<noonien>
how do multiple people work with nixops?
onicrom has joined #nixos
<onicrom>
hello all, im having one helluva time install nix on my ubuntu 18.04 machine... which is also an lxc container ... which im accessing via lxc-attach
m0rphism has quit [Ping timeout: 246 seconds]
kim0 has quit [Quit: Connection closed for inactivity]
<Fare>
Is anyone using Nix with Qubes ?
shibboleth has joined #nixos
jluttine has quit [Ping timeout: 246 seconds]
<onicrom>
error: while setting up the build environment: unable to make '/' private mount: Permission denied
<onicrom>
one issue, if you sudo exits non-szero because you don't have a real tty becaue you're using lxc-attch the install stops, guessing its sh -e
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
hamishmack has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
rcshm has quit [Remote host closed the connection]
rcshm has joined #nixos
doyougnu has joined #nixos
rcshm_ has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
jluttine has quit [Ping timeout: 272 seconds]
pacmeister has quit [Remote host closed the connection]
<jackdk>
pandoc-2.7 just landed in my channel and now hakyll is failing to build (it depends pandoc < 2.7). This is default.nix for my website, but hakyll seems to not jailbreak properly. Anyone know what is going on? https://pastebin.com/8UKYrXMR
<clever>
jackdk: there is a bug in jailbreak-cabal, where it cant jailbreak constaints inside cabal if statements
wfranzini has quit [Remote host closed the connection]
wfranzini has joined #nixos
jtojnar has quit [Quit: jtojnar]
Supersonic has quit [Ping timeout: 252 seconds]
<jackdk>
I want to build nixos vms on my ubuntu machine. I have installed nix and added nixos-18.09 as a channel, but it doesn't appear in $NIX_PATH. How do I reference the local channel when building images? Either a -I param to `nix build` or an `import` expression are what I'm thinking of, but not sure of the correct incantation.
Supersonic has joined #nixos
work_ has quit [Quit: Connection closed for inactivity]
<clever>
jackdk: did you nix-channel --update after adding the channel?
sigmundv has quit [Ping timeout: 246 seconds]
PeterHK has joined #nixos
<{^_^}>
[nixpkgs] @veprbl merged pull request #57444 → pdfgrep: relax platform from linux to unix → https://git.io/fhjBo
<Ashy>
hmm, there's boot.kernel.sysctl, can the scheduler be set via sysctl?
<mconstant>
just installed nixos again and I can't get Intel integrated WiFi for my Lenovo Legion Y520 working with network_manager
boomshroom has joined #nixos
<boomshroom>
Hello!
<slack1256>
mconstant: can you get it working on another distro?
<boomshroom>
I think I managed to get a RedoxOS cross compiler!
<thoughtpolice>
The default IO scheduler is a global only setting (that can be set at boot, maybe sysctl too); drives get hot-plugged later on anyway long after boot so you can only specify a single default. Normally you write the scheduler into... some file... in sysfs after the disk is attached.
<thoughtpolice>
There's no method in the fileSystems configuration AFAIK for setting the scheduler on a particular drive. The normal way is to use systemd, like setting it after the filesystem mount has succeeded, for example.
domogled has joined #nixos
<thoughtpolice>
So you'll probably just have to rely on that. If it's your boot drive you can set it through boot.* -- if it's attached later on you'll need to set Requires= and After= on the mount target for your unit.
<boomshroom>
I'm not trying to make a stdenv to use it. The issue is that I need to make Nix aware of the compilation target, but support isn't upstream so I need a way to tell NixOS to use this compiler rather than upstream.
<thoughtpolice>
Maybe there's something in fileSystems you can use to do a post-mount/post-attach command, and just run 'echo ... > /sys/...' instead? There might be. Then you wouldn't need a custom unit, probably.
<CMCDragonkai>
thoughtpolice: My issue is that the default appears to be mq-deadline now. But I'm trying to reduce the IO priority of a aws sync task. As soon as I run it, it freezes all my other programs atm.
<CMCDragonkai>
thoughtpolice: ionice seems to only work with CFQ
tnks has quit [Ping timeout: 244 seconds]
<CMCDragonkai>
thoughtpolice: do you know of alternatives?
<thoughtpolice>
I think they dropped the non-multiqueue-block layer a release or two back, and unfortunately there's no elevator= argument (yet) for blkmq
<thoughtpolice>
So you can't just add 'elevator=whatever' to boot.kernelParams, unfortunately
<CMCDragonkai>
thoughtpolice: I don't know what elevator means in this case
<thoughtpolice>
It's just the weird name Linux gives to the IO scheduler (it's also in the API, etc).
reallymemorable has quit [Ping timeout: 272 seconds]
Glider_IRC_ has quit [Ping timeout: 246 seconds]
<thoughtpolice>
So you'll probably have to just write a simple systemd unit that does something like 'echo scheduler-name > /sys/block/<BLOCK_DEVICE_HERE>/queue/scheduler' after the device mounts
<thoughtpolice>
(If it's the root device, that target is "-.mount" IIRC)
<thoughtpolice>
Alternatively to all of this: roll back your kernel version a few releases, probably, if that's possible.
<CMCDragonkai>
What about `system.activationScripts.diskScheduler`?
<CMCDragonkai>
I mean you can create your own `activationScripts`. But that runs after each rebuild, not on boot.
<teto>
I have nix crashing on some nix code I really want to have working, how would you run nixos-rebuild with nix-dev ?
<CMCDragonkai>
thoughtpolice: > A set of shell script fragments that are executed when a NixOS system configuration is activated. Examples are updating /etc, creating accounts, and so on. Since these are executed every time you boot the system or run nixos-rebuild, it's important that they are idempotent and fast.
<thoughtpolice>
Sure, you can write an activation script. (Though it might not technically be correct in all cases it'll work for the root device almost certainly.)
<domenkozar>
gchristensen: o/
<domenkozar>
gchristensen: how to update firmware?
<domenkozar>
I sense I'll need windows?
<CMCDragonkai>
thoughtpolice: Do you know if there's an alternative to ionice that works for the new multi-queue schedulers?
hyperfekt has joined #nixos
<thoughtpolice>
CMCDragonkai: No. Deadline inherently does not support ionice "by design". Kyber, the other blk-mq scheduler, could in theory support it IIRC, but does not.
