to take a screenshot of the error?
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
hyphon81 has quit [(Ping timeout: 260 seconds)]
eacameron has joined #nixos
takle has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
<NixOS_GitHub> [nixpkgs] Mic92 pushed 2 new commits to master: https://git.io/vSH1U
<NixOS_GitHub> nixpkgs/master 798e358 Yann Hodique: cayley: init at 0.6.1
<NixOS_GitHub> nixpkgs/master 3f7af64 Jörg Thalheim: Merge pull request #24885 from sigma/pr/cayley...
Itkovian has joined #nixos
<joko_> Hello, is there a way to append init=/bin/bash in NixOS? I have an old VM whose root password I forgot...
<clever> joko_: init=/nix/var/nix/profiles/system/sw/bin/bash
takle has quit [(Ping timeout: 258 seconds)]
<MichaelRaskin> I would mount the root, chroot and run password
<clever> joko_: init=/bin/sh would also work, since the symlink probably still exists
teknico has joined #nixos
hyphon81 has joined #nixos
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
<joko_> clever and MichaelRaskin: I was under the impression that stage 2 is using init already
<clever> joko_: stage1 should obey init=
<clever> but /bin/bash doesnt exist on nixos
<MichaelRaskin> Or boot the VM from minimal ISO, mount/chroot/password
<joko_> MichaelRaskin: unfortunately it's a VM on a cloud service and it's not possible to mount ISOs :(
<clever> i think init=/bin/sh is the fastest way
<clever> try that first and see what happens
<joko_> Btrfs partition and disk went full >_<
<hyphon81> Sorry. My PrtScr button had wireless on/off function... I captured screenshot. where do I paste it if everyone looks here.
<hyphon81> ?
<clever> hyphon81: imgur.com is one place you can upload images
eacameron has quit [(Ping timeout: 252 seconds)]
eacameron has joined #nixos
takle has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
snikkers has joined #nixos
<hyphon81> I forgot to say. I running it VPS with KVM.
<clever> just the fact that its not a camera reveals that
<clever> though i have gotten photos of monitors back from "professional" datacenters...
marsel has joined #nixos
<clever> hyphon81: can you use e to edit the kernel params at grub, and add "boot.shell_on_fail" to the end, then boot the modified config with f10?
Itkovian has joined #nixos
eacamero_ has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
<clever> hyphon81: i suspect the problem is boot.zfs.devNodes
marsel has quit [(Client Quit)]
marsel has joined #nixos
takle has quit [(Ping timeout: 260 seconds)]
<clever> hyphon81: the nixos scripts assume your zfs devices will be within /dev/disk/by-id, but they might not be
pmn has joined #nixos
takle has joined #nixos
eacamero_ has quit [(Ping timeout: 258 seconds)]
endformationage has quit [(Quit: WeeChat 1.7)]
<hyphon81> clever: OK, I start shell with boot.shell_on_fail.
<clever> now check inside /dev/disk/by-id with ls
nico202 has joined #nixos
<hyphon81> Ah, I see only ata-QEMU_DVD-ROM_QM00001.
<clever> yeah, thats probably it then
<clever> try setting boot.zfs.devNodes = "/dev"; in configuration.nix and then re-run nixos-install
<clever> it may also be possible to force it to boot if you dont have the cd anymore
takle has quit [(Ping timeout: 260 seconds)]
<clever> and you can rerun nixos-install without any data loss (makes it much much faster), just mount everything back to the right spot under /mnt, edit the config, nixos-install
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<hyphon81> Yes. There are /dev/vda but there are not in /dev/disk/by-id. I attempt right away. Thanks!
eacameron has joined #nixos
Filystyn has joined #nixos
byteflame has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
Itkovian has joined #nixos
eacameron has joined #nixos
<NixOS_GitHub> [nixpkgs] Mic92 pushed 2 new commits to master: https://git.io/vSHMS
<NixOS_GitHub> nixpkgs/master 77c8470 Benjamin Staffin: latte-dock: 0.5.98 -> 0.6.0
<NixOS_GitHub> nixpkgs/master 3085e95 Jörg Thalheim: Merge pull request #24886 from benley/latte-dock-0.6...
fractal has joined #nixos
byteflame has quit [(Ping timeout: 255 seconds)]
snikkers has quit [(Ping timeout: 240 seconds)]
eacamero_ has joined #nixos
eacameron has quit [(Ping timeout: 258 seconds)]
nico202 has quit [(Ping timeout: 255 seconds)]
takle has joined #nixos
aminechikhaoui has joined #nixos
eacamero_ has quit [(Ping timeout: 258 seconds)]
eacameron has joined #nixos
<hyphon81> clever: OK! I did it! NixOS booted on ZFS in my environment. Many thanks for you!
takle has quit [(Ping timeout: 260 seconds)]
<clever> hyphon81: your welcome :)
<benley> woot zfs
<hyphon81> Thank you for many replying. My problem was solved. :)
filterfish__ has quit [(Ping timeout: 258 seconds)]
<clever> i think i have zfs on 5 systems right now
onethirtyfive has quit [(Remote host closed the connection)]
nschoe has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
thc202 has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
* clever heads to bed
eacameron has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
MP2E has quit [(Quit: leaving)]
filterfish has joined #nixos
indi_ has joined #nixos
takle has joined #nixos
gattler has joined #nixos
oahong has joined #nixos
oahong has quit [(Changing host)]
oahong has joined #nixos
<gattler> Hi all, Is there any way to install nixos from another distro? like i.e. deboostrap
<gattler> domenkozar: thx, but that's somewhat different from debootstrap
<gattler> I'm on gentoo and I
<gattler> I'd like to install nixos in a folder
nico202 has joined #nixos
takle has quit [(Ping timeout: 252 seconds)]
indi_ has quit [(Ping timeout: 240 seconds)]
<c74d> gattler: I think that's `nixos-install --root <directory into which to install>`
<c74d> however, I'd be surprised if you could then just chroot to that directory and have the NixOS installation work
<gattler> c74d: why not? ( I want to boot it later anyways)
<c74d> it may need things at the init system level and/or kernel level
<gattler> so I can't use nixos with a vanilla upstream kernel?
<c74d> I don't know
<c74d> domenkozar might know, but if not, I don't think anyone awake does
<c74d> I shouldn't be awake either — farewell!
<domenkozar> since it works in lxc it should work
<gattler> :P I'll just linger a little longer and if I get bored I'll try the usual way
MichaelRaskin has quit [(Quit: MichaelRaskin)]
eacameron has joined #nixos
<Filystyn> hello
<Filystyn> I need posix copmress how to get it on nix?
DutchWolfie has joined #nixos
<Filystyn> compress* the module is not installed it's part of XSI not sure how to install it
civodul has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
<symphorien> nix-shell --pure -p gcc45 --run "touch a.c; gcc -c a.c"
<symphorien> cc1: error: unrecognized command line option "-fstack-protector-strong"
<symphorien> is there something I missed ?
takle has joined #nixos
pie__ has joined #nixos
byteflame has joined #nixos
pie__ has quit [(Read error: Connection reset by peer)]
pie__ has joined #nixos
takle has quit [(Ping timeout: 240 seconds)]
nico202 has quit [(Remote host closed the connection)]
byteflame has quit [(Ping timeout: 260 seconds)]
aminechikhaoui has quit [(Ping timeout: 260 seconds)]
aminechikhaoui has joined #nixos
<NixOS_GitHub> [nixpkgs] vcunat pushed 2 new commits to master: https://git.io/vSH9l
<NixOS_GitHub> nixpkgs/master 304d30c Vladimír Čunát: all-tarballs: fixup evaluation (and the tarball job)
<NixOS_GitHub> nixpkgs/master 2444451 Vladimír Čunát: Merge branch 'staging'
gattler has quit [(Quit: WeeChat 1.7)]
takle has joined #nixos
snikkers has joined #nixos
<NixOS_GitHub> [nixpkgs] sigma opened pull request #24891: hugo: 0.20 -> 0.20.1 (master...pr/hugo) https://git.io/vSH9i
Wizek_ has joined #nixos
takle has quit [(Ping timeout: 252 seconds)]
<NixOS_GitHub> [nixpkgs] vcunat fast-forwarded staging from 42fd720 to 304d30c: https://git.io/vSHHI
guillaum2 has quit [(Ping timeout: 240 seconds)]
takle has joined #nixos
Filystyn has quit [(Ping timeout: 252 seconds)]
eacameron has joined #nixos
takle has quit [(Ping timeout: 252 seconds)]
pie__ has quit [(Ping timeout: 240 seconds)]
eacameron has quit [(Ping timeout: 260 seconds)]
Filystyn has joined #nixos
Filystyn has quit [(Changing host)]
Filystyn has joined #nixos
bennofs has joined #nixos
<NixOS_GitHub> [nixpkgs] vcunat pushed 1 new commit to master: https://git.io/vSH7S
<NixOS_GitHub> nixpkgs/master da20d0e Vladimír Čunát: murmur service: fix typos from #24830
pie_ has joined #nixos
Avogadro has joined #nixos
takle has joined #nixos
eacameron has joined #nixos
filterfish has quit [(Ping timeout: 240 seconds)]
gattler_ has joined #nixos
pie_ has quit [(Ping timeout: 240 seconds)]
<gattler_> hi all, I have sucessfully installed nixos on a btrfs subvol, now I want to use nixos with my own kernel and initramfs, therfore I did not install grub. Can anyone give me a hint which kernel parameters should be used? exspecially the init= variable
takle has quit [(Ping timeout: 255 seconds)]
eacameron has quit [(Ping timeout: 255 seconds)]
<makefu> gattler_: there is boot.initrd.* with a lot of parameters to customize your init
<gattler_> makefu: no, I'm using my own init
lush has joined #nixos
eacameron has joined #nixos
<makefu> i'd rather advise to let nix/nixos do this customization instead of relying on some blobs you have from previous projects ;)
<gattler_> what's special about nixos that I can't use my own initrd?
takle has joined #nixos
<makefu> keep your initrd up-to-date for example
<makefu> for secrets there is boot.loader.grub.extraInitrd
<domenkozar> gattler_: nixos focuses on making you be able to modify anything
<domenkozar> instead of the traditional approach where everything is just swappable
<gattler_> domenkozar: So can I easily deploy my custom /init script + binaries for my initrd?
<domenkozar> read again what I said
<gattler_> im at a loss here...
bennofs has quit [(Ping timeout: 260 seconds)]
eacameron has quit [(Ping timeout: 258 seconds)]
<gattler_> maybe for now I can just use my initrd to actally boot nixos to try it? All I'm currently missing is the correct boot params
<gattler_> I don't want to messup my current installation so I'd rather not have nixos touch kernel/initrd/bootloader etc...
takle has quit [(Ping timeout: 255 seconds)]
<symphorien> the original nixos initrd is quite complex, because it has to setup /etc/ before launching init (systemd)
<NixOS_GitHub> [nixpkgs] edolstra pushed 4 new commits to release-17.03: https://git.io/vSHdW
<NixOS_GitHub> nixpkgs/release-17.03 f779ca3 Eelco Dolstra: Don't restart systemd-fsck@ units...
<NixOS_GitHub> nixpkgs/release-17.03 aab9d5e Eelco Dolstra: Allow systemd-fsck@.service to find fsck.*...
<NixOS_GitHub> nixpkgs/release-17.03 1126b0f Michael Weiss: gnupg*: Improve the meta set...
<gattler_> ah
<gattler_> symphorien: thx for the explanation, so I actually have to use the nixos initrd
<symphorien> so I think you will have to reproduce all that bahavior in your own initrd
<gattler_> no, then I'll go for the nixos initrd, I was not aware that there is more to it than simply doing an "exec switch_root /mnt"
<NixOS_GitHub> [nixpkgs] edolstra pushed 1 new commit to master: https://git.io/vSHdA
<NixOS_GitHub> nixpkgs/master 7f09c06 Eelco Dolstra: usbredir: Use dev output
Itkovian has joined #nixos
eacameron has joined #nixos
takle has joined #nixos
<gattler_> can I make nixos use an existing grub2 installation?
filterfish has joined #nixos
byteflame has joined #nixos
<symphorien> you can trick nixos in installing grub on another device, probably
eacameron has quit [(Ping timeout: 258 seconds)]
takle has quit [(Ping timeout: 258 seconds)]
<gattler_> hm, I'll give that a try
eacameron has joined #nixos
<gattler_> thx, brb, or not :p ....
gattler_ has quit [(Quit: WeeChat 1.7)]
mkoenig has quit [(Ping timeout: 252 seconds)]
stepcut has joined #nixos
<NixOS_GitHub> [nixpkgs] edolstra pushed 4 new commits to release-17.03: https://git.io/vSHFj
<NixOS_GitHub> nixpkgs/release-17.03 30c8f6e Lengyel Balázs: chromium: flashplayer: 25.0.0.127 -> 25.0.0.148...
<NixOS_GitHub> nixpkgs/release-17.03 c166cc3 taku0: firefox: 52.0.1 -> 52.0.2...
<NixOS_GitHub> nixpkgs/release-17.03 2ddd8b9 taku0: firefox-esr: 52.0.1esr -> 52.0.2esr...
byteflame has quit [(Ping timeout: 258 seconds)]
takle has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
stepcut has quit [(Ping timeout: 260 seconds)]
lewo has quit [(Quit: ZNC - http://znc.in)]
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
lewo has joined #nixos
Itkovian has joined #nixos
freusque has quit [(Quit: WeeChat 1.7)]
<NixOS_GitHub> [nixpkgs] matthiasbeyer opened pull request #24892: mutt: 1.8.0 -> 1.8.1 (master...update-mutt) https://git.io/vSHbM
Avogadro has quit [(Ping timeout: 240 seconds)]
<NixOS_GitHub> [nixpkgs] edolstra pushed 2 new commits to master: https://git.io/vSHNU
<NixOS_GitHub> nixpkgs/master 7260564 Eelco Dolstra: flashplayer: 25.0.0.127 -> 25.0.0.148
<NixOS_GitHub> nixpkgs/master 6a1f168 Eelco Dolstra: flashplayer-standalone: 25.0.0.127 -> 25.0.0.148
<NixOS_GitHub> [nixpkgs] edolstra pushed 2 new commits to release-17.03: https://git.io/vSHNt
<NixOS_GitHub> nixpkgs/release-17.03 08ce4d8 Eelco Dolstra: flashplayer: 25.0.0.127 -> 25.0.0.148...
<NixOS_GitHub> nixpkgs/release-17.03 8521c97 Eelco Dolstra: flashplayer-standalone: 25.0.0.127 -> 25.0.0.148...
gattler has joined #nixos
<gattler> I tried to configure nixos to install grub onto an usb key, unfortunately the grub-install part fails and complains about a missing device map, any suggestions?
eacameron has joined #nixos
<NixOS_GitHub> [nixpkgs] vcunat pushed 1 new commit to release-17.03: https://git.io/vSHAe
<NixOS_GitHub> nixpkgs/release-17.03 07129b4 Aristid Breitkreuz: Merge pull request #24698 from MP2E/ffmpeg_sdl2...
eacameron has quit [(Ping timeout: 258 seconds)]
bennofs has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<Filystyn> how to enable XSI to use compress on nixos
indi_ has joined #nixos
__Sander__ has joined #nixos
pie_ has joined #nixos
<7GHAASMFO> [nixpkgs] edolstra closed pull request #24892: mutt: 1.8.0 -> 1.8.1 (master...update-mutt) https://git.io/vSHbM
<NixOS_GitHub> [nixpkgs] edolstra pushed 2 new commits to master: https://git.io/vSHAN
<NixOS_GitHub> nixpkgs/master 77e71c2 Matthias Beyer: mutt: 1.8.0 -> 1.8.1...
<NixOS_GitHub> nixpkgs/master 3124f5a Eelco Dolstra: Merge pull request #24892 from matthiasbeyer/update-mutt...
blahdodo has quit [(Remote host closed the connection)]
pie_ has quit [(Ping timeout: 260 seconds)]
indi_ has quit [(Ping timeout: 252 seconds)]
blahdodo has joined #nixos
sibi has quit [(Quit: Connection closed for inactivity)]
<NixOS_GitHub> [nixpkgs] vcunat pushed 2 new commits to master: https://git.io/vSHxM
<NixOS_GitHub> nixpkgs/master 769b991 Aneesh Agrawal: openssh: 7.4p1 -> 7.5p1...
