<NixOS_GitHub> [nixpkgs] Hodapp87 opened pull request #22219: RStudio: Optionally allow packages from custom R environment (master...master) https://git.io/vDJOx
<hodapp> \o/
<hpc> what's the current state of nfs4 on nixos?
angerman has quit [(Ping timeout: 258 seconds)]
kampfschlaefer has quit [(Ping timeout: 256 seconds)]
<hpc> the output i get from nixos-rebuild test is "mount.nfs4: mount system call failed"
<NixOS_GitHub> [nixpkgs] mbrgm opened pull request #22220: termite: unbreak by packaging vte-ng (master...fix-termite) https://git.io/vDJ38
MP2E has joined #nixos
<hpc> with this config:
<hpc> fileSystems."/nfsmount" = {
<hpc> device = "nfsbox:/";
<hpc> fsType = "nfs4";
<hpc> };
<hpc> and nfs-utils in my systemPackages
<hpc> any ideas?
froglegstew has joined #nixos
<froglegstew> hey, i'd like to install this https://github.com/NixOS/nixpkgs/pull/22098 pr , i'm running debian, and I have nix installed
<froglegstew> I'd like to test it, i'm just not sure how :D
danharaj has quit [(Ping timeout: 245 seconds)]
<froglegstew> i'll just git clone the nixpkgs and use that as a source i guess
AllanEspinosa has joined #nixos
p0le has quit [(Quit: WeeChat 1.6)]
<MP2E> froglegstew: git clone nixpkgs, then create a new branch via 'git checkout -b branchname' then you can apply the PR easily with this 'curl -L https://github.com/NixOS/nixpkgs/pull/22098.patch | git am -'
<froglegstew> thanks!
<MP2E> from there, building/using derivations works by invoking nix-env with -f followed by the nixpkgs git directory
<MP2E> so 'nix-env -f ./nixpkgs -iA swift' if it's in the current directory in nixpkgs
<froglegstew> what does the A do?
<froglegstew> available?
<MP2E> A means use the attribute name, which is generally more convenient as it's not tied to a specific version
<froglegstew> ah, ok, thanks
<froglegstew> ok, it's building. 32 core machine so it shouldn't take too long
<froglegstew> i'm pushing to get it merged so i'm trying to do the testing required to make sure it works/what i can fix to make it work
Tekmo has joined #nixos
ryantrinkle has joined #nixos
<MP2E> awesome :)
<Tekmo> Is there any way to get Nix to ignore a `preferLocalBuild=true` and force Nix to use a cache?
AllanEspinosa has quit [(Ping timeout: 248 seconds)]
bennofs has quit [(Ping timeout: 240 seconds)]
tarinaky has joined #nixos
<tarinaky> Is there a way to force nix-env to rebuild a package? I think I screwed up by installing lynx before openssl, and now I can't use it to go to a HTTPS resource.
Rotaerk has joined #nixos
erasmas has quit [(Quit: leaving)]
<tarinaky> I'm not sure if rebuild is the right nomenclature.
<tarinaky> Refetch?
<tarinaky> Reparse?
thc202 has quit [(Read error: Connection reset by peer)]
markus1189 has joined #nixos
fresheyeball has quit [(Ping timeout: 255 seconds)]
<avn> Tekmo: override packages with preferLocalBuild=false? (but then you need to build them somewhere, because it affect hash)
markus1199 has quit [(Ping timeout: 260 seconds)]
to266 has quit [(Quit: Page closed)]
eacamero_ has joined #nixos
<Tekmo> avn: So I tried that (i.e. use `pkgs.stdenv.lib.overrideDerivation` and then cache the overriden derivation since the hash changed) but it still didn't work. I'm not entirely sure why, though, but my guess is because it still has (many) dependencies that also set `preferLocalBuild=false`
<Tekmo> The derivation in question that I would like this to work on is the root derivation for a NixOS system
<Tekmo> The workaround that I'm contemplating is doing an explicit `nix-store --export` and `nix-store --import` as a brute force approach, but I was hoping that there might be a better way
<avn> Tekmo: what packages you talk about?
eacameron has quit [(Ping timeout: 276 seconds)]
athan has quit [(Ping timeout: 264 seconds)]
<Tekmo> In this case the derivation produced by `(import <nixos/system> { ...}).system`
<Tekmo> Oops, I mean `<nixpkgs/nixos>`
<MP2E> tarinaky: could try removing the package from the current environment with 'nix-env -e lynx' then run the garbage collector 'nix-collect-garbage -d', then try running nix-env to install it back into the user environment
froglegstew has quit [(Quit: Page closed)]
AllanEspinosa has joined #nixos
athan has joined #nixos
<tarinaky> When nixos-install is run, is there a way to set no root password?
shanemikel has joined #nixos
<tarinaky> (i.e. to disable root)
<tilpner> tarinaky - Did you try --no-root-passwd ?
<tarinaky> I did not. It's not in the manual :P
mudri has quit [(Ping timeout: 240 seconds)]
<tilpner> It is in the manpage though
<tarinaky> (I mean this manual: https://nixos.org/nixos/manual)
shanemikel has quit [(Ping timeout: 240 seconds)]
mudri has joined #nixos
shanemikel has joined #nixos
Tekmo has quit [()]
angerman has joined #nixos
acarrico has quit [(Ping timeout: 258 seconds)]
systemfault has joined #nixos
shanemikel has quit [(Quit: My MacBook Air has gone to sleep. ZZZzzz…)]
<eqyiel[m]> is buildMaven deprecated? I can't seem to find any package in nixpkgs that is using it https://github.com/NixOS/nixpkgs/blob/master/pkgs/build-support/build-maven.nix
<eqyiel[m]> NeQuissimus: what's holding this up? https://github.com/NixOS/nixpkgs/pull/20444
<tarinaky> This section of the manual (https://nixos.org/nixos/manual/index.html#sec-building-cd) says NIXOS_CONFIG must be set before the following command can be run.
<tarinaky> But it doesn't say what it should be set /to/ anywhere in the document.
<tarinaky> man nix-build doesn't contain any mentions of NIXOS_CONFIG
shanemikel has joined #nixos
rmeis has joined #nixos
<Rotaerk> hrm what's this site? where's the root of it?
<gchristensen> yep!
<gchristensen> (hi, I'm gsc)
<Rotaerk> ahh :P
<gchristensen> search.nix.gsc.io runs Etsy's Hound code search engine
<Rotaerk> tarinaky, I don't think it's saying NIXOS_CONFIG is a dependency of nix-build, but of the nix expressions that you would be buildingg
<gchristensen> the only other fun thing there is prs.nix.gsc.io, which is a hydra building all PRs
<Rotaerk> well, not dependency, but something that influences them
<Rotaerk> so looking at the nix-build man page would be the wrong place
<tarinaky> That's probably something that could be improved in the documentation then...
<Rotaerk> a lot of nixpkgs stuff isn't documented; you have to dig into the nix expressions to understand what to do
<Rotaerk> dig and search
* tarinaky is planning on trying nixos on their next server... but needs a livecd accessible over ssh since no keyboard/monitor.
rmeis has quit [(Ping timeout: 258 seconds)]
acarrico has joined #nixos
avn has quit [(Ping timeout: 240 seconds)]
stepcut has joined #nixos
avn has joined #nixos
<tarinaky> Chapter 31, Building your own NixOS CD
<tarinaky> The example command, 'nix-build -A config.system.build.isoImage -I nixos-config=modules/installer/cd-dvd/installation-cd-minimal.nix' doesn't work as written.
<tarinaky> error: getting status of '/home/tarinaky/default.nix': No such file or directory
shanemikel has quit [(Quit: My MacBook Air has gone to sleep. ZZZzzz…)]
<gchristensen> hmm
<corngood> tarinaky: I'm guessing it's meant to be run in nixos/ in the nixpkgs repo or by specifying a path to it
<tarinaky> Don't take this the wrong way: but your documentation is only /slightly/ better than at my day job :P
<corngood> nix-build always defaults to current dir I think
stepcut has quit [(Remote host closed the connection)]
<gchristensen> tarinaky: it is true :(
stepcut has joined #nixos
<corngood> I'm a bit confused about the wiki migration
<corngood> Is everything meant to be in the manual via PRs now?
<simpson> Hopefully yes.
Sonarpulse-Work has quit [(Quit: Leaving)]
mbrgm has quit [(Ping timeout: 256 seconds)]
mbrgm has joined #nixos
<NixOS_GitHub> [nixpkgs] aneeshusa opened pull request #22221: NixOS: Respect nixpkgs.overlays (master...respect-nixpkgs.overlays-for-nixos) https://git.io/vDJ8e
<corngood> I have to admit that's stopped me from fixing thing the odd time I've found out of date instructions, when I probably would have done an edit on e.g. the arch wiki
mkoenig has joined #nixos
<simpson> corngood: The wiki isn't coming back, sorry. It's too much of a spam target.
<gchristensen> the wiki was shut down because it was uncontrollably spammed... :/
<gchristensen> if there was something that prevented that nicely, it could be worth talking about
tnks has quit [(Ping timeout: 240 seconds)]
<corngood> I think I just need to find the right workflow for editing it. I imagine it can all be done WYSIWYG on GitHub?
<simpson> I guess? I don't really do stuff in the browser when I can avoid it.
<simpson> Also, since it's DocBook, WYSIWYG sounds like a high bar.
<corngood> Oh, right. I was thinking of markdown
derjohn_mob has quit [(Ping timeout: 245 seconds)]
<corngood> Next time I find something I'll make a point of fixing it. I don't understand the process enough to think about improving it.
<simpson> $(git grep) for the docs that you don't like, and then edit them.
derjohn_mob has joined #nixos
Kingsquee has quit [(Read error: Connection reset by peer)]
Kingsquee has joined #nixos
iyzsong has joined #nixos
alx741_ has joined #nixos
<Benrob0329[m]> So im trying to make a package, and i have default.nix in the proper folder, and i try to run ```nix-env -f ~/nixpkgs/ -i wxSVG``` it returns
<Benrob0329[m]> ```
<Benrob0329[m]> error: selector ‘wxSVG’ matches no derivations
<Benrob0329[m]> ```
<simpson> Benrob0329[m]: $ nix-env -f path/to/nixpkgs/default.nix -iA wxSVG
alx741 has quit [(Ping timeout: 264 seconds)]
<tarinaky> Okay. So. I've checked https://nixos.org/wiki/Error_Messages and I don't see this error message.
<tarinaky> I'm attempting to rebuild release-small.nix from git, to see if I can.
<Benrob0329[m]> simpson: it tells me ```error: getting information about ‘/home/benrob0329/nixpkgs/pkgs/development/libraries/wxSVG/devault.nix’: No such file or directory``` but the file is there...
<tarinaky> My command is: nix-build -A config.system.build.isoImage -I nixos-config=./release-small.nix
<tarinaky> error: anonymous function at /home/tarinaky/nixpkgs/nixos/release-small.nix:5:1 called with unexpected argument config, at /home/tarinaky/nixpkgs/lib/modules.nix:157:8
mkoenig has quit [(Ping timeout: 255 seconds)]
<simpson> Benrob0329[m]: "default.nix" is not "devault.nix"?
