<cap_sensitive>
Question about installing weechat: With weechat-unwrapped, the build fails at linking guile;
<cap_sensitive>
With weechat, the build fails at linking tcl
<cap_sensitive>
In my overlay, I removed some dependencies: weechat = super.weechat.override { configure = (ps: with ps; [ super.python3 ]); };
<cap_sensitive>
Now it stops at "Linking C static library libweechat_gui_common.a"
dansho has joined #nixos
<cap_sensitive>
What should I do to install weechat? I don't need anything GUI, a command line UI would suffice
<worldofpeace>
cap_sensitive: It sounds like the package is failing to build, which is an issue that would concern the maintainers. Can you open an issue on GitHub?
felixfoertsch23 has quit [Ping timeout: 258 seconds]
dongcarl has quit [Client Quit]
vonfry has quit [Ping timeout: 260 seconds]
dongcarl has joined #nixos
vonfry has joined #nixos
slack1256 has joined #nixos
fragamus has joined #nixos
fragamus has quit [Client Quit]
fragamus has joined #nixos
<vonfry>
hi, I was made a PR for offilenimap on add proxy supporting, and I saw another on offlineimap update by bot. I want to know when it can be reviewed? Furthermore, I find there are a lot of pr. If a pr has been reviewed for a long time, does it won't be merged any longer?
lovesegfault has joined #nixos
<{^_^}>
[nixpkgs] @jtojnar pushed to master « utillinux: no-rebuild clean up »: https://git.io/Jebkd
<{^_^}>
[nixpkgs] @jtojnar pushed 567 commits to staging-next: https://git.io/JebkN
<{^_^}>
[nixpkgs] @ryanorendorff opened pull request #76485 → agda-stdlib: list as not broken on Darwin → https://git.io/JebIK
bvdw has quit [Read error: Connection reset by peer]
bvdw has joined #nixos
dansho has quit [Quit: Leaving]
kvda has joined #nixos
kvda has quit [Client Quit]
doublex_ has quit [Read error: Connection reset by peer]
doublex_ has joined #nixos
felixfoertsch has quit [Quit: ZNC 1.7.3 - https://znc.in]
felixfoertsch has joined #nixos
brnzz has quit [Quit: Ping timeout (120 seconds)]
brnzz has joined #nixos
reallymemorable has quit [Quit: reallymemorable]
kolbycrouch has joined #nixos
dansho has joined #nixos
mac10688 has joined #nixos
Rusty1 has quit [Quit: Konversation terminated!]
ddellacosta has quit [Ping timeout: 258 seconds]
<dongcarl>
hey all, anyone got a minimal example of a working LDAP config for 19.09? Would love to reference
palo1 has joined #nixos
u1f320 has joined #nixos
palo1 is now known as palo
palo has quit [Ping timeout: 258 seconds]
<u1f320>
Hi. I'm booted in stable iso 19.09 and writing a huge/old configuration.nix which nixos-install says syntax error somewhere and it could be because of a parens missing somewhere. To help with this I tried to add that block of code from vim-addon-nix https://nixos.wiki/wiki/Editor_Modes_for_Nix_Files into the iso's configuration.nix in order to get iso's vim to use that addon, but it doesn't seem to have any effect. Any ideas?
<u1f320>
oh wait, it sort of works if I manually use the first vim executable, because of: collision between `/nix/store/1varyg017ghqgfnyhh3ipjdajmp46hj0-vim/bin/vim' and `/nix/store/2jzy04grw8pihdmykjyfp3p49zqj30gb-vim-8.1.1967/bin/vim'
<u1f320>
"sort of works" because I get this now: gzip: /root/.vim/vim-addons/vim-pi/archive/vim-pi.tar.gz: not in gzip format
<notgne2>
cap_sensitive: yeah that's what I expected to see, things in /run/wrappers/bin/ don't actually link to anywhere in the nix store, so I take it that's how Nix deals with this exact problem, I'm just not sure how to enable it
<ashesham`>
notgne2: thanks; what's the difference between the amdgpu and radeon drivers? With these options, the display server fails to start.
<notgne2>
ashesham`: I don't really know, I'
<notgne2>
I've had driver issues in the past but I resolved them by going the other way
<ashesham`>
notgne2: okay
<notgne2>
for me, on an R7 260x, radeon was extremely buggy, so I added the amdgpu driver and then added `["radeon.si_support=0" "amdgpu.si_support=1" "radeon.cik_support=0" "amdgpu.cik_support=1" ]` to my `boot.kernelParams` otherwise radeon will just want to load instead
LLED2_3 has quit [Ping timeout: 240 seconds]
<ashesham`>
yeah; the story of GPUs on Linux.
<ashesham`>
in any case, I think I might need to revert back, the logs on display-manager.service also appear to be useless.
<notgne2>
honestly you might have some luck updating your kernel, that solved some GPU issues for me at one point too somehow
bennofs has joined #nixos
<ashesham`>
I've tried the latest kernel to no avail.
<ashesham`>
I'm at a point where I want to dispose off this GPU on my desktop and look for alternatives; GPU issues in linux are so much shooting in the dark that it isn't worth the heartache, tbh.
