<noonien>
i do wish nix had a better way of importing modules, rather than `{ stdenv }:`, i'd prefer go-like import paths
<jasongrossman>
yl[m]: "I'm considering writing a book given I still have a fresh noob perspective" - Very good way of thinking about it, IMO. I've had several projects that I've found it difficult to write up because by the time I'd thought about them for a while I could no longer remember what the pain points had been when I was first thinking about them (when I was in the same state of mind as my audience).
<slack1256>
we do have a `import` keyword though
<Shados>
noonien: That's not importing a module, that's declaring a formal argument to a function.
stepcut has joined #nixos
<noonien>
i know, but that is the effect
random_yanek has joined #nixos
<yl[m]>
jasongrossman: precisely. Although, I don't really know where to find time for it :)
<slack1256>
let stdenv = import /path/to/stdenv/ ; in
<slack1256>
works
drakonis has joined #nixos
<noonien>
yes, but that's not used to import modules, but just evaluate nix files and store the result
o1lo01ol1o has quit [Remote host closed the connection]
<slack1256>
Mmmmmm either I don't get it or you don't.
<{^_^}>
[nixpkgs] @peti pushed to haskell-updates « hackage-packages.nix: automatic Haskell package set update »: https://git.io/fjfsV
<noonien>
in go, a package (akin to a nixpkg/nixos module) is imported by using the import statement. in nix terms `let lib = import /path/to/lib` would be just `import "lib"`
<noonien>
paths would be translated to actual sources elsewhere, in go's case a go.mod file
<DigitalKiwi>
yl[m]: well you're only working 16 hours and there are 24 hours in a day that leaves 8 hours to write your book
<yl[m]>
DigitalKiwi: lol
<yl[m]>
DigitalKiwi: I don't work there anymore, thank god
<yl[m]>
DigitalKiwi: I had a full physical breakdown after 2.5 years
<DigitalKiwi>
:(
<yl[m]>
and took 2 months sebatical
<slack1256>
ouch
o1lo01ol1o has joined #nixos
lopsided98 has joined #nixos
<yl[m]>
DigitalKiwi: I now do 8-10 a day and best of all no weekend work!!
pie_ has quit [Ping timeout: 245 seconds]
thc202 has quit [Ping timeout: 245 seconds]
<slack1256>
noonien: It's not that bad, plus on nix those files can be either functions or sets or lists. On go they would always be a set of declarations
filterfish has quit [Remote host closed the connection]
filterfish has joined #nixos
<noonien>
Yes, i believe that would induce a bit of change on how modules are handled, perhaps cleaning up the different expressions. There could still be a top-level function to take arguments, just not modules.
filterfish has quit [Remote host closed the connection]
<noonien>
I think the way that modules and parameters are mixed in package files is a bit "unclean".
filterfish has joined #nixos
filterfish has quit [Remote host closed the connection]
<andi->
Have you seen the use-flags PR? I think that tries to clean it up a bit.
<noonien>
hmm, i have not
<noonien>
i remember someone saying there was a problem when "src" was introduced
<noonien>
because "src" was used as an argument to a package
filterfish has joined #nixos
filterfish has quit [Remote host closed the connection]
filterfish has joined #nixos
filterfish has quit [Remote host closed the connection]
<Wulfsta>
Alright, so where do I put nix expressions?
<Wulfsta>
All the docs say pkgs/...
<Wulfsta>
What precedes that path?
<slack1256>
the nixpkgs repo usually
<slack1256>
but you can put your expression on whatever path you like, when you reference them you have to put that path as `let mypkgs = import /my/path/default.nix ; in`
<noonien>
andi-: what PR would that be? i can't seem to find it
ddellacosta has joined #nixos
o1lo01ol1o has quit [Remote host closed the connection]
o1lo01ol1o has joined #nixos
bemeurer has joined #nixos
<bemeurer>
Where does fetchFromGitHub come from? I'm trying to use it for my home-manager config but I keep getting error: attribute 'fetchFromGithub' missing
<Shados>
You have a lowercase 'h'
<{^_^}>
[nixpkgs] @flokli opened pull request #57938 → network-manager: move para about service rename to 19.09 changelog → https://git.io/fjfGe
<bemeurer>
Oh my god
<bemeurer>
duh :P
<bemeurer>
Thanks!
simukis has quit [Quit: simukis]
<bemeurer>
Shados: Is there any easy way to get the sha256? Other than placing a bogus hash, and then getting the real one from the hash
stepcut has quit [Remote host closed the connection]
mekeor has quit [Ping timeout: 252 seconds]
stepcut has joined #nixos
Supersonic has quit [Disconnected by services]
Supersonic112 has joined #nixos
Supersonic112 is now known as Supersonic
stepcut has quit [Ping timeout: 246 seconds]
hamishmack has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
timokau has joined #nixos
Wulfsta has quit [Quit: Page closed]
cyraxjoe has quit [Quit: No Ping reply in 180 seconds.]