<Benrob0329[m]> oops, nvm
mkoenig has joined #nixos
tnks has joined #nixos
<corngood> tarinaky: I think you want just `system.build.isoImage`
<Benrob0329[m]> ```error: cannot auto-call a function that has an argument without a default value (‘stdenv’)```
<tarinaky> corngood: No difference.
johnsona` has joined #nixos
<corngood> Yeah that might have been bad advice. Just taking a look at the source ...
<corngood> Where did you get the idea to set release-small as the config?
<NixOS_GitHub> [nixpkgs] crough opened pull request #22222: Initial commit adding the `polybar' package (master...polybar-3.0.3) https://git.io/vDJ4o
<tarinaky> I may have massively misunderstood something.
<corngood> tarinaky: are you able to build an image using the isoImage example from the manual?
shanemikel has joined #nixos
<tarinaky> corngood: No.
johnsonav has quit [(Ping timeout: 255 seconds)]
<tarinaky> Or... actually... I'm not sure.
<corngood> Not even with nixos/ as your working dir?
<tarinaky> With nixos as my working dir it looks to be working with the snippet from the manual.
jake` has joined #nixos
<corngood> Sorry if I missed it earlier, but are you just trying to make a plain ISO? I assume you're trying to do something custom if you're not just downloading it
<tarinaky> I need SSH access.
<tarinaky> That's, basically, the only change I need over a plain ISO.
<tarinaky> The system I intend to use it with has no keyboard/mouse. So I need SSH access to the installer.
Shou has quit [(Ping timeout: 255 seconds)]
<corngood> You just want sshd running... I guess you'll need a user account or a root password/key for access
<tarinaky> Yup.
shanemikel has quit [(Ping timeout: 276 seconds)]
<jake`> I'm trying to add a package that explicitly uses gcc in the makefile. IT isn't working on my mac because the stdenv only has clang. Should i explicitly include gcc as a dependency, or do i use a patch to override that somehow?
shanemikel has joined #nixos
<NixOS_GitHub> [nixpkgs] shlevy pushed 1 new commit to master: https://git.io/vDJBU
<NixOS_GitHub> nixpkgs/master 9027474 Karn Kallio: racket 6.8...
shanemikel has quit [(Read error: Connection reset by peer)]
shanemikel has joined #nixos
<corngood> tarinaky: seems reasonable. I guess you're comfortable making the actual changes to those iso system configs? Maybe it would actually be useful as an `installation-cd-ssh.nix`? Though you'd really need a user provided secret.
<tarinaky> I'm not really sure what I'm doing :P
<tarinaky> So I set my config .nix's imports array to reference modules/installer/cd-dvd/installation-cd-minimal.nix ?
<tarinaky> Because that's not working >.<
<corngood> I would just `cp installation-cd-minimal.nix installation-cd-ssh.nix`, and add any config you need after the `imports`. if you look at e.g. the gnome one, you can see what I mean
mrkgnao has joined #nixos
<corngood> tarinaky: you'll need at least`services.openssh.enable = true;` and then you can read the nixos manual to see how to add a user with a known password or ssh key (in the wheel', or possibly change the
<corngood> (...)
<corngood> change the root password/key
<tarinaky> Yeah. I've got that down.
Fare has quit [(Ping timeout: 240 seconds)]
athan has quit [(Remote host closed the connection)]
<tarinaky> Weird. I've copied installation-cd-minimal like you said, but when I try to build from it I get: attribute isoImage in selection path config.system.build.isoImage not found
jake` has quit [(Ping timeout: 255 seconds)]
<tarinaky> Oh hang on... it's failing on both of them...
<tarinaky> Ah, derp. Was typing nix-config instead of nixos-config./
alx741 has joined #nixos
<tarinaky> Thanks Rubberduck!
alx741_ has quit [(Ping timeout: 252 seconds)]
gugah has joined #nixos
eacamero_ has quit [(Remote host closed the connection)]
gugah has quit [(Client Quit)]
Elwyn has joined #nixos
<Elwyn> Hello
<Elwyn> I'm trying to setup a rails dev environment but can't get it to work for nothing
<Elwyn> I always fail to build nokogiri
<Elwyn> And I can't find a noob friendly result on Google as well, anyone knows how to simply set it up?
<corngood> tarinaky: that `isoImage` section in the manual is actually a good example of something that could use a little cleanup. The previous chapter mentions `cd /path/to/nixpkgs/nixos`, but that one doesn't. It also says " To build them you have to set NIXOS_CONFIG before running nix-build", but then proceeds to use `-I nixos-config=`. It probably shouldn't mention the env var there at all, since it's redundant, not explained,
<corngood> and not mentioned anywhere else in the manual.
mizu_no_oto has joined #nixos
<tarinaky> Yeah. Sadly I must go to sleep now. Partner wants me to run errands tomorrow. But the iso doesn't quite work :(
<corngood> Did it at least build?
<tarinaky> Yes.
eacameron has joined #nixos
<corngood> Ok, well you're on the right track.
<tarinaky> Just booted it in a VM to test it, but connection closed by remote host.
mkoenig has quit [(Remote host closed the connection)]
<tarinaky> Maybe it's an issue with run levels? IDK
<corngood> yeah, I don't know. the ssh service opens the firewall. as long as it has an address it should be ok
<corngood> try ssh with lots of verbosity
<corngood> Anyway, I'll probably be around over the weekend here and there if you are going to work on it
eacameron has quit [(Ping timeout: 264 seconds)]
k0001 has quit [(Quit: Connection closed for inactivity)]
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
alx741 has quit [(Quit: alx741)]
Elwyn has quit [(Remote host closed the connection)]
alx741 has joined #nixos
Elwyn has joined #nixos
hexagoxel has quit [(Ping timeout: 276 seconds)]
p0le has joined #nixos
lambdael has quit [(Quit: Konversation terminated!)]
mguentner has quit [(Quit: WeeChat 1.6)]
hexagoxel has joined #nixos
mguentner has joined #nixos
eacameron has joined #nixos
Elwyn has quit [(Remote host closed the connection)]
georges-duperon has quit [(Ping timeout: 252 seconds)]
athan has joined #nixos
xwvvvvwx has quit [(Ping timeout: 276 seconds)]
obadz has quit [(Ping timeout: 276 seconds)]
stepcut has quit [(Remote host closed the connection)]
CodingWithClass has quit [(Ping timeout: 252 seconds)]
CodingWithClass has joined #nixos
rmeis has joined #nixos
ertes has joined #nixos
obadz has joined #nixos
rmeis has quit [(Ping timeout: 255 seconds)]
mguentner has quit [(Read error: Connection reset by peer)]
mguentner has joined #nixos
Fare has joined #nixos
eacameron has quit [(Remote host closed the connection)]
<NixOS_GitHub> [nixpkgs] aszlig pushed 2 new commits to nixpart: https://git.io/vDJEz
<NixOS_GitHub> nixpkgs/nixpart 9482c1d aszlig: blivet: Update patch for UUIDs to latest version...
<NixOS_GitHub> nixpkgs/nixpart 0a18f59 aszlig: nixpart: Update to latest master version...
mrkgnao has quit [(Ping timeout: 264 seconds)]
eacameron has joined #nixos
mrkgnao has joined #nixos
Fare has quit [(Quit: Leaving)]
<mrkgnao> how do I react to an inscrutable "File does not exist" message?
<mrkgnao> I built the Write You A Scheme executable from the repo (https://github.com/write-you-a-scheme-v2/scheme), but trying to run the generated executable just fails with that error
<mrkgnao> I'm guessing this is akin to some kind of linker error...
eacameron has quit [(Ping timeout: 255 seconds)]
shanemikel has quit [(Quit: Textual IRC Client: www.textualapp.com)]
<corngood> Could you post the output of `file` on the executable ?
calvinx has joined #nixos
<corngood> And the exact error message?
<mrkgnao> The error message is just that: "File does not exist"
<mrkgnao> I did run file. Wait.
<mrkgnao> .stack-work/dist/x86_64-linux-nix/Cabal-1.24.0.0/build/scheme/scheme: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /nix/store/q3wx1gab2ysnk5nyvyyg56ana2v4r2ar-glibc-2.24/lib/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, not stripped
<mrkgnao> do I have to do something like run patchelf?
<corngood> Looks like a nix elf at least
<corngood> Assuming that interpreter path exists
AllanEspinosa has quit [(Ping timeout: 240 seconds)]
<mrkgnao> it does
<corngood> You said you built it yourself so that makes sense.
<corngood> I guess I'd just run it on strace next
eacameron has joined #nixos
<mrkgnao> I did, it seems it can't find ... glibc?!
<mrkgnao> and it refers to the exact interpreter path in the output of file. weird.
<mrkgnao> I guess I'll try running it in a glibc-aware nix shell
<mrkgnao> someone really needs to fix nix-stack-cabal-* integration. nix is like "omfg my laptop is a spaceship" half the time and "I guess I should copy-paste nix store paths in for now" otherwise.
<mrkgnao> are there guides available? I think some people actually use cabal-install in nix-shell environments, don't they?
t7 has quit [(Remote host closed the connection)]
sellout- has quit [(Quit: Leaving.)]
sellout- has joined #nixos
sdothum has quit [(Quit: ZNC - 1.6.0 - http://znc.in)]
cgdub has joined #nixos
<mrkgnao> corngood: It's actually an error from the program itself, not some NixOS thing. Sorry for the noise.
<mrkgnao> strace led me astray, in my defence
<avn> mrkgnao: cabal2nix works well (usually)
<mrkgnao> avn: what do I do with the generated expression?
<mrkgnao> the problem with having started on Haskell very recently is that I can't use cabal to save my life; I'm used to just doing stack build
<mrkgnao> that's unrelated, actually. but do I use that expression with nix-shell or nix-build?
<avn> for both (is an option there)
<avn> mrkgnao: https://gist.github.com/89ec44f136bde987940db80cbf9b9491 (assuming project.nix -- is generated by cabal2nix from xxx.cabal)
endformationage has quit [(Quit: WeeChat 1.6)]
Ralith_ has quit [(Ping timeout: 256 seconds)]
Ralith_ has joined #nixos
raydeo has quit [(Quit: until next time!)]
cgdub has quit [(Remote host closed the connection)]
mrkgnao has quit [(Ping timeout: 260 seconds)]
panostimos has joined #nixos
<Myrl-saki> Is there a way to limit the download speed?