<notgne2>
apparently radeon is actually the default, so you could try what I did
<clever>
ive been using amdgpu lately and havent had any issues with it
<notgne2>
or potentially just use the modesetting driver but I'm not really sure how that works
<ashesham`>
clever: amdgpu is _extremely
<ashesham`>
_ buggy with me.
<clever>
it may vary from card to card
afics has joined #nixos
<notgne2>
what card do you have ashesham`
<ashesham`>
yeah, my card is almost unusable; once it has booted up things are okay, but I have numerous issues with sleep-wake, boot (blank screens, blank screens with just a mouse pointer)
<clever>
i never put the machine into sleep, and i rarely reboot
<notgne2>
clever: I had the exact same until recently, I also have `03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Oland XT [Radeon HD 8670 / R7 250/350]` on my wife's machine, which is running with amdgpu perfectly (and had issues with radeon, and ofc no vulkan etc)
<ashesham`>
So far, really disappointed with its support on linux.
<ashesham`>
clever: do you keep your desktop running all the time?
<notgne2>
ashesham`: amdgpu really should be the driver for a card that new, you could try amdgpu-pro through in case there are any important differences (pro as in proprietary)
<clever>
ashesham`: 07:05:42 up 53 days 15:04, 2 users, load average: 0.00, 0.00, 0.00
<notgne2>
also I wouldn't really know what else to suggest in terms of GPUs honestly, I would've gotten the same for myself had it been any cheaper at the time, and NVidia has pretty awful support (no wayland, etc)
<notgne2>
clever: `04:05:58 up` here but that's because I just installed my UPS that long ago lol
<ashesham`>
clever: I'd be worried about my electricity bill with my machine's power consumption.
<clever>
notgne2: my ups failed a while back, so ive been at the whims of random 2 second blips lately
<clever>
ashesham`: .... i leave 3 desktops, 1 rack case, and a laptop on, 24/7
<ashesham`>
clever: and how much do you pay for electricity :-)
<ashesham`>
clever: but in any case, I'd like to drill down to the root of this problem and replace/dispose the appropriate hardware if needed.
<clever>
ashesham`: its fairly flat year-round, between 600 and 800 kWh
<clever>
and has been that flat for the full 3 years that i can see in the online billing page
<ashesham`>
clever: I see, I'd still like to be able to put the pc on sleep.
<ashesham`>
and overall, having things this unstable is also scary; and quite frustrating when you've shelled out a good sum of money to get yourself a nice machine.
tsrt^ has joined #nixos
<clever>
the previous drivers did all kinds of wonky things, it mostly looked like use before set bugs
<clever>
when i first login on linux, i would see the WINDOWS WALLPAPER on the desktop, lol
bennofs_ has joined #nixos
<ashesham`>
is there any GPU that is kind of reliable on linux?
<ashesham`>
(I)
<evils>
my wx 3100 is of the same architecture, i've had no issues with amdgpu, though amdgpu-pro doesn't seem to build
<ashesham`>
(I wouldn't want to buy anything intel, so no integrated graphics for me)
<evils>
as far as i know the only difference between amdgpu and amdgpu-pro is opencl support, for which i have higher hopes with rocm
<ashesham`>
evils: same for me, I'm getting ```error: unsupported xorg abiCompat 1.19 for xorg-server-1.20.5``` when trying to build with amdgpu-pro.
<evils>
btw clever, since you asked, i've since seen that your GPU is explicitly not supported by rocm
<clever>
evils: dang
bennofs has quit [Ping timeout: 260 seconds]
<evils>
ashesham`: i keep thinking "surely i've had an issue with this gpu" i think that was it, i spend some time trying to get amdgpu-pro to work, but it's not my forte
<ashesham`>
evils: getting GPU's to work on linux at this time, imo, is just guesswork.
<ashesham`>
and most of them out there are garbage; that includes ATI's and NVidia's.
<ashesham`>
and I think I'm just about to give up and toss this GPU out and get one that gives me the least trouble
bvdw has quit [Read error: Connection reset by peer]
bvdw has joined #nixos
<notgne2>
I think my favourite part of buggy GPU drivers is opening steam to see your friend's face plastered over literally every single texture
<evils>
how do you know this isn't one of the least trouble models?
Thra11 has joined #nixos
<ashesham`>
evils: googling? That's the best way to go I guess
<notgne2>
I honestly would've expected it to be, are you sure it's not just a dodgy card
<ashesham`>
dodgy?
<notgne2>
as in, with a manufacturing fault
<ashesham`>
notgne2: I have no idea, tbh.
<notgne2>
I had an NVidia Quadro FX 3800 which I assumed to be unusable due to drivers, but I then swapped it for an identical card which worked fine
<notgne2>
I then later realized that the broken GPU smells like burnt cheese, not sure if that's related
<ashesham`>
in any case, looking into that seems like a hassle to me; most likely my assembler is going to ask for the defective GPU back after which I'm likely to wait for weeks before I have a replacement.
<evils>
do you have anything else in your config related to the GPU?
<ashesham`>
evils: none that I can spot
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
alexherbo2 has quit [Ping timeout: 258 seconds]
<ashesham`>
notgne2: yeah I guess, maybe I should follow up for a replacement.
Thra11 has quit [Quit: WeeChat 2.6]
<ashesham`>
But then, not sure how to manage without a desktop, which is also my livelihood.
chenhq2005 has joined #nixos
<evils>
just to make sure, your CPU supports IOMMU and you're not trying to do PCI-e passthrough?