teej has quit [Quit: Connection closed for inactivity]
cyraxjoe has joined #nixos
work_ has joined #nixos
timokau has quit [Ping timeout: 245 seconds]
slack1256 has quit [Remote host closed the connection]
nD5Xjz has quit [Ping timeout: 246 seconds]
nD5Xjz has joined #nixos
MightyJoe has joined #nixos
cyraxjoe has quit [Ping timeout: 250 seconds]
countingsort has joined #nixos
cyraxjoe has joined #nixos
MightyJoe has quit [Ping timeout: 272 seconds]
buffet has quit [Ping timeout: 245 seconds]
<teto>
trying to use remote builds but I keep getting "Host key verification failed". I've cleaned up /root/.ssh/known_hosts already, Whose known_hosts is used ?
rardiol1 has left #nixos [#nixos]
hellrazor has joined #nixos
hellrazo1 has quit [Ping timeout: 255 seconds]
timokau has joined #nixos
timokau has quit [Ping timeout: 272 seconds]
mkoenig_ has quit [Remote host closed the connection]
<mojjo>
hi! a question: at one point in my nixos config I specify some `extraUsers`. Is it possible to refer to those in another part, e.g. create a cronjob for each? the naive approach (just accessing config) produces an infinite recursion...
<mojjo>
oh, it does not..
o1lo01ol1o has quit [Remote host closed the connection]
evanm has quit [Ping timeout: 256 seconds]
drakonis has quit [Quit: WeeChat 2.3]
adetokunbo has quit [Quit: This computer has gone to sleep]
<tdammers>
so I enabled virtualbox guest additions (I think) in nixos running inside a VM, and configured a shared folder on the host side; is the shared folder supposed to appear somewhere on the nixos side? if so, where is it?
<hyper_ch2>
it will appear as network share, you have to mount it
johann__ has joined #nixos
<hyper_ch2>
at least with windows guests it was like that that you had server share virtualbox or something and inside that server share you had the shared folders
nD5Xjz_ has quit [Ping timeout: 250 seconds]
johanot has quit [Quit: WeeChat 2.4]
nD5Xjz has joined #nixos
<tdammers>
what kind of network share? where?
linarcx has quit [Ping timeout: 255 seconds]
winem_ has joined #nixos
<tdammers>
(maybe worth noting that I don't have x11 enabled, and am not planning to, so no DE and no automounter either)
<hyper_ch2>
dunno
<hyper_ch2>
as I said, in windows hosts it appears as server in the network
nD5Xjz_ has joined #nixos
linarcx has joined #nixos
<tdammers>
on the host side, it's a plain old directory that I pointed virtualbox to, the host side isn't the issue
nD5Xjz has quit [Ping timeout: 245 seconds]
<tdammers>
I want to know how I can use it on the guest side
<hyper_ch2>
yeah, I meant windows guest
<hyper_ch2>
just setup a windows guest vm and see how it appears
agander_ has joined #nixos
<hyper_ch2>
there's probably also documentation for it
<tdammers>
that won't help though, will it
dnlkrgr has joined #nixos
agander is now known as Guest43030
Guest43030 has quit [Killed (barjavel.freenode.net (Nickname regained by services))]
agander_ is now known as agander
<hyper_ch2>
it will show if the sharing is actually working
<tdammers>
it works fine with windows guests, I've done that before, but that doesn't tell me how I'd make it work in nixos
<tdammers>
oh wait, here we go
<hyper_ch2>
then you'd just have to figure out how to do it with a linux guest
* tdammers
tries mount -t vboxsf [-o OPTIONS] sharename mountpoint
freeman42y has quit [Ping timeout: 252 seconds]
<tdammers>
oh wait, it's already mounted in /media
<clever>
tdammers: `mount` should have shown that
<clever>
`df -h` may hide it, depending on what it says about total size
<tdammers>
right, and it does
<tdammers>
heh
<tdammers>
between the other 29 entries in the output of mount, I must have read past that entry
agander_ has joined #nixos
agander_ is now known as agander
agander has quit [Killed (asimov.freenode.net (Nickname regained by services))]
mkoenig has joined #nixos
johann__ has quit [Quit: WeeChat 1.9.1]
__monty__ has joined #nixos
work_ has quit [Quit: Connection closed for inactivity]
johann__ has joined #nixos
agander_ has joined #nixos
<Myrl-saki>
Oh boi. Do assertions not work with submodules?
<Myrl-saki>
The option `services.duplicity-backup.archives.asciivn.assertions' defined in `/home/myrl/Development/nix-duplicity-backup/duplicity-backup-common.nix' does not exist.
<Myrl-saki>
It kinda makes sense
<Myrl-saki>
I guess I'll have to import the assertions system?
<Myrl-saki>
Is this recommended?
<Myrl-saki>
# impl of assertions is in <nixpkgs/nixos/modules/system/activation/top-level.nix>
<Myrl-saki>
Uhh, maybe not.
wfranzini has quit [Remote host closed the connection]
<sb0>
is it replicated somewhere on a faster server?