<Myrl-saki> I'm guessing there's some curl arguments I can use.
xwvvvvwx has joined #nixos
p0le has quit [(Quit: WeeChat 1.6)]
xwvvvvwx has quit [(Ping timeout: 240 seconds)]
nhooyr55 has joined #nixos
raydeo has joined #nixos
raydeo has quit [(Remote host closed the connection)]
systemfault has quit [(Quit: Bye!)]
pikajude has quit [(Quit: ZNC 1.6.3 - http://znc.in)]
pikajude has joined #nixos
raydeo has joined #nixos
raydeo has quit [(Remote host closed the connection)]
raydeo has joined #nixos
nhooyr55 has quit [(Quit: Page closed)]
raydeo has quit [(Remote host closed the connection)]
raydeo has joined #nixos
<NixOS_GitHub> [nixpkgs] vbgl opened pull request #22223: mathcomp 1.6.1 and related fixes (master...mathcomp-1.6.1) https://git.io/vDJ28
raydeo has quit [(Remote host closed the connection)]
raydeo has joined #nixos
danielrf has quit [(Ping timeout: 248 seconds)]
rmeis has joined #nixos
<nhooyr> trying to create a nix package from a node project (thelounge). I installed node2nix and downloaded thelounge. then I ran node2nix inside thelounge and it generated the nix expressions. I placed these into `/etc/nixos/nixpkgs/pkgs/servers/irc/thelounge`. Now I modified `/etc/nixos/nixpkgs/top-level/all-packages.nix` and added `thelounge = callPackage ../servers/irc/thelounge { };`. Unfortunately, this is giving me
<nhooyr> https://gist.github.com/nhooyr/f570c1d205ab001e9f5de6f82ca671ec but it's not clear what went wrong.
eacameron has quit [(Remote host closed the connection)]
rmeis has quit [(Ping timeout: 240 seconds)]
iyzsong has quit [(Quit: bye.)]
<NixOS_GitHub> [nixpkgs] vcunat pushed 3 new commits to master: https://git.io/vDJaO
<NixOS_GitHub> nixpkgs/master 1d6c853 Marius Bergmann: termite: unbreak by packaging vte-ng...
<NixOS_GitHub> nixpkgs/master 2cbf509 Vladimír Čunát: gnome3.vte-ng: nitpick: gperf is a native input
<NixOS_GitHub> nixpkgs/master 781c791 Vladimír Čunát: Merge #22220: termite: unbreak by packaging vte-ng
<NixOS_GitHub> [nixpkgs] vcunat closed pull request #22220: termite: unbreak by packaging vte-ng (master...fix-termite) https://git.io/vDJ38
calvinx has quit [(Ping timeout: 240 seconds)]
calvinx has joined #nixos
yorick has quit [(Read error: Connection reset by peer)]
xadi has joined #nixos
yorick has joined #nixos
<nhooyr> ok, I got it to install by using generate.sh in `/etc/nixos/nixpkgs/pkgs/development/node-packages`.
<nhooyr> how can I print the directory of some package, say `pkgs.nodePackages.thelounge`? like, where it is right now on the disk.
xadi has quit [(Ping timeout: 255 seconds)]
phant0mas has joined #nixos
<srhb> nhooyr: find /nix/store -name "*thelounge*" ?
xadi has joined #nixos
<nhooyr> srhb: i was wondering if there was a way with the nix tools, it just seems like there would be a way. to get the latest version.
<NixOS_GitHub> [nixpkgs] nhooyr opened pull request #22224: fzf: from 0.15.1 to 0.16.2 (master...fzf) https://git.io/vDJab
xadi1 has joined #nixos
<srhb> nhooyr: nix-env --query --out-path --attr pkgs.nodePackages.thelounge
<srhb> Hmm, not quite.
xadi has quit [(Ping timeout: 276 seconds)]
<srhb> I can't figure out how to query by attribute path, but it works by name..
<nhooyr> weird
jmeredith has quit [(Quit: Connection closed for inactivity)]
<srhb> But yeah, nix-env -q --out-path packagename works
eacameron has joined #nixos
<nhooyr> srhb: I get nothing for that, but I think it's because my package is declaratively installed.
eacameron has quit [(Ping timeout: 252 seconds)]
rly has joined #nixos
calvinx has quit [(Ping timeout: 276 seconds)]
calvinx has joined #nixos
<srhb> nhooyr: Ah, right.
Mercuria1Alchemi has quit [(Ping timeout: 255 seconds)]
MP2E has quit [(Quit: sleep)]
<srhb> nhooyr: Well, I suppose the answer is to follow the symlink then.
<srhb> In the system profile.
<srhb> If there is one. :)
<nhooyr> ok
jgertm has joined #nixos
jgertm is now known as Guest79363
hyper_ch has left #nixos ["Konversation terminated!"]
hyper_ch has joined #nixos
<NixOS_GitHub> [nixpkgs] bachp opened pull request #22225: Glusterfs service (master...glusterfs-service) https://git.io/vDJwc
Ivanych has quit [(Quit: Leaving.)]
taarnet has joined #nixos
<srhb> Can I set a default path for nix-env to look for nixpkgs? Like nix.nixPath = [ "nixpkgs=/foo" ]; for the global config
taarnet has quit [(Client Quit)]
afics has quit [(Ping timeout: 240 seconds)]
<srhb> If I set the same option in ~/.nixpkgs/config.nix, it seems to be ignored.
<srhb> And nix-env doesn't seem to inherit the global setting.
eacameron has joined #nixos
afics has joined #nixos
<srhb> I guess I can symlink <nixpkgs> into ~/.nix-defexpr, but that sounds a bit hacky
eacameron has quit [(Ping timeout: 245 seconds)]
ixxie has joined #nixos
kampfschlaefer has joined #nixos
dbmikus has quit [(Ping timeout: 255 seconds)]
Heffalump has joined #nixos
<Heffalump> is it normal that the hash returned by nix-prefetch-git doesn't work as the sha256 in pkgs.fetchFromGitHub ?
Ivanych has joined #nixos
<ixxie> Heffalump: not if you correctly also entered the corresponding revision
<Heffalump> I did (I've had this problem twice now). A bit more careful googling found a mailing list post - turns out you need to use -prefetch-zip to get the right URL
<pikajude> can I make a fileSystem optional?
<pikajude> or even better, mounted on demand
<pikajude> I want to mount my external HD at /var/lib/mpd when it's connected
<pikajude> because mpd cannot cooperate with udiskie
<pikajude> and more importantly because even if I do chgrp to mpd for the entire tree of /run/media/pikajude/etc etc etc, i don't know if udiskie will preserve it next time it's plugged in
<Yaniel> using $out in cmakeFlags just adds it verbatim, is there any way I can have it expand it instead?
ixxie has quit [(Ping timeout: 255 seconds)]
<NixOS_GitHub> [nixpkgs] domenkozar pushed 1 new commit to master: https://git.io/vDJoz
<NixOS_GitHub> nixpkgs/master 0056a3a Domen Kožar: haskellPackages.hspec-expectations-pretty-diff: fix build
dbmikus has joined #nixos
nhooyr has left #nixos ["Kicked by @appservice-irc:matrix.org"]
nhooyr has joined #nixos
<NixOS_GitHub> [nixpkgs] domenkozar pushed 1 new commit to master: https://git.io/vDJK4
<NixOS_GitHub> nixpkgs/master fa18c37 Domen Kožar: haskellPackages.cryptonite-openssl: redistribute
<NixOS_GitHub> [nixpkgs] domenkozar pushed 2 new commits to release-16.09: https://git.io/vDJKE
<NixOS_GitHub> nixpkgs/release-16.09 abb7d15 Domen Kožar: haskellPackages.hspec-expectations-pretty-diff: fix build...
<NixOS_GitHub> nixpkgs/release-16.09 4281374 Domen Kožar: haskellPackages.cryptonite-openssl: redistribute...
nhooyr has quit [(Remote host closed the connection)]
nhooyr has joined #nixos
eacameron has joined #nixos
rmeis has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
rmeis has quit [(Ping timeout: 245 seconds)]
thc202 has joined #nixos
<ben> how do i get into a shell where echo '#include "zlib.h"' | gcc -E - succeeds/finds zlib.h? :/
t7 has joined #nixos
<Yaniel> by putting zlib.h in the working directory
<srhb> nix-shell -p zlib -p gcc --pure ?
<Yaniel> unless you mean '#include <zlib.h>' in which case add zlib to the buildInputs of your shell.nix
<nhooyr> does anyone develop in Go using nix? What sort of workflow do you have?
<srhb> Yaniel: What difference do the angle brackets make? both seem to work fine here.
<Yaniel> angle brackets: search system include dirs, quotation marks: search project dir
<Yaniel> compilers are usually forgiving, except when they are not
<srhb> Yaniel: Why does it work with quotation marks even when its not in the current directory?
<srhb> Ah, okay.
<srhb> I hate it when they do that. :)
<Yaniel> and that part is what makes development with c++ so awful
<Yaniel> *is part of
<srhb> Fail hard, fast and up front.
<Yaniel> also when they always include some headers by default
<srhb> :|
<Yaniel> "joe, your code doesn't build" "Works On My Machine™" "you didn't include stdio again, did you"
<srhb> Oodles of fun, I imagine... :-)
<Yaniel> "invalid declaration of function 'malloc'" "joe go fuck yourself, and while you are at it, learn to use includes"
<Yaniel> I have one friend whose code I end up having to fix every 5 commits or so
<Yaniel> and every.single.time it's the includes
<ben> apparently my mistake was having gcc-wrapper in my user profile and just saying nix-shell -p zlib
eacameron has joined #nixos
<tarinaky> So. I'm trying to generate a custom iso that'll allow me to log in over SSH as root to perform the installation on a headless server.
<tarinaky> I have a modified nix I'm trying to nix-build, but it doesn't appear to actually be modifying the generated system.
<tarinaky> I've got permitRootLogin="yes"; in my nix file.
<NixOS_GitHub> [nixpkgs] kevincox opened pull request #22226: Revert "libmsgpack_1_4: remove" (master...revert-msgpack) https://git.io/vDJiG
<tarinaky> But when I run the iso, and inspect /etc/ssh/ssh_config there's no entry for permitRootLogin.
eacameron has quit [(Ping timeout: 240 seconds)]
<Yaniel> IIRC that goes in sshd_config
<Yaniel> not ssh_config
<tarinaky> There is no sshd_config.
<tarinaky> Unless I'm looking in the wrong place.
<tarinaky> brb, need to run an errand.
<NixOS_GitHub> [nixpkgs] joachifm pushed 2 new commits to master: https://git.io/vDJii
<NixOS_GitHub> nixpkgs/master b18ff7a Joachim Fasting: torbrowser: make additional media playback support optional...
<NixOS_GitHub> nixpkgs/master 6303d2b Joachim Fasting: nixos: add sysstat to module list...
morsicus1 has joined #nixos
<regnat> tarinaky: I think sshd_config isn't copied to /etc, it's only in the nix store
<Yaniel> can I somehow make nix evaluate $out in cmakeFlags?
morsicus1 has quit [(Read error: Connection reset by peer)]
morsicus1 has joined #nixos
athan has quit [(Ping timeout: 256 seconds)]
morsicus1 has quit [(Ping timeout: 240 seconds)]
morsicus1 has joined #nixos
morsicus1 has quit [(Read error: Connection reset by peer)]
morsicus1 has joined #nixos
MichaelRaskin has joined #nixos
i_rony has quit [(Read error: Connection reset by peer)]
ertesx has joined #nixos
ertes has quit [(Ping timeout: 252 seconds)]
ertesx is now known as ertes
chrishill has joined #nixos
<chrishill> Hey guys.
<NixOS_GitHub> [nixpkgs] vcunat pushed 2 new commits to master: https://git.io/vDJPN
<NixOS_GitHub> nixpkgs/master c09690b Kevin Cox: Revert "libmsgpack_1_4: remove"...