<ashesham`>
evils: how can I check that?
<cap_sensitive>
I don't think there's an option to wrap 'ping' for nix-darwin. I'm going to file a report to nix-darwin
<evils>
ashesham`: i think you'd know if you were attempting passthrough to a VM, for CPU support, what model is it?
bennofs_ has quit [Remote host closed the connection]
<evils>
ashesham`: have you looked through your motherboard settings? after that, maybe try with iommu disabled?
smolboye[m] has joined #nixos
<ashesham`>
evils: I have yes, and I don't even know what iommu means :-)
Thra11 has joined #nixos
dansho has joined #nixos
<ashesham`>
also, how will that help, from a brief readup I see that iommu deals with giving each pci device its own "domain" within the memory.
<dansho>
,locate api-sanity-checker
<{^_^}>
Couldn't find in any packages
<evils>
ashesham`: from your dmesg output it looks like the issue here is iommu related, if the iommu can just be disabled, it could bypass the problem
<dansho>
does ,locate search perl packages?
<evils>
dansho: i think it searches all outputs in hydra, so probably
<ashesham`>
evils: and disabling iommu can be done via the bios menu?
<evils>
ashesham`: i think there are options there for that, though you may need to pass the kernel a flag for it too, i'm unfamiliar with it
domogled has quit [Ping timeout: 240 seconds]
<ashesham`>
evils: okay, I'll try that now
<ashesham`>
evils: also, for my learning, what makes you think the issue is iommu related? (from the logs?)
<ashesham`>
evils: I've rebooted with iommu=off set as a kernel param.; isn't iommu supposed to add some sort of security for hw?
<evils>
ashesham`: jup, not sure how much advantage that has for a GPU, but we're just trying to narrow down the problem
<ashesham`>
evils: maybe I should've mentioned this earlier though: the random crashes on boot only happen on some boots
domogled has joined #nixos
<ashesham`>
once in a while, I'd be greeted with a blank screen and just a mouse pointer, sometimes just a blank screen; in the former case I'd be able to ssh into the machine but attempting to shut it down would freeze and crash the system, in the latter, the system would be completely unresponsive.
<evils>
wpcarro: sry, i missed the emacs part, i'm unfamiliar with it, are the errors critical?
<wpcarro>
evils: I'm attempting to nixify my Emacs setup, and being able to create a
<wpcarro>
terminal inside of Emacs is pretty critical to my workflow. I'd love to
<wpcarro>
understand more about the problem.
<wpcarro>
Before packaging everything with Nix, I could create terminals w/o these problems.
<wpcarro>
Someone told me that X and GL stuff is linked in a problematic way when
<wpcarro>
information though.
<wpcarro>
running a nix-built Emacs on a non-Nix system. I'm unsure what to do with this
<evils>
wpcarro: oh, you're not on NixOS?, try running emacs via `nix-shell -p emacs --pure`?
<wpcarro>
evils: I'd like to run the Emacs that my emacs.nix expression built. Can I
<wpcarro>
do that with `nix-shell -p ... --pure`? It's in `./result/bin/wpcarros-emacs`
<evils>
wpcarro: nix-shell -I . -p emacs --pure
<wpcarro>
evils: okay I'm in a `nix-shell` now.
<evils>
now run emacs and see if the issue occurs
<wpcarro>
Now shall I run `./result/bin/wpcarros-emacs` from within here? Or just `emacs`?
<evils>
ok, maybe i should haev asked what `wpcarros-emacs` is :P
<wpcarro>
evils: Okay. Good question. It's the binary I'm outputting from
<wpcarro>
terribly complicated. (I hope).
<wpcarro>
`emacs.nix`. I'll send over the expression if you're interested. It's not
Ridout has joined #nixos
dansho has quit [Quit: Leaving]
<evils>
so the `nix-shell -I . -p emacs --pure` command gave you a shell with only declared dependencies, from nix, and the package called by emacs in all-packages.nix
<evils>
if you changed that emacs for yours, just run emacs, executing the result/bin/wpcarros-emacs could work though
cjpbirkbeck has joined #nixos
<ashesham`>
evils: just had a crash
<ashesham`>
and it looks like its a hard lock up, I cannot ssh into the machine either
<ashesham`>
this is _really_ frustrating :-( )
<evils>
ashesham`: tried with the latest kernel yet?
<evils>
and do undo the kernel flag i think you added
<ashesham`>
evils: yes, just as unstable (I've tried a few days ago)
<ashesham`>
is there a place I can look for a kernel panic etc. after a hard reset?
imdoor has quit [Quit: imdoor]
<evils>
ashesham`: i don't know...
<ashesham`>
there are also issues when I hit reboot on the system; the usb seems to stay on (keyboard etc.) but the display goes off.
<ashesham`>
could it be an issue with the psu itself?
<evils>
ashesham`: i'm still looking through your full dmesg dump, there's some mention of VM, sure you're not doing anything VM related?
<ashesham`>
evils: not that I know of, tbh.
<evils>
ashesham`: your keyboard should still get power, your monitor probably just goes blank when it loses signal
<ashesham`>
there's the docker daemon that starts up, and AFAIK that's the closest thing I have to VMs on this machine.
zebrag has quit [Ping timeout: 240 seconds]
inkbottle has joined #nixos
<ashesham`>
Yeah, but at times the keyboard goes off when typing poweroff, once in a while it doesn't, neither does the fan etc.