<thoughtpolice>
You can throttle a number of other ways though, like cgroups, which should work
<thoughtpolice>
But I don't think there's a nice command for that
<thoughtpolice>
Or alternatively just roll back linuxPackages, which will give you the old non-mq schedulers, unless you need a newer feature. That's at least a bandaid (and the LTS kernels with non-blk-mq should last for quite a while)
<etu>
In there I pull in the old stable channel and pick out a few packages that I overlay :-)
<eyJhb>
Hmm I see, but I think I am forced to go with running stable, as it is not the package itself that is broken, but the packages around it that broke it :/
Ariakenom has joined #nixos
talqu has joined #nixos
reinhardt has joined #nixos
agander has joined #nixos
drakonis_ has joined #nixos
bsima has quit [Quit: Bye!]
wfranzini has joined #nixos
drakonis has quit [Ping timeout: 252 seconds]
periklis has joined #nixos
<{^_^}>
[nixpkgs] @volth closed pull request #56953 → cc-wrapper: give -march=... flag less priority → https://git.io/fhpsH
jbgi has joined #nixos
lord| has quit [Read error: Connection reset by peer]
<Orbstheorem>
There are still some commands with manpages not found
__monty__ has joined #nixos
fusion809 has joined #nixos
<das_j>
Were the haskell packages moved from all-packages?
<das_j>
Or was there some other move?
<Orbstheorem>
For instance, I can access zathura's man page, but I get “zathura: nothing appropriate.” when I run `man -k zathura`
PeterHK has joined #nixos
<PeterHK>
i finally got containerbuild working with dockertools for a django app... its super messy :/ can someone review and leave some comments? i can share on pastebin
zupo_ has joined #nixos
zupo has quit [Read error: Connection reset by peer]
<Taneb>
I've got a Hydra job failing with "Output limit exceeded" but I think it was good to succeed. Is it possible to up the output limit
<Taneb>
?
fusion809 has quit [Remote host closed the connection]
sinner has joined #nixos
sinner is now known as Guest52151
<unacceptable>
be less noisy
<Taneb>
That's certainly one solution
<Taneb>
But not so easy, I'm using proprietary build tools and when things go wrong I may want all of the output as clues
<srhb>
Taneb: Are you talking about the log size or the output size? The latter is max_output_size
<Taneb>
case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[5]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[6]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[3]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[4]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[0]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[1]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|synchrotron_core_instructionstore_coreinstrstore_internal_
<Taneb>
ds11|synchrotron_core_inputbuffer_undefinedinputbufferinstr1_ds6|clash_components_fifo_dcfifodf1_1_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_1_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[2]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|clash_interface_circuit_fixc1_3_ds5_0|synchrotron_multicor
<Taneb>
e_corerow_1_result_0|\map_r_0:0:synchrotron_stn_router_serialising_router_0|synchrotron_stn_router_serialising_deeperrorx_ds10|clash_components_fifo_dcfifodf1_3_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_3_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[5]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|clash_interface_circuit_fixc1_3_ds5_0|synchrotron_multicor
<Taneb>
e_corerow_1_result_0|\map_r_0:0:synchrotron_stn_router_serialising_router_0|synchrotron_stn_router_serialising_deeperrorx_ds10|clash_components_fifo_dcfifodf1_3_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_3_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[6]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|clash_interface_circuit_fixc1_3_ds5_0|synchrotron_multicor
<Taneb>
e_corerow_1_result_0|\map_r_0:0:synchrotron_stn_router_serialising_router_0|synchrotron_stn_router_serialising_deeperrorx_ds10|clash_components_fifo_dcfifodf1_3_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_3_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[3]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<Taneb>
Critical Warning (19519): The instance assignment "POWER_UP_LEVEL" on register "u_pr|vfpga|synchrotron_synchrotronstratix3x6_ds4|synchrotron_core_core_internal_ds9_case_scrut|synchrotron_multicore_coregrid_ds8|synchrotron_multicore_coregrid_0_ds5|clash_interface_circuit_fixc1_0_ds6|synchrotron_multicore_corerow_result_0|clash_interface_circuit_fixc1_2_ds5_0|synchrotron_multicore_corerow_0_result_0|clash_interface_circuit_fixc1_3_ds5_0|synchrotron_multicor
<Taneb>
e_corerow_1_result_0|\map_r_0:0:synchrotron_stn_router_serialising_router_0|synchrotron_stn_router_serialising_deeperrorx_ds10|clash_components_fifo_dcfifodf1_3_internal_ds10_case_scrut|clash_primitives_common_fifo_dcfifo_3_ds1|\dcfifo_block:dcfifo_inst|auto_generated|wrptr_g[4]", in reconfigurable partition "u_pr", is ignored. Initial condition is not guaranteed during partial reconfiguration.
<srhb>
Taneb: Oops
<Taneb>
Big oops
<Taneb>
I'm... very sorry about that
<srhb>
Happen :-P
<srhb>
Taneb: Anyway, max_log_size for logs, max_output_size for outputs.
<Taneb>
srhb: thank you
PeterHK has quit [Ping timeout: 256 seconds]
Guest52151 has quit [Read error: Connection reset by peer]
sinner has joined #nixos
rauno has quit [Remote host closed the connection]
<srhb>
mvnetbiz: Your inner self (pythonPkgs) is shadowing your outer self (nixpkgs)
<srhb>
mvnetbiz: Rename one or the other
<__monty__>
Hmm, this causes infinite recursion, adding super.python27Packages... causes an anonymous function called with unknown argument packageOverrides error: http://ix.io/1DhS
<srhb>
mvnetbiz: I like naming schemes like: pyself: pysuper: ... to disambiguate
<mvnetbiz>
oh yeah. I think I have seen that naming pattern before.
jtojnar has joined #nixos
Dr8128 has quit [Ping timeout: 246 seconds]
agander has quit [Ping timeout: 252 seconds]
<mvnetbiz>
https://hastebin.com/ejuxawurim.http now I get that the crccheck dep of the other package is undefined. I think I get why but I need to think about how to reorder it. I couln't find any good exapmles of this kind of thing.
<__monty__>
Hmm, override on python27 works but that doesn't seem to affect python27Packages : /
<srhb>
mvnetbiz: Sorry, I didn't quite understand what error you're getting
<mvnetbiz>
error: undefined variable 'crccheck'
<srhb>
mvnetbiz: Oh, right
<mvnetbiz>
on line 10
<srhb>
mvnetbiz: It's a python package?
<mvnetbiz>
Yeah.
<srhb>
mvnetbiz: You have with self, not with pyself; or similar
<mvnetbiz>
:/
<srhb>
mvnetbiz: So you're referring to nixpkgs.crccheck instead of (morally) nixpkgs.pythonPackages.crccheck
<srhb>
mvnetbiz: Makes sense?
<mvnetbiz>
Yeah, think so. I will probably watch one of the presentations on it.
jtojnar has quit [Remote host closed the connection]
Dr8128 has joined #nixos
<srhb>
mvnetbiz: In this case, all you really need to consider is that your outer self (and super) refers to nixpkgs and your inner pyself and pysuper refers to pythonPackages
<srhb>
mvnetbiz: So while pyself.crccheck exists, self.crccheck does not :)
<srhb>
mvnetbiz: In other words, with pyself; [ crccheck ]; is what you want
<srhb>
(or, more clearly [ pyself.crccheck ];
<mvnetbiz>
Yeah I got it working. I just don't have the right attributes for fetchhg but that's easy to figure out
sigmundv has joined #nixos
<mvnetbiz>
I don't get how overlays operate on each layer of the nixpkgs attribute set. Like the "pyself: pysuper:" overlay function is assigned to super.python3.override.packageOverrides
<mvnetbiz>
is there a nixpkgs.override.packageOverrides?