<NixOS_GitHub> nixpkgs/master 5b3f807 Vladimír Čunát: Merge #24179: openssh: 7.4p1 -> 7.5p1
<NixOS_GitHub> [nixpkgs] vcunat closed pull request #24179: openssh: 7.4p1 -> 7.5p1 (master...update-openssh-to-7.5p1) https://git.io/vypcO
eacameron has joined #nixos
Itkovian has joined #nixos
<NixOS_GitHub> [nixpkgs] vcunat closed pull request #23990: openssh_hpn: build current version (master...openssh_hpn) https://git.io/vyQhf
eacameron has quit [(Ping timeout: 240 seconds)]
byteflame has joined #nixos
metaphysician has joined #nixos
bennofs has quit [(Ping timeout: 260 seconds)]
<NixOS_GitHub> [nixpkgs] vcunat closed pull request #24637: gcc: include dylibs for darwin build (staging...darwin-gcc) https://git.io/vSHpr
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
griff_ has quit [(Quit: griff_)]
byteflame has quit [(Ping timeout: 260 seconds)]
<NixOS_GitHub> [nixpkgs] ttuegel closed pull request #24882: Backport nm-openvpn fix to 17.03 (release-17.03...nm-openvpn-cherrypick) https://git.io/vS9pJ
DutchWolfie has quit [(Quit: Konversation terminated!)]
<NixOS_GitHub> [nixpkgs] ttuegel closed pull request #24813: nixos: Add nm-openvpn to the networkmanager group (master...nm-openvpn) https://git.io/vSPO5
agjacome has joined #nixos
drp has joined #nixos
Itkovian has joined #nixos
Jackneill has joined #nixos
<drp> anyone know where you find out the option list for these .nix files?
<gchristensen> nixos.org/nixos/options.html
vaibhavsagar has joined #nixos
<nixy> You can also use `nixos-option` to browse them locally on a NixOS machine, but its not as nice or searchable
<drp> gchristensen thanks, this is nice
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<drp> nixy: true but nixos-option looks gives me a full list which is handy
<drp> so thank you as well!
<gchristensen> drp: or man configuration.nix
<nixy> gchristensen: I did not know about that one. That is one unwieldy man page
<gchristensen> its a big ella
<gchristensen> fella*
<MoreTea> ooh, I like `man configuration.nix`. It's probably faster than `nixos-option`
<NixOS_GitHub> [nixpkgs] peterhoeg pushed 1 new commit to master: https://git.io/vSQeQ
<NixOS_GitHub> nixpkgs/master 862fc1b Peter Hoeg: tensor: init at (post) 0.3
<gchristensen> :)
pie_ has joined #nixos
<MoreTea> hi pie_ are *you* able to compile Idris packages on nixos?
<LnL> depends, I prefer nixos-option for some stuff
pie_ has quit [(Ping timeout: 252 seconds)]
m0rphism has quit [(Quit: WeeChat 1.7)]
<MoreTea> ah, scared him away again.
griff_ has joined #nixos
justbeingglad has joined #nixos
griff_ has quit [(Client Quit)]
snikkers has quit [(Ping timeout: 258 seconds)]
PeterHoeg[m] has joined #nixos
justbeingglad has left #nixos []
byteflame has joined #nixos
Avogadro has joined #nixos
betaboon has joined #nixos
xadi has joined #nixos
byteflame has quit [(Ping timeout: 258 seconds)]
bennofs has joined #nixos
eacameron has joined #nixos
acarrico has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
JagaJaga has joined #nixos
CMCDragonkai has joined #nixos
<CMCDragonkai> Can the networking/nat.nix module support UDP port forwarding?
eacamero_ has joined #nixos
<lush> What should I do if a package has got a fix at github but it isn't fixed in the repos?
<lush> package: lxqt
eacamero_ has quit [(Ping timeout: 252 seconds)]
<copumpkin> what's your goal? you could make a PR to nixpkgs adding the fixed version, or you can make a quick local override to avoid waiting, or both
<lush> pull request sounds nice
abcrawf has quit [(Remote host closed the connection)]
<lush> local override would mean I'd manually change the file?
Itkovian has joined #nixos
<copumpkin> nah
eacameron has joined #nixos
<copumpkin> you can override most things pretty easily in nix
gal_bolle has quit [(Read error: Connection reset by peer)]
florent_ has joined #nixos
florent_ is now known as Guest40759
Itkovian has quit [(Client Quit)]
<drp> oh lel is this configuration.nix manpage literally everything? It's looking like it
indi_ has joined #nixos
<domenkozar> humm, nixops doesn't report service failures correctly anymore
<domenkozar> anyone else noticed this?
<domenkozar> maybe it doesn't report error, but just failed
<domenkozar> nonexistent -> errored wasn't shown
eacameron has quit [(Ping timeout: 258 seconds)]
Avogadro has quit [(Ping timeout: 258 seconds)]
PeterHoeg[m] is now known as peterhoeg
jensens has quit [(Ping timeout: 252 seconds)]
peterhoeg_ has joined #nixos
peterhoeg_ has quit [(Remote host closed the connection)]
<copumpkin> niksnut: would you expect Nix 1.11.8 to be able to handle a DB created by 1.12?
<copumpkin> I know you added some forward compatibility, but I'm seeing some weird issue
indi_ has quit [(Ping timeout: 260 seconds)]
<copumpkin> error: creating statement: no such table: FailedPaths
<lush> copumpkin: I didn't use github that much yet. What is the simplest way to create a pull request to just take one commit from master and put it into 17.03?
<Filystyn> how can i tell nix to install fmod lib?
<copumpkin> lush: git clone && git checkout release-17.03 && git cherry-pick <commit ID you want>, then git push to a fork of the repo and create a pull request from there
<lush> ty :)
nh2 has joined #nixos
vaibhavsagar has quit [(Ping timeout: 255 seconds)]
gattler has quit [(Quit: WeeChat 1.7)]
<niksnut> copumpkin: yes
<copumpkin> niksnut: yeah, I see the stuff trying to hit FailedPaths is guarded by if schema < 9
<copumpkin> but somehow I'm still seeing it!
<copumpkin> I'l poke at it some more and see if I can track down where it's coming from
eacameron has joined #nixos
<Filystyn> hello, how to install library for example fmod
<Filystyn> could be any
nh2 has quit [(Ping timeout: 252 seconds)]
<Filystyn> need fmod and not sure how to add it
Tarinaky has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
eacameron has joined #nixos
Jackneill has quit [(Remote host closed the connection)]
<drp> looks like fmod got removed due to them putting source behind a paywall
drasich has joined #nixos
<drp> you might be able to download that and install using that .nix file?
<copumpkin> niksnut: I like that you have a big fat lock called big-lock
betaboon has quit [(Quit: This computer has gone to sleep)]
eacameron has quit [(Ping timeout: 268 seconds)]
eacameron has joined #nixos
<copumpkin> so here's an interesting nix puzzle for someone
<copumpkin> figure out why pinentry_mac isn't showing up as a valid job in Hydra
<copumpkin> there's clearly some sort of weird evaluation issue going on
<copumpkin> but I can't figure out why
byteflame has joined #nixos
<schoppenhauer> hi. is there a channel for nix-related off-topic talk?
<copumpkin> (nixpkgs/trunk)
calvertvl has quit [(Quit: Lost terminal)]
<MoreTea> @schoppenhauer, this channel will do just fine ;)
<gchristensen> schoppenhauer: you're in it!
<MoreTea> brb
<schoppenhauer> ok. so. what do you think of Gobo Linux. It appears to have a somewhat similar concept to nixos, at least in some respect.
<copumpkin> schoppenhauer: yeah, similar idea, except human-controlled names rather than machine-controlled names
<copumpkin> one of my favorite things about nix is that I don't have to think about names anymore
eacameron has quit [(Ping timeout: 240 seconds)]
<copumpkin> but it does seem like it'll have many of the same advantages. I do love the ability to randomly create an on-the-fly derivation for anything though
<copumpkin> without having to worry about name collisions
<schoppenhauer> copumpkin: names?
eacameron has joined #nixos
<gchristensen> in nix, names are computed by the hash of the input, in gobo it is human-driven, no?
<schoppenhauer> ah, yes.
<schoppenhauer> and afaik Gobo doesn't have a central configuration and procedurally generated programs
<copumpkin> yeah, so if I need a random intermediate buildEnv for something, or a silly config file as a substituteAll
<copumpkin> I don't give it a second thought in Nix because I know it won't collide
byteflame has quit [(Ping timeout: 240 seconds)]
<schoppenhauer> yes
<copumpkin> overall it makes things more composable I think
MoreTea has quit [(Ping timeout: 260 seconds)]
<copumpkin> I like to think of the nix store as the heap in your average program
<copumpkin> (with deduplication for free)
<copumpkin> but heap as in I don't give malloc a name when I ask for memory
<schoppenhauer> deduplication isn't the default setting I think.
<copumpkin> it just gives me an address and I use it
<domenkozar> right, so this one will be tricky: https://github.com/NixOS/nixpkgs/issues/24895
<copumpkin> schoppenhauer: we get source-based dedup, not output-based
<schoppenhauer> at least yesterday I activated this feature to optimize the store, and it saved me a few ten gigabytes
eacameron has quit [(Remote host closed the connection)]
<copumpkin> schoppenhauer: yeah, that's the other one
<copumpkin> I just mean that if I write the exact same expression in two places, I'll get the same output
eacameron has joined #nixos
<copumpkin> the hard linking is more granular and doesn't require exact matches
freusque has joined #nixos
<NixOS_GitHub> [nix] edolstra pushed 8 new commits to master: https://git.io/vSQIO
<NixOS_GitHub> nix/master 1673c37 Eelco Dolstra: nix-daemon: Don't die if the user sends an unknown setting
<NixOS_GitHub> nix/master 01dcdfc Eelco Dolstra: nix-daemon: Don't set untrusted-* settings...
<NixOS_GitHub> nix/master 3872371 Eelco Dolstra: Minor cleanup
eacameron has quit [(Ping timeout: 260 seconds)]
<NixOS_GitHub> [nixpkgs] tholzschuh opened pull request #24896: lxqt: fix packaging errors (release-17.03...release-17.03) https://git.io/vSQIs
<copumpkin> niksnut: did you figure out the gcc bug?
<domenkozar> encrypted store paths, recursive nix, new ui cli
<domenkozar> we can't possibly fix nix in 2017, that's too fast
<copumpkin> I still really like my proposal for secrets in the nix store :(
<copumpkin> but unfortunately it would require a NixFS
<copumpkin> and probably a lot more than that :)
<gchristensen> domenkozar: haha
<domenkozar> copumpkin: did you write it down?
<copumpkin> domenkozar: I think in one of the various threads on private nix stores, but it's only sort of serious because I don't know how to implement it sensibly in current OSes :P the basic idea is that "you can have something from the store if you can provide an expression that evaluates to the same hash you're asking for" :)
<copumpkin> there's all sorts of tricky corner cases there, and then how to actually handle user programs
<copumpkin> but it feels like the right basic criterion
<domenkozar> copumpkin: but how would that solve remote deployments? You'd have to ship expressions source code, leaking password anyway
<niksnut> copumpkin: I'm still in the process of trying to report the bug
<copumpkin> niksnut: seeing some really strange behavior with the schema issue. This is the error message I'm getting back from nix-instantiate: "error: creating statement: no such table: FailedPaths"
<niksnut> wow, I just ran across a 1.6 MB .drv file
<copumpkin> oh never mind
<copumpkin> not confused, but still don't know why it's barfing :)
<copumpkin> domenkozar: the idea is just that if you have something that produces the secret, you can have the secret. So the thing you'd have to ship is the secret itself, which means that nothing is gained
<copumpkin> (I think?)
<copumpkin> I might just be wrong :)
mkoenig has joined #nixos
<copumpkin> niksnut: who produced it?
<niksnut> __Sander__: have you ever noticed that node-env.nix generates huge .drv files? mostly because it emits gazillions of copies of the $ ll /nix/store/sghym9ppzpjvnjksc4racjxas5fmq4s9-node-dependencies-logicblox-0.1.0.drv
<niksnut> -r--r--r-- 1 root root 1620076 Jan 1 1970 /nix/store/sghym9ppzpjvnjksc4racjxas5fmq4s9-node-dependencies-logicblox-0.1.0.drv
<niksnut> oops
<copumpkin> :)
<niksnut> __Sander__: have you ever noticed that node-env.nix generates huge .drv files? mostly because it emits gazillions of copies of the composePackage body
lush has quit [(Quit: Konversation terminated!)]
cpennington has joined #nixos
<bennofs> do the platform values in lib/platforms.nix match the values of the currentSystem nixpkgs attr?
<bennofs> also, what would be the correct way to conditionally depend on some package if the target platform (where the program will be run) matches some given platform=
<niksnut> __Sander__: in fact, from a quick look at node-env.nix, I suspect there may be an exponential (worst case) blowup there because includeDependencies/composePackage don't check whether a package has already been handled
vandenoever has joined #nixos
<gchristensen> niksnut: I figured out the kvm issue, we're calling qemu for i386. I have a patch to test now :)
freusque has quit [(Quit: WeeChat 1.7)]
<NixOS_GitHub> [nixpkgs] FRidh closed pull request #24841: pythonPackages.bleach: declare unix support (master...bleech-unix) https://git.io/vSMwc
eacameron has joined #nixos
<NixOS_GitHub> [nixpkgs] grahamc pushed 2 new commits to unstable-aarch64: https://git.io/vSQLk
<NixOS_GitHub> nixpkgs/unstable-aarch64 b86499b Graham Christensen: qemu: support qemu-system-aarch64 in the system wrapper
<NixOS_GitHub> nixpkgs/unstable-aarch64 69cff1f Graham Christensen: kernel: delete fewer files, which are necessary on aarch64
freusque has joined #nixos
eacameron has quit [(Ping timeout: 255 seconds)]
<copumpkin> debugging VM tests is 100% miserable
<copumpkin> just felt like putting that out there
<niksnut> how so?
<domenkozar> copumpkin: do you know you can get into the vm?
eacameron has joined #nixos
<copumpkin> yeah I do, but I need to make changes to it to figure out what's going on
<domenkozar> oh the feedback loop is small
<domenkozar> I know, I was working on openstack for months
<domenkozar> probably most of that time was waiting on qemu
<copumpkin> $ nix-build -k --option build-use-sandbox true -j8 nixos/tests/installer.nix -A simple
<copumpkin> that's what I run
<copumpkin> it takes about 5 minutes, then gives me an error
<__Sander__> niksnut: yes the strategy is quite inelegant
<copumpkin> then I add another patch to the nix package, and rerun it
<copumpkin> (I'm trying to add debugging info to Nix itself because somehow it thinks I have the wrong schema version)
<copumpkin> so I've been adding prints to its source code via a patchPhase
<domenkozar> copumpkin: what I'd do is lunch multiple VMs
<domenkozar> :D
<__Sander__> niksnut: better than the previous approach, but still quite evil
<domenkozar> launch*
<armijn> niksnut evil? naaah.
<niksnut> armijn speaks!
<copumpkin> domenkozar: why multiple VMs?
<__Sander__> armijn: I was referring to myself
<__Sander__> I came up with this crazy hackery :P
<armijn> __Sander__: yes, you are evil ;)
<__Sander__> armijn: yes, but it's not my fault
<__Sander__> because of the NPM developers I turned to the dark side
eacameron has quit [(Ping timeout: 268 seconds)]
<NixOS_GitHub> [nixpkgs] viric pushed 3 new commits to release-17.03: https://git.io/vSQLN
<NixOS_GitHub> nixpkgs/release-17.03 5cd8be4 Lluís Batlle i Rossell: vlock: fix EINTR handling...
<NixOS_GitHub> nixpkgs/release-17.03 cb0ae53 Lluís Batlle i Rossell: vlock: fix ENOPATCH...