<NixOS_GitHub> nixpkgs/master 2200532 Vladimír Čunát: Merge #22226: Revert "libmsgpack_1_4: remove"...
<chrishill> I want to install NixOS 16.09, 32 bits. But the download page only seems to link to the 64 bit version:
<NixOS_GitHub> [nixpkgs] vcunat closed pull request #22226: Revert "libmsgpack_1_4: remove" (master...revert-msgpack) https://git.io/vDJiG
<chrishill> In particular, the “full” installation (graphical), not the minimal one.
<chrishill> Where can I find the full installation for 32 bits?
<MichaelRaskin> I don't think it is built
<clever> yeah, graphical is only done for 64bit
<clever> but if you have nix on a machine, you can just checkout a copy of 16.09 nixpkgs, edit that line, and have nix-build generate the ISO
<chrishill> clever: So if I have Nix installed, I can build the ISO on that machine
<MichaelRaskin> Yes
<MichaelRaskin> And the graphical iso doesn't provide a graphical installer, by the way
<chrishill> So it’s just a matter of checking it out in Git, editing the line, and running that command? Cool. Do you have a link to some reference I could peruse while I try that?
panostimos has quit [(Ping timeout: 260 seconds)]
<clever> chrishill: nix-build ~/apps/nixpkgs/nixos/release.nix -A iso_graphical.i686-linux should build the 32bit graphical ISO
<clever> once that line has been edited
<clever> if you make 135 match 129, then it will have support to do either 32 or 64bit
<chrishill> Nice, thank you! I’ll try that
eacameron has joined #nixos
<clever> chrishill: and you will want to use the nixos-16.09 branch of this repo: https://github.com/NixOS/nixpkgs-channels/tree/nixos-16.09
iyzsong has joined #nixos
eacameron has quit [(Ping timeout: 240 seconds)]
<gchristensen> clever: what version of netcat do you have?
<clever> gchristensen: my desktop has version 1.105 in nixpkgs, not currently in $PATH
<globin> gchristensen: mine is openbsd
<clever> «derivation /nix/store/1q85vpfxsgla9wwg0h4kn8hkn2fk37il-netcat-openbsd-1.105.drv»
<gchristensen> ok so why do I have gnu?
<clever> nix-repl> netcat-gnu
<clever> «derivation /nix/store/y69vxqggl2yc2bzlp5jzllbh4rf4gyq0-netcat-gnu-0.7.1.drv»
<clever> how did you install it?
<gchristensen> I didn't
<clever> where does "which netcat" say it is?
<gchristensen> /run/current-system/sw/bin/netcat -> /nix/store/xq2kr63qr69hd00zs5m42wmrziz8m01n-netcat-gnu-0.7.1/bin/netcat
<clever> nixpkgs/nixos/modules/virtualisation/libvirtd.nix: [ pkgs.libvirt pkgs.netcat-openbsd ]
<clever> nixpkgs/nixos/modules/config/system-path.nix: pkgs.netcat
<clever> if you enable libvirtd, then the openbsd version is added to systemPackages
<gchristensen> the default was changed between 16.09 and 17.03
<clever> ah
<clever> ive switched over to socat, it has a lot more options
<gchristensen> I can't
<clever> socat can handle unix sockets, and inter-mix things
<clever> ive used socat before to convert /tmp/.X11-unix/X0 into a tcp socket, allowing remote X11 clients
<clever> for some reason, xorg refuses to listen on tcp now, even if you remote -nolisten tcp
<gchristensen> ok, so ... with netcat(-gnu) in nixos/modules/config/system-path.nix's systemPackages, is there a way to override it and force netcat-openbsd?
<clever> system-path just uses netcat
<gchristensen> yeah but for me that is the gnu package and I need it to be the openbsd package
<clever> so you can just make an override that sets netcat = pkgs.netcat-openbsd
<gchristensen> that breaks other software :(
simendsjo has joined #nixos
<clever> then prepend ${netcat-openbsd}/bin/ to the PATH of whatever cares
<clever> this is why i avoid installing such things globaly, always insert them into PATH and you get exactly the right version, without causing others trouble
<gchristensen> that will be very difficult
<gchristensen> there is no way to override netcat in systemPackages?
<clever> what is it breaking?
<clever> yeah, no way to change that, i had to use a copy when building not-os
<clever> you would have to clone nixpkgs and edit it locally
<gchristensen> yikes
<MichaelRaskin> I just have a script wrapper that prepends a package to PATH before running the command given in the arguments
<gchristensen> this would have to happen when a user SSH'd in, it'd have to be prepended automatically
chrishill has quit [(Quit: Page closed)]
<clever> gchristensen: nix-env stuff does take priority over systemPackages
<clever> gchristensen: so you could just nix-env -iA nixos.netcat-openbsd and your done
chrishill has joined #nixos
<gchristensen> the systems being ssh'd in to are nixos-containers and are created and deleted rapidly with automatino
<clever> ah
<clever> just override netcat = pkgs.netcat-openbsd; and see what breaks?
<clever> maybe nothing will?
<gchristensen> a few things do but I can fix that with more overrides ;)
chrishill has quit [(Client Quit)]
<clever> yeah, you can also foo = foo.override { netcat = pkgs.netcat-gnu; }; to undo it to some things
chrishill has joined #nixos
<gchristensen> yeah
<gchristensen> alright. thank you for brainstorming :)
panostimos has joined #nixos
<gchristensen> slightly annoying that it is in systemPackages though. I wonder why
Guest79363 has quit [(Ping timeout: 258 seconds)]
<clever> i do feel that the default systemPackages is a bit fat
<clever> why do i need nano installed?
simendsjo has quit [(Read error: Connection reset by peer)]
<gchristensen> I think I found a weird bug in gnu netcat though where it closes early and fails to send some data in a pipe
<gchristensen> where it was `datagenerator | nc` which was somehow solved via `datagenerator | { cat; sleep 5 } | nc`
simendsjo has joined #nixos
<NixOS_GitHub> [nixpkgs] lsix opened pull request #22227: glpk: 4.56 -> 4.61 (master...update_glpk) https://git.io/vDJXN
<clever> gchristensen: ah, ive run into similar bugs in irssi with /exec -o
<gchristensen> oh?
<clever> people assume that waitpid/SIGCHILD means there is no more data
<clever> and dont attempt one last read of the stdout pipe
JagaJaga has joined #nixos
<gchristensen> yes!
<clever> so commands like "/exec -o df -h" that run very fast and output a lot, just get silently truncated
<gchristensen> that is exactly what happened!
<MichaelRaskin> And even if they do one last input, it is not guaranteed that their buffer will get all the data in one read…
<clever> yeah
<clever> keep reading until EOF
<tarinaky> Okay. So I'm trying to find the generated sshd_config to debug whether or not it's correct.
<clever> tarinaky: read /etc/systemd/system/sshd.service to find its path
<tarinaky> I've found sf3q*-openssh-7.4p1/etc/ssh/sshd_config
<clever> [clever@amd-nixos:~]$ grep sshd_config /etc/systemd/system/sshd.service
<clever> ExecStart=/nix/store/ii0q2jbxzfgp3sw9mmk09cs1zykcnkma-openssh-7.3p1/bin/sshd -f /nix/store/z75qn23w186ihdy0w49w0yibc23gls5j-sshd_config
<tarinaky> But it also doesn't contain the PermitRootLogin line I'm looking for.
<clever> because the /nix/store will contain dozens or more copies of sshd_config
<clever> you want to look at the current one, via the systemd unit file
<tarinaky> Right. So. PermitRootLogin yes
<tarinaky> Looking in /etc/shadow, my root user has a password set.
<tarinaky> So, when I try to ssh -vvv root@127.0.0.1 -p2222 (redirected to the VM) ... Why doesn't it try keyboard-interactive
<clever> ps aux | grep sshd
<clever> is the currently running sshd using that config file?
<clever> "journalctl -f -u sshd", what do the logs say?
<tarinaky> sshd isn't running. That'll do it. journalctl -f -u sshd just says "Logs begin at 2017-01-28 15:38:11 UTC"
teh__ has joined #nixos
<clever> tarinaky: is this a normal nixos install or the ISO image?
<tarinaky> The ISO image.
<clever> sshd is configured to not run automaticaly
<clever> you need to manualy start it with systemctl
sdothum has joined #nixos
<tarinaky> Is there any way to make it start automatically when the ISO boots?
<avn> tarinaky: rebuild iso as I know
<tarinaky> I did.
<NixOS_GitHub> [nixpkgs] gpyh opened pull request #22228: rustNightlyBin: 2016-12-29 -> 2017-01-26 (master...rustNightlyBin) https://git.io/vDJ1i
<tarinaky> I set openssh.enable = true and rebuild the image.
<tarinaky> Then tried to boot the image in a VM to see if it worked.
<clever> systemd.services.sshd.wantedBy = mkForce [ "multi-user.target" ];
<clever> i have used this in my configuration to re-enable running on bootup
tarinaky has quit [(Ping timeout: 240 seconds)]
astsmtl has quit [(Ping timeout: 258 seconds)]
<avn> clever: ty for hit, would be useful, when I take on install headless machine
<clever> avn: i have used this recently on a headless install: https://github.com/cleverca22/nix-tests/blob/master/kexec/configuration.nix
<clever> avn: the biggest problem i ran into, was keeping the hardware raid controller happy, and configuring the static ip right (the datacenter doesnt use dhcp right)
astsmtl has joined #nixos
astsmtl has quit [(Changing host)]
astsmtl has joined #nixos
tarinaky has joined #nixos
<tarinaky> Still doesn't work :(
<tarinaky> Still doesn't work :(
<tarinaky> And I have to go out again -.-
<tarinaky> bbiab
<tarinaky> ssh-agent appears in my processes though.
<tarinaky> But no SSHD
<corngood> Did you try `ssh localhost` from in the VM?