<mvnetbiz>
I was testing if disabling specter mitigations made a difference but that option should probably be removed
<gchristensen>
oooh with docs too. thank you
<gchristensen>
!
rprije has quit [Ping timeout: 246 seconds]
<mvnetbiz>
Not a lot of well done studies on how much the stuff like pinning CPUs and using Huge pages affects gaming performace as far as I could find. But it was recommended by threads I read through a lot on level1techs and reddit
<mvnetbiz>
There is a paravirtualized evdev keyboard mouse driver for windows which is really cool. You can switch between host and guest by pressing left&right ctrl together.
<mvnetbiz>
It's really low latency I don't think I can notice it when playing CS:go
<{^_^}>
[nixpkgs] @nlewo opened pull request #57505 → dockerTools.buildImage.runAsRoot: preserve layers ordering at image unpacking → https://git.io/fhjKj
<Taneb>
What's the protocol for adding a package that you need but really don't want to maintain? Is there a "needs maintainer" thing anywhere?
<srhb>
Taneb: Put up a PR with no maintainer and be really secretive about not wanting to maintain it. :P
<srhb>
Better, find a maintainer!
<mvnetbiz>
gchristensen, games like gta5 run surprisingly not terrible installed on an nfs share on the host lol
<mvnetbiz>
Would that module be too off topic for nixpkgs?
linarcx has quit [Ping timeout: 246 seconds]
<gchristensen>
mvnetbiz: might be better to put like in the wiki
<vasarmilan>
hi everyone! I use NixOps with virtualbox backend and I'd like to be able to write to a shared folder from the guest. I have the following in the config: fileSystems."/shared" = { device = "shared"; fsType = "vboxsf"; }; however "touch /shared/test still returns "Read-only file system"
<vasarmilan>
do you have any idea why is that?
reallymemorable has joined #nixos
<vasarmilan>
sorry I meant to copy shared = { hostPath = "..."; readOnly = false; };
<vasarmilan>
so I have the readonly attribute set to false.
Neo-- has joined #nixos
reallymemorable has quit [Ping timeout: 252 seconds]
<symphorien>
pmiddend: there is also separateDebugInfo for a more long term solution
<{^_^}>
[nixpkgs] @bjornfor pushed to release-19.03 « deadpixi-sam: remove duplicate alias »: https://git.io/fhjiY
<joepie91>
symphorien: that was just a simplified repro; I do actually have a shell.nix, but gcc5 is only specified in the environment's build inputs, and that previously worked... exactly how would I modify this shell.nix to work like you described? https://gist.github.com/joepie91/2f9b0fa228c2f3ac6cd70dc88558156d
<{^_^}>
[nixpkgs] @rixed opened pull request #57510 → junkie: init at 2.8.0 → https://git.io/fhjiO
<symphorien>
joepie91: iirc, use (overrideCC stdenv gcc5) instead of stdenv and remove gcc5 from buildInputs
<siraben>
How do I change the font size in the TTY console?
<joepie91>
symphorien: that appears to have worked
<joepie91>
thanks!
<joepie91>
symphorien: is this on the wiki somewhere?
<etu>
siraben: I think one of the news in linux 5.0 was that it would support HiDPI consoles or something. Otherwise you are on the right track with changing font
<symphorien>
there is no result for overrideCC
<symphorien>
maybe starting a "packaging faq" page could be a good idea
jboy has joined #nixos
<{^_^}>
[nixpkgs] @bjornfor pushed to release-18.09 « deadpixi-sam: remove duplicate alias »: https://git.io/fhjiZ
<siraben>
etu: so no HiPDI support?
<siraben>
For TTY?
<etu>
siraben: It's just a matter of changing font
<siraben>
etu: where?
<__monty__>
Hmm, http://ix.io/1Dir but I have xcode 10.something, any ideas what I'm doing wrong? My channel is nixpkgs-unstable.
<gchristensen>
hmm I think I've found a regression in 19.03
<gchristensen>
lots of "/etc/terminfo directory contains user files. Symlinking may fail. at /nix/store/amhpx347b7n6m2bq24siw0qq27w3np8n-setup-etc.pl line 114." -- anyone else seeing this?
<linarcx>
Hi. i have a simple questionl. Can we use another init system like openrc or runit instead of using systemd in nixos?
<srhb>
gchristensen: Think we already have an issue on that?
<__monty__>
linarcx: Last I heard there was an effort to seperate nixos from systemd, not sure what it was called anymore though. But they hadn't succeeded yet. So I'd expect it to be hard if at all possible.
rardiol1 has joined #nixos
<__monty__>
linarcx: I'd expect GuixSD to be NIH enough about init systems to not be tied to systemd if that's important to you.
<linarcx>
Guys, a qustion about generations. when i query: `sudo nix-env --list-generations --profile /nix/var/nix/profiles/system`, this list generations from number 33 to 45. And when i use: `nix-env --list-generations --profile /nix/var/nix/profiles/per-user/linarcx/profile`, it list generations from number 45 to 56. when i reboot the machine i can see only first list. why?
<avn>
lewo: looks cool.
glasserc has quit [Ping timeout: 244 seconds]
<musicmatze>
linarcx: because you're booting into the system, not into the user. The second command shows only the generations which are available for the user, not for the system
<musicmatze>
You boot a system generation and your local user also "has" that system generation because it is system-wide (duh!)
<musicmatze>
your user maintains an own set of packages installed only for the user ... and those packages are also organized in generations - the generations for your user, ... :-)
<musicmatze>
I hope that clears things up a bit :-)
<linarcx>
musicmatze: Ok, it means when i use gc it only affects the generations of root user? and deleteing generations for each profile is seperate task? right?
orivej has joined #nixos
<musicmatze>
no, `nix-collect-garbage` collects everything that is not reachable. But yes, deleting generations is only for one profile at a time
<linarcx>
musicmatze: Thank you so much.
<musicmatze>
it has nothing to do with the root user though, only with system and local user.
<{^_^}>
[nixpkgs] @aszlig opened pull request #57519 → nixos: Add 'chroot' options to systemd.services → https://git.io/fhjXk
<musicmatze>
system != root user.
<linarcx>
musicmatze: I'm confused a little bit! what `system` contain?
<musicmatze>
so if you remove generation 45 for your user and generation 33 of the system and then `nix-collect-garbage`, everything installed in user/45 and system/33 will be "gone",...