<NixOS_GitHub> nixpkgs/release-17.03 1ff2818 Lluís Batlle i Rossell: vlock: improve eintr patch...
eacameron has joined #nixos
<NixOS_GitHub> [nixpkgs] armijnhemel opened pull request #24897: Icoutils 0.31.3 (master...icoutils-0.31.3) https://git.io/vSQLj
snikkers has joined #nixos
<__Sander__> niksnut: maybe I have to think about a better strategy that reuses more code
<armijn> niksnut: yes, I was a bit busy...slowly digging myself out of a hole
<__Sander__> I have to generate the entire NPM dependency tree ahead of time which I also do with the corresponding package build
<__Sander__> but the latter can be done in a smarter way
<gchristensen> cstrahan: ping?
<__Sander__> then the corresponding builder script won't have to become this big
<gchristensen> its about our favorite topic, ruby :)
eacameron has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
<bennofs> __Sander__: is there any documentation on how node packaging works btw?
<__Sander__> bennofs: the README.md of node2nix describes how to do it
<bennofs> __Sander__: hmm that seems to be more about development. what if I just want to add an app from npm to nixpkgs?
<__Sander__> ah
<__Sander__> well that's quite easy
cpennington has quit [(Ping timeout: 258 seconds)]
<__Sander__> in pkgs/development/node-packages
eacameron has quit [(Ping timeout: 260 seconds)]
<__Sander__> there is a generate shell script
<__Sander__> so simply add your package to the JSON config file and then run this shell script
<bennofs> __Sander__: and what if that package depends on native libraries?
<bennofs> __Sander__: will the auto-gen handle that
<__Sander__> no
<__Sander__> because there is no way to detect those
<__Sander__> but you can override packages and supply them manually if you must
<bennofs> __Sander__: where is the node build machinery? I have no idea how this stuff works, when I find packages using nodejs i generally just give up packaging them for nixpkgs :)
<__Sander__> bennofs: you mean the Nix expression providing the builder that deploys a package?
MoreTea has joined #nixos
<bennofs> __Sander__: yeah the builder that builds the pkg
<__Sander__> it's in node-env.nix in the same folder
<__Sander__> but the script is being copied from node2nix
<__Sander__> that's where it originates from
eacameron has joined #nixos
<bennofs> __Sander__: oh I just took a look at node-packages-v4.nix, is it really necessary to reify the whole node dependency chain in nix? :/
ugjka has quit [(Quit: ByE ByE)]
<__Sander__> bennofs: it's the only solution that works accurately enough I'm afraid
<copumpkin> Sander: on that topic :) https://github.com/NixOS/nixpkgs/issues/23259
<__Sander__> at least as far as I can think of
<__Sander__> not that the solution is elegant
<__Sander__> but
<bennofs> __Sander__: have you tried an approach like the rust packaging uses, where you pin the registry and compute a hash over the node_modules ?
<__Sander__> all my projects work including some really crazy ones I'm working on at my employer
ugjka has joined #nixos
<bennofs> or is NPM not deterministic even with pinned registry?
<__Sander__> bennofs: one of my blog posts discusses the issue why I generate the dependency tree ahead of time: http://sandervanderburg.blogspot.com/2016/02/managing-npm-flat-module-installations.html
<__Sander__> basically what node2nix does is the third major revision
<__Sander__> we did other strategies in the past
<__Sander__> each having their own problems
byteflame has joined #nixos
eacameron has quit [(Ping timeout: 258 seconds)]
<bennofs> __Sander__: what is the problem with not giving each node module its own store path, but instead having a single node_modules per app that you let NPM build?
<__Sander__> bennofs: it is not always a problem
<__Sander__> but
<__Sander__> in some cases
<__Sander__> you have situations in which for example a dependency's dependency resolves to something different depending on the context
<__Sander__> also
<__Sander__> cyclic dependencies are allowed in NPM, but not in Nix
<__Sander__> you need to break them
eacameron has joined #nixos
<__Sander__> and in NPM 3.x or higher you need to apply a deduplication strategy which is imperative
<bennofs> __Sander__: yeah well I would not manage the deps at all. I would basically run npm to create node_modules and take the hash of that, so that there can be a fixed output derivation that builds node_modules for a given app
<gchristensen> npm can't be trusted to do it twice :(
<bennofs> gchristensen: even if you pin the registry?
<gchristensen> I believe so
<bennofs> wait, what? ugh.
erasmas has joined #nixos
k2s has joined #nixos
<gchristensen> even npm's shrinkwrap isn't really that good IIRC
<__Sander__> nope
<__Sander__> it works good if everything comes from the same source
<bennofs> gchristensen: but if it runs against a freezed npm index, how could it behave differently?
<niksnut> __Sander__: you may want to have a look at builtins.genericClosure, it could to remove the recursion from composePackage
<bennofs> gchristensen: except if it uses some randomness like time or such internally
byteflame has quit [(Ping timeout: 240 seconds)]
<niksnut> i.e. compute a flat list of dependencies
ebzzry has quit [(Ping timeout: 260 seconds)]
<gchristensen> bennofs: I'm out of my depth here but I think it manages
<__Sander__> niksnut: ok I'll check
k2s has quit [(Remote host closed the connection)]
cpennington has joined #nixos
eacameron has quit [(Ping timeout: 260 seconds)]
<bennofs> __Sander__: at this point, do we even need to use the `npm` tool? It seems like we mostly re-implemented everything it does anyway
<__Sander__> bennofs: npm also does build management
<__Sander__> e.g. it executes script directives
<__Sander__> those still need to be carried out
ebzzry has joined #nixos
Guest82186 has quit [(Ping timeout: 240 seconds)]
<gchristensen> __Sander__: you're a saint for having this working
<__Sander__> yes
<bennofs> gchristensen: hmm I guess if it uses threads it could depend on scheduling
<__Sander__> well
eacameron has joined #nixos
<__Sander__> I'm happy to a have a solution with a great compatibility
<__Sander__> but
<__Sander__> it's of course all but an elegant solution
<gchristensen> npm is not an elegant problem
<bennofs> i think after easter I will go through every language with build support in nixpkgs and try to write up challenges / solutions / design of it
<bennofs> this is really needed IMO
<gchristensen> yes please
<gchristensen> one I'd love to see is for ruby ;)
<bennofs> i have a pretty good grasp on python (well not so up to date) and know haskell packaging, but I don't know much about the rest (yet)
<bennofs> probably going to ask a lot of questions then :)
eacameron has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
<nixy> Turning questions asked into documents written is always a ++
<gchristensen> niksnut: unstable-aarch64 has had a pending evaluation for 47 minutes, everything ok? :)
<gchristensen> should I be more patient?
<ben> What might I have broken if ~/.config/nixpkgs/config.nix doesn't seem to be reflected by what nix-env does?
<bennofs> ben: do you have a ~/.nixpkgs/config.nix
mizu_no_oto has joined #nixos
<ben> no
eacameron has quit [(Ping timeout: 240 seconds)]
<ben> Maybe I'm misunderstanding how nixpkgs works. If I say nix-repl or nix-shell or whatever, ~/.config/nixpkgs/config.nix gets evaluated, if I say nix-env -iA blahblah, it doesn't
<gchristensen> how do you know?
<ben> i replaced the contents with let x = x; in x
eacameron has joined #nixos
<gchristensen> hrmm I think nix-env has an escape hatch if your config fails to evaluate... not 100% sure though
<ben> I also put some trace statements in and didn't see them either :/
<gchristensen> let's try something else. what are you trying to accomplish?
<metaphysician> notmuch (email reader) package can be updated to new upstream release.
<ben> I've had a packageOverrides thing in there to grab those non-nixpkgs-packaged rust nightlies in there, now I can't install them as expected
<ben> looks like my ~/.nix-defexpr is wonky
<ben> oh i see, my nix-env invocation was so wrong it died before it got to evaluating nixpkgs
nh2 has joined #nixos
<ben> whoops.
<gchristensen> :)
<gchristensen> glad you sorted it
<NixOS_GitHub> [nixpkgs] vbgl pushed 1 new commit to master: https://git.io/vSQYN
<NixOS_GitHub> nixpkgs/master 31fef05 Vincent Laporte: ocamlPackages.lwt: 2.6.0 -> 2.7.1
lezed1 has joined #nixos
<lezed1> Hi, is there a way to change the version of GCC being used? I want to use GCC 6, but installing the package doesn't produce a binary I can use.
<copumpkin> niksnut: okay, so somehow it thinks my schema version is 0 o.O
<copumpkin> and that's why it's triggering the FailedPaths code
indi_ has joined #nixos
<copumpkin> yet,
<copumpkin> boot-after-install: must succeed: (cat /nix/var/nix/db/schema; echo) >&2
<copumpkin> boot-after-install# 10
<copumpkin> oh, I might have an idea for why
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
hamishmack has quit [(Ping timeout: 240 seconds)]
<copumpkin> niksnut: yeah, if this gets triggered: https://github.com/NixOS/nix/blob/1.11.8/src/libstore/local-store.cc#L313 then you never call getSchema() and the value is left at at the default
<copumpkin> aha, readOnlyMode at the top too
<cstrahan> gchristensen: hey
indi_ has quit [(Ping timeout: 260 seconds)]
<gchristensen> hi there cstrahan
simukis_ has joined #nixos
<gchristensen> cstrahan: I'm trying to debug why the grpc rubygem won't build, it complains of openssl but I'm providing it openssl. it seems to build in $out, so there is no way to access mkmf.log. is there a good way to debug why a gem won't build?
<copumpkin> niksnut: created an issue; easier to track than random IRC comments :) https://github.com/NixOS/nix/issues/1332
nh2 has quit [(Read error: Connection reset by peer)]
mkoenig has quit [(Remote host closed the connection)]
spinus has joined #nixos
<copumpkin> (I think that's the last obstacle to my killing fakechroot)
mkoenig has joined #nixos
<NixOS_GitHub> [nixpkgs] risicle opened pull request #24898: Backport of mozilla sops to 17.03 (release-17.03...r17.03-moz-sops) https://git.io/vSQ3e
<copumpkin> hmm, does nix 1.12 not auto-upgrade my schema anymore? I just ran nix-build from 1.12 on my 1.11 schema and it didn't upgrade
<cstrahan> gchristensen: quick hack to get you unstuck: out=$PWD at the beginning of the build, then add -K to keep the build tree as usual. I can revisit the build directory soonish.
<copumpkin> oh I know what
<copumpkin> there we go
<gchristensen> cstrahan: I'd do this by editing bundler-env/default.nix, yeah?
CMCDragonkai has quit [(Ping timeout: 260 seconds)]
<gchristensen> I mean gem/default.nix
byteflame has joined #nixos
nh2 has joined #nixos
CMCDragonkai has joined #nixos
CMCDragonkai is now known as Guest41414
drasich has quit [(Ping timeout: 260 seconds)]
Avogadro has joined #nixos
byteflame has quit [(Ping timeout: 258 seconds)]
mjhoy has joined #nixos
takle has quit [(Remote host closed the connection)]
Arguggi has quit [(Quit: Bye)]
takle has joined #nixos
takle_ has joined #nixos
Guest41414 has quit [(Ping timeout: 260 seconds)]
takle has quit [(Ping timeout: 245 seconds)]
takle_ has quit [(Ping timeout: 260 seconds)]
byteflame has joined #nixos
dridus has joined #nixos
snikkers has quit [(Remote host closed the connection)]
<gchristensen> big news everyone, as of today, after ages of work -- Packet.net can deploy NixOS via their panel, and have included a $25 coupon to play: http://weekly.nixos.org/2017/06-i-never-wished-a-man-dead.html
magnetophon has quit [(Ping timeout: 258 seconds)]
<copumpkin> \o/
<hodapp> blaaargh, some R package I need is broken and I'm getting some really cryptic C errors
<hodapp> "xlib-backend.c:366:57: error: 'caddr_t' undeclared (first use in this function)"
mojjo has joined #nixos
dbmikus has joined #nixos
takle has joined #nixos
dbmikus has quit [(Client Quit)]
dbmikus has joined #nixos
<hodapp> or, can anyone else get "rEnv = pkgs.rWrapper.override { packages = with self.rPackages; [ Cairo ]; };" to install?
dbmikus has quit [(Client Quit)]
dbmikus has joined #nixos
<mojjo> does anybody know, why fetchGitPrivate returns a path to the nix-store, whereas fetchFromGitHub returns a derivation? Say the git repo contains a default.nix. In the former case one could just use callPackage for building. But how would this be done for the latter?
takle has quit [(Ping timeout: 260 seconds)]
<cstrahan> gchristensen: you should be able to add a preBuild to the gemconfig arg to bundlerEnv.
magnetophon has joined #nixos
Guest40759 has quit [(Quit: Konversation terminated!)]
ixxie has joined #nixos
<cstrahan> Dunno if that's enough guidance. My day is just getting started, so I'm AFK getting things in order.
nh2 has quit [(Remote host closed the connection)]
<copumpkin> I don't think even Hetzner allows such easy deployment of NixOS, right?
<copumpkin> like nixops knows how to do it
ambro718 has joined #nixos
<gchristensen> copumpkin: nixops reboots in to the rescue environment, copies over a set of install tools, does an install, and reboots
<copumpkin> yeah
<bennofs> mojjo: you can call toString on a derivation and it'll return a path
<copumpkin> so still not really "easy" from a casual user standpoint
<gchristensen> yeah
<gchristensen> and difficult to debug
<NixOS_GitHub> [nixpkgs] 7c6f434c pushed 1 new commit to master: https://git.io/vSQWL
<NixOS_GitHub> nixpkgs/master b6f8809 Armijn Hemel: icoutils: update to 0.31.3
<bennofs> mojjo: like let source = fecthFromGitHub ...; in callPackage "${source}/default.nix"
<copumpkin> so I think packet.net is at same level as EC2 now for ease of deployment
<bennofs> hodapp: I can install it
filterfish has quit [(Ping timeout: 258 seconds)]
<bennofs> hodapp: on 17.09pre105032.7c36657d7d
<__Sander__> niksnut: is there any documentation for builtins.genericClosure?
<__Sander__> or perhaps a code example?
<niksnut> pkgs/tools/typesetting/tex/nix/default.nix has an example for finding latex dependencies
<hodapp> bennofs: ohh, hmm, I'm somewhere near git head now
<niksnut> but basically, it takes a "startSet" (a list of initial graph nodes) and a "operator" function that must return nodes reachable from a node
<__Sander__> niksnut, copumpkin: cool, thanks
<niksnut> nodes are represented as attribute sets with a "key" attribute
xadi has quit [(Ping timeout: 258 seconds)]
<niksnut> so "key" could be the package name in this case
<mojjo> bennofs: alright!
<niksnut> any other attributes are passed through
<copumpkin> niksnut: is there some builtin that gives me the schema version?
xadi has joined #nixos
<copumpkin> I don't see anything
aneeshusa has joined #nixos
MoreTea has quit [(Remote host closed the connection)]
takle has joined #nixos
<niksnut> copumpkin: no
<bennofs> hodapp: also builds on nixos-unstable channel
<niksnut> and that should not be exposed to nix expressions
<hodapp> bennofs: interesting, I wonder what's borked up in my setup then
<copumpkin> niksnut: oh I definitely agree; I was just hoping, because I'm trying to make a repro of the bug and am struggling to get it in a controlled environment
<bennofs> copumpkin: can you not use runCommand to get it?
fresheyeball has joined #nixos
MoreTea has joined #nixos
<fresheyeball> hey folks
<copumpkin> I want to learn what nix thinks its schema version is, not just query /nix/var/nix/db/schema
<bennofs> oh
<fresheyeball> so I just had a miniheart attack
<copumpkin> bennofs: because I think there's a bug in it
<fresheyeball> I tried to turn off kde5 since I'm not actually using it
<fresheyeball> and restarted my computer
<fresheyeball> to get a grey screen of nothing after login
cpennington has quit [(Ping timeout: 258 seconds)]
<fresheyeball> I was able to use the [all configurations] in the boot menu to load an earlier config and the system came right back up
<fresheyeball> problem is, my configuration.nix file is just the same as the poison one
<fresheyeball> how can I retrieve the configuration.nix file for the old version that works?
<bennofs> fresheyeball: configuration.nix is not saved with system config by defaukt
<fresheyeball> so when I loaded an old configuration what happened?
gleber_ has quit [()]
gleber_ has joined #nixos
<copumpkin> fresheyeball: you can't really (there's an option that would've copied your configuration.nix before but if you didn't have it turned on, it won't help)
<bennofs> fresheyeball: configuration.nix is only used to build the system profile, once it is build, the "recipe" for building it (configuration.nix) is no longer needed
<fresheyeball> so what do I do now?