<LnL>
hmm that still works, thought hydra downloads where disabled
agander_ has quit [Ping timeout: 246 seconds]
<LnL>
get the store path from the details and nix-store -r /nix/store/y8w6v09pa4zv2xqqx440c3v6fqpsy3vi-nixos-sd-image-18.09.2362.b6050c5f438-aarch64-linux.img
<sb0>
aha
agander__ has quit [Ping timeout: 246 seconds]
<LnL>
that will download it from the cache just like any other nix package and should be much faster
<sb0>
okay that seems to work, but it's downloading a bunch of dependencies as well
<{^_^}>
[nixpkgs] @vbgl opened pull request #57953 → dune: move into ocamlPackages → https://git.io/fjflp
iyzsong has joined #nixos
<manveru>
Orbstheorem: there's a package for it
<Orbstheorem>
manveru: Right, thanks!
linarcx has quit [Ping timeout: 245 seconds]
<Orbstheorem>
manveru: I installed it, restarted firefox, but it can't execute host app
<Orbstheorem>
Have you had this issue before?
linarcx has joined #nixos
<manveru>
i haven't tried it, sorry
<manveru>
but all it does is make this share/passff-host/passff.py
<Orbstheorem>
The path to the python app should be in the json file
<Orbstheorem>
I'm assuming it creates this file in firefox's share path
<manveru>
apprently you have to link it to "$HOME/.mozilla/native-messaging-hosts"
<manveru>
according to the install_host_app script
<manveru>
or well, depends on your browser
hamishmack has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<manveru>
and the package seems outdated, latest is 1.1.0 while this is 1.0.2
<noonien>
out of curiosity, how would one solve `chown user:users $(which sudo)`?
<manveru>
noonien: "solve"?
<noonien>
i had assumed that the binary is actually in a read-only fs
<manveru>
it is
<noonien>
and a friend was asking how to solve chown-ing / recursively, so i just chowned sudo
ubert has joined #nixos
<clever>
noonien: nix doesnt allow setuid binaries in /nix/store/, so nixos must create a wrapper on bootup
<noonien>
well, it is not, `sudo` actually points to a wrapper
<manveru>
oh :)
<clever>
noonien: if you reboot, it will fix sudo
<noonien>
clever: ah, so basically a reboot?
<clever>
the sudo wrapper, is on a tmpfs
<clever>
so its lost at shutdown
<clever>
and recreated on bootup
<spacekookie>
When I try to install qt applications via nix (on non-nixos) with nix-shell it crashes because the qt platform can't be loaded. What do I need to include to actually make that work?
<spacekookie>
I would have assumed that any qt package would provide the platform?
<manveru>
i forgot about the wrapper thing...
<clever>
tmpfs on /run/wrappers type tmpfs (rw,nodev,relatime)
<noonien>
interesting!
<clever>
spacekookie: QT_PLUGIN_PATH must be set, to a path like ${super.qt5.qtbase.bin}/lib/qt-${super.qt5.qtbase.version}/plugins
<clever>
spacekookie: nix-env solves that by impurely jamming it into ~/.nix-profile/lib/ and then breaking any qt with an incompatible version
<spacekookie>
Oof
<noonien>
is there anoter way, apart from rebooting?
<spacekookie>
I would have expected nix-shell do set that for me
wfranzini has quit [Remote host closed the connection]
<clever>
noonien: nixos-rebuild switch, with a change in configuration.nix
<clever>
noonien: assuming you have root somehow
wfranzini has joined #nixos
eeva has joined #nixos
<Orbstheorem>
what must an applicate do in order to install files in /run/current-system/sw/share/passff-host/ ?
<fmsbeekmans>
As part of a build process in nix I need to parse and modify a JSON file, is this something that I would want to do in NIX or should I outsource json modification it to a different language/tool?
fendor has joined #nixos
<noonien>
If anyone is interested, I managed to fix the permissison for my `sudo` binary, by using `sudoedit` on VBoxHeadless(a binary that happened to have suid), then backed up and overwritten the temporary file created by sudoedit with a binary that chowned and chmodded sudo, and ran it
<noonien>
(i also happened to be in the vboxusers group)
troydm has joined #nixos
Avaq has quit [Ping timeout: 255 seconds]
agander has joined #nixos
johann__ has quit [Quit: WeeChat 1.9.1]
<manveru>
fmsbeekmans: if it's simple, you can do it in nix, otherwise you can make a derivation that's doing it via `jq` or similar
<eyJhb>
Would anybody care to review #57915? :p Should be fairly easy
<betaboon>
anyone has a PR for updating elasticsearch to 6.6.2 already ?
<slabity>
Sorry, what I mean is do you want to just install wxPython in your user environment? If so, you can just use `nix-env -iA nixos.pythonPackages.wxPython`
<slabity>
If you're just using wxPython as a test for learning pypi2nix, that's a little more involved
<kandinski>
slabity: I'm trying to install the dev environment from work, which includes wxPython
<kandinski>
I was hoping to be able to create it with pypi2nix from the requirements*.txt files, because well, everything else works so well
rycwo has quit [Quit: WeeChat 2.2]
rycwo has joined #nixos
<slabity>
Huh. I was gonna point you to https://nixos.wiki/wiki/Python, but it looks like the pypi2nix section is completely blank
<kandinski>
I'm already reading both that and the pypi2nix instructions, and coming up blank. That's why I wondered of someone had an example already working.
<slabity>
Does `pypi2nix -r requirements.txt` not work?
<kandinski>
nope, fails when building the native dependencies
<slabity>
kandinski: Can you post the requirements.txt?