<musicmatze>
linarcx: "system" is the machine itself, like the packages installed in `configuration.nix` for example
<musicmatze>
"user" is `nix-env -iA nixos.somepackage` installed stuff
<linarcx>
musicmatze: i've got it.
<musicmatze>
:-)
JonReed has joined #nixos
ma27 has quit [Quit: WeeChat 2.2]
glasserc has joined #nixos
johanot has quit [Quit: WeeChat 2.2]
martingale has joined #nixos
ma27 has joined #nixos
<rardiol1>
Anyone here runs nixos on a multiseat setup?
<__monty__>
rardiol1: You've never rendered your machine unusable by running a compile? Now imagine 3 people who might kick off such a compile whenever, usually when they won't be bothered by it but the others might be.
<__monty__>
adamt: Yeah, but that gets awful close to containers.
<adamt>
__monty__: You did ask about containers as well?
<rardiol1>
__monty__: well, it's for my home, so I could just yell to stop hogging :). And I'm the only one who would be compiling
<__monty__>
yorick: I was very confused about what your beef with hedgewars was : )
<__monty__>
Ah, was imagining a small business scenario.
<adamt>
__monty__: Oh sorry, now I understand. Nvm.
<__monty__>
np
agander_ has joined #nixos
agander has quit [Ping timeout: 246 seconds]
evilmonads_ has joined #nixos
Philonous_ has joined #nixos
TweyII has joined #nixos
iqubic has quit [Remote host closed the connection]
waleee has quit [*.net *.split]
pjan has quit [*.net *.split]
FRidh has quit [*.net *.split]
PLPD-Bot has quit [*.net *.split]
jbgi has quit [*.net *.split]
Twey has quit [*.net *.split]
ixiqem has quit [*.net *.split]
statusbot has quit [*.net *.split]
Philonous has quit [*.net *.split]
buckley310 has quit [*.net *.split]
johnw has quit [*.net *.split]
leothrix has quit [*.net *.split]
kgz has quit [*.net *.split]
sam_d has quit [*.net *.split]
tshaynik has quit [*.net *.split]
makefu has quit [*.net *.split]
xantoz has quit [*.net *.split]
ciil has quit [*.net *.split]
bollu has quit [*.net *.split]
veverak has quit [*.net *.split]
rsa has quit [*.net *.split]
eon` has quit [*.net *.split]
delroth has quit [*.net *.split]
aminechikhaoui has quit [*.net *.split]
lrvick has quit [*.net *.split]
mulderr has quit [*.net *.split]
shiver has quit [*.net *.split]
Ankhers has quit [*.net *.split]
oleks has quit [*.net *.split]
evilmonads has quit [*.net *.split]
lohfu has quit [*.net *.split]
Ankhers_ is now known as Ankhers
TweyII is now known as Twey
<adamt>
Oh, the wonders of modern chat protocols
bollu has joined #nixos
aminechikhaoui has joined #nixos
ilmu has joined #nixos
aminechikhaoui5 has joined #nixos
Fare has quit [Ping timeout: 255 seconds]
aminechikhaoui5 has quit [Client Quit]
aminechikhaoui has quit [Remote host closed the connection]
aminechikhaoui5 has joined #nixos
mvnetbiz has quit [Ping timeout: 246 seconds]
Xal has quit [Ping timeout: 246 seconds]
fractal has quit [Ping timeout: 246 seconds]
<rardiol1>
How many years until the foss world abandons irc?
<lewo>
Is there a NixOS module to copy the /nix/store to another disk and remount it? I would like to boot a VM in the cloud with the /nix/store on a volume.
<Cale>
rardiol1: I don't know, but all the new chat services don't seem to be helping anything. It just means I have to be on many other chat services as well as IRC.
<etu>
adamt: awdamn, a wednesday. I wish I could make it :(
* etu
should roll his own nixos meetups in stockholm
<adamt>
etu: Stockholm seems terrible far away from Copenhagen
cryptomonad has joined #nixos
<Taneb>
I'm just too far for the London meetup to be achievable :(
linarcx has quit [Ping timeout: 252 seconds]
<etu>
adamt: We almost have a nix usergroup going on in stockholm every monday, because we meet up and talk computers and stuff. And several of us run nixos nowadays :p
linarcx has joined #nixos
<adamt>
etu: Lucky you. That's what we want in Copenhagen as well. Maybe not weekly, but still. :-)
vidbina has joined #nixos
<jtojnar>
Mic92: now I got delivery notice from gmail; this was the server response: 450 4.7.1 <=?us-ascii?q?nixos1+somehash?=@discoursemail.com>: Recipient address rejected: Internal error
<jtojnar>
the somehash matches the reply-to header of the op
_kwstas has joined #nixos
_kwstas has quit [Remote host closed the connection]
<Mic92>
jtojnar: I guess the only thing we can do if that happens more often to ask discourse directly. However it might be the case that they already see those errors in their logs.
ggpeti has joined #nixos
<ggpeti>
any nixpkgs maintainers hanging out here on this beautiful day/morning/evening/night?
<ggpeti>
how can I get your attention to a pull request of a simple new package?
<Taneb>
ggpeti: "Can someone look at [link to PR]" normally works
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
pie_ has quit [Ping timeout: 252 seconds]
<siers>
I'm trying to find why there's a cache miss after my change. I'm looking at nix-diff $old $new and patches seem to be different… (big green block) but when I open up the .drv files, the patches seem very similar
<clever>
siers: line endings?
<clever>
siers: what does the `file` program say about both?
<tilpner>
That doesn't seem to contain the location, yl
pmiddend has quit [Ping timeout: 252 seconds]
<srhb>
srid: It'll be something like /root/.nix-defexpr/channels/nixos/nixpkgs
<srhb>
srid: Don't rely on that though
<srhb>
(use <nixpkgs>)
<yl>
tilpner: location online or on your machine?
<ar>
,locate bin/zip
<{^_^}>
Found in packages: zip, plan9port
<ar1a>
so i have a pr i want to make, and i want it to go to both nixos-unstable and backport it to 18.09. Do I need two prs for that?
<srhb>
ar1a: You can start with one against master
<clever>
ar1a: yeah
<srhb>
ar1a: Usually a maintainer will be kind and backport it for you if you ask for it in the PR. However, it's fine to do it yourself too, using cherry-pick -xe
<ar1a>
ty
<srhb>
ar1a: But if you end up having to make fixes to the one targeting master, you'll have to redo work
<tilpner>
yl: srid said "local location", which I interpreted to be "on your machine"
<srhb>
(for that reason I tend to open backport PRs after the first one is in master)
xkapastel has quit [Quit: Connection closed for inactivity]
ThatDocsLady has joined #nixos
<srid>
yl: Interesting. what does the () and <> signify in nix syntax?