<fresheyeball> I'm not sure what change I made borked the system
<copumpkin> fresheyeball: I thought you just turned off kde5?
<fresheyeball> I thought that is what did it
<fresheyeball> but I turned it back on and the problem persisted
ryanartecona has joined #nixos
<copumpkin> ah
<fresheyeball> so it must have been something else I did
<copumpkin> could check the journal to see what happens in the other boot
<fresheyeball> do tell
<copumpkin> journalctl will show you all the system logs
darlan has quit [(Quit: Communi 3.5.0 - http://communi.github.com)]
cpennington has joined #nixos
<copumpkin> you'll want to find stuff from the previous boot
<copumpkin> I don't really know more specifically than that
<cstrahan> fresheyeball: a symlink was replaced to point at your previous system state, and an activation script (shell) run to touch up imperative stuff. That's what happens when you switch back to a previous configuration (therefore it doesn't need to know what your original configuration.nix was)
<fresheyeball> wow, I would for sure make a feature request to remember the file as well
<hodapp> huh, okay, it works on latest nixpkgs, but 'imager' is still broken... gah
<fresheyeball> having a link back is nice, but not being able to actually restore sucks
<bennofs> fresheyeball: there was a recent discussion about that on the ML iirc
<copumpkin> fresheyeball: system.copySystemConfiguration = true will help in future
<copumpkin> there's a better version of that in progress on Nix itself, but the work stalled a bit
<copumpkin> (I know because I'm the one who stopped working on it :P)
magnetophon has quit [(Ping timeout: 260 seconds)]
<niksnut> __Sander__: here is an example: https://pastebin.com/zLy8tB4z
<niksnut> it traverses a derivation via the nativeBuildInputs attribute to get dependencies
<copumpkin> the issue is that copying configuration.nix doesn't really give you the full power to revert to the old system, because configuration.nix could import or pull all sorts of random stuff from the environment it was evaluated in
<copumpkin> so the Nix work I was doing addresses that
<copumpkin> but copying configuration.nix is generally a good first step regardless
<copumpkin> niksnut: is geeqie your go-to example package? :)
aminechikhaoui has quit [(Ping timeout: 258 seconds)]
<niksnut> I would go for hello, but I don't think it has many dependencies :-)
<__Sander__> niksnut: thanks
fresheyeball has quit [(Quit: WeeChat 1.7)]
<gchristensen> I use `tree` and `lzop` as examples
<cstrahan> fresheyeball: there's an option to copy the configuration.nix into the store, but it's off by default. There's a PR (maybe accepted now?) to make nixos-install add the option to the default config template. I think it's off by default because you might put passwords in plain text in your config, and so putting the config in the store will make it world
<cstrahan> readable -- which would be as if that silently happened without your knowledge.
<cstrahan> *which would be bad if that silently happened without your knowledge*
<cstrahan> Typing from phone. Autocorrect is shit.
<gchristensen> you're doing great, cstrahan
<copumpkin> are there really very many passwords you could put in configuration.nix that don't end up in the store through some other means?
<copumpkin> we don't have any password hashing function builtins
<copumpkin> so I'm struggling to imagine how you'd have a password in configuration.nix that wouldn't end up somewhere else in the store
<hodapp> huh. all 'imager' needed was x11 as a native build input.
<bennofs> copumpkin: well at least they would be gone after GC
<cstrahan> That's a good point.
<hodapp> guess I'll be doing a PR soon...
<cstrahan> And ditto for bennofs
<copumpkin> :) yeah
Avogadro has quit [(Ping timeout: 245 seconds)]
mojjo has quit [(Read error: Connection reset by peer)]
stepcut has joined #nixos
aneeshusa has quit [(Quit: WeeChat 1.7)]
davidak has joined #nixos
magnetophon has joined #nixos
jensens has joined #nixos
nschoe has quit [(Quit: Program. Terminated.)]
alx741 has joined #nixos
digitus has joined #nixos
<copumpkin> can someone double check my logic in https://github.com/NixOS/nix/issues/1332?
<copumpkin> curSchema is 0 in that case, right?
<copumpkin> I haven't been able to repro the bug outside of my VM yet so I'm starting to question myself
<Filystyn> guys
<Filystyn> Could not open MIDI out device
<Filystyn> I_InitSound: Initializing FMOD
<Filystyn> System::init returned error code 62
<Filystyn> Sound init failed. Using nosound.
bennofs1 has joined #nixos
<Filystyn> can somethign missing in configuation.nix can cause this?
<Filystyn> i have one program that behaves like this. no problems with sounds anywhere else
<avn> strace is your friend, but I think it looks like pulseaudio vs something conflict
<Filystyn> ok. big program but ill starce it ;-)
<Filystyn> thx avn
bennofs has quit [(Ping timeout: 240 seconds)]
<avn> -e trace=open + some grep should help
<avn> If I remember right -- FMOD is something closedsource, and idk what it use for sound
indi_ has joined #nixos
zeus_ has joined #nixos
<copumpkin> niksnut: lest you think I'm manipulating votes on that nixos-install thread, someone tweeted a link to it so now I have lots more thumbs... :) anyway, I'm wrapping up the fakechroot-less version and I think it'll be quite pretty, if you wouldn't mind taking a look at it a bit later (the Nix 1.11.8 thing is biting me right now though)
indi_ has quit [(Ping timeout: 260 seconds)]
<NixOS_GitHub> [nixpkgs] 7c6f434c closed pull request #24897: Icoutils 0.31.3 (master...icoutils-0.31.3) https://git.io/vSQLj
teknico has left #nixos []
civodul has quit [(Quit: ERC (IRC client for Emacs 25.1.1))]
vaibhavsagar has joined #nixos
Ivanych has quit [(Quit: Leaving.)]
<joko_> Hm, apparently it's hard setting init to /bin/sh or /nix/var/nix/profiles/system/sw/bin/bash, the system is not bootable :/
<gchristensen> copumpkin: if you come in to a few minutes to help out with the gem, I'd really appreciate it. if not, no worries of course :)
<gchristensen> do'h
<gchristensen> cstrahan: if you come in to a few minutes to help out with the gem, I'd really appreciate it. if not, no worries of course :)
<cstrahan> hey, back at my desk.
<cstrahan> gchristensen: ^^^
<gchristensen> oh, what timing!
<cstrahan> Here's an example of what I mentioned earlier: pkgs/applications/office/ppl-address-book/default.nix
Ivanych has joined #nixos
<gchristensen> yeah, I sorted that out -- on to new problems :P
<cstrahan> oh, cool. what's new?
<gchristensen> I'm making a gist
<cstrahan> cool.
jensens has quit [(Ping timeout: 252 seconds)]
<copumpkin> niksnut, shlevy: does nix-instantiate always open the store? judging by the code it seems to, but I can't get it to print my random debug statements
fresheyeball has joined #nixos
<fresheyeball> so I figured out which package caused my gray screen of death
<fresheyeball> it was gnome3.nautilus
<fresheyeball> putting that in configure.nix borked it
<fresheyeball> but installing with nix-env is fine
blorps has joined #nixos
<blorps> does anyone know how to install pulseaudio on nixos without needing to do a nixos rebuild? I just want to run it as a user. There doesn't seem to be a pulseaudio derivation, at least not at top level
<pmade> blorps: What's wrong with doing a nixos rebuild? If you do that you can have pulseaudio run as a user service.
<pmade> hardware.pulseaudio.enable = true;
<blorps> I just tend to avoid doing nixos rebuilds in general because they involve altering the state of my machine
<gchristensen> installing pulseaudio as a user also alters the state of the machine
<blorps> Yes, but in a much more limited fashion...
<simpson> blorps: It's NixOS. You can *revert* to an older version if the rebuild doesn't work.
<blorps> also when I build and run pulseaudio myself, I know what I'm doing, but when I just copypaste something into my configuration.nix I don't really know what's happening
<gchristensen> you can look up what it is doing if you'd like
<blorps> it just magically starts working, or mysteriously doesn't work, and I'm none the wiser hah
freusque has quit [(Quit: WeeChat 1.7)]
<pmade> blorps: With that logic you wouldn't enable any service in configuration.nix
<simpson> blorps: I strongly doubt that you actually know what PA is doing. Very few people do, and most PA debugging consists of searching for obscure error messages on the Internets.
<blorps> pmade: yes, in general I prefer running stuff in userland if I can
<pmade> I'm super confused because setting hardware.pulseaudio.enable = true; does indeed run pulseaudio in your user env via systemd user services.
<blorps> pmade: well that's interesting, how does it set up a user systemd unit?
<pmade> It sounds like you'd be more happy using ~/.config/systemd and just defining all your own service units. And then just stop using NixOS
<gchristensen> pmade: please don't encourage people to stop using nixos :P
<blorps> I think I can make my own decisions about what OS to use, but thanks
<pmade> gchristensen: Sorry, that was stupid ;)
<simpson> I would rather that people not use NixOS if they're not enjoying it. The revolution is not tomorrow and we will not be starving the non-Nix users.
<simpson> But if you're going to use NixOS, you might as well *use* NixOS.
<blorps> this is strange. I'm asking a very simple question and I'm getting lectures about what OS to use
<copumpkin> a nice part of NixOS is being able to use random packages without them affecting the entire system though :)
<pmade> That's what I was trying to say in an asshole sort of way. Avoiding configuration.nix on NixOS doesn't make sense to me.
systemfault has joined #nixos
<copumpkin> blorps: anyway, I think libpulseaudio is the attribute you want
<gchristensen> pulseaudioFull I think?
<blorps> yeah, pulseaudioFull is what I ended up going with
<simpson> blorps: Your question is not simple. You're asking how to do something which is not supported directly, and we're all a little surprised and confused because the thing that *is* supported is pretty reasonable.
<copumpkin> oh yes
<gchristensen> cstrahan: it seems the grpc_c.so (second file in this gist) isn't in the right place: https://gist.github.com/grahamc/0cc3b0ed2a90b4580eba5267d398a6f2
<blorps> simpson: Maybe I'm just curious? Maybe I want to try running some stuff directly or see how it's set up by hand, so that I can learn some more? I mean, why is it so strange that I might want to build the program and run it?
<cstrahan> gchristensen: Interesting. I'm taking a look.
<simpson> blorps: Mostly it suggests to me that you haven't internalized the role of Nix as sole build tool.
<blorps> .... I'm building PA with nix
<gchristensen> let's not hate on blorps here
<simpson> blorps: After all, Nix provides an embarrassment of tooling for examining builds if you want to dig under the hood.
marsel has quit [(Ping timeout: 260 seconds)]
<blorps> This is weirdly aggressive
<simpson> gchristensen: You think I'm hateful?
<gchristensen> I think we're being, as blorps said, weirdly aggressive about this
exarkun has quit [(Read error: Connection reset by peer)]
<copumpkin> simpson: I think it's pretty clear that blorps just wants to get the thing they asked for done and doesn't want any kind of deeper analysis of their behavior
<fresheyeball> relax
<simpson> blorps: I'm trying to figure out exactly what part of the PA build and configuration you feel that NixOS is hiding from you.
<blorps> Sorry I haven't "internalized" things to your degree of satisfaction, because I'm asking what expression builds pulseaudio
<gchristensen> blorps: pulseaudioFull should be what you're looking for
dgpratt has quit [()]
<cstrahan> blorps: simpson: I suspect you two are talking past each other. doubt either intends any malice.
<blorps> gchristensen: thanks, yeah I got that :) taking a look at it
dgpratt has joined #nixos
simpson has left #nixos ["WeeChat 1.0.1"]
<gchristensen> cool
<copumpkin> I think it's just easy for us to gush about how much we love all our tooling :)
<NixOS_GitHub> [nixpkgs] LnL7 pushed 1 new commit to master: https://git.io/vSQaj
<NixOS_GitHub> nixpkgs/master 9a10e5a Lancelot SIX: nano: 2.8.0 -> 2.8.1...
<copumpkin> except for the UI, of course :P
<copumpkin> but that's all changing in 1.12
<blorps> I really didn't intend to offend or quarrel with anyone, I just find I learn better when I get a sense of a program is meant to be invoked and run etc. Nixos configuration is super convenient but it's very high-level so it's harder to see how something works
<gchristensen> makes sense
<copumpkin> blorps: if you want to dig in at some point, we'd be happy to answer questions. the module system looks daunting at first but ends up being quite simple in how it composes together once you stare at it a bit
<pmade> blorps: I apologize, I misunderstood what you were asking.
<gchristensen> if you'd like some help learning how the module system works, and how to understand what is behind the `enable` options, we're happy to help
<blorps> awesome, thanks guys. No worries
<cstrahan> gchristensen: It looks like the so is ending up here: /nix/store/1d7xln660bqnv01xay543vsz605qp0yn-ruby2.3.3-grpc-1.2.0/lib/ruby/gems/2.3.0/gems/grpc-1.2.0/src/ruby/ext/grpc/grpc_c.so
<gchristensen> cstrahan: indeed! I added `cp $out/lib/ruby/gems/2.3.0/extensions/x86_64-linux/2.3.0/grpc-1.2.0/grpc/grpc_c.so $out/lib/ruby/gems/2.3.0/gems/grpc-1.2.0/src/ruby/lib/grpc/` to my postInstall for grpc and that worked
<NixOS_GitHub> [nixpkgs] Ericson2314 pushed 2 new commits to master: https://git.io/vSQVo
<NixOS_GitHub> nixpkgs/master 6f69681 David Johnson: Add hardening, bump hash to HaLVM
<NixOS_GitHub> nixpkgs/master eec35cb John Ericson: Merge pull request #24889 from formaltech/halvm-bump...
<pmade> blorps: This is the file that creates the puluseaudio systemd unit: https://github.com/NixOS/nixpkgs/blob/master/nixos/modules/config/pulseaudio.nix
<cstrahan> gchristensen: I think this is the issue we're looking for: https://github.com/grpc/grpc/issues/9998
<pmade> If you `man configuration.nix` and search for pulse you can see that file is where options are coming from.
<cstrahan> gchristensen: I'd recommend leaving a message there. I will, too.
<blorps> pmade: ah, interesting, I didn't know about the man configuration.nix
<pmade> It's very, very useful.
<blorps> so, I put `hardware.pulseaudio.enable = true` and did a nixos-rebuild switch, but I don't see pulseaudio running
<cstrahan> gchristensen: Actually, from that issue, might already be fixed: "It seems the fix made it into 1.2.2. Yay! https://github.com/grpc/grpc/blob/v1.2.2/grpc.gemspec"
<gchristensen> yeah, I see! interesting ... time to try upgrading etcdv3's
<blorps> is there a way for nixos-rebuild to tell you the systemd units its creating, or the new ones at least?
<copumpkin> there's also nixos-option
<copumpkin> which can be super handy
exarkun has joined #nixos
<copumpkin> it normally prints them out at the end of each rebuild, I think
<nixy> Is it possible to setup commands to run after you exit nix-shell? I know shellHook runs commands before you enter it, but I want it to run pre exit
<copumpkin> nixy: there are bash traps for trapping exit events and the like
<blorps> sudo nixos-rebuild switch
<blorps> building Nix...
<blorps> building the system configuration...
<blorps> updating GRUB 2 menu...
<blorps> activating the configuration...
<blorps> setting up /etc...
<blorps> that's all it says :|
<nixy> copumpkin: I figured that would be a way to do it, setup the traps in the shellHooks but I was hoping there was a nix way to do it
<copumpkin> I can't think of anything
<nixy> Thanks ya'll that was what I needed <3
<blorps> well, gotta run to lunch but hopefully some people who know about pulseaudio are here when I get back... hah
vaibhavsagar has quit [(Ping timeout: 240 seconds)]
<pmade> Sometimes when updating the OS and you have a running user session you have to `systemctl --user daemon-reload'
<avn> blorps: you want system-wide PA?
<pmade> And then you can `systemctl --user status pulseaudio`
MoreTea has quit [(Ping timeout: 258 seconds)]
__Sander__ has quit [(Quit: Konversation terminated!)]
<LnL> vcunat: ping
<Filystyn> guys is this ok https://pastebin.com/raw/3jt35vRi?
earldouglas has quit [(Quit: Lost terminal)]
bennofs1 has quit [(Ping timeout: 252 seconds)]
<NixOS_GitHub> [nixpkgs] vcunat pushed 3 new commits to master: https://git.io/vSQKm
<NixOS_GitHub> nixpkgs/master 6617c8b Vladimír Čunát: A fixed version of commit eb750f9a7b2...