<globin> niksnut, ikwildrpepper: staging builds are aborting "Hydra failure" https://hydra.nixos.org/eval/1326825
Kingsquee has quit [(Quit: https://i.imgur.com/qicT3GK.gif)]
<avn> clever: is a simple case, little barebone machine with hdmi output (which should be new router), and I have only one hdmi/dp capable display ;)
<clever> avn: ah, if you can get display during install, its trivial
<avn> clever: I prefer to install headless if possible (won't take display from my own work machine)
<clever> avn: yeah, you can always try that first, thats what i was doing with the racklodge machine
<globin> gchristensen: I'd recommend running the prs after commenting on the PR as member to trigger the builds, there are so many nitpicks that should be reviewed first before running the build as many of them will trigger fixing after rebuilds
<globin> LnL, ^
<gchristensen> could be, I wonder how long minor fixes take to evaluate / build
<LnL> we could create a disabled jobset first, and perhaps enable it when it's approved or something
<LnL> indeed, I don't think there's a load issue with the current build machines
<globin> gchristensen: there will be stuff triggering mass-rebuilds that needs fixing
<gchristensen> true
<globin> and multiple mass-rebuilds for one pr is a lot of load
<globin> (rust does the same btw)
<unlmtd[m]> I want to install on zfs without legacy mount, but nixos-install complains I havent set the root filesystem
<unlmtd[m]> so .. I set it but then all my datasets didnt mount
<unlmtd[m]> except root
<unlmtd[m]> since nix/store uses its own dataset the boot didnt get far
<gchristensen> globin: that makes sense. what we have now is basically the simplest thing that could possibly work... I would like someone to contribute such a feature
<gchristensen> right now for example it even triggers if you edit the title of the PR :|
pi3r has joined #nixos
<globin> gchristensen: I'll have a look if I can merge domens/fpletz/my bot with your stuff
<clever> unlmtd[m]: it needs something like this to tell it what zfs to mount where, and that it needs zfs tools in the initrd: https://gist.github.com/cleverca22/70a08dfbaf95bf34ffd515779be638eb
<clever> unlmtd[m]: both datasets have mountpoint=legacy
<gchristensen> globin: I specifically avoided fanciness (like detecting if the change was a commit change, or a text change) in the interest of getting it up and running
<unlmtd[m]> clever: Iv done this before, but now I want to do it without legacy mounts
<unlmtd[m]> so I set zfs=pool in kernelParams, and it should just work
<unlmtd[m]> but nixos-install refuses to install unless I set "/"
rmeis has joined #nixos
<fpletz> has anybody gotten nixos tests working on a self-hosted hydra? our build machines do have the kvm feature but no builds seem to be scheduled :/
<globin> gchristensen: I'll see what I can do
<gchristensen> fpletz: do you have big-parallel?
<fpletz> gchristensen: yup
<gchristensen> hmm
<unlmtd[m]> I guess I shoulda asked if it could work before I spent the whole day trying
<fpletz> the behaviour is exactly the same as if a feature was missing, but I double checked our configuration and the code
rly has quit [(Ping timeout: 258 seconds)]
rly has joined #nixos
<fpletz> noooo
<unlmtd[m]> Id really prefer to let zfs take care of the mounts
<fpletz> there is a nixos-test feature? \o/
<gchristensen> haha
<gchristensen> aww
<fpletz> ok, let me check
<fpletz> thanks a lot!
Jackneill_ has joined #nixos
<unlmtd[m]> Im leaning towards using nix on another system if I have to mount zfs legacy style
<gchristensen> fpletz: I thought I "borrowed" this list from nixos-org-configurations or something but now I can't find it there
<fpletz> gchristensen: yeah, just saw that, I only looked at makeTest in that file which only requires the kvm feature
<unlmtd[m]> very little google can come up with on that
<gchristensen> "Amazing work :) Can't wait to see Github PRs getting a binary cache." nooo, domenkozar, I can't serve it!
Guest79363 has joined #nixos
Guest79363 is now known as jgertm
rmeis has quit [(Ping timeout: 240 seconds)]
zarel has joined #nixos
<unlmtd[m]> can I force nixos-install ?
<unlmtd[m]> specifically without "/" set
taeradan has quit [(Remote host closed the connection)]
<clever> unlmtd[m]: one min
<clever> unlmtd[m]: what is the exact error its giving?
<unlmtd[m]> I dont have it handy sorry, but I would try again later
<unlmtd[m]> like . fileSystem for root must be set
<teh__> gchristensen: the packet.net announcement is just awesome! Do you have a specific list of contributions / improvements you are looking for? Maybe point people at some meta-bug?
<clever> modules/system/boot/stage-1.nix: message = "The ‘fileSystems’ option does not specify your root file system.";
<unlmtd[m]> yes
<clever> unlmtd[m]: it is checking that at least one of the entries in the filesystems array has a mountpoint of /
<unlmtd[m]> can I bypass ip
<unlmtd[m]> it
<clever> unlmtd[m]: simplest thing i can think of is to edit a copy of nixpkgs to just delete that check
<unlmtd[m]> ok cool
<unlmtd[m]> at least I know its doable, I need to go for a walk now, fuming a bit. thanks
<clever> but it sounds like you want the kernel to mount the rootfs for you
<clever> but that wont happen when using an initrd
<unlmtd[m]> hmm
<unlmtd[m]> so nixos' limits are in sight
<clever> you would need to heavily modify stage-1 to bypass that
<unlmtd[m]> current* limits
<clever> oh, and also
<clever> config = mkIf (!config.boot.isContainer) {
iyzsong has quit [(Quit: bye.)]
<clever> if you set this to true, it will bypass the rootfs check for you
<clever> but it also wont include a kernel/initrd
<clever> so its up to you to make it boot, somehow
<unlmtd[m]> what was nix originally developped on, anyone knows?
<unlmtd[m]> clever: thanks a lot, really helpful
<unlmtd[m]> clever: I hope we can work on fixing that one day
<unlmtd[m]> I reckon the redoxos rewrite should come someday in the near future, so Ill look at porting nix, then doing a nixos-style redoxos
<clever> unlmtd[m]: i have done something similar with not-os
<unlmtd[m]> ya Id love for that project to be favored
<clever> unlmtd[m]: https://github.com/cleverca22/not-os/blob/master/default.nix#L12-L28 i used a small sub-set of nixos modules, and a few custom ones
<clever> this creates a distro that is massively stripped down, and in the current config, it compiles down to a ~47mb squashfs
<unlmtd[m]> bah, I dont think I have time to use an imperative OS right now. Guess ill legacy mount zfs for now...
<clever> it sucessfully boots under x86-qemu with testcases, and i have it acting as a nix build slave on a raspberry pi 3
<unlmtd[m]> heres my 2 cents: I want to put as little effort in linux as possible
<unlmtd[m]> id be all for porting nix to bsd thol
<unlmtd[m]> thou
<unlmtd[m]> or minix
<viric> nix should work in bsd, cygwin, darwin, ...
<unlmtd[m]> I meant nixos
<unlmtd[m]> to bsd
<unlmtd[m]> like clever's not-os
<viric> do it
<MichaelRaskin> Minix should be hard, I am not sure it has enough of POSIX compatibility implemented
civodul has joined #nixos
<unlmtd[m]> ya minix is dark art
<viric> I tried to build nix in minix3 years ago
<viric> I forgot the troubles I had
<unlmtd[m]> ahh!
<viric> there may be even a github issue about that
<unlmtd[m]> has anyone forayed into a C rewrite of nix?
<unlmtd[m]> anyhow, off for a walk
ixxie has joined #nixos
<viric> unlmtd[m]: you sound like those who say a lot of "I will do I will do" and will not do
AveryC has joined #nixos
eacameron has joined #nixos
<avn> unlmtd[m]: write (port) some code, and make PR. It be better illustration of your intentions
<gchristensen> hrm. my hydra server says it is out of space. it is not out of space. "Unable to create temporary file '/tmp/git-checkout-tmp-KrVN1jxY/git-export/.git/objects/pack/tmp_idx_XXXXXX': No space left on device"
<gchristensen> any ideas?
<clever> gchristensen: is /tmp on a tmpfs?
<avn> gchristensen: environment.TMPDIR = "/var/buildroot"; (and put buildroot to ZFS with good ARC/L2ARC)
<gchristensen> I can't use zfs
<avn> sad ;( I found it perfectly fits builder
<gchristensen> probably so clever
<gchristensen> good idea
<avn> chromium build twice faster
<avn> gchristensen: anyway, example above will work anywhere, if you add preStart = ''mkdir -p /var/buildroot'' ;)
nckx has quit [(Read error: Connection reset by peer)]
nckx has joined #nixos
bennofs has joined #nixos
eacameron has quit [(Ping timeout: 276 seconds)]
teh__ has quit [(Ping timeout: 240 seconds)]
<srhb> Is there a nice way to patch fonts with custom glyphs?
sellout- has quit [(Ping timeout: 248 seconds)]
<gchristensen> avn: will that change where hydra checks out these exports?
<gchristensen> /tmp/nix-9898-767 etc.
<avn> gchristensen: if hydra haven't them hardcoded ;)
<avn> gchristensen: but it work for nix-daemon (from where I take example)
<gchristensen> hmm
<gchristensen> is there, instead, a way to make the tmpdir biger, clever? I have tons of ram.
taeradan has joined #nixos
<avn> clever: LnL: btw I finish my ZFS migration out of LVM/MD ;) All except soft-lookups is great -- http://list.zfsonlinux.org/pipermail/zfs-discuss/2017-January/027254.html (details, if anyone interesting in them)
mkoenig has joined #nixos
<gchristensen> avn: also that'll move the tmpdir for everything, including posgresql ... hmm...
<gchristensen> I think I'd rather fix /tmp
stepcut has joined #nixos
<NixOS_GitHub> [nixpkgs] michaelpj opened pull request #22230: arbtt: multi-user.target does not exist in user systemd (master...services/arbtt-fix-wanted-by) https://git.io/vDJ9f
<avn> gchristensen: if you not interesting in persisting postgresql... ;)
<avn> gchristensen: anyway, I think better to have dedicated tmpfses here with different limits
<gchristensen> eh? pgsql persists elsewhere
<LnL> gchristensen: change the TMPDIR for the nix-daemon
<gchristensen> hmm yeah
<avn> gchristensen: I am kidding a bit, imagining that you have whole pgsql on /tmp ;)
<gchristensen> the other thing though is will hydra prune its own tmp? I'm not sure
itsN1X has joined #nixos
itsN1X has quit [(Client Quit)]
itsN1X has joined #nixos
civodul has quit [(Read error: Connection reset by peer)]
civodul has joined #nixos
itsN1X has quit [(Read error: Connection refused)]
itsN1X has joined #nixos
<clever> avn: i'm also being plauged by frequent soft-lockups
<clever> avn: they typicaly recover if i replug a usb keyboard a couple times
sellout- has joined #nixos
itsN1X has quit [(Remote host closed the connection)]
itsN1X has joined #nixos
itsN1X has quit [(Client Quit)]
<avn> clever: also on io?
<clever> avn: the weird thing, is that the system just gives up and doesnt do any IO
<clever> avn: something like updatedb takes under a minute if i have tons of ram free, but can take 10-15minutes if i have 6gig of ram actualy in-use
<clever> avn: even if the arc size is the same, and there is another gig free
<avn> clever: I also got some complains in dmesg (look my mailing list post)
<clever> i havent had anything like that in dmesg
<clever> [159790.096294] sd 4:0:0:0: [sdd] tag#1 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
<clever> but now that i check dmesg, i do see some fishy things
<clever> that is half of my zfs mirror
civodul has quit [(Ping timeout: 248 seconds)]
teh__ has joined #nixos
<clever> currently, the arc is at 2.6gig, no heavy memory usage, running a GC
georges-duperon has joined #nixos
civodul has joined #nixos
<clever> running nix-collect-garbage is causing upwards of 30k reads/sec in arcstat.py
<gchristensen> clever: so I haven't enabled tmp on tmpfs... my full config is here: https://github.com/grahamc/hydra-prs/ and my `df -ha` as root didn't report anything getting even close to over 50%
<gchristensen> are there per-user limits setup by systemd I'm not aware of?
<clever> none that i'm aware of
<avn> clever: looks like your disk closer to death than my
<clever> and default.nix refers to a directory that doesnt exist
<gchristensen> eh?