<evils>
ok, dare i propose this heretical suggestion?, see if it works on ubuntu
<ashesham`>
the only kernel that seemed stable to me was 5_1, which was removed from 19.09
magneticduck has quit [Ping timeout: 248 seconds]
<ashesham`>
evils: that would entail reinstalling the whole system?
<ashesham`>
or booting off the usb.
<ashesham`>
and tbh, I really don't know why the non-lts kernels were just removed
<evils>
ashesham`: start with usb, if that works, maybe install ubuntu proper, reinstalling nixos shouldn't be too much hassle if you save your config (or just install ubuntu to a separate partition)
<ashesham`>
__monty__: seems overly aggressive to me, tbh, it should be up to the user to decide which kernel to run.
<ashesham`>
I've tried to switch back to 5_1 to no avail.
<__monty__>
The idea is that if you want a kernel that's supported for longer you have to run LTS. nixpkgs has no intention of providing support for any kernels that upstream doesn't.
<evils>
ashesham`: i'm afraid a non ad-hoc way of diagnosing hardware involves knowing what you're doing xD
<ashesham`>
evils: yeah, I understand; and I don't know what I'm doing here tbh.
<ashesham`>
just that I wish there was a place where the kernel would put out some information before a hard crash.
<evils>
ashesham`: fixing it for others in the future :P
<ashesham`>
BSOD doesn't seem like a bad idea at this point :-)
<evils>
ashesham`: if the problem is big enough that the kernel dies, it's not unreasonable that it can't save some info
<evils>
you wouldn't happen to have a serial adapter and another device?
<ashesham`>
I have another laptop, but not a serial adapter.
chloekek has joined #nixos
<__monty__>
ashesham`: clever mentioned the "crash kernel" yesterday which sounds like it does exactly that. It runs in a reserved part of memory.
<ashesham`>
actually, a reasonable thing here to do would be to switch back to 5_1; even if it is not available on 19.09, maybe via a custom derivation?
<ashesham`>
and I can probably also just copy the expr. from the nixpkgs using the earlier tag, right?
<__monty__>
Maybe just overlay nixpkgs with the kernel... exactly.
magneticduck has joined #nixos
<notgne2>
ashesham`: you can just add 19.03 as a nix channel and import it specifically for that package
<ashesham`>
notgne2: I'm a nix noob, so I'll need to look that up
<{^_^}>
[nixpkgs] @raboof opened pull request #76494 → Allow using vscode extensions in vscodium → https://git.io/JebO2
<evils>
ashesham`: that's the assert that failed
<ashesham`>
as I read it, the assertion is tht if the kernel is not null, then to asser that the kernel.version is older than 4.15, right?
<ashesham`>
and in that case, why would it fail with 5_1?
<__monty__>
Because 5 > 4.15?
<evils>
ashesham`: i'm not sure what `->` means there xD but if that's the right interpretation, yes, 5.1>4.15
<ashesham`>
__monty__: ah yeah, right, I was reading "older" as "later" for some reason, brainfart.
<__monty__>
Maybe you want an earlier version of 19.09 rather than 19.03?
<ashesham`>
but to me that assertion still doesn't make sense.
<ashesham`>
the assertion seems to state that if a kernel value is set, and is not null, then to ensure than the version is older than 4.15? And that assertion comes from where exactly? And why?
<evils>
makes no sense to me either, that assert is still the same on master
<wpcarro>
Is it reasonable to equate Nix "attributes" with directories and files
<wpcarro>
containing Nix expressions?
<__monty__>
Ok, that's perplexing. I assumed it was a statement of the maintainers that they didn't support any newer kernels for a particular release.
<__monty__>
wpcarro: Without more info I'd have to say "No."
<wpcarro>
__monty__: I ran `nix-env -iA <attribute> -f .` and was perplexed about how
<wpcarro>
Nix resolved <attribute>.
<ashesham`>
overall, I also think that the decision on which kernels to allow in nixpkgs is overreaching. Many people may choose to run particular versions of the kernel (in my case for example); due to hw compat. issues that are impenetrable to any kind of diagnosis.
<raboof>
on nixos, is there a convenient way to enable `allowUnfree` only for specific packages? i.e., for now I still want dropbox, but I don't want to accidentally install further unfree software?
<raboof>
for now I imported the nixpkgs again as 'nixpkgs-unfree' and set `config.allowUnfree` just for that import - that seems to work
<__monty__>
That sounds like a good way.
<__monty__>
wpcarro: And attribute here is?
<__monty__>
ashesham`: The idea is that either you're sticking to something older that works, LTS. *Or* latest will work for you.
<evils>
ashesham`: i think they want as few kernels to support as possible, and expect people that want something else to know how to set that up for themselves
<ashesham`>
basically, I have 2 choices it seems: dig into this issue and possibly waste the whole day on this when I had set out to do something else on the computer; or rollback to 19.03 completely.
<wpcarro>
__monty__: emacs. And I have a file called `emacs.nix` defined at `./emacs.nix`
LLED2_ has joined #nixos
<ashesham`>
evils: Maybe some kernels can be marked as unsupported instead of completely being removed?