<NixOS_GitHub> nixpkgs/master 4aec96b Vladimír Čunát: Merge the two "same" merges...
<NixOS_GitHub> nixpkgs/master 2090aa4 Vladimír Čunát: Merge: fixup a bad merge...
<niksnut> gchristensen: any idea what's up with this: http://hydra.nixos.org/build/51684786
<niksnut> (I added a CI job for Nix on aarch64)
<niksnut> so glibc fails to compile on release-17.03
<niksnut> that used to work didn't it?
<niksnut> or does it only work on master?
<drp> odim madduck
<drp> MrBougo edward konubinix andschwa_ matled cargill boudiccas kier olaf
<drp> johnw M-hrjet Demosthenex Xof stooj mattcen YoungFrog tbm dj_ryan_ Ralith_
<drp> MacCoaster xnox_ mstypa
<drp> *** klink (~klink@67.12.70.115.static.exetel.com.au) has quit: Read error:
<drp> Connection reset by peer [19:47]
<drp> *** klink (~klink@67.12.70.115.static.exetel.com.au) has joined channel
<drp> #ledger
<drp> *** JuanDaugherty (~juan@98.4.124.117) has joined channel #ledger [20:07]
drp has quit [(Remote host closed the connection)]
<johnw> what are you doing dpr?
<niksnut> paste accident?
<gchristensen> accidental pasting I assume
takle has quit [(Remote host closed the connection)]
aminechikhaoui has joined #nixos
<gchristensen> niksnut: hmm I'm not sure, I know I'm usually building aarch64 against unstable. I'd try retrying the failed builds and see if it helps?
<gchristensen> the build logs smell of a host failure
<niksnut> ok
<NixOS_GitHub> [nixpkgs] vcunat pushed 1 new commit to staging: https://git.io/vSQKS
<NixOS_GitHub> nixpkgs/staging f7a4f14 Vladimír Čunát: Merge branch 'master' into staging...
<niksnut> nah it's giving the getcwd errors again
<niksnut> strange error
<gchristensen> very strange
<gchristensen> unfortunately I can't dig deeper right now
<niksnut> ok, not urgent
<copumpkin> this is what my nix build looks like right now
<copumpkin> machine# schema version: 7
<copumpkin> machine# no really, schema version: 7
<copumpkin> :P
takle has joined #nixos
<gchristensen> niksnut: ok :) I'll take a bit more of a look later. can you take a look at the unstable-aarch64 build? it has been pending eval for 4 hours
<niksnut> link?
<niksnut> thanks
<gchristensen> thank you!
takle has quit [(Read error: Connection reset by peer)]
takle_ has joined #nixos
<copumpkin> I think the standalone nixos VM test driver doesn't like it when you restart a machine a few times during a test
takle has joined #nixos
* copumpkin fffffuuuuuuuu
takle_ has quit [(Ping timeout: 240 seconds)]
<NixOS_GitHub> [nixpkgs] vcunat pushed 1 new commit to release-17.03: https://git.io/vSQ6x
<NixOS_GitHub> nixpkgs/release-17.03 27cc923 risicle: sops: init at version 2.0.8 (#24898)...
<fresheyeball> hey so I'd like to buy a webcam for this desktop
<fresheyeball> is there one yall would recommend for nixos?
takle has quit [(Ping timeout: 268 seconds)]
<joko_> fresheyeball: Logitech C920 is working fine for me
edvorg has quit [(Ping timeout: 258 seconds)]
darlan has joined #nixos
eacameron has quit [(Remote host closed the connection)]
<dash> mine is apparently a... Logitech C310
<copumpkin> niksnut: does my latest 1.11.8 bug make sense to you? I haven't been able to repro it but the code seems fairly easy to follow to find the bad case. I don't know if I should keep looking for a repro or if it's good the way it is
hamishmack has joined #nixos
<fresheyeball> dash: does it work on nixos?
<dash> fresheyeball: yes
sellout- has joined #nixos
byteflam1 has joined #nixos
byteflame has quit [(Read error: Connection reset by peer)]
<NixOS_GitHub> [nixpkgs] dezgeg pushed 1 new commit to master: https://git.io/vSQPX
<NixOS_GitHub> nixpkgs/master 3ed0d7e Tuomas Tynkkynen: kernel-config: Explicitly enable CONFIG_NETFILTER...
Filystyn has quit [(Quit: Konversation terminated!)]
byteflame has joined #nixos
byteflam1 has quit [(Ping timeout: 260 seconds)]
aminechikhaoui has quit [(Ping timeout: 260 seconds)]
calvinx has quit [(Quit: calvinx)]
fresheyeball has quit [(Quit: WeeChat 1.7)]
eacameron has joined #nixos
takle has joined #nixos
marsel has joined #nixos
<gchristensen> is there any handy function in nixpkgs which accepts a path / drv / etc. + a list of patches, and produces a new path / drv / etc. which is the result of the two?
byteflame has quit [(Ping timeout: 260 seconds)]
takle has quit [(Ping timeout: 255 seconds)]
takle has joined #nixos
aminechikhaoui has joined #nixos
<niksnut> gchristensen: the unstable-aarch64 jobset was disabled :-)
<gchristensen> hrm
<gchristensen> d'oh :P
joko_ is now known as joko
<gchristensen> I meant to Enable it and Disable polling :)
<gchristensen> thank you
joko has quit [(Changing host)]
joko has joined #nixos
Avogadro has joined #nixos
<joko> Anyone using systemd-networkd?
MichaelRaskin has joined #nixos
ertesx has joined #nixos
indi_ has joined #nixos
ertes has quit [(Ping timeout: 255 seconds)]
ertesx is now known as ertes
indi_ has quit [(Ping timeout: 252 seconds)]
<NixOS_GitHub> [nixpkgs] pjones opened pull request #24900: plex: Don't overwrite primary database on restart (master...pjones/plex-service) https://git.io/vSQy9
<gchristensen> niksnut: hmm does the aarch64 machine have the nixos-test feature? still isn't building tests
<niksnut> yes
<copumpkin> "aarch64-linux:kvm,nixos-test" : { "lastActive" : 1492173251, "runnable" : 0, "running" : 0 }
bennofs has joined #nixos
<copumpkin> I think that means it doesn't think there's anything to do, right?
<gchristensen> well I don't know heh
<gchristensen> oy
hamishmack has quit [(Quit: hamishmack)]
<NixOS_GitHub> [nixpkgs] Hodapp87 opened pull request #24901: imager (r-modules): add pkgs.x11 to fix build, unmark imager/ForestTools as broken (master...master) https://git.io/vSQHi
<hodapp> \o/
ryanartecona has quit [(Quit: ryanartecona)]
<NixOS_GitHub> [nixpkgs] dezgeg pushed 2 new commits to release-17.03: https://git.io/vSQQm
<NixOS_GitHub> nixpkgs/release-17.03 1ff02f8 Tuomas Tynkkynen: bash: Set bash_cv_getcwd_malloc=yes when cross compiling...
<NixOS_GitHub> nixpkgs/release-17.03 a98a835 Tuomas Tynkkynen: stdenv: aarch64: Update bootstrap tarballs...
<Dezgeg> niksnut: ^ that should fix the aarch64 glibc build
<gchristensen> nice
<gchristensen> Dezgeg: did you see the packet announcement?
<niksnut> ah thanks
ryanartecona has joined #nixos
mizu_no_oto has joined #nixos
peti has quit [(Quit: WeeChat 1.7)]
<NixOS_GitHub> [nix] copumpkin opened pull request #1333: Ensure that curSchema is set before opening the DB (1.11-maintenance...fix-schema-version) https://git.io/vSQQQ
<copumpkin> niksnut: if you can merge https://github.com/NixOS/nix/pull/1333 that'll unblock all my other work :)
<copumpkin> niksnut: oh I guess we'd probably need yet another 1.11 release too :/
<benley> anyone know who/what I should wait for before merging this chromium PR? https://github.com/NixOS/nixpkgs/pull/24887
<gchristensen> benley: merge & backport to 17.03 please
<copumpkin> niksnut: on the plus side, another 1.11 release would allow me to take out the shitty patchPhase I added to the 1.11 expression in nixpkgs :D
<benley> gchristensen: aye aye
<copumpkin> cap'n?
<gchristensen> thank you, benley :)
<NixOS_GitHub> [nixpkgs] benley closed pull request #24887: chromium: 57.0.2987.110 -> 57.0.2987.133 [security] (master...chromium) https://git.io/vSHGt
rodarmor has quit [()]
rodarmor has joined #nixos
justan0theruser has quit [(Ping timeout: 240 seconds)]
justanotheruser has joined #nixos
<NixOS_GitHub> [nixpkgs] benley pushed 1 new commit to release-17.03: https://git.io/vSQ72
<NixOS_GitHub> nixpkgs/release-17.03 7714553 Benjamin Staffin: chromium: 57.0.2987.110 -> 57.0.2987.133 [security]...
<benley> Done and done
<gchristensen> !m benley
<[0__0]> You're doing good work, benley!
<benley> ^___________^
teknico has joined #nixos
byteflame has joined #nixos
teknico has left #nixos []
<NixOS_GitHub> [nix] edolstra pushed 2 new commits to 1.11-maintenance: https://git.io/vSQ5L
<NixOS_GitHub> nix/1.11-maintenance ad9e603 Dan Peebles: Ensure that curSchema is set before opening the DB...
<NixOS_GitHub> nix/1.11-maintenance 1ab5cc0 Eelco Dolstra: Merge pull request #1333 from copumpkin/fix-schema-version...
<copumpkin> niksnut: omg thanks
<niksnut> no thank you :-)
Gravious has joined #nixos
nalice has joined #nixos
byteflame has quit [(Ping timeout: 258 seconds)]
ryanartecona has quit [(Quit: ryanartecona)]
<contrapumpkin> weird, matrix seems to not be forwarding my messages
<contrapumpkin> niksnut: would a 1.11.9 be a PITA? I can keep patching the nixpkgs expression but it's awkward
<copumpkin> is it a PITA to do a 1.11.9 too? I can keep patching in the nixpkgs expression but it's a little awkward
<copumpkin> is matrix blocked?
<gchristensen> matrix is working again copumpkin
<copumpkin> oh I see :)
<niksnut> copumpkin: no
AllanEspinosa has joined #nixos
<nalice> i'm trying to get nixos installed with zfs as root filesystem. not going too well, not sure how to troubleshoot. anyone got nixos 17.03 running on zfs?
globin has quit [(Quit: WeeChat 1.6)]
<gchristensen> yep, nalice
<gchristensen> can you share what you've done?
<gchristensen> (and the problems you're having)
takle has quit [(Read error: Connection reset by peer)]
takle has joined #nixos
globin has joined #nixos
codeape has joined #nixos
<codeape> Every time I launch nix-shell it redownloads the dependencies, can someone help me understand what I've broken to cause this?
<LnL> do you run the garbage collector often?
<nalice> nice, at least then i know it *should* work if done correctly. just going to get my notes in order. i'm preparing on a VM so i can install on a laptop later that i've ordered.
<gchristensen> nalice: what did you order?
ebzzry has quit [(Ping timeout: 258 seconds)]
<nalice> very cheap thing from china with a monitor that should be OK
<gchristensen> gotcha
<nalice> "chuwi lapbook 14.1"
<copumpkin> matrix is still being weird
<codeape> LnL: I do run GC from time to time, but I don't think that's the issue. I can launch the shell, exit it, then relaunch it and it will still reinstall dependencies.
<NixOS_GitHub> [nixpkgs] chris-martin opened pull request #24902: idea.idea-{community,ultimate}: 2017.1 -> 2017.1.1 (master...pr/idea-2017-1-1) https://git.io/vSQNp
<gchristensen> copumpkin: irc is still going strong ;)
<gchristensen> codeape: does your shell.nix use ./. as its source or something?
Itkovian has joined #nixos
<LnL> codeape: are the paths the same?
<LnL> does it happen *every* time, or just often?
<gchristensen> lol I'm trying to send you an example nalice but my browser keeps crashing.
<gchristensen> such is osx
<LnL> :)
<nalice> this is how i set it up from the boot CD, https://pastebin.com/JdHC0was
<codeape> gchristensen: Oh, yes it does (this is a friends project I'm working on), what does src = ./. mean?
<gchristensen> now it crashed while reporting the crash :(
<gchristensen> codeape: it means the current directory is the source
<gchristensen> so if you change ./ as part of the nix-shell, it will rebuild next time
<nalice> but i will only be using one disk
<codeape> LnL: every time.
<codeape> LnL: what do you mean by "are the paths the same"?
<gchristensen> codeape: hrm. why use zfs?
<codeape> gchristensen: I'm not using zfs...?
<gchristensen> sorry, nalice ^
<nalice> and it builts fine, but when trying to boot i get "cannot import tank: no such pool available"
Filystyn has joined #nixos
Filystyn has quit [(Changing host)]
Filystyn has joined #nixos
<LnL> codeape: src = ./.; will copy the current state of the project and build based on that. this means that if any file in there changes, everything will rebuild
xadi has quit [(Ping timeout: 240 seconds)]
<nalice> gchristensen: learning mostly.
<codeape> LnL: ah! That helps a lot then, thank you :)
<LnL> that includes potential cache/tempfiles
<nalice> finding if zfs is practical on a laptop. running something that tries to make sure your the same things you put in are the same that comes out, but isn't btrfs.
<codeape> LnL: so should src really just refer to the source code?
<LnL> codeape: you can use filterSource to exclude certain files to reduce the amount of rebuilds
<gchristensen> nalice: zfs is not really intended for that
<gchristensen> nalice: I find ext4 to be very good
<codeape> LnL: Awesome! Thank you again :)
<contrapumpkin> niksnut: cool, thanks, let me know when you release it and I'll rebase and all will be amazing :)
<nalice> gchristensen: i'm aware, it's a bit of an experiment. i use ZFS already on my NAS, but on it i use ext4 on the separate boot drive, because it was simple.
<gchristensen> ok
<nalice> and also it's a cheap chinese laptop. it would be nice to have some confirmation it's not eating away at my data. :)
<gchristensen> not sure how much zfs will help with that if you only have one disk
<LnL> nalice: you'll probably get better/more consistent performance with other filesystems on a laptop
<nalice> it won't be able to repair a corrupt file, but it would scream
<gchristensen> so will ext4 fwiw
<nalice> then i'd know to restore from a backup, or trash the drive
<gchristensen> how much ram will you have?
<nalice> gchristensen: ext4 doesn't checksum data though?
<nalice> only 4 GB
<gchristensen> oh man, you're going to have a bad time with zfs
<zetok> nalice: use btrfs
<gchristensen> what about fsck?
<NixOS_GitHub> [nixpkgs] pjones opened pull request #24903: atomicparsley: 0.9.0 -> 0.9.6 (master...pjones/atomicparsely) https://git.io/vSQxa
<LnL> should be fine if you tweak the parameters, but speeds are pretty inconsistent with a small amount of ram
<contrapumpkin> nalice: I think it has basic checksum on data doesn't it, just not a cryptographic hash or hashing of metadata?
<zetok> nalice: just beware of performance drop on older fs
<zetok> contrapumpkin: it doesn't
<nalice> i don't know, maybe. it's just a laptop, not a multiuser system, system requirements you'd find in ZFS documentation doesn't really apply i think. maybe.
davidak has quit [(Quit: Leaving.)]
<zetok> that being said, there was added feature to ext4 that adds checksums, but only on newest kernels and it's not used anyway
<gchristensen> I know the ZFS channel doesn't like helping people who install it on underspecced systems
<c74d> ZFS can repair files even if it has only a single disk, if one sets copies=2
<c74d> certainly it's not *as* safe as multi-disk redundancy, but I've always run ZFS on single-disk systems
<c74d> (including laptops)
<nalice> c74d: right, but my laptop only ships with 60 GB of disk, so that ain't gonna happen. also i suspect that would be quite slow.