<kandinski>
slabity: sure
nomeata has joined #nixos
agander has quit [Ping timeout: 255 seconds]
<nomeata>
Hi. I started using a remote builder (a server of mine) that I access via `ssh`. But often I get random aborts sayin “error: unexpected end-of-file”. Any idea where to look and how to debug that?
<hexa->
does anybody have a dns01 challenge via dynamic updates (rfc2136) setup up and running?
<hexa->
looking for inspiration
<__monty__>
nomeata: Maybe the connection's getting dropped?
<slabity>
kandinski: Which python version do you need to build this for?
stepcut has joined #nixos
<nomeata>
__monty__: maybe, hard to tell … I never had problems with other tools connecting to that server over ssh.
<kandinski>
3.7 or 3.8, it's a new project
<Ankhers>
Does anyone have experience running kdeconnect? I'm trying to get it running through home-managers service, but I am getting an error in the indicator saying with the following error -- https://gist.github.com/ankhers/6f598e98e6a6e90f05d69559cd98c510. Any ideas?
<{^_^}>
[nixpkgs] @LinArcX opened pull request #57967 → mcos-dark-xfce-edition: init at f0ec359 → https://git.io/fjf0S
stepcut has quit [Remote host closed the connection]
nckx has quit [Quit: Updating my GNU Guix server — https://guix.info]
<goibhniu>
I'm running it Ankhers
stepcut has joined #nixos
<kandinski>
slabity: it's 2am here, I'm going to bed now. Please let me know how you go if you give it a try.
<kandinski>
slabity: (and thanks!)
<musicmatze>
Ankhers: I don't know about kdeconnect, but I ran into an issue the other day where a Qt application has to be actually installed instead of just nix-build'ed and ./result/bin/app called. ... I don't know how home-manager works, but maybe try nix-env -iA'ing it to compare?
<goibhniu>
Ankhers: I have it installed in my systemPackages. By the way, it's also important to open some ports: networking.firewall.allowedTCPPortRanges = [{ from = 1714; to = 1764; }]
zupo has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<Ankhers>
goibhniu: I just disabled my firewall for now just to test it out. Do you have some sort of service running, or are you starting it manually?
<emily>
but I think cache.nixos.org is Fastly now?
<emily>
and has some known issues for some people
<samueldr>
right, it's fastly
<teozkr>
oh, good point
<samueldr>
and it's highly likely that then ext time you try (like now?) it will work
<samueldr>
:/
<teozkr>
found some old issue that mentioned AWS and got overexcited :|
<srhb>
samueldr: No, I can reproduce it
<teozkr>
and nope, still broken samueldr
<samueldr>
I said likely :)
<teozkr>
:P
<srhb>
Anyway, since we get 503, it _might_ be AWS? Like, fastly can't get the object from the bucket, say.
<srhb>
Hard to say since both providers claim to be healthy..
<samueldr>
oh, that one looks like an AWS one
linarcx has quit [Ping timeout: 268 seconds]
<samueldr>
or at least, I have headers
<samueldr>
but not 503, curl: (23) Failed writing body (1671 != 2759)
<samueldr>
with different sizes each run
<srhb>
"ERROR 503: first byte timeout.
<srhb>
(from wget)
<samueldr>
curl
linarcx has joined #nixos
<emily>
AWS is famous for claiming nothing's wrong when there is
<teozkr>
hey, their health check example file is probably fine :P
<emily>
"a meteor could hit us-east-1 and it'd be reported as 'increased error rates in some regions'", etc.
<teozkr>
haha
<samueldr>
if I knew more about AWS I would try curling directly from there to see if it's somehow fastly that's failing, but yeah, that's a different 503 :/
<srhb>
It's funny that we see different errors though. I just get zero ytes from them.
<emily>
for what it's worth, it works fine here
<samueldr>
points towards fastly somehow having trouble?
<srhb>
Yup. It _sounds_ as though fastly is not getting the first byte from aws, but who knows...
<srhb>
Maybe AWS realized it would be a competitive edge to block all other CDNs.
<srhb>
:-)
<teozkr>
haha
<andi->
to me it seems more like it times out on larger files that they try to write to their cache first before delivering it and then timeout on some other part that is user-facing
<teozkr>
curiously I had the same issue trying to download mongodb
<teozkr>
.. but that seems to have resolved itself in the meantime
<srhb>
andi-: Surely they stream first-hits?
stepcut has quit [Ping timeout: 255 seconds]
* srhb
shrugs
<srhb>
As long as they fix it :-)
<andi->
srhb: It doesn't seem like it. I tried the above nar.xz and it tool SECONDS to report HTTP/2 200 on a HEAD
<andi->
now it is fast(er)
<srhb>
andi-: Wow.
<andi->
I think they are just chaining a bunch of webservers causing the 503 to the client..
<andi->
so the timeout we are able to configure is probably just the time to the response header from their backend?! All wild guesses at this point.
<gchristensen>
fwiw AWS's status page will be green even if their entire datacenter has burned to the ground -- a few servers being slow definitely won't make it to the status page.
wfranzini has quit [Remote host closed the connection]
<gchristensen>
they have not chained webservers -- we have their "shielding" feeature (which does this) disabled
kisonecat has quit [Ping timeout: 246 seconds]
<andi->
When Debian moved to fastly they also had similar issues for MANY users. Not sure how they solved that.