<clever> avn: my impression is that ZFS is spinlocking the cpu, causing it to not be capable of receiving a reply from the hdd
<clever> avn: leading to timeouts
<gchristensen> clever: fixed
<avn> so it freeze all processes scheduled to these cpu? I have six cores, so other 5 cores should go I think
<viric> what's the size of your nixpkgs ".git" subdir?
<clever> avn: i have had it freeze all 8 cores at once before
<gchristensen> viric: 1.6M
<viric> sorry?
<gchristensen> is the size of my nixpkgs .git subdir
<clever> avn: under normal conditions, a full nix GC took 4 minutes 53 seconds, and the arc rose to a peak of 3.6gig, but has since dropped to 2.8gig
<viric> I had 1GB. I did "repack -a" and now it is 1.4GB
<viric> damn it.
<gchristensen> :o
<viric> ah my fault
<viric> -a takes more, sure
<clever> avn: a second run of GC with no garbage took 32 seconds, and the arc rose to 3.1gig
<avn> clever: Surprising, that no one from zfs-devs not catch this case
<clever> avn: putting some strain on the system now...
myheadhurts_ has quit [(Quit: Connection closed for inactivity)]
<gchristensen> oh wtf the problem isn't disk space, clever
<clever> gchristensen: inodes?
<gchristensen> yep.
<clever> ive had that once or twice
<clever> df -i
<gchristensen> I thought it was suspicious /tmp was only 10mb
<clever> forgot about it this time
<gchristensen> yeah 100%
xadi1 has quit [(Quit: Leaving.)]
<viric> gchristensen: du --inodes :)
<clever> viric: oh nice
madonius has joined #nixos
<madonius> greetings
<clever> avn: system strain increased, 6gig in use by a single app, arc has gone down to 1.9gig
<madonius> i am trying to install borg backup on my nixos and the os claims that it does not know borg
<viric> I have 3% inodes usage... but I always format ext4 with 10x more inodes than the default
<clever> hydra must have pushed more things over, GC found something to eat
<clever> avn: gc took 41 seconds, cant reproduce the issue at this instant
jmiven has quit [(Quit: co'o)]
<gchristensen> viric: "du --inodes -s /nix" is taking a "long time", so
<gchristensen> I suspect the auto-gc doesn't account for inodes
<viric> no wonder
<gchristensen> well crap, I wish I had remembered to set 10x inodes at format time :|
<viric> gchristensen: everyone learns ext4 the hard way
<viric> ext2 and ext3 were the same
<clever> all of my ZFS boxes are at 1% inode usage
<viric> That's why I used reiserfs for long
<clever> the ext4 hydra is at 30%
jmiven has joined #nixos
<slyfox> inode limit on ext4 was one of the primary reasons i've switched to btrfs
* gchristensen would rather deal with ext4
<slyfox> yesterday i had another fun issue with ext4 :)
<slyfox> one of our mailbots went bizarre and sent 1,8M emails to a target maildir
<viric> I didn't know ext4 had encryption
<slyfox> even after deletion of all emails directory took 125MB on dist
<slyfox> to fix it you need to delete and recreate an empty dir
<gchristensen> 13945751/nix
<gchristensen> yep ... that's all my inodes :)
<clever> gchristensen: one sec
<gchristensen> ok
<clever> avn: and the problem is recreated
<clever> [root@amd-nixos:/nix]# du --max=1 --inodes
<clever> avn: i ran this, arc is using 1.8gig, top says 300mb free, and all io just stops dead
<avn> clever: by your logs you pasted above, I think it looks like hdd issue.
<clever> avn: unplugging a usb keyboard instantly resumes all io and it continues like nothing happened
<clever> even if 1 hdd was bad, it should be able to run on the other
<avn> Would be funny if it some internal conflict betweek zfs io and usb guts
<clever> thats the whole point of a mirror
<avn> clever: what kernel you have?
<gchristensen> clever: what about it?
<clever> 4.4.36
<clever> gchristensen: trying to get a better cmd for your inodes, but /nix always cripples my system
<clever> gchristensen: so i'm also gathering data on my zfs problem at the same time
<avn> clever: I am on 4.9.5 atm
<gchristensen> hah
<gchristensen> o
<gchristensen> ok*
<clever> avn: i can barely even irc on that machine now, its freezing that badly
<avn> clever: usually one of 4 irsii locked up, when I have lockups -- possible on log writting
<slyfox> a silly 'nix-channel --update' question: every time that command download 8 MB file (nixpkgs https://nixos.org/channels/nixpkgs-unstable). Can I teach nix-channel to pull from git repo instead if that exists?
<clever> avn: the irc client is on a different machine, screen + ssh
<avn> so all lockups happens on write something like logs or sync -- systemd-journald, syslog-ng, irssi, browsers (chromium) complain on locking cache
<clever> avn: top says du is using 76% of the cpu, it has not output a single line yet
<clever> and the machine locked up in the middle of typing that
<avn> how long it in "freezed" state?
<avn> My cases was -- 2-3 minutes
<clever> i have gone go bed, then woke up to find it was locked up for the last 5 hours
<slyfox> does zfs support defragmentation?
<clever> avn: and replugging a usb keyboard instantly fixed it
<clever> avn: it almost never recovers on its own
<clever> gchristensen: and du has its first line of output!
<clever> that was about 7 minutes to run do over /nix/store
jgertm has quit [(Ping timeout: 240 seconds)]
<avn> all stuff resumes in 2-3 min, it why I name "softlocks" not deadlocks
<gchristensen> LnL, clever: I might just erase and start this hydra over :P
<clever> attempting to open the journal pushed it over the edge
<clever> replugging a keyboard brought it back
<clever> gchristensen: almost have the cmd ready
<gchristensen> ok
<avn> will try next time, although is not very useful (a lot cables around)
<clever> [clever@laptop:/nix/store]$ du --max=1 --inodes | sort -n
<clever> gchristensen: this will sort every storepath by how many inodes it contain
<clever> avn: main reason i discovered that usb helps, is because i have repurposed a box meant for sharing 1 usb printer to 4 pc's
<clever> avn: its sharing the keyboard/mouse between a desktop/laptop
nckx has quit [(Quit: ☭ + 🐧 + GNU Guix: https://gnu.org/s/guix)]
<clever> and one day, when the desktop locked up, i switched the keyb/mouse to the laptop to ease debuging
<clever> and the problem just ran away
nckx has joined #nixos
<clever> avn: "journalctl -f" was open for ~2 minutes, during 3 or 4 lockups, and the only thing it has is usb activity, and various other things timing out
<clever> alsa, vpn
<clever> i have released the memory pressure (closed the 6gig app), and now du runs much faster
<clever> and the arc has increased in size to 2.6gig
nckx has quit [(Client Quit)]
<clever> avn: 55 seconds to run du on the store now, arc heaped at 3.5gig
jgertm has joined #nixos
<clever> which implies it only needed another ~1gig of ram to do the job
<clever> why was it not just pushing junk into swap?
<gchristensen> lol
<gchristensen> this conversation is why I'm not using zfs
<clever> why did ALL IO just halt?, and bring the cpu to the ground with it?
rly has quit [(Ping timeout: 240 seconds)]
<gchristensen> I just don't have enough disks to justify zfs
nckx has joined #nixos
<clever> gchristensen: none of my other zfs machines do this, only the desktop
<avn> clever: `time du -sh /nix` is ok?
<clever> [root@amd-nixos:/nix/store]# time du --max=1 --inodes -l | sort -n
<clever> avn: this one finished in 16 seconds now that the arc has cached things
<LnL> gchristensen: can't you just change the TEMPDIR?
<avn> running, but no any system responsivity degraded
<gchristensen> LnL: I'm out of inodes
<LnL> oh
<avn> gchristensen: it why you should use zfs, they not use inodes ;)
<avn> clever: du -sh /nix 1,51s user 25,65s system 25% cpu 1:44,73 total
<LnL> ^ was just going to point that out :)
<LnL> gchristensen: ;)
<clever> real 0m16.899s
<clever> user 0m2.054s
<clever> time read dmis dhit dh% mrug mru mfu mfug mread c ph% pm% mm% miss hits mhit mh% mmis hit% eskip arcsz
<clever> 11:33:30 6.6K 227 6.4K 96 24 2.7K 3.7K 12 297 3.8G 29 70 35 266 6.4K 191 64 105 95 0 3.8G
<slyfox> i think reiserfs and xfs also don't have an inode limit
<clever> avn: restoring the strain on the system
<clever> slyfox: xfs does have inodes last i looked, but it allocates a lot more of them, and each bank of inodes is for data at a certain position in the disk
ixxie has quit [(Ping timeout: 252 seconds)]
<clever> slyfox: and one issue i have ran into, is that data stored past the 2tb position on the block device, will have an inode number larger then 4 billion (32bits)
pi3r has quit [(Ping timeout: 276 seconds)]
calvinx has quit [(Quit: calvinx)]
<slyfox> is it a problem? is inode number a 32-bit thing in some structures?
<clever> slyfox: only when running a 32bit userland
<clever> slyfox: and in my case, i was nfs mounting the 4tb XFS on a 32bit raspberry pi
<slyfox> aha
<clever> on a 64bit system, getdent() (used to list files in a dir) uses 64bit fields all around
<slyfox> TIL typedef unsigned long int __ino_t;
<clever> on a 32bit system, getdent() uses 32bit fields, and getdent64 uses 64bit fields
<clever> so you have to specialy compile the program to use the 64bit compat syscalls
<slyfox> even if _LARGE_FILE is specified?
<clever> i forget which -D flag it was, but it switches on 64bit compat
<slyfox> right
<clever> the problem, is that it changes the size of a field a lot of people assume is 32bits
xwvvvvwx has joined #nixos
<clever> and if you happen to use that field in your applications binary files
<slyfox> in Gentoo we have a elf checker right after package is being merged that it does not use legacy 32-bit glibc calls
<clever> and the secondary problem, a lot of programs dont correctly handle ESIZE from getdents()
<clever> so things like the console font utils, claim file not found
<clever> when it clearly does exist
<slyfox> thus maintainers are encouraged to tweak a package
<clever> that can easily be added to nixpkgs
<clever> i have done similar to prevent armv7 assembly from being used in armv6
<tnks> hey, I did this in Nix, and it seems to work; I wouldn't mind some review: https://gist.github.com/shajra/afed560778a556e487b201b5b4c2ad52
<tnks> the comment at the bottom sets up context.
nckx has quit [(Quit: ☭ + 🐧 + GNU Guix: https://gnu.org/s/guix)]
<clever> avn: strain is back, 6gig of the 16gig of ram is in use by a single app, du is r
<clever> avn: and it locked up
<clever> unning
nckx has joined #nixos
<clever> replugging a keyboard twice made it recover
<clever> avn: and du finished after 3 minutes, with a solid 1m 23s of system usage
<clever> gchristensen: looks like it needs some more aggresive GC'ing
<avn> tnks: should work, as quick-n-dirty, but also possible to look for `static` switch of ghc, and better way to deal datafiles (I am not so professional haskeller to tell how to do it fast)
<gchristensen> yeah
<clever> gchristensen: maybe auto-optimize has reduced the usage of those so far, that its not triggering your auto-gc stuff
<gchristensen> clever: or "many" ;)
<clever> gchristensen: the symlinks are cheap, and getting under the radar
<gchristensen> I'm not auto-gcing at all,actually, i think, on the master
<clever> ah
<clever> try manualy running a gc
<gchristensen> yeah
<clever> real 2m37.544s
<clever> sys 1m48.650s
Fare has joined #nixos
<clever> real 2m30.022s
<clever> sys 2m8.274s
<clever> avn: yikes, almost all in kernel space!
xwvvvvwx has quit [(Ping timeout: 248 seconds)]
<MichaelRaskin> I actually happen to use online shrinking of BtrFS a lot (for /nix/store)
<avn> MichaelRaskin: I not benchmarked, but looks like zfs outperforms btr on reading
<clever> MichaelRaskin: i have managed to kill btrfs with hydra before
<clever> MichaelRaskin: creating ~20,000 files of ~500 bytes each
xwvvvvwx has joined #nixos
<gchristensen> clever: still gc'ing
<gchristensen> bajillions of .drv's
<MichaelRaskin> Well, I have killed BtrFS a few times but I don't care
<clever> gchristensen: due to --optimize, you will basicaly regain none of the free space, until the very last step where it cleans up the .links dir
<MichaelRaskin> It has stopped being that easy to kill, though
<gchristensen> clever: I don't use optimise I think
<MichaelRaskin> It is not that to outperform BtrFS, actually…
<MichaelRaskin> not that hard
<clever> gchristensen: what does "stat /nix/store/.links" say?