<nalice> it does do 2 copies for metadata though, by default
<wak-work> that's fine though
<wak-work> metadata is relatively very small
cheecheeo has joined #nixos
<c74d> 4 GB of RAM sounds suboptimal but acceptable to me; 60 GB of disk sounds painful if one doesn't like storing most everything in the cloud
<nalice> i'm not too worried about the RAM usage. i don't want to get into that, at least not before i got it working at all, but i think the common wisdoms about zfs are not all correct. such as, should never be used with one drive, or without a large amount of ram, or even that ecc ram is required.
<nalice> i got all those things on my NAS, but that's a NAS
<zetok> nalice: if you like pain, sure
<gchristensen> by all means
<gchristensen> have fun :)
<c74d> nalice: your hardware doesn't sound all that much worse than mine on which I run ZFS
<nalice> i'm totally open to the possibility that it may be painful, but i'd like to try
<c74d> the ZFS On Linux FAQ explains why ECC RAM is good but (implicitly) contests the myth that it's required
<nalice> i think i read something about this from a ZFS developer, i'll see if i can find it
<c74d> I've never had multiple disks in a zpool, and I like ZFS still
<timclassic> nalice: I think you'll be fine
<timclassic> nalice: And I agree about the wisdom surrounding a single disk--it really depends on what your requirements are. Personally, I want the integrity features.
aminechikhaoui has quit [(Ping timeout: 252 seconds)]
<nalice> by the way, how do you handle the kernel version on nixos when using ZFS? will a compatible kernel be selected, should you lock it to the latest kernel tested by ZoL, or get on the latest from upstream, or how do you handle that?
<nalice> this applies to anything on nixos that could be very dependant on the kernel version i suppose
aminechikhaoui has joined #nixos
AllanEspinosa has quit [(Ping timeout: 255 seconds)]
AllanEspinosa has joined #nixos
<gchristensen> nixos doesn't do package "selection" like that, but if it doesn't compile you won't accidentally break your system
<zetok> err
<gchristensen> IIRC some kernel packages are flagged as not working with zfs
<zetok> but system can break at runtime when nixos fails to build?
<nalice> but since zfs is a module the kernel would always compile, and may still be broken when booting, i guess?
<gchristensen> no
<gchristensen> and no
<zetok> i.e. nixos kills all running processes if you misconfigure
marsel has quit [(Ping timeout: 260 seconds)]
<gchristensen> no
<zetok> Yes.
<avn> gchristensen: #zfsonlinux is enough friendly for us ;)
<avn> I found a lot of wisdom there ;)
<zetok> gchristensen: I wouldn't be saying that if I didn't test it
<gchristensen> ok
<gchristensen> trying to build a new system where the kernel modules don't compile against the kernel, it won't allow you to switch to the new, broken version
<avn> nalice: It works well, althoug I suggest 4.10 kernel for zfs unstable, and 4.9 for 6.5 branch
<avn> Also always copy kernels, boot from /nix/store won't work
<avn> and turn dedup off, unless you have TONS of ram
<nalice> gchristensen: ok, i see what you mean
<gchristensen> avn: would you like to vet my configuration?
<gchristensen> I would find that helpful
<zetok> avn: wouldn't it be better to just use btrfs then?
byteflame has joined #nixos
<nalice> avn: friends don't let friends enable dedup :)
<gchristensen> dedup is a good feature if you have matching requirements and resources
griff_ has joined #nixos
<nalice> zetok: side by side, for this type of computer btrfs looks better on paper. i just don't have much trust for it. it may not be ccompletely rational.
<avn> so 16G and six cores is not enough for dedup
<zetok> nalice: it works for storing data
<nalice> but it is a fact that ZFS has a good record of being stable and has been for a long time. btrfs, more controversial.
<zetok> nalice: there are some problems, but they are in areas of some features that you probably wouldn't use
<nalice> i know, i won't use redundancy either way
<zetok> i.e. creating & deleting thousands of snapshots per second
<Ralith> personally I'm hype for bcachefs
<zetok> and trying to send/receive btrfs snapshots (don't even bother, just use rsync)
<nalice> Ralith: oh yes, i hope that takes off. linux really could use a better fs.
<nalice> zetok: ok, i'll keep that in mind if i have to fall back to btrfs.
byteflame has quit [(Ping timeout: 252 seconds)]
<avn> gchristensen: ashift should be picked from actual block size (9 or 12), idk what sizes there
<zetok> Ralith: in like 10 years it might be good for use :/
<Ralith> why so pessimistic? just don't trust a FS that hasn't been tested for a very long time?
<nalice> gchristensen: my config looks much like yours but i do not have the grub.zfsSupport set
<avn> also I have grub on efi
<zetok> Ralith: kinda
<gchristensen> avn: the block size is reporting 4096
<zetok> Ralith: it's just a matter of making sure that most of the corner cases have been checked & fixed
<MichaelRaskin> I am OK with an FS going belly-up, if it doesn't _silently corrupt_ anything
<avn> gchristensen: well, better check docs (and may be check real sizes on install time)
<MichaelRaskin> I mean, all the storage devices that fit into a laptop and are produced in the last decade are just a joke of reliability anyway
<gchristensen> avn: can do :) thanks for the tip
<nalice> does grub.zfsSupport mean i can install grub on the ZFS volume, or that i can boot my ZFS volume (which i currently cannot, and got sidetracked :) )?
<nalice> the wiki page didn't mention it
<hyphon81> Good morning from japan.
Guest60 has joined #nixos
<hyphon81> Perhaps, zpool not found when NixOS stage1 ?
<gchristensen> hello hyphon81, the people in #packethost are impressed by your work :)
<zetok> Ralith: btrfs was started like 10 years ago ;)
<nalice> hyphon81: that's right
<Ralith> I'm aware
<Ralith> it's not an unreasonable stance
<nalice> this is the error message https://i.imgur.com/syYaOoo.png
<nalice> when i boot the live CD back up, i can import the pool fine manually with "zfs import tank"
<hyphon81> gchristens: Oh, really? I'm interested in the ARM server. However it can't use NixOS now.
<nalice> and then "mount -t zfs tank/root/nixos /mnt"
<gchristensen> last I saw that I didn't have kernel modules to address the storage
<gchristensen> hyphon81: you can't use nixos, or you can't make an arm server on packet?
<gchristensen> hyphon81: you can't use nixos, or you can't make an arm server on packet running nixos?
<jack[m]1> nalice: boot.supportedFilesystems = [ "zfs" ... ]
<gchristensen> hyphon81: because if you need an arm server running nixos, I can help with that
<nalice> jack[m]1: yup, i have that set.
<nalice> this is what i've done https://pastebin.com/JdHC0was
<jack[m]1> nalice (IRC): Also, I have a ... larger ... pool, so I have a NVMe SSD that nixos lives on, and then I use dmcrypt for all of the pool devices. This way I can remote boot the system (without the pool) and log in to decrypt everything..
<nalice> actually "boot.loader.grub.device" isn't correct there, i changed to to /dev/sda, because that didn't make sense
<hyphon81> nalice: I met same problem yesterday. And I solved with "boot.zfs.devNodes = "/dev";"
cpennington has quit [(Ping timeout: 252 seconds)]
<jack[m]1> Hrm. Did I ever submit the zfs boot tests?
<nalice> hyphon81: oh, really? that looks like it could be related!
<nalice> yes, this is probably the issue
<nalice> the default is /dev/disk/by-id
<nalice> but my drive doesn't show up there! only the CD.
<hyphon81> gchristens: Ah, I can't make an arm server on packet. I can't chose NixOS with Type 2A plan.
<nalice> why is that?... it's listed under by-path, among others
<hyphon81> nalice: Yes, that is same situation when I had problem.
<gchristensen> hyphon81: it isn't good enough yet for them to release to everyone, but I can help you provision one automatically.
<nalice> hyphon81: "Name of directory from which to import ZFS devices. This should be a path under /dev containing stable names for all devices needed, as import may fail if device nodes are renamed concurrently with a device failing." it may be a good idea to use one of the other /dev/disk/by-* that works then rather than straight by /dev/, possibly.
justanotheruser has quit [(Ping timeout: 240 seconds)]
<ajp> I'm having some trouble with the new overlays. I added a trivial one to test, and it definitely gets pulled in by /etc/nixos/configuration.nix (because it errors if I add a syntax error), but any new packages I introduce aren't available from the pkgs attribute
<gchristensen> hyphon81: (I built their nixos integration, so it'd be almost the same as if they provisioned it)
<ajp> any idea why that might be the case?
Wizek_ has quit [(Read error: Connection reset by peer)]
Wizek_ has joined #nixos
cheecheeo has quit [(Ping timeout: 260 seconds)]
Wizek_ has quit [(Remote host closed the connection)]
indi_ has joined #nixos
<hyphon81> gchristens: Oh, that's good news. I guess because of the NixOS isn't enough suit for arm processor, so I can't chose NixOS on packet.net Type 2A.
Wizek_ has joined #nixos
<gchristensen> hyphon81: right, but you can get nixos on their Type 2A server anyway
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<benley> nalice: I use ZFS on my laptops (with nixos, and atop LUKS) and it works quite nicely, fwiw
endformationage has joined #nixos
<benley> I even managed to get it working with /boot being an encrypted zfs dataset, which frankly amazes me
justanotheruser has joined #nixos
<nalice> benley: whoa. did you document that by any chance? :)
<NixOS_GitHub> [nixpkgs] benley opened pull request #24904: latte-dock: backport to 17.03 (release-17.03...latte-dock-backpot) https://git.io/vS7JU
<benley> nalice: I have my configs online at least - I ought to write up a blog post detailing how to do it. It's not as hard as it sounds like it would be.
<benley> (on a UEFI system, even)
Filystyn has quit [(Ping timeout: 240 seconds)]
<benley> nalice: https://github.com/benley/dotfiles/blob/master/machines/totoro/configuration.nix is the machine's config. (hardware-configuration.nix is in the same directory in that repo)
indi_ has quit [(Ping timeout: 255 seconds)]
cpennington has joined #nixos
<benley> the next thing I want to get working is UEFI Secure Boot
<benley> I haven't seen anybody do that on nixos yet.
<gchristensen> I'm not sure we can do SB without getting MS to sign our signing key?
<MichaelRaskin> By enrolling your own keys or by using one of The Shims?
<NixOS_GitHub> [nixpkgs] rycee pushed 1 new commit to release-17.03: https://git.io/vS7JM
<NixOS_GitHub> nixpkgs/release-17.03 0a24814 Robert Helgesson: haskellPackages.bench: 1.0.2 -> 1.0.3...
systemfault has quit [(Quit: Bye!)]
<benley> MichaelRaskin: I'm not sure yet. I was thinking probably enrolling keys?
<benley> I once got an Ubuntu install working with secure boot by using one of the shims
<MichaelRaskin> him
<MichaelRaskin> I think one of the shims can be used with other distributions
<ToxicFrog> Argh.
<ToxicFrog> I updated my server from the 16.09 channel to 17.03
<ToxicFrog> It no longer builds, complaining that it cannot coerce a function to a string
<benley> MichaelRaskin: the thing I'm not sure I understand about the shims is if it subverts the whole purpose of secure boot
<MichaelRaskin> (it comes signed by MS key, and it requires significant but simple and guided user interaction to enroll a new chainloaded key)
<ToxicFrog> But none of my code is in the stack trace, so tracking this down is going to really suck
<benley> MichaelRaskin: it's a signed shim that chain loads into grub, right? How does it ensure the integrity of the grub code that it loads?
<benley> MichaelRaskin: oh maybe you just explained that.
<hyphon81> packet.net with NixOS, now, gives us coupon code. I saw it twitter. So, I accessed there and I'm interested in arm server has many cores and reasonable price(expensive for my wallet XD).
<benley> <sarcasm>Gee, if only nixos had some sort of wiki where notes on these things could be collected...</sarcasm>
<MichaelRaskin> PRs to the manual seem to be processed, though
<benley> yes totally :)
<gchristensen> hyphon81: join #packethost
<benley> there's still a bit of a mental barrier between adding some notes to a wiki page vs writing formal documentation
<NixOS_GitHub> [nixpkgs] ndowens closed pull request #24219: nitrokey-app: 0.5.1 -> 0.6.3 (master...nitrokey-app-0.6.3) https://git.io/vSe8w
<benley> anyway I don't need to rant about the wiki right now
Guest60 has quit [(Quit: My Mac Pro has gone to sleep. ZZZzzz…)]
Supersonic112 has quit [(Ping timeout: 252 seconds)]
Itkovian has joined #nixos
<hyphon81> OpenStack is good at the server has Many core and big memory.
Supersonic112 has joined #nixos
<nalice> wooo, success! boot.zfs.devNodes = "/dev"; did the trick
<nalice> thanks
<timclassic> I'm playing with the new overlay functionality. Why is it important that I use super instead of self to refer to functions defined in nixpkgs?
<benley> nalice: woot. Do you have any idea why that's necessary? I've never used it
<timclassic> Is it something to do with fixed point?
<nalice> yes, i do. the default is /dev/disk/by-id, but my disk doesn't show up there. why, i have no idea. it's a just a vmware image. hyphon81 had the same issue so it's not unheard of.
byteflame has joined #nixos
<nalice> although, "This should be a path under /dev containing stable names for all devices needed, as import may fail if device nodes are renamed concurrently with a device failing.", so it's not an optimal solution.
codeape has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<nalice> (shouldn't matter for a laptop though, if the single drive fails, well...)
<gchristensen> :lickability:
<gchristensen> +:information_source:
<gchristensen> +:btrain:
<gchristensen> +:t:
<gchristensen> +:atrain:
<gchristensen> +:rtrain:
<gchristensen> +:dtrain:
<gchristensen> +:ctrain:
<gchristensen> +:ophiuchus:
<gchristensen> ugh
* gchristensen goes to fix that now
<clever> lol
<LnL> :p
<gchristensen> for whatever reason my keyboard won't unmap that key, and I have to read the docs on how to mount its config filesystem to fix it :(
<NixOS_GitHub> [nixpkgs] ndowens opened pull request #24905: speedtest-cli: 1.0.3 > 1.0.4 (master...speedtest-cli) https://git.io/vS7TB
pie_ has joined #nixos
plumps has joined #nixos
<ToxicFrog> `Ok, this is weird
<pie_> is there a way to put my cpu into a lower power state so my fans arent so loud?
<ToxicFrog> I have a bunch of systems on nixos unstable
<ToxicFrog> All of them have environment.extraOutputsToInstall = ["doc" "devdoc" "man"]
<ToxicFrog> On one of them, that setting causes collisions on 'nixos-rebuild'
Gravious has quit [(Remote host closed the connection)]
<plumps> Hey guys, anyone an idea why /nix/var/nix/gcroots/profiles/per-user/plumps/channels/ keeps a channel which I deleted a long time ago
<clever> ToxicFrog: if programs.man is enabled, man will already be in that config
mizu_no_oto has quit [(Quit: ["Textual IRC Client: www.textualapp.com"])]
<pie_> gchristensen, just temporarily though :I
<clever> plumps: did you nix-channel --update after deleting it?
<pie_> ah well ok
<clever> plumps: and double-check nix-channel --list
<plumps> @clever sure
<plumps> there I have nixpkgs and monte only
<pie_> gchristensen, how do i know what governors i have?
<ToxicFrog> clever: ok, programs.man is on by default, so it should be on on everything
<clever> plumps: and your running nix-channel as the plumps user?
<plumps> si
<gchristensen> pie_: what cpu do you have?
<clever> ToxicFrog: but that wont cover doc and devdoc, i dont see those in defaults
<ToxicFrog> clever: but the collisions all look like this:
<ToxicFrog> collision between `/nix/store/sbhqs7d7j7n7mx0zmf7174gxybbj22b4-perl-5.22.2-devdoc/share/man/man3/diagnostics.3.gz' and `/nix/store/0miiph0pp9gnxdc3wf37fypdzrqzcc6n-std-man-pages-4.4.0/share/man/man3/diagnostics.3.gz'
<ToxicFrog> So devdoc and man have started providing the same files, and it doesn't like that
<clever> ToxicFrog: oh, thats because both perl and std-man-pages provide the same man page
<pie_> gchristensen, some intel i5
<pie_> i think
<nalice> avn: what was that about copying kernels? i need to do something in addition to what nixos-rebuild provides in case i end up with a system that doesn't boot?