<samueldr>
gchristensen: fire gets a blue square
<gchristensen>
we have several configurable timeouts
<srhb>
samueldr: "unconfirmed reports of latency"
wfranzini has joined #nixos
<srhb>
samueldr: :-)
<srhb>
Meanwhile, screams in the background.
<teozkr>
okay that's curious, I get the same result as andi- for HEADs
<teozkr>
(first request takes ages but eventually 200s, subsequent requests 200 instantly)
<gchristensen>
S3 is definitely having trouble.
<teozkr>
and now.. regular GETs work too?
<srhb>
teozkr: Yeah, I see the same.
<andi->
can't we get someone to work at AWS to debug S3 for us?
<gchristensen>
lol
knupfer has joined #nixos
<teozkr>
haha
stepcut has joined #nixos
<srhb>
gchristensen: How can you tell?
<srhb>
gchristensen: Maybe just my location was good..
<gchristensen>
AWS probably took out a sick node or something
<gchristensen>
I'll increase the first byte timeout from 15000ms to 20000ms
<emily>
very polite of them to use a conservative connections limit to protect S3 from overload
<srhb>
gchristensen: I doubt it'll help
<gchristensen>
okay, I won't :) I'm going to lunch, back in a bit
Fare has quit [Ping timeout: 252 seconds]
<srhb>
gchristensen: Enjoy :)
<teozkr>
alright, enjoy
johanot has quit [Quit: WeeChat 2.4]
<gchristensen>
emily: good point. I'll increase that from 200 to like 10 billion or something :P
simukis has joined #nixos
o1lo01ol1o has joined #nixos
o1lo01ol1o has quit [Ping timeout: 250 seconds]
reinhardt has quit [Quit: Leaving]
kisonecat has joined #nixos
<linarcx>
What is this error: `error: anonymous function at /home/linarcx/nixpkgs/pkgs/tools/networking/persepolis/default.nix:1:1 called without required argument 'buildPythonPackage', at /home/linarcx/nixpkgs/lib/customisation.nix:69:12`
<andi->
you probably didn't use callPackage to "import" that package expresssion?
<{^_^}>
[nixpkgs] @aanderse opened pull request #57973 → monitoring-plugins: compile missing mysql, ldap, dig, procs, and dbi executables → https://git.io/fjfuR
drakonis_ has quit [Ping timeout: 246 seconds]
jbgi has joined #nixos
linarcx has quit [Ping timeout: 246 seconds]
<disasm>
fpletz: we need the prometheus 2 stuff in IOHK. jbgi is willing to work on it since I see the original submitter hasn't followed up with your requests in https://github.com/NixOS/nixpkgs/pull/49802. Should we open a new PR since he can't push commits to the original authors branch?
<{^_^}>
#49802 (by azazel75, 19 weeks ago, open): Add Prometheus 2 service in parallel with 1.x version
linarcx has joined #nixos
<disasm>
fpletz: also, if you have any other suggestions of chunks in the mayflower nixpkgs that should be added to the PR, please let me know.
rprije has quit [Read error: Connection reset by peer]
rprije has joined #nixos
goibhniu has joined #nixos
<slabity>
kandinski: I got most of the way using `-E "pkgconfig zlib libjpeg openjpeg libtiff freetype lcms2 libwebp tcl gtk3 libGLU gstreamermm which glibc libcxx libstdcxxHook"` as a flag, but the wxPython tests are looking for a hardcoded /lib/cpp path. Normally a nix expression would pass `--replace CPP=/lib/cpp CPP=${stdenv.cc}/bin/cpp` to direct it to the actual compiler, but I'm not sure how to get pypi2nix to do so.
<slabity>
If you don't need to pip-provided wxPython package, I would recommend using the Nixpkgs one. It sounds like a headache to get it working
stepcut has quit [Ping timeout: 255 seconds]
stepcut has joined #nixos
<{^_^}>
[nixpkgs] @xaverdh opened pull request #57975 → WIP: Redshift add early startup option, rewrite to use toINI → https://git.io/fjfuM
<srhb>
rycwo: It also has things like: /nix/store/p8nr27r69v2lk8v165b7k5f7zk748xq4-firmware-linux-nonfree-2018-12-13/lib/firmware/ath10k/QCA988X/hw2.0/firmware-5.bin
<srhb>
rycwo: (note: ath10k)
<srhb>
It's a big bunch of stuff.
ilmu has joined #nixos
<rycwo>
right, that's what I'm looking for
<rycwo>
I'll dig a little then come back...
<rycwo>
thanks!
<srhb>
rycwo: hardware.enableRedistributableFirmware installs the package. You may need to bump it, I suppose.
martinb has left #nixos ["Leaving"]
timor has joined #nixos
<linarcx>
:Gyus, in neovim, when i want to try :terminal, its open up a black buffer, not a real terminal. why?
<linarcx>
*blank
<slabity>
linarcx: Is it not executing your shell?
<srhb>
rycwo: So I'm under the impression that there's firmware and there's kernel driver. Firmwares get.. loaded onto the "board" at runtime, but one lives on the "board" at all times. Without newer firmwares, the old one is loaded? Or something. I may be talking out of my arse and this is just how cpu microcode works.