<raboof>
clever: for `allowUnfreePredicate`? I tried `config.allowUnfreePredicate = (pkg: false);` but that already doesn't work.
LLED2_ has quit [Ping timeout: 240 seconds]
<__monty__>
ashesham`: That's not an option afaics. We've had discussions about SLiM being dropped recently and that was *not* an option. (Even though it still works just like it did in 19.03.)
<raboof>
(which is likely because I don't know what I'm doing ofc ;) )
<clever>
raboof: which file did you put that in? how did you test it?
<raboof>
clever: in /etc/nixos/configuration.nix, with 'nixos-rebuild switch
<__monty__>
raboof: That predicate would not allow *any* unfree packages.
<__monty__>
Try the next example.
domogled has quit [Ping timeout: 260 seconds]
<raboof>
__monty__: yes - the next step would be to allow just what I want, but this already fails with "error: Module `/etc/nixos/configuration.nix' has an unsupported attribute `boot'. This is caused by assignments to the top-level attributes `config' or `options'."
rardiol has joined #nixos
<clever>
raboof: you want to set nixpkgs.config.allowUnfreePredicate
<evils>
__monty__: btw, setting the python executable and patchShebanging everything .py didn't help with my kicad on debian issue
<raboof>
cool! that indeed works!
is_null has joined #nixos
<raboof>
though I have to say I'm pretty confused by the difference between `config` and `nixpkgs.config`
Rusty1 has joined #nixos
<clever>
raboof: all nixos modules, must contain 3 keys and only 3 keys, config, imports, and options
<ashesham`>
__monty__ I see.
<clever>
raboof: if a module is missing all 3 of those, the module system will wrap it with { config = { ... }; } for you
<ashesham`>
I guess the only way forward for me to try is the _latest kerne.
<clever>
raboof: and if imports winds up inside config, it will get moved out, becoming { config = { ... }; imports = [ ... ]; }
<clever>
raboof: then, there is also the nixpkgs config, which is at the nixos option of nixpkgs.config
vidbina has joined #nixos
<clever>
raboof: and behind the scenes, everything has an extra config. on it, so your file turns into this: { config = { boot = ...; nixpkgs.config = ...; }; }
<evils>
ashesham`: do post here or a PR if you find a solution
<ashesham`>
will do; I've given up on trying to switch to 5_1; but I'm now going to reboot with latest and see how the system behaves. From what I remember, I've tried this before and the system was quite unstable.
<ashesham`>
from what I understand; the _latest kernel is not a fixed version right?
wpcarro has quit [Remote host closed the connection]
<dirkx_>
During a `nixos-rebuild test' -- I get an build aborting error: ``error: cannot connect to daemon at '/nix/var/nix/daemon-socket/socket': Connection refused'' -- is that normal ? Or should I set a flag to stop the 'test' from doing things like that ?
romildo has joined #nixos
chenhq2005 has quit [Read error: Connection reset by peer]
<__monty__>
Though I'm not sure how quickly it's updated for a stable release.
<{^_^}>
[nixpkgs] @prusnak opened pull request #76497 → nixos-generate-config: i18n.consoleFont has been renamed to console.font → https://git.io/Jeb3V
<lirzhv>
also, what is the most appropriate way to install python pip. i see it is in nixos.python37Packages.pip but i suspect this package will pin me to python 3.7 which may be a problem when python38Packages.pip comes out
<__monty__>
python3Packages.pip?
<__monty__>
,locate python3Packages
<{^_^}>
Couldn't find in any packages
<__monty__>
Huh, I thought this existed.
<__monty__>
,locate pythonPackages
<{^_^}>
Couldn't find in any packages
<__monty__>
Ah, maybe it does but they're hidden from locate.
romildo has quit [Ping timeout: 268 seconds]
<__monty__>
> python3Packages.pip.name
<{^_^}>
"python3.7-pip-19.3.1"
<__monty__>
There you go, lirzhv
<lirzhv>
i discovered this handy command-not-found tool which tells me which package provides a certain binary, used it on pip3
<lirzhv>
__monty__: i think that python3Packages.pip.name doesn`t exist
<__monty__>
{^_^} says it does ^
<__monty__>
I only used ".name" so it wouldn't complain about evaluating the entire expression.
<__monty__>
Leave that off if you want to install pip.
<__monty__>
Also though, it's kinda doubtful you want to install pip. That's used for dirty dirty imperative package management ; )
zebrag has joined #nixos
inkbottle has quit [Ping timeout: 265 seconds]
<lirzhv>
wowowwow wait i can do pip stuff with nix? :o
<ashesham`>
__monty__: it seems to me that the kernel _latest has moved forward since the last time I had crashes by thistime after a boot.
<__monty__>
ashesham`: Definitely possible. I think it's supposed to track upstream but I'm not sure how closely.