<MichaelRaskin> But I want shrinking, and I want to be able to upgrade the kernel easily
Shou has joined #nixos
<clever> avn: ok, now things are not making any sense, i have 3gig of ram free in "free -m", and the arc is using the same 3gig it had before, but the du command now takes twice as long
<clever> avn: because i still have a single program with 6gig in its RSS open
<MichaelRaskin> Shrinking is very useful when I am forgetting to GC: grow the store, finish the builds, reboot, do a GC, shrink back
<avn> browser? ;)
<clever> avn: i just closed all of chromium to get that 3gig of ram free
<clever> [root@amd-nixos:/proc/26667/task]# kill -stop *
<gchristensen> ... still GCing
<clever> ok, now the program hogging 6gig of ram is 100% unable to touch the cpu
xwvvvvwx has quit [(Ping timeout: 255 seconds)]
<clever> avn: acording to arcstat.py, i'm getting hit rates as low as 9% still
fnljk has joined #nixos
<pierron_> copumpkin: can you give more detail on what you mean by making modules overrideable?
<tarinaky> So. I'm trying to build a variant iso where sshd is enabled from boot.
<avn> clever: 79% hit rates, and 32% hit to L2ARC
<clever> avn: no L2 right now, and the whole array is on an SSD
<clever> avn: only way an L2 could help is if it was nvme possibly
<tarinaky> I've added the line systemd.services.sshd.wantedBy=lib.mkForce["single-user.target" "multi-user.target"];
<tarinaky> But when I boot from the iso, and ps aux|grep sshd
<tarinaky> There's no sshd process.
<clever> tarinaky: how are you building the iso?
<tarinaky> nix-build -A config.system.build.isoImage -I nixos-config=modules/installer/cd-dvd/installation-cd-ssh.nix
<clever> tarinaky: try making your own configuration.nix, with imports = <nixpkgs/nixos/modules/installer/cd-dvd/installation-cd-ssh.nix>;
<clever> that can make things simpler to edit
<tarinaky> I did try that (installation-cd-minimal.ssh is the one from the git repo)
<tarinaky> Someone else suggested I make a copy of it and edit that.
<clever> avn: the system should have even less load then before now, but du refuses to get anywhere near its past numbers, its holding steady at 33sec, when it previously got 16sec
<tarinaky> As I was having trouble getting the import to work
<avn> clever: I have 2x1tb spinning drives, and 64g ssd splitted for ZIL and L2ARC
<clever> avn: a pair of 240gig SSD's in a zfs mirror, no ZIL or L2
<clever> avn: but this pool did start out on a smaller magnetic, i converted it into a magnetic+ssd mirror, then to a single ssd, then back to an ssd+ssd mirror
<avn> my pool started as vg0/tank volume on lvm
<clever> avn: one oddity i noticed when doing the conversion, when i made it into a magnetic+ssd+ssd mirror (with 1 "bad" ssd having to resilver), it performed horibly
Wizek has joined #nixos
<clever> avn: it was splitting the reads 50/50 between the 2 good drives, the magnetic and the ssd
<clever> obviously, the magnetic couldnt keep up
<avn> clever: I had a lot of horrigying lockups when I resilver from vg0/tank on degraded mdraid1 to bare /dev/sdb2 ;)
<clever> avn: i also ran into the occasional write error/timeout during my initial resilver
<clever> avn: and zfs is overly cautious, and restarted the entire device each time
<avn> clever: I see, and it explain why lockups happens, drive is enough old and do some sectors remap or likewise, timeouts, and zfs resets it
pi3r has joined #nixos
ktosiek has quit [(Quit: No Ping reply in 180 seconds.)]
simendsjo has quit [(Ping timeout: 258 seconds)]
ktosiek has joined #nixos
johnsona` has quit [(Ping timeout: 255 seconds)]
ertes has quit [(Ping timeout: 276 seconds)]
mizu_no_oto has joined #nixos
<NixOS_GitHub> [nixpkgs] Ma27 opened pull request #22231: geogebra: 5.0.271.0 -> 5-0-320-0 (master...hotfix/update-geogebra) https://git.io/vDJd5
bennofs1 has joined #nixos
bennofs has quit [(Ping timeout: 255 seconds)]
rmeis has joined #nixos
mizu_no_oto has quit [(Ping timeout: 255 seconds)]
derjohn_mob has quit [(Ping timeout: 245 seconds)]
<tarinaky> Hurray, it works.
jensbin has joined #nixos
<corngood> Do you just needed the single-user -> sshd dependency because the ISO boots to single-user?
<corngood> *so
jensbin has quit [(Client Quit)]
rmeis has quit [(Ping timeout: 240 seconds)]
nckx has quit [(Quit: ☭ + 🐧 + GNU Guix: https://gnu.org/s/guix)]
nckx has joined #nixos
nckx has quit [(Client Quit)]
sellout- has quit [(Ping timeout: 260 seconds)]
<clever> systemd.services.sshd.wantedBy = mkForce [ "multi-user.target" ];
<clever> corngood: ive had it work with just this
<corngood> Wouldn't that be the normal behaviour of sshd?
<clever> corngood: not on the ISO
<clever> 's
mkoenig has quit [(Remote host closed the connection)]
<NixOS_GitHub> [nixpkgs] teh opened pull request #22232: Dask update (master...dask-update) https://git.io/vDJFS
danharaj has joined #nixos
<corngood> clever: ah, I see. I never saw this module.
hotfuzz_ has joined #nixos
Shou has quit [(Ping timeout: 240 seconds)]
<clever> corngood: heh, i should try out line 30 of that file next time i do an install
<corngood> That's amazing
hotfuzz has quit [(Ping timeout: 240 seconds)]
<clever> corngood: it can be interesting to just start at a file like this, and read every imports recursively
danharaj has quit [(Remote host closed the connection)]
S0rin has quit [(Quit: WeeChat 1.6)]
S0rin has joined #nixos
<corngood> Yeah I never really ventured into profiles/ before
danharaj has joined #nixos
jgertm has quit [(Ping timeout: 258 seconds)]
johnsonav has joined #nixos
ttuegel has joined #nixos
<tarinaky> corngood: I got it to work :D
<ttuegel> Does anyone know how I can get a version of nixos-install that uses nixUnstable from the minimal installer? Short of building a new installer, I mean.
<clever> ttuegel: the linode guide could be abused, https://nixos.org/wiki/Install_NixOS_on_Linode
chrishill has quit [(Quit: Page closed)]
<clever> ttuegel: this section explains how to get nixos-install and friends via nix-env -i
<clever> at which point, you can pick any nixpkgs you want
<ttuegel> Oh ho! That's clever, clever!
mizu_no_oto has joined #nixos
<clever> this guide was meant for jamming nix into another distro (a rescue shell) and then using it to nixos-install
<ttuegel> I actually don't need a different nixpkgs; I custom built this installer from exactly the version I need, I just forgot that I needed nixUnstable. [shame]
<clever> but i have reused it for many things, like "nixos-rebuild build-vm" on gentoo
nckx has joined #nixos
<NixOS_GitHub> [nixpkgs] qknight pushed 1 new commit to master: https://git.io/vDJNq
<NixOS_GitHub> nixpkgs/master 61e4313 Yacine Hmito: rustNightlyBin: 2016-12-29 -> 2017-01-26 (#22228)...
<ttuegel> FWIW, I discovered that the installer has its own /etc/nixos/configuration.nix, which I edited and then ran `nixos-rebuild switch`, which has the advantage of giving me the unstable nix-daemon, too.
<clever> ttuegel: yep, ive used that to add zfs support sometimes, but it can sometimes lead to downloading the world if you have changed the channel
hpc has left #nixos []
<corngood> tarinaky: good to hear
alp__ has joined #nixos
<justanotheruser> trying to install tensorflow, getting "protobuf-3.0.0-beta-2 not supported for interpreter python3.5m", how do I fix this? https://paste.debian.net/plain/910852
nckx has quit [(Quit: ☭ + 🐧 + GNU Guix: https://gnu.org/s/guix)]
ixxie has joined #nixos
alpounet has quit [(Quit: WeeChat 0.4.2)]
<NixOS_GitHub> [nixpkgs] vbgl pushed 1 new commit to master: https://git.io/vDJAn
<NixOS_GitHub> nixpkgs/master 23894b7 Vincent Laporte: ocamlPackages.dolmen: init at 0.2...
pi3r has quit [(Quit: Nettalk6 - www.ntalk.de)]
alp__ has quit [(Remote host closed the connection)]
alp_ has joined #nixos
alp_ has quit [(Remote host closed the connection)]
<copumpkin> gchristensen: how much would a reduced redundancy S3 binary cache be? any ideas/predictions?
<copumpkin> I guess niksnut probably knows best
<NixOS_GitHub> [nixpkgs] jpierre03 opened pull request #22234: perl::List-Compare: 0.39 -> 0.53 (master...pr-cpan-list-compare) https://git.io/vDJA9
systemfault has joined #nixos
<simpson> With nixops, what is the right way to upsize?
<simpson> Oh, I guess that I need to ask about upsizing on GCE specifically.
gilligan_ has joined #nixos
Wizek has quit [(Ping timeout: 255 seconds)]
<clever> gchristensen: oh, hows the GC going?
<NixOS_GitHub> [nixpkgs] rycee pushed 1 new commit to master: https://git.io/vDJxZ
<NixOS_GitHub> nixpkgs/master 44eedb9 Robert Helgesson: perl-List-Compare: 0.39 -> 0.53...