<clever>
srid: <nixpkgs> is a search path, that tells nix to search NIX_PATH (and -I prepends to NIX_PATH)
statusbot has joined #nixos
<clever>
srid: () are used to convince nix-eval that its an expression to eval
Ariakenom has quit [Ping timeout: 246 seconds]
<ar1a>
so i have discord installed via home-manager, but i want to install a local derivation of discord tempoarily, until its in nixos-unstable. Do I need to uninstall discord in home-manager or can nix wrangle the versions for me
<eyJhb>
Do I nede to specify both nixos and nixpkgs in my channels?
<ar1a>
not exactly what i meant
<clever>
eyjhb: the nixos channel includes nixpkgs
<clever>
eyjhb: and the nixpkgs channel includes nixos (but without as much testing)
o1lo01ol1o has quit [Remote host closed the connection]
<eyJhb>
So I only need to have the nixos in my `nix-channels
<eyJhb>
list`?
<clever>
eyjhb: correct
<eyJhb>
... I hate that.. But thanks :D
Erasmus has joined #nixos
<srhb>
ar1a: I believe they'll conflict because of the way home-manager composes its packages into your user profile
<tilpner>
eyjhb: What do you hate about it?
<ar1a>
srhb: thanks, ill just comment it out temporarily then
o1lo01ol1o has joined #nixos
<srhb>
ar1a: If it's a well-behaved application, maybe nix-shell will suffice?
<ar1a>
maybe if i run it with &!
<eyJhb>
tilpner: not the channel thing. I have a plugin to weechat the converts 2x'-' to a two seperate messages... Usefull for facebook chatting, not writing commands
<srhb>
ar1a: Worth a sho.
<srhb>
...t
<srhb>
>_>
<{^_^}>
[nixpkgs] @hedning pushed 122 commits to gnome-3.32: https://git.io/fhjSw
<tilpner>
Ah, okay
<srhb>
domenkozar: Nevermind my ping, I opened an issue and pinged you there instead.
<clever>
eyjhb: you are reading a variable called allowUnfree (which doesnt exist) and getting the .betterlockscreen attribute of it
<eyJhb>
Yeah
<eyJhb>
I fucked up soooo bad
<eyJhb>
Sorry....
Guest18194 has quit [Read error: Connection reset by peer]
Guest72641 has joined #nixos
johanot has joined #nixos
scientist has joined #nixos
Glider_IRC_ has joined #nixos
drakonis has joined #nixos
o1lo01ol1o has joined #nixos
rcshm has quit [Remote host closed the connection]
<scientist>
hey guys, quick question, what happens if I switch back to generation: 1 nix.configuration file, does it bring back nixos to the state where i first installed or does it keep some changes ?
<clever>
scientist: a few minor things are kept, but most of it will be rolled backwards
rcshm has joined #nixos
Glider_IRC__ has quit [Ping timeout: 272 seconds]
<matthewbauer[m]>
paulyoung: this was fixed at some point in nixpkgs...
<scientist>
clever: ok, I am having problems with x window system so I think I am going to go back to that config
Erasmus has quit [Read error: Connection reset by peer]
scientist has quit [Quit: WeeChat 2.2]
<tilpner>
I have an option with type = attrs, and I want to provide a default for a few attributes of it
o1lo01ol1o has quit [Ping timeout: 246 seconds]
<tilpner>
But I can't use mkDefault, because the whole attrset is replaced when it is set with a higher priority
<sphalerite>
Taneb: oh you mean the spam? srhb was the one who stopped that :p
<sphalerite>
I don't have ops here
lawlesseel has quit [Ping timeout: 256 seconds]
<ddellacosta>
folks, I'm trying to check my understanding of callPackage. It seems like it is structured as a function closure over the nixpkgs, which, because of the inherent laziness of the nix language doesn't cost anything until we actually need to evaluate the closure it returns. I'm feel a bit confused as to _why_ it's structured that way vs. simply returning an overwritten/overloaded derivation set. This has
<ddellacosta>
something to do with propagating dependencies with overrides?
<ddellacosta>
I should say, a function closure over a package set
lawlesseel has joined #nixos
<Taneb>
sphalerite: I was having a bit of a panic attack, you helped me see the funny side
<sphalerite>
Taneb: oh good :D
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
<sphalerite>
glad I could help
rpg has joined #nixos
o1lo01ol1o has joined #nixos
lawlesseel has quit [Remote host closed the connection]
chpatrick has joined #nixos
<elvishjerricco>
How do you make a filesystem inessential to booting? I've got an external disk I want configured in hardware-configuration.nix, but I don't want my machine to care if it's not plugged in when I boot.
<sphalerite>
elvishjerricco: add nofail to the mount options
<tilpner>
elvishjerricco: neededForBoot seems to be off by default
<elvishjerricco>
I thought neededForBoot just meant it needed to be mounted by initrd, not that the boot wouldn't fail if it can't be mounted in stage 2
<sphalerite>
tilpner: clever neededForBoot means it gets mounted from initrd
<sphalerite>
yeah
<sphalerite>
nofail is what you're looking for
<elvishjerricco>
cool. Thanks
agander_ is now known as agander
rcshm has quit [Read error: Connection reset by peer]
rcshm_ has joined #nixos
Izorkin has joined #nixos
linarcx has joined #nixos
<{^_^}>
[nixpkgs] @veprbl opened pull request #57531 → [19.03] qt511.qtwebkit: fix on darwin → https://git.io/fhj9d
o1lo01ol1o has quit [Remote host closed the connection]
paulyoung has quit [Quit: Page closed]
o1lo01ol1o has joined #nixos
maxmaxmax has joined #nixos
rcshm_ has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
o1lo01ol1o has quit [Ping timeout: 272 seconds]
o1lo01ol1o has joined #nixos
ma27 has quit [Quit: WeeChat 2.4]
pie___ has joined #nixos
rcshm has quit [Read error: Connection reset by peer]
rcshm has joined #nixos
pie_ has quit [Ping timeout: 252 seconds]
rcshm has quit [Remote host closed the connection]
ma27 has joined #nixos
rcshm has joined #nixos
<reilithion>
Can I take a /nix/store/hash-name.drv and modify it and then build that derivation somehow?
<infinisil>
ddellacosta: I don't really get your simple alternative
reivilibre has quit [Remote host closed the connection]
<ddellacosta>
infinisil: it may not actually be a feasible alternative, as I'm a bit stuck in my understanding. Basically, it comes from me messing around with a simple default.nix. If I return the result of calling mkDerivation at the end (a set) nix-build/nix-instantiate builds a derivation just fine. If I call `callPackage` with that derivation it also works, although that seems to be returning a closure vs. a set. I
<ddellacosta>
also get an error when I try to assign the return value of `callPackage` and poke at it--it complains about there not being a default.nix in what is apparently the derivation's $out, which I don't understand either
reivilibre has joined #nixos
orivej has quit [Ping timeout: 268 seconds]
waleee1 has quit [Quit: WeeChat 2.4]
<ddellacosta>
so my assumption was that if you can return a derivation as a set, why do we need callPackage--just to expose a simpler API for overriding derivation attributes?