<ToxicFrog> clever: yeah, but apparently they didn't used to
<clever> nalice: if /boot is on a different filesystem, it should automaticaly copy
<benley> hrm. It would be so cool if you could put a git repository url in NIX_PATH and have that magically work.
<ToxicFrog> I'm wondering if I can turn off doc and devdoc now without losing the man pages that I originally turned them on to get
<clever> ToxicFrog: perl may have added that manpage recently
<gchristensen> try `cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor`
<clever> ToxicFrog: you will loose some of the perl man pages, but not the std-man-pages
<pie_> gchristensen, says performance
<ToxicFrog> Yeah, it wasn't the perl man pages I wanted devdocs for, but I don't remember what I *did* want them for
* ToxicFrog checks history
<nalice> clever: i did create a ext4 partition for /boot, so i suppose i'm in the clear
<gchristensen> pie_: try echo "powersave" | sudo tee : you will loose some of the perl man pages, but not
<pie_> gchristensen, yeah i was about to, thanks
<ToxicFrog> Past me wrote a completely useless commit message
<ToxicFrog> That jerk
<NixOS_GitHub> [nixpkgs] ndowens closed pull request #24905: speedtest-cli: 1.0.3 > 1.0.4 (master...speedtest-cli) https://git.io/vS7TB
<pie_> gchristensen, uh i have 4 cores tho
<gchristensen> pie_: yeah but did it work?
<clever> plumps: not sure why it would be doing that
<plumps> tried to delete it manually
<pie_> gchristensen, also apparently this is a thing cat /sys/devices/system/cpu/cpufreq/policy0/scaling_available_governors
<pie_> i have powerave and performance htere
<gchristensen> pie_: I don't have that ,but neat
<gchristensen> ok powerManagement.cpuFreqGovernor = "powersave"; ):)
<pie_> gchristensen, havent tried yet, i can govern cores individually? 0.o
<gchristensen> yep
<clever> pie_: i think that depends more on the hardware
<pie_> not that thats surprising...because we arent in the 2000 anymore, but im surprised
<gchristensen> I powersave 47 cores on one of my boxes and pin a particular process to the 48th core and set it to performance
<NixOS_GitHub> [nixpkgs] ndowens opened pull request #24906: speedtest-cli: 1.0.3 > 1.0.4 (master...speedtest) https://git.io/vS7kt
<plumps> @clever I will file an issue
magnetophon has quit [(Ping timeout: 240 seconds)]
<clever> gchristensen: 48 cores isnt a normal machine, lol
<pie_> um.
<pie_> :P
<NixOS_GitHub> [nixpkgs] adamruzicka opened pull request #24908: cjdns: 18 -> 19.1 (master...cjdns-v19.1) https://git.io/vS7kY
<pie_> gchristensen, well, i set it dunno if it worked
<pie_> i mean in /sys
<gchristensen> I don't do it on my 96 core machine tho ;)
<pie_> lol
lezed1 has quit [(Remote host closed the connection)]
<MichaelRaskin> clever: «48 cores is not a normal machine» — you forgot «yet»
byteflame has quit [(Ping timeout: 240 seconds)]
cpennington has quit [(Remote host closed the connection)]
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
dbmikus has quit [(Quit: WeeChat 1.7)]
<pie_> well powersave certainly didnt clock to minimum
<pie_> according to cpufreq this can clock down to 500mhz :O
<pie_> 500000 actually
<pie_> idk what that is
<pie_> 500khz?
<MichaelRaskin> benley: I think the user-interaction shim works by the logic that UEFI settings are password-protected, and so the loader has to be the same, and the same user who installed it in the first place was asked to show the good key, and now you need to pass a password check to change the configuration
<clever> pie_: when i first saw cpufreq, there was no kernel governers, only a userland governer
<MichaelRaskin> pie_: cpufreq measures in kiloherz, I think
<pie_> ah looks like theres a 1k multiplier since max is 2700000 presumably 2.7ghz
<clever> pie_: plus powernowd, that monitors cpu usage and writes orders into /sys
<jack[m]1> gchristensen: The skylake-EX machines look interesting. ;)
<benley> MichaelRaskin: does that protect against the second stage bootloader being changed on disk out-of-band?
<MichaelRaskin> benley: if you chainload unprotected GRUB, well, you have chosen that and jumped enough hoops. If you sign your own GRUB image with needed modules and forbid loading more modules, you get nicer protection
<gchristensen> jack[m]1: on packet?
<gchristensen> jack[m]1: boy howdy are they a treat
<benley> aha, ok
<benley> so really what I want, if I'm going to go to the trouble of setting up secure boot, is signing my own grub image right?
<MichaelRaskin> You can load signature-checking modules into the protected GRUB2 image
<benley> assuming my goal is end-to-end firmware security
<clever> benley: another thing to keep in mind, if somebody does bypass secureboot, they can trivialy lie to your OS and say secureboot is still enabled
<MichaelRaskin> If you want end-to-end, you want to enroll your own keys…
<clever> benley: its just a function in the EFI that returns boolean and has no security checks on it
<jack[m]1> gchristensen; 28 core/56 thread, 3.9GHz per socket, 8-socket.
<benley> Argh, so it's fundamentally not securable?
<gchristensen> jack[m]1: *droooool*
<benley> I was hoping to get integrity guarantees roughly on par with what ChromeOS has.
<clever> benley: everything relies on the bios not being reconfigured to disable secureboot
<clever> benley: but if somebody can yank out the hdd and run it in a vm, or modify the bios config, your hosed
<jack[m]1> gchristensen; kinda like a GPU, except x86 cores. ;)
<gchristensen> and way faster than the ~600MHz gpus operate at :)
<wak-work> jack[m]1: but not like 8 gpus
<gchristensen> I have high end GPUs sitting in my closet that I can't even use, because I have no use for them :(
<clever> benley: and as an example, if you have a signed grub, signed linux kernel, but dont have signed kernel modules, an attacker could insmod a custom module, that now has ring0 and hijacks the machine
<clever> benley: and that custom module could then re-boot the entire OS under a hypervisor, and claim secureboot is working, but spy on everything
<benley> bleh, yeah
<jack[m]1> wak-work: heh, yeah, only like 165W a piece instead of 350W. ;)
<clever> benley: for this reason, MS demands that all signed code will only ever execute code that has also been signed
<gchristensen> jack[m]1: or 350W each...
<gchristensen> oh you said a piece :)
<clever> benley: and given how easily you can rebuild kernel modules in nix, good luck signing those
Gravious has joined #nixos
<benley> clever: but they signed the bootloader shim that you can use to execute unsigned grub?
Itkovian has joined #nixos
<clever> benley: they forgot to remove a debug feature, and now the core secureboot key for M$ is useless :P
<benley> oy
<clever> attempting to revoke that key would brick every install cd out there
<MichaelRaskin> Not really
<benley> okay so ... do you think I should bother with secure boot, or is it a pointless exercise?
<pie_> ugh, how can i run a graphical application with sudo?
<MichaelRaskin> I mean, there is this shim that can be used to execute unsigned GRUB if the user actively wants that
<MichaelRaskin> Basically, what is your threat model?
<gchristensen> !m MichaelRaskin
<[0__0]> You're doing good work, MichaelRaskin!
<benley> MichaelRaskin: make it more difficult for someone who gets temporary physical access to my laptop to hijack it by compromising the bootloader
codeape has joined #nixos
<MichaelRaskin> gchristensen: that really feels keyword-triggered!
<benley> and help improve the security features of nixos in general.
<maurer> benley: So, the easiest way to do that is not actually secureboot
<gchristensen> MichaelRaskin: discussions o threat model are good!
<maurer> benley: You want to look at measured boot + TPM
<clever> benley: one thought is TPM measured boot
<benley> huh, I am not familiar with measured boot.
<maurer> benley: basically what you do is you put part of your disk key in the TPM
<maurer> benley: Then, when you boot, the TPM basically hashes the trace of the boot process
seagreen has quit [(Ping timeout: 252 seconds)]
<clever> benley: i believe it works by having each stage of the bootloader writing the hash of the next stage to the TPM, as it boots
<maurer> (not literally, but this is a good analogy without getting into impl)
<benley> neat.
<clever> benley: as a crude example, when the bios runs grub, the hash(grub) gets written to the tpm
<clever> benley: and when grub loads linux, it writes hash(linux) to the tpm
<maurer> Then it gives you access to the any keys that match
<pie_> huh that sounds cool
<maurer> It's not too bad to set up _once_, the thing I never managed to get working myself is the handoff
<maurer> which is very important
<clever> benley: so if all of the binaries are un-modified, the same series of hashes gets written to the tpm, and you can unlock the key for decrypting the hdd
<pie_> um
<maurer> by handoff, I specifically mean "I'm upgrading the kernel, how do I make the new kernel able to get the old keys"
<benley> neat.
<pie_> but what if you just swap the drive
<pie_> and use the same binaries
byteflame has joined #nixos
<pie_> and then get the ke
<MichaelRaskin> The only minor problem is that this sequence changes every time you rebuild the initrd
<maurer> pie_: If you use the same binaries, you don't get the key
<pie_> uh, nevermind.
goibhniu has joined #nixos
<clever> pie_: those binaries would have to be secured to never reveal the key
<maurer> the software does, and then you can't log in :P
<maurer> Also, that's why I didn't say "put the disk key in the TPM"
<MichaelRaskin> I guess handoff is even funnier with rollbacks
<clever> MichaelRaskin: yeah, the power of nixos also makes many things harder
<maurer> and instead said "put part of the disk key into the TPM"
<maurer> the idea is that you perform key derivation either direct from the TPM key, or the TPM key + a password
<maurer> to get the disk crypto key
<maurer> so the thing that's actually in memory after boot is not the TPM secret
<maurer> In any case, tl;dr hardware can do what you want, configuring it to actually work involves a lot of chicken blood and dancing naked in the moonlight
<MichaelRaskin> I have a suspicion that chicken blood doesn't count unless _the chicken_ has danced plucked in the moonlight
seagreen has joined #nixos
<gchristensen> hehe
<benley> I maurer I do like the sound of this
<pie_> also may or may not involve losing your virginity ot a sacrificial virgin
<maurer> It would be really cool if you could get this on nixos, since aiui it is not a default available feature in any other distro
<benley> maurer: I do like the sound of this
<pie_> which may or may not be a chicken
<maurer> which would be a pretty cool claim to fame
<pie_> depending on the phase of the moon
<maurer> but I also know from experience that it's difficult and brittle
<benley> I dunno if I'll be able to dedicate a lot of time to this but I'm going to look into it
<maurer> and making it not so would probably require a bit of effot
<maurer> Cool!
<maurer> good luck!
<gchristensen> maurer: it is extra hard on nixos IIRC
<maurer> gchristensen: Due to generation swapping you mean?
<gchristensen> because our build process is so transparent, we have to means to hide keys
<pie_> well, the good thing about nixos is maybe it can be gotten to work consistently :D
<gchristensen> have no means*
<wak-work> right but you can sign the binaries after the fact
<wak-work> so you don't need the build process to do it
<NixOS_GitHub> [nixpkgs] ck3d opened pull request #24909: dhcpcd service: clear exit code of exitHook (master...dhcpcd-clear-exit-code) https://git.io/vS7L3
<pie_> theres been a discussion on a github thread for a while now about a way to have secrets yeah?
<wak-work> you just need an out of band signing process
<gchristensen> yeah we just don't really have anything that can do that
<maurer> gchristensen: This is not secure boot
<maurer> gchristensen: it's Intel TXT
<wak-work> and also you can add impure paths to the nix builder
<maurer> gchristensen: there's no signing keys
<wak-work> and require that a key for signing be one of those
<gchristensen> ahhh
Guest60 has joined #nixos
* gchristensen takes it all back
<maurer> Secure Boot (the signing-key mechanism) is present on other distros (ubuntu+redhat) already
<gchristensen> we can do whatever we want when the rubber meets the road at activation time :P
<maurer> since they do central distribution
<maurer> benley: https://wiki.gentoo.org/wiki/Trusted_Boot may be a good starting point
* benley scribbles notes into a text file
drasich has joined #nixos
<wak-work> the crappy thing is how expensive TPM's are for some boards
<maurer> wak-work: basically every modern intel chip has an onboard tpm
<maurer> You probably have one and just aren't aware
<maurer> s/onboard/ondie/g
<ToxicFrog> ok, the problem is with python27.withPackages
<wak-work> huh?
<gchristensen> holy guacamole nixos' docker test passed on aarch64
<copumpkin> \o/
<maurer> gchristensen: Time to install nixos on iphones everywhere?
<gchristensen> lol sure
<maurer> wak-work: You mentioned that TPMs were expensive - I was saying that you usually have one anyways
<maurer> wak-work: because it comes inside the CPU
<wak-work> my e5-1650v4's have them?
<wak-work> and my e5-2670's
<copumpkin> maurer: slightly different signed boot process on there :(
<gchristensen> "Intel® Trusted Execution Technology for safer computing is a versatile set of hardware extensions to Intel® processors and chipsets that enhance the digital office platform with security capabilities such as measured launch and protected execution. It enables an environment where applications can run within their own space, protected from all other software on the system." on the 2670
<gchristensen> not sure if measured launch is the same as measured boot
Avogadro has quit [(Ping timeout: 260 seconds)]
<wak-work> you still need a TPM for that though
<wak-work> TXT support comes from the external TPM + CPU
<ToxicFrog> All I want is to be able to 'import mutagen' from python without using nix-shell, why is it this hard >.<
<gchristensen> ToxicFrog: pythonWithPackages? :/
<ToxicFrog> gchristensen: is that a thing? last time I asked someone pointed me at python27.withPackages()
<gchristensen> maybe I'm thinking python35.withPackages
<maurer> wak-work: I could have sworn the TPM had been moved onboard
<maurer> err, on die
MP2E has joined #nixos
<maurer> because if it wasn't, you can replay measured boot readings to the TPM
<maurer> unless there's some other communication channel stuff I'm unaware of
vandenoever has quit [(Ping timeout: 252 seconds)]
<ToxicFrog> But either it doesn't play nice with configuration.nix or I'm holding it wrong
<gchristensen> I think you're holding it wrong
<maurer> but everything I'm reading about TXT availability is assuming the TPM lives on the board, so I guess you're right?
<wak-work> maurer: im pretty sure that's an attack vector
<gchristensen> ToxicFrog: can you paste info?
<ToxicFrog> gchristensen: do you know where I can find a working example of using it with systemPackages (as opposed to in a package derivation)?
<pie_> thanks for the help earlier btw
<ToxicFrog> Not much to paste; it's just: environment.systemPackages = with pkgs; [ ... python27.withPackages (ps: [ps.mutagen ps.websocket_client]) ... ];
<copumpkin> gchristensen: now I'm wishing I was on a T2
<gchristensen> oh hehe
<pie_> also found cpupower on the arch wiki, ended up using that
<copumpkin> gchristensen: someone pushed some sort of kernel update that's not in hydra yet to master :)
pie_ has quit [(Quit: Leaving)]
<clever> ToxicFrog: needs another ( and ) around the whole function call
<clever> ToxicFrog: lists in nix are picky
<gchristensen> [ ... (python27.withPackages (ps: [ps.mutagen ps.websocket_client])) ... ]
<gchristensen> ToxicFrog: ^
<ToxicFrog> ....ahaaaaaa
* ToxicFrog mashes face into keyboard
<gchristensen> yup
<ToxicFrog> Yep, that fixed it
<ToxicFrog> Thank you
<gchristensen> great :) you're welcome
<gchristensen> copumpkin: :o
<ToxicFrog> I guess that gets parsed into [package, package, ..., function, function, package, ...]