<samrose>
anyone have insight into the error /nix/store/pathto/openssl/lib/libssl.so.1.0.0 no version information available ?
<samrose>
I only have one version of openssl installed, but still seem to see this error
_kwstas has joined #nixos
Fare has joined #nixos
hodor123456 has joined #nixos
rcshm has quit [Remote host closed the connection]
_kwstas has quit [Client Quit]
<hodor123456>
hello world
rcshm has joined #nixos
<samrose>
ah I see it can mean that versions don't match what is expected in executable
dnlkrgr has quit [Ping timeout: 252 seconds]
<hodor123456>
How does nix package manager deal with multiple versions?
<srhb>
hodor123456: You probably need to specify a bit further your use case. Basically, every "package" lives in /nix/store/somehash-package-version -- so nothing ever clashes at a low level
<srhb>
(Because the "somehash" will always differ)
<hodor123456>
deal with dependencies correctly when multiple Octave and packages co-exist
ubert has quit [Quit: Leaving]
<hodor123456>
srhb, my target isto deal with dependencies correctly when multiple Octave and packages co-exist.
<srhb>
hodor123456: That should Just Work, as long as you don't install them into the same profile
<hodor123456>
srhb, sorry I didn't get that
<srhb>
hodor123456: Have you read any resources about nix yet? Trying to figure out where to start explaining :)
<slabity>
hodor123456: Yea, you can't install v1 and v2 of a package in the same profile. You need to install them in separate profiles
fusion809 has quit [Remote host closed the connection]
Ariakenom has joined #nixos
<slabity>
tbf the manuals are pretty... Heavy
<srhb>
Yeah, just trying to calibrate.
<slabity>
But they're definitely worth a read
<hodor123456>
srhb, actually I want to implement 'pkg' octave package manager. which would have following use cases.
<hodor123456>
keep multiple versions of the same package installed side-by-side
<hodor123456>
keep multiple versions of Octave in a system using the same installed packages
<hodor123456>
I am totally lost in dependency hell, would you mind to explain it all.
hooo has joined #nixos
erasmus has quit [Quit: END-AUMF]
<srhb>
hodor123456: The short answer is: Don't install multiple versions into the same profile. We use nix-shells for this purpose usually
<srhb>
For instance, giving a python example...
<slabity>
hodor123456: Well, for example I could run 'nix-shell -p gcc5' and that shell now has a profile with gcc on it. Then I can run `nix-shell -p gcc7` in another shell and they will both work
<hodor123456>
srhb, I am here from ubuntu guys, they headed me down hre
<srhb>
hodor123456: nix-shell -p 'python3.withPackages (ps: with ps; [ requests six ])'
fusion809 has joined #nixos
<srhb>
hodor123456: OK, you're in for a ride ;-)
<srhb>
hodor123456: So what we generally do is make it possible to cheaply spin up "virtual environments" (nix-shells) wherein certain package versions are available
<srhb>
hodor123456: So I could spin one up with one version of octave, and in a different shell, spin up one with another version of octave
<srhb>
hodor123456: This is bread and butter for nix.
<srhb>
The example above spins up a python3 shell with the packages requests and six
<srhb>
I could just as well spin up a python2 version, or one with different packages.
ikitat has quit [Ping timeout: 246 seconds]
<srhb>
They're not "installed" in the usual /usr/bin sense, but I can make them available in my shell at any time.
<srhb>
I think that's the ultra short version...
<hodor123456>
srhb, what do you mean by profile?
<srhb>
A profile is a symlink of packages that are immediately available on your system or for your user without invoking any nix-shell shenanigans. They look like /bin/python /share/doc/man... etc and live in ~/nix-profile and similar places. These do not allow conflicting definitions in general.
<{^_^}>
[nixpkgs] @4z3 opened pull request #57979 → nixos/nginx: use nginxfmt and gixy → https://git.io/fjfgn
<slabity>
~/.nix-profile; hidden folder
jbgi has quit [Ping timeout: 246 seconds]
<srhb>
A symlink tree* I meant to say..
<hodor123456>
srhb, so basically we can switch from one version to another.?
<srhb>
hodor123456: Yeah, very quickly. Usually we tie this to development "projects" (say, a git repo) so that you can just invoke nix-shell when you enter your project, and all dependencies in the correct version are available
<slabity>
hodor123456: It also allows you to do things like rollback to a 'safe' profile if a package breaks or your system becomes unstable
<hodor123456>
I am still struggling understanding the solution of dependency problem.
<hodor123456>
slabity, that is awesome.
<slabity>
Because your system has a profile, and your user has a profile
<srhb>
hodor123456: I guess the Nix solution is really to not have the problem in the first place..
slack1256 has joined #nixos
Dr8128 has quit [Ping timeout: 245 seconds]
zupo has joined #nixos
<hodor123456>
can nix update/autoupdate from repo?
<slabity>
Yea
zupo has quit [Client Quit]
<hodor123456>
can it update a specific version?
<slabity>
Each time you update, it creates a new profile. If it successfully builds, then it atomically switches to the new one
<slabity>
If it fails, then you're not left with a half-way installed system
<hodor123456>
ulike updating every version.