<ddellacosta>
as always, there's probably something basic I'm missing
<infinisil>
ddellacosta: callPackage takes 2 arguments, not only one, so when you just give it one and evaluate it, that's just a function/closure, and that's not an error, it just doesn't even evaluate the derivations
Ariakenom has joined #nixos
<ddellacosta>
yeah, I'm glossing over the second argument of override attributes, sorry
<infinisil>
What's your code exactly?
<ddellacosta>
I guess I'm still not being clear in what I'm confused about. Basically I just don't understand what good callPackage is other than to allow for easily overriding arguments, per the Nix Pill about the "callPackage pattern"
<infinisil>
Yeah that' exactly it
<ddellacosta>
infinisil: one sec., let me post what I've been messing with
<infinisil>
(It could be done differently though)
<reilithion>
Hmm. Can I somehow get hold of the end derivation of a package like gcc and override its configuration environment variable? JUst for testing purposes. I want to see if a particular configuration will successfully build.
<ddellacosta>
infinisil: okay--so then my question becomes, why is one a closure (the return value of callPackage), but I can also return a set (output of mkDerivation) (in terms of what nix-build/nix-instantiate are able to produce a store derivation with)? Here's what I've been messing with btw, really stupid, meant to just expose stuff as obviously as possible:
<slabity>
reilithion: Override the configuration environment after it's already built? Like modifying the nix-store directly?
<infinisil>
reilithion: nix-shell -A gcc, then unpackPhase, cd to the source, and then run the normal phases, then you have a working dev environment for gcc, probably
<infinisil>
ddellacosta: that's not how callPackage is supposed to be used, you shouldn't pass it a derivation
rcshm_ has joined #nixos
<infinisil>
Oh but i guess it doesn't throw an error
<infinisil>
I'd have thought it would give an error that it expects a function
<LnL>
yeah, this isn't how callPackage is usually used in nixpkgs
jb55 has joined #nixos
<ddellacosta>
I get that this isn't the normal usage--I'm specifically trying to understand what it accepts and returns now
<infinisil>
> import {}
<{^_^}>
cannot coerce a set to a string, at (string):222:1
rcshm has quit [Ping timeout: 246 seconds]
<ddellacosta>
and it seems like calling it with `import ...` as the second arg. vs. any other nix expression shouldn't really matter (other than because it's not particularly useful)
<ddellacosta>
but, that may be another thing I'm wrong about
<infinisil>
ddellacosta: Um, evaluating drv' doesn't throw an error!
<infinisil>
?*
<ddellacosta>
yeah it does here--it fails on `builtins.trace "what about this? ${drv'}"`
<ddellacosta>
`error: opening file '/nix/store/pb3fglb6s7h6j1mri5smlqhhd5pnf1gg-type-guided-tour/default.nix': No such file or directory`
<reilithion>
By "end derivation" I intended to mean the derivation of the package after it's applied to all its arguments and is just a set.
<ddellacosta>
infinisil: yeah, sorry it's not clear at all just from the gist
<ddellacosta>
LnL: I understand how to get a working derivation and etc.--I'm not stuck on that, but on understand more core details of how and why stuff is implemented
<ddellacosta>
I definitely don't need help with the surface-layer syntax stuff, but thanks
<LnL>
ddellacosta: your example is the file case
zupo has joined #nixos
<infinisil>
ddellacosta: check out the source of callPackage(With) here: https://github.com/NixOS/nixpkgs/blob/master/lib/customisation.nix In case the argument (drv' in your case) isn't a function (which it isn't), it tries to (import drv') with drv' being made into a string, and because drv' is a string in your case, it builds the derivation and uses its resulting path as string
<LnL>
so it's the same as let f = import "${drv}"; in f { some; arguments }
<infinisil>
ddellacosta: then because import uses /default.nix of directories, that's the result
knupfer has joined #nixos
<infinisil>
s/and because drv' is a string/and because drv' is a derivation
linarcx has quit [Quit: WeeChat 2.4]
<ddellacosta>
infinisil: okay one sec, let me absorb all of that, but sounds like what I was looking for
<infinisil>
In short: it's a weird error you get by not using callPackage like it was intended to
xkapastel has joined #nixos
<ddellacosta>
huh so I guess I was assuming referential transparency where it does not exist?
<ddellacosta>
in this particular case
linarcx has joined #nixos
<ddellacosta>
infinisil: okay so if we were going to attach a faux type sig to `callPackage`, then I guess it would be (Haskell-style) callPackage :: FunctionOrCoercibleToFunction -> OverrideSet -> StoreDerivation ?
<ddellacosta>
(being very hand-wavey about OverrideSet)
phdoerfler has joined #nixos
<ddellacosta>
yeah okay, I see what bad assumptions I was making now--thanks infinisil
<guest874329>
Hi all, just looking if anyone knows how to add a local nixpkgs repository to your channels
linarcx has quit [Quit: WeeChat 2.4]
<guest874329>
As in, I'm working on creating a couple packages that I'll submit PRs for in the future, but while I'm working on them I'd ideally like to just call nixos-rebuild switch the same as always, with any packages not being found in the official repo being looked for in my personal one
<guest874329>
Is there a better approach for this kind of workflow?
linarcx has joined #nixos
rumle23 has joined #nixos
rumle23 has quit [Client Quit]
<srk>
guest874329: I'm using a branch with my commits until they land in master. you can also use overlays with your new packages
<srk>
,overlay guest874329
<{^_^}>
guest874329: Overlays look like `self: super: { foo = ...; }`. Use the self argument to get dependencies, super for overriding things and library functions (including callPackage). More info: https://nixos.org/nixpkgs/manual/#sec-overlays-install
<infinisil>
slabity: I'm keeping a separate, private repository as a submodule in my configuration
Makaveli7 has quit [Quit: Leaving]
<infinisil>
slabity: In there I define a set of options under `config.private`, which defines all the private things
rpg has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<infinisil>
slabity: Or alternatively, if I don't want to make it public that I set a certain option, I can just set it in the private repo
<infinisil>
(this really works with any private file storing way, not only private git repos)
<clever>
back
<guest874329>
Thanks for the help. I'm not really understanding how I could use overlays for this case?
<guest874329>
Or are you just saying that you've changed your channel to be your repo and you just manually keep it up to date until your PRs are accepted?
<clever>
guest874329: overays are meant to let you change things, without having to swap out all of nixpkgs
<guest874329>
ah, yes, in that case I'm familiar. I've used it before to customize some already existing packages
<srk>
you can use callPackage to add new packages as well
<guest874329>
So in the same way that I add lines to top-level, I could just do that in my configuration pointing to my personal repo for packages that haven't been accepted yet?