<infinisil>
Well it is called _latest, indicating it's the latest kernel version, which of course changes over time
knupfer has joined #nixos
knupfer has quit [Remote host closed the connection]
reallymemorable has quit [Quit: reallymemorable]
knupfer has joined #nixos
is_null has quit [Ping timeout: 260 seconds]
vidbina has quit [Ping timeout: 260 seconds]
<__monty__>
infinisil: The opaque part is not whether it changes but how often/soon it changes.
vidbina has joined #nixos
romildo has quit [Ping timeout: 268 seconds]
knupfer has quit [Ping timeout: 258 seconds]
doublex_ has quit [Read error: Connection reset by peer]
doublex_ has joined #nixos
<ptrcmd_>
Hi, I am trying to build webkitgtk from nixpkgs rev f24f7010b719f84805215cab7af308e7700d1240
<ptrcmd_>
and it failed because it seems that some of the mirrors specified in that nixpkgs rev is unreachable
rardiol has quit [Ping timeout: 260 seconds]
jmeredith has joined #nixos
<ashesham`>
__monty__, infinisil: moreover, latest could also just mean the latest kernel at the time of 19.09's release.
<evils>
ashesham`: have you tried amdgpu from nixos-unstable?
<ashesham`>
I'm a nixos noob but considering the fact that nix's package management is supposed to be declarative, Id probably also expect a given package to not change.
<ashesham`>
evils: are there fixes upstream that might fix this? I did a bit of googling and it sems like the RX 580 is pretty bad on linux at this time: there have been similar reports of crashes etc. that I could spot.
<ashesham`>
and it seems like what one might suspect (and my worst fear): sh*t drivers
<evils>
ashesham`: just trying the latest everything is way easier than looking through the changelogs
<ashesham`>
I'll note that down in the list of things to try if there's another crash. Its been more than 1h on the _latest kernel and things are stable for now.
<ashesham`>
evils: and in any case, after a while of tinkering with this (months) where I've not been able to find out the real culprit, I'm inclined to just replace the GPU with something older and more supported (hopefully)
<evils>
ashesham`: i'd definitely try building everything from nixos-unstable before throwing hardware out...
shibboleth has joined #nixos
reallymemorable has quit [Quit: reallymemorable]
reallymemorable has joined #nixos
<{^_^}>
[nixpkgs] @kolaente opened pull request #76502 → hugo 0.61.0 -> 0.62.0 → https://git.io/JebsQ
<dirkx_>
Doing a `nixos-rebuild build'; seeing a new httpd.conf getting created in a new /nix/store/hash-path -- and a httpd service reload happeing -- but whatever systemd restart still points to the oldest /nix/store configured value.
<dirkx_>
Is there some command to force this too to change ?
<dirkx_>
systemctl show httpd shows that ExecStart* have indeed the wrong path set.
zupo has joined #nixos
<infinisil>
dirkx_: Well nixos-rebuild build doesn't apply the new system
<infinisil>
Is `nixos-rebuild switch` what you meant to use?
<dirkx_>
Ah 'k - had assumed that what you were running was indeed that - aws the first nixos-build did restart/create that serviced ?
kim0 has quit [Quit: Connection closed for inactivity]
pie_ has joined #nixos
bvdw has quit [Read error: Connection reset by peer]
bvdw has joined #nixos
jjakob has quit [Ping timeout: 248 seconds]
<lirzhv>
can someone help me understand this "with" construct in nix? i read that what it does is to "Add all attributes from the given set to the scope"... but i`m not sure i get it
<lirzhv>
is this some kind of list concatenation?
<lirzhv>
or why is it used here - python35.withPackages (ps: with ps; [ numpy toolz ])
<simpson>
lirzhv: It's a kind of dynamic scope action. In your example, `with ps;` is taking all of the attribute names from ps, all of the Python package names, and making them available as local names.
<samueldr>
like python35.withPackages (ps: [ ps.numpy ps.toolz ])
<simpson>
> with Python35Packages; [ numpy toolz ] # expecting a list of Python packages
<{^_^}>
[ <CODE> <CODE> ]
<samueldr>
`withPackages` is not related to `with`
<samueldr>
(in case there is confusion)
<simpson>
^^ Good point! You do not need to use the `with` keyword; it is a space-vs-understanding tradeoff.
<lirzhv>
samueldr: no, i did not understand that numpy and tools are actually coming from ps but not it makes sense
<lirzhv>
and ps is an argument to that function yes? am i reading the syntax correctly?
<samueldr>
> with { a = 42; b = 9001; }; b + a
<{^_^}>
cannot coerce a list to a string, at (string):271:29
<samueldr>
yes lirzhv
<simpson>
Yes.
<lirzhv>
nice, thanks guys
<samueldr>
uh...
<samueldr>
> with { a = 42; b = 9001; }; (b + a)
<{^_^}>
cannot coerce a list to a string, at (string):271:30
<samueldr>
nix-repl> with { a = 42; b = 9001; }; b + a
<samueldr>
9043
<samueldr>
weird discrepancy in the eval bot I guess
<infinisil>
samueldr: a and b are defined to something different with a let in
<infinisil>
> a
<{^_^}>
[ 2 3 5 7 11 ]
Grauwolf has joined #nixos
<samueldr>
hm, that's something I didn't know, that a previously scoped variable from a let would be stronger than the scope of a with
<Grauwolf>
hi! is there a way to tell nix to step off of configuring a service so that i can configure it myself? just a general question
dckc has quit [Ping timeout: 268 seconds]
<simpson>
Grauwolf: In general, sure; just don't set up the service in NixOS. Which specific service are you having trouble with?