<hodor123456>
slabity, that's nice.
<srhb>
I'm not sure what's meant by "autoupdate from repo"
<hodor123456>
srhb, I meant if any update is available in the repo, it would be done by nix automatically.
periklis has quit [Remote host closed the connection]
<hodor123456>
I meant update.
<srhb>
hodor123456: Which repo? nixpkgs?
<hodor123456>
srhb, yeah, I suppose.
<srhb>
You can do that, yes. I advise against it though.
<hodor123456>
I meant, can nix do that automatically?
PLPD-Bot has quit [Remote host closed the connection]
<srhb>
Yes, it can.
<hodor123456>
<hodor123456> can it update a specific version?
<srhb>
Like, one package to a specific version?
<hodor123456>
yes
<hodor123456>
nope
Myhlamaeus has joined #nixos
<hodor123456>
one package_with_specificVersion to latest version.
<catern>
does anyone know why we have this crazy setup where gnupg1 points to gnupg2 with some additional symlinks?
<catern>
gnupg1 and gnupg2 are not syntax compatible
<srhb>
hodor123456: I'm really unsure what you mean by latest version. The latest in our repo, or the latest upstream version?
<srhb>
hodor123456: I mean, both are sort of possible in varying degrees, but... At this point there's a lot of details to discuss before I just answer yes or no :P
<slabity>
hodor123456: So what do you mean update a specific version to latest version? If you want a specific version you can do so. If you want the latest version you can do so
PLPD-Bot has joined #nixos
<catern>
and if you want actual version 1 of gnupg, you have to use gnupg1orig - that seems really dumb
<hodor123456>
can I have the whole working mechanism in a summary?
<srhb>
hodor123456: No :P
<hodor123456>
*of nix
<hodor123456>
:(
<srhb>
hodor123456: There's, like, a whole PhD paper about it. And a ton of huge manuals :)
<slabity>
And building it using `nix-build '<nixpkgs/nixos>' -A config.system.build.isoImage -I nixos-config=iso.nix`
<slabity>
linarcx: It's likely neovim is not interacting with your terminal properly if it won't respond to `:terminal /bin/sh`. What's your TERM variable?
goibhniu has quit [Ping timeout: 245 seconds]
zakkor has joined #nixos
<zakkor>
yo
<linarcx>
slabity: xterm-256color and my terminal is kitty. Is it a bug?
<zakkor>
what's the best way to set a different refresh rate for my display?
<slabity>
linarcx: Possibly. Might be a config issue, but I can't tell.
<linarcx>
slabity: Thank you. i also try: `nvim --noplugin`, but nothing changed. Maybe i have to try another terminal.
<slabity>
I just realized I'm running the nixos-unstable-small channel
<tilpner>
It happens on 19.03 too
<slabity>
Would that cause me not to use any caches when building this?
<tilpner>
So no
<tilpner>
It also doesn't build without the overlay, for some reason
* tilpner
is gone for a few minutes
reinhardt has joined #nixos
<catern>
disasm: ty
pie_ has quit [Ping timeout: 272 seconds]
kisonecat has quit [Ping timeout: 268 seconds]
kisonecat has joined #nixos
dashkal has quit [Ping timeout: 255 seconds]
<catern>
tfw a piece of software says "we want to make this software easy to install and deploy" and has release tarballs
linarcx has quit [Quit: WeeChat 2.4]
<catern>
but their makefiles have a hard dependency on running out of a git repo, and don't work on the release tarballs
ym555 has joined #nixos
<{^_^}>
[hydra] @disassembler opened pull request #643 → allow using a shorter context and increase hydra-notify debug → https://git.io/fjf2B
<Dandellion[m]>
has anyone had problems launching vscode 1.32.2?
englishm_ has joined #nixos
stepcut_ has quit [Ping timeout: 246 seconds]
dashkal has joined #nixos
rauno has joined #nixos
<Dandellion[m]>
`/nix/store/9psx6gyndawpw0m5pjcjsqzcgr41snyn-vscode-1.32.2/bin/.code-wrapped: line 35: /nix/store/9psx6gyndawpw0m5pjcjsqzcgr41snyn-vscode-1.32.2/lib/vscode/bin/../code: No such file or directory`
<tilpner>
slabity: I don't know either. Please create an issue for this (and link it here afterwards)
ng0 has joined #nixos
stepcut has joined #nixos
linarcx has joined #nixos
fendor has quit [Ping timeout: 252 seconds]
<vikingman_>
still not managed to make my laptop gpu working i got a gt335 and im not able to make optimus working well and im not sure that bumblebee is doing the right thing
<slack1256>
nvidia propietary driver or noveau?