<guest874329>
with callPackage that is
<srk>
yup
scaroo has joined #nixos
<guest874329>
okay, that is perfect. I think that's exactly what I was looking for
<guest874329>
In general though I'm still trying to get used to a workflow around developing new packages. Is there a best practice approach I should be following?
grizwako has quit [Remote host closed the connection]
<guest874329>
Thanks for the help btw
<guest874329>
I've been on this channel a couple of times over the last week and it's been very helpful
<scaroo>
hi chan.I am looking for a way to write a nix expression to a file from within a nix definition. A writeExprToFile( { ... }); . Any pointer? Thanks
<infinisil>
scaroo: What is your actual problem? What are you trying to achieve?
jeromegamez has joined #nixos
<worldofpeace>
<freenode_jto "Do we still want to merge all th"> jtojnar: Any other update but that one I wouldn't have merged.
<scaroo>
infinisil: writing a packages set definition from an xml file. The idea is to install a node's or haskell's xml2json, then using fromJSON, then transforming the structure. Then write it
<scaroo>
would be a static generator to run periodically
<infinisil>
scaroo: Write the json directly
<scaroo>
infinisil: dont control the source
thomashoneyman has joined #nixos
<worldofpeace>
Huh matrix made my response look weird :)
<infinisil>
scaroo: But you just said you need to write a nix file?
<infinisil>
Just write a json file instead
<infinisil>
Oh, you mean you don't control the source that reads the file?
<thomashoneyman>
elvishjerricco: i'm not sure if i sent a message about it (no history), but your installation guide ended up working and blacklisting 'nouveau' from nvidia got the graphics working
<scaroo>
indeed, I could fetch it periodically
<scaroo>
and read it at build time
<infinisil>
scaroo: What's this xml2json thing do?
<elvishjerricco>
thomashoneyman: Oh, neat. Thanks for the update :)
<thomashoneyman>
elvishjerricco: now I have to figure out how to install Windows to dual boot, too :|
<scaroo>
infinisil: well it kinda in the name :) Turns an xml hierarchie to a json object
<thomashoneyman>
ordinarily i think windows is installed first, but as I couldn't get any desktop / display / window manager to load i wanted to try pure nixorsts fi
<thomashoneyman>
whoa
<thomashoneyman>
*pure nixos first
<scaroo>
and wrie it on fs
<infinisil>
scaroo: Well what's described in this xml?
<thomashoneyman>
by the way: does anyone have a link to documentation on how to have a command-line argument always passed to a binary?
<thomashoneyman>
i installed spotify but it looks TERRIBLE on 4k
<scaroo>
infinisil: no need to write the nix expr
<scaroo>
infinisil: zill be recreatedeach build time, indeed. Thanks
<thomashoneyman>
but running `spotify --force-device-scale-factor=2.5` fixes it
<thomashoneyman>
i'd like for the application spotify to always be run with that flag in place
<thomashoneyman>
i was thinking there is a way to override the package so this flag is always present or something like that
<infinisil>
scaroo: I didn't get that, but no problem I guess :P
<thomashoneyman>
how is this sort of thing usually accomplished in Nix?
<scaroo>
thomashoneyman: you mighr want to add a 'additionalArgs' attribute to spotify, and add the flags in the wrapper script or .desktop file
<infinisil>
scaroo: Ehh, I wouldn't recommend that
<scaroo>
thomashoneyman: *etraArgs
<thomashoneyman>
there is no .desktop file for spotify, surprisingly
<infinisil>
The usual way is to define an alias in your shell
schneid3306 has joined #nixos
<thomashoneyman>
having installed with nix-env -i spotify
<infinisil>
alias command="command --extra-argument"
<infinisil>
Or in NixOS configuration.nix: `environment.shellAliases.command = "command --extra-argument";`
<thomashoneyman>
thanks infinisil
<thomashoneyman>
and scaroo
<thomashoneyman>
i'm a little confused as to what happens when i select an application in the KDE Plasma desktop
<thomashoneyman>
does that simply call the binary?
zupo has joined #nixos
<scaroo>
infinisil: yeah sounds simpler and less invassive. sorry for misleading :s
<thomashoneyman>
like -- after the first run of Firefox it showed up in the 'favorites' section of the little start menu
<thomashoneyman>
and i can click that to run firefox again
<infinisil>
thomashoneyman: How it usually works is that packages define a .desktop file, you sure spotify doesn't have one?
<thomashoneyman>
i couldn't find one
<infinisil>
thomashoneyman: Oh, well that's just a firefox thing probably, or a plasma thing, nothing nixos does
<thomashoneyman>
oh! wait, it looks like there _is_ one now
<infinisil>
Are you telling me it suddenly appeared? :o
<thomashoneyman>
ah, no
<thomashoneyman>
well, i don't know
<infinisil>
They should be in /run/current-system/sw/share/applications
<thomashoneyman>
grep -i spotify.desktop . in the root doesn't find it
<thomashoneyman>
but it does pick up my bash history and some plasma config stuff referencing a spotify.desktop so maybe i'm being blind
schneid3306 has quit [Quit: Quit.]
<infinisil>
Is it not in that folder?
<thomashoneyman>
nope. not there
<thomashoneyman>
neither is firefox, incidentally
<infinisil>
Oh, you installed it with nix-env
<infinisil>
Right?
<thomashoneyman>
yes
<thomashoneyman>
is that not the way to go?
<infinisil>
Then it should be in ~/.nix-profile/share/applications
<thomashoneyman>
well, firefox was installed in configuration.nix
<infinisil>
Well then it should be in /run/current-system/sw/share/applications too :/
<thomashoneyman>
ah!
<thomashoneyman>
yep, the file _is_ there
scientist has joined #nixos
<thomashoneyman>
ah! looks like `Exec=` in here has what i'm looking for
<thomashoneyman>
that seems to be the command that's run to start
<infinisil>
Yup
<thomashoneyman>
thanks!
pie_ has joined #nixos
<infinisil>
:)
knupfer has quit [Ping timeout: 240 seconds]
<thomashoneyman>
hmm
<thomashoneyman>
can i just edit this file directlyL
<infinisil>
Nope, why?
<thomashoneyman>
oh, never mind
pie___ has quit [Ping timeout: 252 seconds]
<thomashoneyman>
if i have applied the alias
<thomashoneyman>
then it will apply to the Exec here
<thomashoneyman>
i was thinking of changing the exec
<thomashoneyman>
alias is better
<infinisil>
Well, an alias would only be used in your shell, so when you double click firefox it wouldn't apply the alias
<thomashoneyman>
that makes sense
<thomashoneyman>
lets say I want double clicking firefox to apply the alias
<scientist>
hello you guys, I have a quick question, I finally finished setting up my own custom keyboard layout and fixing all the errors, everything works except for the shift keys and when I try to restart the display-manager my processor spins like crazy(I think it's an infinite loop), do you guys have any tips on how to fix these two problems ?