<gchristensen> yup
plumps has quit [(Quit: Textual IRC Client: www.textualapp.com)]
seagreen has quit [(Ping timeout: 252 seconds)]
mjhoy has quit [(Ping timeout: 260 seconds)]
simukis_ has quit [(Ping timeout: 260 seconds)]
byteflame has quit [(Ping timeout: 252 seconds)]
ambro718 has quit [(Ping timeout: 240 seconds)]
shadowx has joined #nixos
ryantrinkle has quit [(Ping timeout: 258 seconds)]
ryanartecona has joined #nixos
<nalice> if anyone's interested, i found the commentary on zfs memory requirements i was thinking about https://www.reddit.com/r/DataHoarder/comments/5u3385/linus_tech_tips_unboxes_1_pb_of_seagate/ddqzfew/
<nalice> (and no, it has nothing to do with linus tech tips!)
shadowx has quit [(Quit: Leaving)]
<clever> nalice: i have recently been considering using an NVME drive as an L2ARC in my NAS
AllanEspinosa has quit [(Ping timeout: 240 seconds)]
shadow-x has joined #nixos
<gchristensen> neat, nalice
AllanEspinosa has joined #nixos
<gchristensen> LTT is fun to watch sometimes, anywho :)
<clever> gchristensen: his april fools video involved using hairspray to blast a flame thrower out of the data closet air vent and faking a server "room" fire, lol
<gchristensen> lol
<gchristensen> "It appears that Linus is using gluster only because he does not know how to connect 100 drives to a single machine." or maybe they want redundancy across machines
<gchristensen> "only because" implies incompetency, which may not be the case
dpren has joined #nixos
<clever> i have had many problems with NBD before, but not much with iSCSI
<gchristensen> "Also, Gluster ought to run on the Raspberry Pi." ok I no longer believe this person's posts
<clever> gchristensen: have you seen this board i'm looking into? http://www.banana-pi.org/r1.html
hiratara has quit [(Ping timeout: 258 seconds)]
dpren has left #nixos []
<gchristensen> oohh
<clever> gchristensen: dual core 912mhz, proper sata, proper gigabit ethernet
<clever> gchristensen: and i have nixpkgs u-boot running on it: https://gist.github.com/cleverca22/8ff5bd6a322c45f5a3bf7e6109e03e7a
<clever> it just needs a partition table, kernel, initrd, and rootfs, and it should be good to go
<gchristensen> nice...
<NixOS_GitHub> [nixpkgs] bjornfor pushed 2 new commits to release-17.03: https://git.io/vS7Oe
<NixOS_GitHub> nixpkgs/release-17.03 9f64797 romildo: lxqt: fix qtsvg attribute...
<NixOS_GitHub> nixpkgs/release-17.03 76c6313 romildo: lxqt-panel: fix for `explicit` (a C++11 keyword) being used as variable...
<gchristensen> $50?
<NixOS_GitHub> [nixpkgs] bjornfor closed pull request #24896: lxqt: fix packaging errors (release-17.03...release-17.03) https://git.io/vSQIs
<clever> cant remember how much it was, too forever to get it into canada
<clever> i had it shipped to a friend in america, and then we had shipping complications and i wound up picking it up when i flew down to bayhac
systemfault has joined #nixos
<gchristensen> yikes
<clever> gchristensen: http://www.banana-pi.org/m2u.html another model, without the 5 ethernet ports
<gchristensen> less interesting IMO :)
<clever> yeah, depends on if your wanting a router or a nas
<nalice> just don't take linus storage tips seriously ;)
<clever> i am seeing a lipo connector on a lot of these
<clever> i think they have internal battery backup
<nalice> LTT has several videos of their homebrew storage solutions breaking down
<clever> and they did have a homebrew watercooled server mounted OVER the petabyte storage array
<clever> and one day, he stepped on a power bar and cut power to half the rack
<nalice> it's entertaining storage only :)
<clever> just a standard office powerbar with a switch on it
<clever> gchristensen: the R2 also have mini-pcie!
<gchristensen> O.O
aminechi1haoui has joined #nixos
hiratara has joined #nixos
griff_ has quit [(Quit: griff_)]
<nalice> gluster doesn't run on rpi? that would be really neat to stick a board to single drives connected them over the network. you'd probably would get to make another "lost my data" video later, but neat.
<nalice> maybe not rpi in particular... no sata.
ryantrinkle has joined #nixos
<nalice> does it use a lot of memory?
aminechikhaoui has quit [(Ping timeout: 260 seconds)]
<clever> nalice: http://www.banana-pi.org/m2u.html real sata and ethernet on this board
<ToxicFrog> Argh
<ToxicFrog> Finally got the server updated and everything seemed to go fine, except now the windows machine can't connect to any of the smb shares
<copumpkin> this started happening to my install test as soon as I rebased onto master
<gchristensen> not good, copumpkin ...
<MichaelRaskin> I wonder at what point ARM boards become a cost-effective alternative to desktops from raw power point of view. I mean, fifteen m2u is a nice amount of power, even if cluster only, but even for compilation raw MHz count is not a reliable measure, especially cross-arch
<NixOS_GitHub> [nixpkgs] grahamc pushed 2 new commits to unstable-aarch64: https://git.io/vS736
<NixOS_GitHub> nixpkgs/unstable-aarch64 944380c Graham Christensen: netboot: don't appear to neeed syslinux / grub
<NixOS_GitHub> nixpkgs/unstable-aarch64 b4f0f26 Graham Christensen: netboot: aarch64 compatability
<hyphon81> BPI-R2 has mini PCIE interface... If the exchange to PCIE x1 connector will be used, perhaps we can use Nvidia video cards? Maybe, power source is lackage...
<gchristensen> MichaelRaskin: for many builds, the 96-core box has been great. if I'm working on a single compilation over and over and over, it it brutal
<NixOS_GitHub> [nixpkgs] armijnhemel opened pull request #24910: RPM: 4.13.0-rc1 -> 4.13.0.1 (master...rpm) https://git.io/vS73D
<clever> ive also found ram and other things to complicate the problem
<clever> if you make -j96, and you hit a derivation that lacks enableParallelBuild = true;, it will waste time on 1 core
stepcut has quit [(Remote host closed the connection)]
<gchristensen> but making squash FSs in parallel with 96 threads? darling
<clever> and in the case of my rpi, if i make -j4 with a max of 1 job, and it hits glibc, a single task within that build needs ~3gig of ram
<clever> and it fights 3 other gcc's over it, and swaps for hours
<MichaelRaskin> Which is why we should try having an option to force parallel builds unless they are _known_ to be problematic, but oh well
AllanEspinosa has quit [(Ping timeout: 260 seconds)]
<clever> i need to force a single step to be done without any contention, and parallel the rest of the build
<MichaelRaskin> Even for parallel stuff, is raw integer IPS up to deskop i7's?
<MichaelRaskin> Last when I checked it was a long way to go
ToxicFrog has quit [(Quit: WeeChat 1.5)]
<MichaelRaskin> But maybe in the couple of years that have passed…
<gchristensen> load average: 81.55, 159.49, 206.58
<ixxie> does anybody here run NixOS with a Haskell shell xD
goibhniu has quit [(Ping timeout: 260 seconds)]
<barrucadu> A Haskell shell?
<hyphon81> Hmm. BPI-R2 has sata. However, How can we get power source for HDD or SSD?
<MichaelRaskin> Shouldn't it kill your willpower with constant IO monad lifting?
<gchristensen> hehe MichaelRaskin
<clever> hyphon81: yeah, the R2 lacks the sata power connector, so you need an external supply that powers both the hdd and the R2
<clever> hyphon81: the R1 has a full data+power connector like a laptop, and you just slot the whole drive in
AllanEspinosa has joined #nixos
<hyphon81> Oh, it has a SATA power connector on the board.
<NixOS_GitHub> [nixpkgs] grahamc pushed 1 new commit to unstable-aarch64: https://git.io/vS7sF
<NixOS_GitHub> nixpkgs/unstable-aarch64 b13e7c8 Graham Christensen: fixup netboot
<ixxie> barrucadu: apparantly, somebody made this : https://github.com/Gabriel439/Haskell-Turtle-Library
ToxicFrog has joined #nixos
<clever> hyphon81: yeah, the R1 seems simpler for powering the drive: http://www.banana-pi.org/images/bpi-images/R1/r5.jpg
<clever> ixxie: the gist i linked replaces the entire OS with a single haskell binary
<hyphon81> The power is 12V/2A... Perhaps, can we use a graphic card....?
<clever> ixxie: so the haskell prog gets ran as pid 1 at bootup, and that does everything
<ixxie> clever: is that for your minimal OS?
<clever> ixxie: a different minimal os i threw together in one night
indi_ has joined #nixos
ryantrinkle has quit [(Ping timeout: 252 seconds)]
<hyphon81> I would like to attempt. Someday I will attempt it. :)
<ixxie> clever: one day you will write an OS in a one-line shell command
<ixxie> clever: I have to say I admire the elegance of this, even though I cannot understand it yet (thinking of learning Haskell soon)
mjhoy has joined #nixos
<clever> ixxie: yeah, i'm still learning haskell as well, had some help from dmj` and taktoa when writting that part of it
indi_ has quit [(Ping timeout: 240 seconds)]
hiratara has quit [(Quit: ZNC - http://znc.in)]
hiratara has joined #nixos
<NixOS_GitHub> [nixpkgs] grahamc force-pushed unstable-aarch64 from b13e7c8 to e0a17a3: https://git.io/vSSlm
<NixOS_GitHub> nixpkgs/unstable-aarch64 eab1f52 Graham Christensen: qemu: support qemu-system-aarch64 in the system wrapper
<NixOS_GitHub> nixpkgs/unstable-aarch64 53b42cb Graham Christensen: kernel: delete fewer files, which are necessary on aarch64
<NixOS_GitHub> nixpkgs/unstable-aarch64 b219a39 Graham Christensen: netboot: don't appear to neeed syslinux / grub
<gchristensen> hrm. that might have been a mistake :)
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
boxofrox has joined #nixos
<boxofrox> is there a way to create a disabled user account in configuration.nix? i.e. set user shell to /bin/false or /bin/nologin.
<clever> boxofrox: should be as simple as just setting the shell to "/run/current-system/sw/bin/false" i believe
<gchristensen> yeah that rebase was a mistakee :(
<boxofrox> clever: i'll give it a try. i didn't think that would work if the program wasn't listed in /etc/shells. thanks.
sellout- has quit [(Quit: Leaving.)]
codeape has quit [(Quit: Textual IRC Client: www.textualapp.com)]
mjhoy has quit [(Ping timeout: 240 seconds)]
<hyphon81> Now, mastodon active instances are 628 in the world. Everybody are favorit mastodon.
<gchristensen> wow, 600 new passing jobs on aarcch64... that was actually a very good rebase.
<hyphon81> The nixpkgs issue for mastodon had been risen. https://github.com/NixOS/nixpkgs/issues/24600
<NixOS_GitHub> [nixpkgs] dezgeg pushed 1 new commit to master: https://git.io/vS7na
<NixOS_GitHub> nixpkgs/master 28f87e4 Tuomas Tynkkynen: stdenv: ARM bootstrap: Update bootstrap tarballs to latest hydra-built ones...
dridus has quit [(Quit: znc? wassat?)]
<hyphon81> gchristens: That's amazing.
<NixOS_GitHub> [nixpkgs] benley closed pull request #24904: latte-dock: backport to 17.03 (release-17.03...latte-dock-backpot) https://git.io/vS7JU
ryanartecona has quit [(Quit: ryanartecona)]
<NixOS_GitHub> [nixpkgs] benley closed pull request #24891: hugo: 0.20 -> 0.20.1 (master...pr/hugo) https://git.io/vSH9i
digitus has quit [(Quit: digitus)]
<NixOS_GitHub> [nixpkgs] benley pushed 2 new commits to master: https://git.io/vS7cW
<NixOS_GitHub> nixpkgs/master 3f69d10 Peter Jones: atomicparsley: 0.9.0 -> 0.9.6
<NixOS_GitHub> nixpkgs/master aeebfd9 Benjamin Staffin: Merge pull request #24903 from pjones/pjones/atomicparsely...
<NixOS_GitHub> [nixpkgs] benley closed pull request #24906: speedtest-cli: 1.0.3 > 1.0.4 (master...speedtest) https://git.io/vS7kt
<NixOS_GitHub> [nixpkgs] benley closed pull request #24902: jetbrains.idea-{community,ultimate}: 2017.1 -> 2017.1.1 (master...pr/idea-2017-1-1) https://git.io/vSQNp
Gravious has quit [(Remote host closed the connection)]
Guest60 has quit [(Quit: My Mac Pro has gone to sleep. ZZZzzz…)]
<NixOS_GitHub> [nixpkgs] benley pushed 2 new commits to master: https://git.io/vS7cS
<NixOS_GitHub> nixpkgs/master c1ef227 Tristan Helmich: radarr: 0.2.0.535 -> 0.2.0.596
<NixOS_GitHub> nixpkgs/master f399ea9 Benjamin Staffin: Merge pull request #24842 from mayflower/radarr_596...
takle has quit [(Remote host closed the connection)]
<NixOS_GitHub> [nixpkgs] benley closed pull request #24883: fanficfare: init at 2.9.0 (master...fanficfare) https://git.io/vSHeZ
markus1189 has joined #nixos
hamishmack has joined #nixos
markus1199 has joined #nixos
<NixOS_GitHub> [nixpkgs] benley pushed 3 new commits to master: https://git.io/vS7Cq
<NixOS_GitHub> nixpkgs/master 39e95aa Vince v. Oosten: add disk storage to libvirt
<NixOS_GitHub> nixpkgs/master 00c3bd1 Vince van Oosten: add parted to build inputs
<NixOS_GitHub> nixpkgs/master cdbcaa5 Benjamin Staffin: Merge pull request #24822 from techhazard/add-disk-storage-to-libvirt...
markus1219 has quit [(Ping timeout: 258 seconds)]
markus1209 has quit [(Ping timeout: 258 seconds)]
agjacome has quit [(Remote host closed the connection)]
sorin has joined #nixos
sorin is now known as Guest44006
Guest44006 has quit [(Client Quit)]
sophiag has joined #nixos
<sophiag> is there really not a nix pkg for libboost-test-dev? i figured it would have been included in the regular boost package since it's a fairly common dependency
<sophiag> come to think of it, cmake's error messages make it a bit unclear as to whether it could even locate boost at all. sometimes i feel like rebuild switch isn't enough and i need to reboot after installing certain packages. can anyone shed light on this?
hamishmack has quit [(Quit: hamishmack)]
takle has joined #nixos
bennofs has quit [(Ping timeout: 260 seconds)]
takle has quit [(Ping timeout: 268 seconds)]
erasmas has quit [(Quit: leaving)]
justanotheruser has quit [(Ping timeout: 258 seconds)]
<clever> sophiag: boost-test is inside the main boost package
<clever> sophiag: just add boost to buildInputs and it should work
mjhoy has joined #nixos
justanotheruser has joined #nixos
<sophiag> clever: thanks, that's what i thought. it seems the issue is likely that i don't even have buildInputs configured
stepcut has joined #nixos
ixxie has quit [(Ping timeout: 240 seconds)]
takle has joined #nixos
mjhoy has quit [(Ping timeout: 240 seconds)]
systemfault has quit [(Quit: Bye!)]
takle has quit [(Ping timeout: 258 seconds)]
<sophiag> is the idea you add stdenv.mkDerivation for every package you need to build manually?
cpennington has joined #nixos
<clever> yeah
<clever> make a default.nix with something like this: with import <nixpkgs> {}; stdenv.mkDerivation { name="foo"; buildInputs = [ boost ]; src = ./.; }
<sophiag> i'm also building something that *is* already in nix packages yet didn't seem to install most of what i need (and has no maintainer listed for me to ask). i'm wondering if perhaps it just didn't fully install since i didn't already have all the dependencies declared?
<clever> sophiag: what package is missing what files?
<sophiag> it's simgrid, so not exactly common yet seeing as i try very hard to avoid manual builds i didn't have cmake, boost, or python3 installed
<sophiag> oh and it's missing the header files for the C interface (at least, maybe it's missing everything?)
<clever> when you install anything with nix-env -i, it will ignore the headers
<clever> thats how nix works
<clever> the headers will only be available inside nix-build and nix-shell
<sophiag> i never use nix-env -i. i just maintain everything in configuration.nix and call nixos-rebuild switch
<clever> systemPackages will also ignore header files
<sophiag> ah ok
<clever> if you add simgrid to the buildInputs and run nix-build, it will just find the headers
<sophiag> oh ok
<sophiag> so literally just simgrid in the buildInputs and i don't have to worry about the dependencies like boost, etc.?
takle has joined #nixos
<clever> yeah
<clever> you should never have to put dependencies into systemPackages or nix-env -i