<vikingman_>
when i check my current gpu its always tell me its the intel one thats running
<vikingman_>
with the nouveau driver
<vikingman_>
buts its a legacy hardware
neminis has joined #nixos
stepcut has quit [Ping timeout: 245 seconds]
sinner has joined #nixos
sinner is now known as Guest44467
<vikingman_>
its seem to work using nouveau and bumblebee
<vikingman_>
since steam is starting
<vikingman_>
but im not sure because its reporting an intel gpu
<{^_^}>
#57984 (by Slabity, 30 seconds ago, open): Building ISO takes excessive amounts of time+memory
<tilpner>
(optirun may work too)
<tilpner>
slabity++
<{^_^}>
slabity's karma got increased to 1
<vikingman_>
let me take a look at it
hodor12345 has quit [Ping timeout: 255 seconds]
<{^_^}>
[hydra] @disassembler opened pull request #644 → improve the error messages when invalid declarative jobsets are defined → https://git.io/fjf25
jbgi has joined #nixos
Izorkin has quit [Remote host closed the connection]
hodor123456 has joined #nixos
Izorkin has joined #nixos
<vikingman_>
it return GLXBadContext
ikitat has quit [Ping timeout: 246 seconds]
<tilpner>
vikingman_: Did you try both (optirun, primusrun)?
* tilpner
isn't familiar with the differences, but sometimes one seems to work better than the other
<{^_^}>
[hydra] @disassembler opened pull request #645 → redo the notify events between eval and queue-runner → https://git.io/fjf2A
<tilpner>
Is bumblebeed running? Is it reporting errors/warnings?
<vikingman_>
where the error should be located
englishm_ is now known as englishm
<vikingman_>
both optirun and primusrun gave me the same thing
<tilpner>
journalctl -eu bumblebeed
<vikingman_>
no entries
<tilpner>
systemctl status bumblebeed
<tilpner>
Does it say "active"?
<vikingman_>
Unit bumblebee.service could not be found
<tilpner>
You forgot a "d"
ThatDocsLady_ has quit [Ping timeout: 245 seconds]
<vikingman_>
ho
<vikingman_>
Mar 20 15:33:49 UFO bumblebeed[738]: [XORG] (EE) /dev/dri/card0: failed to set DRM interface version 1.4: Permission denied
<vikingman_>
last entree
<vikingman_>
and a few other error
<tilpner>
But is is "active"?
<tilpner>
Paste the other erros to e.g. gist.github.com
<scientist>
hey guys i want to zero out my usb with this command: sudo dd if=/dev/zero of=/dev/sdb status=progress , but it takes way to long, what do i change to make it faster ?
leat2 has quit [Remote host closed the connection]
<slabity>
If you do 'set shell=/usr/bin/env zsh' it should use whichever zsh shell is in your path
<slabity>
scientist: bs=1M
swflint has quit [Ping timeout: 264 seconds]
leat2 has joined #nixos
<slabity>
Put that right after the of=/dev/sdb arg
<scientist>
slabity: thanks
<scientist>
slabity++
<{^_^}>
slabity's karma got increased to 2
<slabity>
Two karma in one day :)
fendor has quit [Ping timeout: 252 seconds]
domogled has joined #nixos
dhess has joined #nixos
<tilpner>
vikingman_: I don't understand what you mean. It doesn't matter which driver you run while you do a nixos-rebuild
<vikingman_>
if i put services.xserver.videoDrivers = [ "nvidiaLegacy340" ];
<vikingman_>
it fail
<tilpner>
What's the error?
<vikingman_>
it compile for a while then fail
<slabity>
vikingman_: Fails with...?
<vikingman_>
let me run it
fendor has joined #nixos
<noonien>
how can i check if adding something into `services.xserver.monitorSection` ha made its way into my xorg config?
<noonien>
i have to filter the option names somehow
<buffet>
noonien: just the option names, without any description?
<noonien>
yes
stepcut has quit [Ping timeout: 246 seconds]
<buffet>
can i have man output escape sequences even if stdout is not a tty?
boogiewoogie has quit [Remote host closed the connection]
Ariakenom has quit [Quit: Leaving]
<noonien>
hmm, is there a way of running something after X has started? i'm trying to run `xrandr --output DP-0 --mode 3440x1440 --rate 120.00` since xorg doesn't seem to set 120hz on my monitor
<foldingcookie>
noonien: .xinitrc and/or your window manager's configuration are the usual answers
<{^_^}>
[nixpkgs] @worldofpeace pushed 3 commits to master: https://git.io/fjfw4
alex`` has joined #nixos
<buffet>
matthewbauer: yeah theres too many split lines etc
<catern>
anyone know the correct approach to building an application (not a library) for Perl? is there is something comparable to buildPythonApplication?
chreekat has quit [Ping timeout: 255 seconds]
<gchristensen>
"tar (child): xz: Cannot exec: No such file or directory" this means what I think, right? that `xz` isn't installed?
<buffet>
gchristensen: should yeah
knupfer has joined #nixos
knupfer has quit [Client Quit]
<gchristensen>
that is strange, b/c I'm pretty sure xz is installed :P thanks
knupfer has joined #nixos
<tilpner>
gchristensen: strace -vfefile will tell if you're not sure
<pemeunier>
Hi! the following Nature paper, released today, uses Nix to ensure reproducibility of their (somewhat convoluted) system to encode boolean circuits into actual DNA strands: https://www.nature.com/articles/s41586-019-1014-9
<haslersn[m]>
Whats the difference between (a) running `${package-name}/bin/binary-name}` in a build command and (b) adding package-name as `buildInput` and running `binary-name` instead? Does the latter involve a build during evaluation time while the former does not?