<{^_^}>
[nixpkgs] @worldofpeace pushed 49 commits to gnome-3.32: https://git.io/fhj7E
<thomashoneyman>
or "apply the alias"
<infinisil>
thomashoneyman: Then you need something like this
<thomashoneyman>
infinisil: ah! this looks like exactly what i'm looking for
ng0 has joined #nixos
<infinisil>
thomashoneyman: I recommend the first symlinkJoin way
<scientist>
do any of you guys have any tips to help me solve these two problems I am having ?
<thomashoneyman>
infinisil: where would i put this file?
Neo-- has quit [Ping timeout: 240 seconds]
<thomashoneyman>
like where do you store these things?
<infinisil>
scientist: Don't be so impatient, people are less likely to help you if you ask every 3 minutes
<scientist>
infinisil: sorry, frustration
<infinisil>
thomashoneyman: You can put something like `let wrappedFirefox = pkgs.symlinkJoin { ... }; in { environment.systemPackages = [ ... wrappedFirefox ]; }` in your configuration.nix
<thomashoneyman>
i see. thanks!
guest874329 has quit [Ping timeout: 256 seconds]
schneid3306 has joined #nixos
schneid3306 has quit [Client Quit]
<reilithion>
All of a sudden, --dry-run for nix-shell has stopped working. O_o
<infinisil>
scientist: You need to figure out what is causing the loop. Look at htop to figure out what process it is and/or look a while at `journalctl -f`
<scientist>
infinisil: https://hastebin.com/zolerutohi.xml , everything seems to be working except for the shift keys, earlier I had set left shift to be a sticky key and then removed it, i'm not sure if that's why the shift keys are not working ? and i'm not sure if I need to reset the dispay-manager for some of the changes to take effect(darn infinite loops) ?
johnnyfive has quit [Quit: I quit now.]
thomashoneyman has quit [Ping timeout: 256 seconds]
<reilithion>
Ok, I think I was just doing something stupid there. Keeping clam and carrying on now.
<infinisil>
scientist: You first need to fix the infinite loop before you can worry about some keys not working
johanot has quit [Quit: WeeChat 2.4]
<{^_^}>
[nixpkgs] @matthewbauer pushed 2 commits to staging: https://git.io/fhj7j
<scientist>
systemctl restart display-manager , when i run this command my processor goes crazy and my pc stays frozen, how do i see the errors that are causing this ?
_d0t has joined #nixos
<{^_^}>
[nixpkgs] @xeji pushed commit from @marsam to master « syncthing: 1.0.1 -> 1.1.0 (#57364) »: https://git.io/fhj5L
<scientist>
noonien: x11 was using up 100% of the cpu
rcshm has quit [Remote host closed the connection]
<scientist>
noonien: i think it is an infinite loop
<fendor>
everytime i try to mount an external hard drive, formatted as ntfs, it is mounted as a read only file system and I can not get to use it to copy data. how can i mount it with write permissions?
<clever>
scientist: depending on your gpu drivers, just scrolling thru journalctl can max out the cpu
pie_ has quit [Ping timeout: 255 seconds]
xkapastel has quit [Quit: Connection closed for inactivity]
<clever>
fendor: -o rw
<fendor>
clever, well, doesnt seem to work
<clever>
fendor: -o rw -v, what does it say?
rcshm has joined #nixos
Ariakenom has quit [Quit: Leaving]
<fendor>
clever, mount: /dev/sdb1 mounted on /run/media/baldr/seagate.
<scientist>
clever: if i switch back to an older nix.configration , will x window(x11) be fixed ?
<clever>
scientist: but which broke it, changes to nixpkgs, or changes to your configuration.nix
linarcx has quit [Quit: WeeChat 2.4]
<_d0t>
hi. How do I make flycheck in emacs use the include paths from nix sandbox for C/C++ code? I'm using nix-emacs, but it doesn't seem to work in this case.
<scientist>
clever: well I havent done anything to configuration.nix so it must be the nixpkgs then,
fosskers has quit [Remote host closed the connection]
pie_ has joined #nixos
<immae>
As far as I know there is no "single" place to get documentation about everything, saddly
<immae>
(I may be wrong though, I just don’t know about such place)
pie___ has quit [Ping timeout: 252 seconds]
reilithion has quit [Quit: Leaving]
<steve-chavez>
immae: Thank you! I'd like to know more about that function, I've been told it has a `pname` attribute that I don't understand it. Somehow it generates a name that appends the version.
<immae>
Well, in that case I can only suggest to go have a look at the source code
<immae>
(maybe someone over there knows better)
<steve-chavez>
Do you know where the source of that function is located? I'm lost on the nixpkgs repo.
ng0 has quit [Remote host closed the connection]
<immae>
pkgs/stdenv/generic/make-derivation.nix
ng0 has joined #nixos
drakonis has joined #nixos
iqubic has joined #nixos
<steve-chavez>
immae: That's just what I needed! Thank you for your support.
<immae>
I see no mention of pname though, but I know there is such thing for buildFooPackage where Foo is some programming language, are your sure you’re not confusing with other things?
<iqubic>
I really hate when Discord updates.
rcshm has quit [Read error: Connection reset by peer]
<iqubic>
Why the hell do they force you to use the latest version of Discord in order to connect using the stand-alone client?
rcshm has joined #nixos
fosskers has joined #nixos
<iqubic>
Discord is a chat client.
ddellacosta has quit [Ping timeout: 250 seconds]
<fosskers>
Sorry, I may have missed any reso2019-03-12 14:57 *** fosskers JOIN
<fosskers>
The hashes mentioned by "wanted" and "got" don't match anything I have referenced, particularly the image hashes of the image I'm successfully pulling down
<petersjt014>
anyone know where I can find documentation for config.nix (the one at ~/.config/nixpkgs/config.nix)? I'm trying to specify caches for cross-compiling so I can build cachix for armv6l but I don't know If I have the syntax right. This is what I got rn: http://0x0.st/zXzh.nix
<{^_^}>
[nixpkgs] @andrewchambers opened pull request #57549 → Add the Janet programming language. → https://git.io/fhjbv
<{^_^}>
[nixpkgs] @jgeerds pushed to master « misc: Remove myself from list of maintainers »: https://git.io/fhjbf
pie___ has joined #nixos
o1lo01ol1o has quit [Ping timeout: 245 seconds]
civodul has quit [Quit: ERC (IRC client for Emacs 26.1)]
<mankyKitty>
are there extra incantations required to make the services for `bolt` start up? the derivation drops them in the store but doesn't seem to set them up