chloekek has joined #nixos
o1lo01ol1o has quit [Ping timeout: 260 seconds]
ddellacosta has joined #nixos
<Grauwolf>
simpson: no specific one. was just a general question. thanks! :)
philr has quit [Ping timeout: 260 seconds]
ashesham` has quit [Ping timeout: 252 seconds]
pie_ has quit [Ping timeout: 260 seconds]
ddellacosta has quit [Ping timeout: 265 seconds]
reallymemorable has quit [Quit: reallymemorable]
pie_ has joined #nixos
<{^_^}>
[nixpkgs] @bignaux opened pull request #76508 → kmahjongg: init at 19.08.3 → https://git.io/Jebne
<genesis>
:)
Guanin has joined #nixos
<{^_^}>
[nixpkgs] @glasserc closed pull request #61194 → Pass CARGO_HOME on more invocations of cargo → https://git.io/fjCmU
dckc has joined #nixos
joepie91 has quit [Remote host closed the connection]
chloekek has quit [Ping timeout: 268 seconds]
jjakob has joined #nixos
vidbina has quit [Ping timeout: 248 seconds]
selfsymmetric-pa has joined #nixos
vidbina has joined #nixos
kleisli has joined #nixos
Ridout has quit [Ping timeout: 258 seconds]
pie_ has quit [Ping timeout: 268 seconds]
selfsymmetric-pa has quit [Remote host closed the connection]
<mpickering>
Anyone know about the error "error: the lock file /build/geckodriver.zip/testing/geckodriver/Cargo.lock needs to be updated but --frozen was passed to prevent this"
<{^_^}>
[nixpkgs] @danbst pushed commit from @d-goldin to master « nixos/manual: fixing links for awstats release note (#76500) »: https://git.io/JebcU
diego_ has joined #nixos
Ashjkaell has quit [Ping timeout: 248 seconds]
Chiliparrot has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
rxF is now known as torque
<diego_>
Hi! I am using lightdm with the mini greeter and xmonad. I used to have the default greeter. Now that I changed the cursor also changed to a cross, not only during the login screen but also when I hover over my xmobar and an empty desktop. Does anyone have a clue of what could be happening?
<afics>
Hi, I'd like to use ip vrf exec which fails because it isn't allowed to load BPF programs. Does anybody know if there's sysctl settings or something else for that?
<{^_^}>
[nixpkgs] @danbst merged pull request #76485 → agda-stdlib: list as not broken on Darwin → https://git.io/JebIK
<{^_^}>
[nixpkgs] @danbst pushed commit from @ryanorendorff to master « agda-stdlib: list as not broken on Darwin (#76485) »: https://git.io/JebcC
<Thra11>
I have set services.udev.extraHwdb, but it hasn't created 10-local.hwdb anywhere, and the contents I set don't appear anywhere in /etc/udev
mexisme has joined #nixos
<gagbo>
Hello, I'll trnasition soon to nixos and I'm wondering about partitioning : I'd like to have /boot clear and / encrypted, do I have to run the cryptsetup step before installing nixos ? I'm not sure to understand the manual correctly :(
jjakob has quit [Ping timeout: 260 seconds]
<Thra11>
It looks like the documentation for services.udev.extraHwdb says it'll write 10-local.hwdb, but nixos/modules/services/hardware/udev.nix looks like it's going to write extraHwdb into /etc/udev/hwdb.d/99-local.hwdb instead.
jjakob has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
shibboleth has joined #nixos
shibboleth has quit [Remote host closed the connection]
Heirlung has quit [Ping timeout: 260 seconds]
shibboleth has joined #nixos
lord| has quit [Ping timeout: 268 seconds]
diego_ has quit [Remote host closed the connection]
Heirlung has joined #nixos
<Thra11>
Looks like Nikolay Amiantov updated both the actual paths from 10 to 99, but only one of the paths in the docs.
<Thra11>
I'm assuming it's the docs which are wrong?
<evanjs>
networking.enableIPv6 is all I need for ipv6, right? Like the rest of it is going to be on me, router-side, correct?
gagbo__ has joined #nixos
gagbo has quit [Quit: I'm out !]
gagbo__ is now known as gagbo
<{^_^}>
[nixpkgs] @bignaux opened pull request #76513 → ksudoku: init at 19.08.3 → https://git.io/JebCy
Thra11 has quit [Quit: WeeChat 2.6]
<genesis>
:)
mexisme has quit [Ping timeout: 260 seconds]
<genesis>
next !
shibboleth has quit [Quit: shibboleth]
lord| has joined #nixos
Thra11 has joined #nixos
lord| has quit [Read error: Connection reset by peer]
lord| has joined #nixos
thomashoneyman has joined #nixos
selfsymmetric-pa has joined #nixos
<thomashoneyman>
quick question: i don't have any nix channels set up, as far as i can tell (~/.nix-channels has nothing, and nix-channel --list shows nothing)
<thomashoneyman>
(i'm on NixOS)
<thomashoneyman>
yet i seem to be able to install things just fine in home-manager and configuration.nix
<thomashoneyman>
but not the latest things, only things perhaps from the 19.09 release
<Guanin>
thomashoneyman: the channels is inherited from root, iirc
<thomashoneyman>
i have no idea what version of nixpkgs i'm on
<thomashoneyman>
ooooooh that's right, that's what i did
<thomashoneyman>
ok right i'm on 19.09 in the root
<duairc>
If I have a derivation that simply assembles a config file using pkgs.runCommand, but some of the paths in the config file are the outputs of other derivations, do I need to add those derivations to somethings like buildInputs?
<gchristensen>
thomashoneyman: `nix-info -m` (`nix-info` for a shorter output)
<thomashoneyman>
i need to write this down somewhere, i regularly forget
<thomashoneyman>
gchristensen: also useful, thank you
niksnut has quit [Remote host closed the connection]
cosimone has quit [Quit: Terminated!]
<duairc>
And if so... it's failing, because stdenv tries to source the outputs of derivations I'm adding, which fails because they're binary files...
<thomashoneyman>
another quick one: i'm installing postgres locally to muck around with a bit with some tests. adding it to configuration.nix and enabling the service sorta works: psql is available
<thomashoneyman>
however, there's no role for my user, nor am i able to use the 'postgres' user (psql -U postgres) as that has a 'peer authentication error'
<duairc>
Okay it seems that stdenv is just hardcoded to treat inputs which are single files as setup hooks. I guess I'll just have to restructure the dependencies to create a directory and output a single file inside that directory instead
joepie91 has joined #nixos
thomashoneyman has quit [Quit: Lost terminal]
thblt has joined #nixos
<evanjs>
clever: think I found the culprit of the random tarball that keeps stopping things. it looks like the nixos-upgrade service fails to download a tarball, and the service right now seems to be in an "activating" state?
<thblt>
Is there a way to disable a given “tunable” with powerManagement.powertop.enable? Specifically, autosleep for the receiver of my wireless mouse, which makes it lag for a half second or so every time I grab it.
<evanjs>
unsurprisingly, the process that owns the file is `nix-build <nixpkgs/nixos> --no-out-link -A system --no-build-output -I nixpkgs=nixos/nixexprs.tar.xz`
<prusnak>
"nix-prefetch-url --type sha256 --unpack https://example.com/foo.zip" shows different hash than expected by fetchzip - do you know where's the culprit? how do i calculate the correct hash?
o1lo01ol1o has joined #nixos
domogled has joined #nixos
tilpner has joined #nixos
justanotheruser has quit [Ping timeout: 260 seconds]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
opthomasprime has quit [Remote host closed the connection]
<chris__>
I'm trying to build a nixos config with `nix-build` instead of nixos-rebuild, to work around an error relating to using additional channels (like unstable). I'm trying the command `nix-build -I nixos-config=/mnt/etc/nixos/configuration.nix '<nixos>' -A system` but this gives me an error of `expression does not evaluate to a derivation (or a set or list of those)`. Any ideas what is wrong with my command?
<symphorien>
You can use nixos-rebuild build
<chris__>
Ohh, my bad. I'm on a live nixos USB - so I'm trying to work around `nixos-install`.. nixos-rebuild build wouldn't target the config from /mnt would it?
o1lo01ol1o has quit [Remote host closed the connection]
<duairc>
If you did nixos-enter to chroot into /mnt, I don't see why it wouldn't work on the target
skrzyp has joined #nixos
<chris__>
Hmm I've only got the nix config files in /mnt/etc/nixos/... So I don't think the chroot would work? (I'm trying to install nixos on an empty disk). At least `nixos-enter` comes back with `/mnt is not a NixOS installation`.
<infinisil>
chris__: replace <nixos> with <nixpkgs/nixos> in your original command
<chris__>
infinisil: thank you! I've gotten further now.
<elvishjerricco>
Is there a way to measure actual power usage? I've got a dell poweredge r710 and want to know how much I'm spending on electricity for the thing.
ddellacosta has quit [Read error: Connection reset by peer]
ddellacosta has joined #nixos
skrzyp has quit [Quit: WeeChat 2.6]
skrzyp has joined #nixos
dsx has joined #nixos
<evils>
elvishjerricco: your powersupplies probably report it somewhere, otherwise you can use something like a kill-a-watt
Rusty1 has quit [Remote host closed the connection]
<pikajude>
is it possible to write a stateless lexer for nix?
ng0 has quit [Quit: leaving]
<sphalerite>
pikajude: I'm not sure, not familiar enough with bison to understand the nix lexer, but maybe you could look at that and tell from there if it can be made stateless easily (or is already stateless)?
<pikajude>
yeah...
<sphalerite>
pikajude: I suspect it may not be possible because of URLs. Not sure though.
<sphalerite>
haven't given a lot of thought to it
<pikajude>
i tihnk it's impossible because of string splicing
justanotheruser has joined #nixos
<pikajude>
yeah, in nix itself the parser is stateful
Rusty1 has joined #nixos
justanotheruser has quit [Read error: Connection reset by peer]
chris__ has quit [Quit: Lost terminal]
ddellacosta has joined #nixos
Thra11 has quit [Quit: WeeChat 2.6]
cjpbirkbeck has joined #nixos
<evanjs>
via home-manager or nixos, how do I ensure dependencies normally provided to xmonad in a cabal file get picked up and used in my config?
<evanjs>
is xmonad.*Packages enough?
cmacrae has joined #nixos
<evanjs>
currently getting plenty of 'Could not find module' errors
<cmacrae>
lassulus: Just wanted to drop by and say thank you again for your guidance on generating some config with Nix! I managed to get it working how I wanted, and went on to write another function to do another type of config \o/ https://git.io/Jeb84