<NixOS_GitHub> [nixpkgs] rycee closed pull request #22234: perl::List-Compare: 0.39 -> 0.53 (master...pr-cpan-list-compare) https://git.io/vDJA9
<gilligan_> any Haskell folks in here that use ghc-mod ?
wangoe has joined #nixos
Fare has quit [(Quit: Leaving)]
copumpkin has quit [(Quit: My MacBook Pro has gone to sleep. ZZZzzz…)]
<NixOS_GitHub> [nixpkgs] rycee pushed 1 new commit to release-16.09: https://git.io/vDJx1
<NixOS_GitHub> nixpkgs/release-16.09 55d3730 Robert Helgesson: perl-List-Compare: 0.39 -> 0.53...
_rht has joined #nixos
wangoe has left #nixos ["WeeChat 1.5"]
rly has joined #nixos
<NixOS_GitHub> [nixpkgs] rycee pushed 2 new commits to master: https://git.io/vDJp3
<NixOS_GitHub> nixpkgs/master cf7f34a Robert Helgesson: perl-IO-CaptureOutput: 1.1103 -> 1.1104...
<NixOS_GitHub> nixpkgs/master 588f4b7 Robert Helgesson: perl-HTML-Tiny: add meta section
wangoe has joined #nixos
dbmikus has quit [(Ping timeout: 256 seconds)]
<NixOS_GitHub> [nixpkgs] aneeshusa opened pull request #22235: mosh: use ssh from nixpkgs (master...mosh-use-ssh-from-nixpkgs) https://git.io/vDJp2
<ttuegel> I've had nixos-install running for almost an hour now and its stuck at "building the system configuration...". There is no CPU or disk activity. The process does not respond to ^C. Any ideas how to debug?
<Ralith> if it's reproducible, strace is often insightful
<MichaelRaskin> strace can be connected to running processes, too
<MichaelRaskin> no CPU and no IO and no reaction to Ctrl-c may be a stuck non-interruptable syscall
bennofs1 has quit [(Quit: WeeChat 1.6)]
endformationage has joined #nixos
<viric> ttuegel: you can dump kernel stack traces
<viric> for a syscall
<viric> in /proc/PID/stack
<viric> that's where you use to find a filesystem deadlock :)
phant0mas has quit [(Remote host closed the connection)]
pi3r has joined #nixos
<clever> viric: you can also do it with sysrq, either show-blocked-tasks(w) or show-backtrace-all-active-cpus(l)
<viric> clever: right, I couldn't find the sysrq thing
<clever> [root@amd-nixos:/nix/store]# echo ? > /proc/sysrq-trigger
<clever> [root@amd-nixos:/nix/store]# dmesg | tail
<clever> [225741.334066] sysrq: SysRq : HELP : loglevel(0-9) reboot(b) crash(c) terminate-all-tasks(e) memory-full-oom-kill(f) kill-all-tasks(i) thaw-filesystems(j) sak(k) show-backtrace-all-active-cpus(l) show-memory-usage(m) nice-all-RT-tasks(n) poweroff(o) show-registers(p) show-all-timers(q) unraw(r) sync(s) show-task-states(t) unmount(u) force-fb(V) show-blocked-tasks(w) dump-ftrace-buffer(z)
<viric> clever: I read the (h) and couldn't find it
<viric> I know
<clever> ah
<viric> I read it twice and I couldn't find it
<viric> :)
bennofs has joined #nixos
<clever> viric: something ive been thinking about recently, is to make a program that is designed to eat an entire core
<clever> in theory, if i flag a program as real-time, and never yield the cpu, zfs cant spinlock the core and steal it
<clever> and then it can continue to analyze the problem, even when the system deadlocks
<Ralith> what can I put in configuration.nix to specify a custom nixpkgs path?
<clever> Ralith: you need to use -I nixpkgs= with nixos-rebuild
<clever> it is nixpkgs that loads configuration.nix, so by the time its being read, its too late to change the nixpkgs
<Ralith> clever: that's what I do now, but I want to avoid having to do that every single time
MarkusBarthlen has joined #nixos
<Ralith> I want to make that sticky
<clever> Ralith: there is nix.nixPath, but this takes effect AFTER the config is built, not during the build
<Ralith> yes, I understand
<clever> that provides a default for $NIX_PATH
<Ralith> looks like just the thing, thanks
<NixOS_GitHub> [nixpkgs] FRidh pushed 9 new commits to master: https://git.io/vDJjs
<NixOS_GitHub> nixpkgs/master b90737a Tom Hunger: sortedcollections: init at 0.4.2
<NixOS_GitHub> nixpkgs/master aaf5a2b Tom Hunger: cloudpickle: 0.2.1 -> 0.2.2
<NixOS_GitHub> nixpkgs/master dfc098c Tom Hunger: zict: 0.0.3 -> 0.1.1
<bennofs> Ralith: I have set my NIX_PATH to /run/current-system/nixpkgs, and my configuration symlinks the nixpkgs it is build against into the output environment. This means that if I run nixos-rebuild switch -I nixpkgs=/path/to/nixpkgs, then the next time I run nixos-rebuild, it'll use that as nixpkgs (i.e. it always uses the nixpkgs of the currently active system profile)
<clever> bennofs: ah neat, so it can only upgrade if you -I nixpkgs=, and from then on, it sticks
<Ralith> that sounds clever, but for my case I'm making my nixpkgs fork a submodule of my nix config repo, so a fixed path makes more sense
<bennofs> clever: yeah
<clever> bennofs: ive done something similar, https://gist.github.com/cleverca22/d70a6826db80d0ee2ae9074d83690e98
<clever> this causes the entire config repo to be included inside the build
pi3r has quit [(Ping timeout: 240 seconds)]
<corngood> Sort of related... I was trying to sync up my nixpkgs config between nix-env and nixos, so I ended up having config.nix read the nixpkgs attribute from configuration.nix (nixos). Anyone else done this or have a better way?
<avn> bennofs: neat trick
<bennofs> clever: ah, that's nice as well.
<clever> bennofs: it also ate the .git directory as well, so i can "git diff" and see what i changed and didnt commit yet, in any build
<clever> [root@amd-nixos:/run/current-system/nixcfg]# git status
<clever> modified: amd-nixos.nix
<clever> ...
<bennofs> corngood: that's exactly why I've done what I've just described. I don't use channels, instead, I have a symlink to /run/current-system/nixpkgs in my ~/.nix-defexpr
<bennofs> corngood: so nix-env uses the nixpkgs of the currently active system profile
<NixOS_GitHub> [nixpkgs] viric pushed 4 new commits to master: https://git.io/vDJjg
<NixOS_GitHub> nixpkgs/master aeb1dff Lluís Batlle i Rossell: Update scantailor
<NixOS_GitHub> nixpkgs/master 20465be Lluís Batlle i Rossell: Updating shotcut and melt.
<NixOS_GitHub> nixpkgs/master 91686f2 Lluís Batlle i Rossell: Add fixed cinelerra...
<corngood> bennofs: Not for nixpkgs itself. I used defexpr (?) For that, but for package configs.
<corngood> Unless I'm misunderstanding
<bennofs> corngood: oh, nixpkgs config. Well, I just put import /etc/nixos/nixpkgs.nix into my ~/.nixpkgs/config.nix
<bennofs> corngood: and then use nixpkgs.nix from my configuration.nix as well
<bennofs> clever: hmm, I have a script that does (sort of) git pull and nixos-rebuild, so I never build my system from uncomitted changes
<corngood> OK, so similar. I just wanted to keep that stuff in ~
<clever> bennofs: i usualy edit localy, and confirm things work before i commit
<bennofs> clever: well, I use git amend a lot :)
<clever> bennofs: and sometimes i just dont commit for weeks, and then i commit all the changes everywhere and sync the machines up
AveryC has quit [(Ping timeout: 240 seconds)]
eacameron has joined #nixos
alp_ has joined #nixos
alp_ is now known as alpounet
eacameron has quit [(Ping timeout: 260 seconds)]
* Ralith watches a big rebuild on his server spend 90% of its time negotiating HTTP connections
<Ralith> would be neat if nix could pipeline requests
mizu_no_oto has quit [(Quit: ["Textual IRC Client: www.textualapp.com"])]
bennofs has quit [(Quit: WeeChat 1.7)]
bennofs has joined #nixos
<NixOS_GitHub> [nixpkgs] Baughn opened pull request #22236: apache-httpd: Add 'extensions' config option (master...mediawiki) https://git.io/vDUeJ
<toogley> is there a way how i can fix https://paste.fedoraproject.org/538542/63160814/raw/ ?
eacameron has joined #nixos
nckx has joined #nixos
<madonius> hey
<madonius> I am a noob and somehow i have issues installing stuff on my nixos
<madonius> I want to install 'pass', the appropriate package seems to be password-store
eacameron has quit [(Ping timeout: 252 seconds)]
<gchristensen> clever: I had to leave for a protest, it finally finished at some point.
<madonius> nix-env returns that name, when I copy it to the config nix claims to now know it
<madonius> any pointers?
joko has quit [(Ping timeout: 258 seconds)]
<LnL> I think that's the package name, for expressions you habs to use the attribute
joko has joined #nixos
<clever> 43 password-store-1.6.5 (nixos.pass)
<clever> from "nox pass"
<clever> madonius: have you tried pkgs.pass ?
<madonius> hmm I will try the above
<madonius> ok pass works
<madonius> this is utterly confusing for me /o\
<ixxie> madonius: have you seen some example configurations?
<madonius> yeah i looked at some on github
<madonius> but then again, i might be missing the point, since nixos is on my system only since yesterday
<LnL> madonius: you can list all the package atteibutes and names with nix-env -qaP
<LnL> and you can also install things. y attribute with nix-env -iA
<ixxie> madonius: perhaps your confusion is due to lack of familiarity with the nix expression language
<madonius> ixxie: very possible
<ixxie> for me this has been the main learning curve
<LnL> I think nix-env just has some confusing differences compared to all the other things
<ixxie> but LnL, I think madonius is doing declarative configuration.nix
<madonius> i would expect that the package name is what is to be entered in the systempackages list
<LnL> yes but nix-env -iA matches the names you expect, the default nix-env -i doesn't
<clever> madonius: the nix files always need attribute paths, and nothing else
<clever> madonius: nix-env is a weird tool, and will search the name attributes on things by default
jgertm has joined #nixos
<ixxie> madonius: it is important to understand that each nix dotfile is a nix expression (i.e. a function) and it takes the dependencies as arguments
<ixxie> this is why you have {config, pkgs, ... } on top
<NixOS_GitHub> [nixpkgs] viric pushed 1 new commit to master: https://git.io/vDUeM
<NixOS_GitHub> nixpkgs/master 5762551 Lluís Batlle i Rossell: Adding nuttcp.
<ixxie> pkgs is your nixpkgs collection
<ixxie> config is your nixos configuration as you define it in the your configuration moduiles
<ixxie> so if you are including a package from nixpkgs, you need to take it from the pkgs set - i.e. pkgs.yourpackage
<LnL> and pkgs.yourpackage.name could contain something different
<LnL> that's what nix-env looks at by default
<clever> nix-repl> pkgs.pass.name
<clever> "password-store-1.6.5"
zarel has quit [(Ping timeout: 255 seconds)]
<ixxie> madonius: for multiple packages, you can use the `with` pattern to add the pkgs set into a scope of - for example - a list like I did here https://github.com/fluxcraft/fluxstack/blob/master/env/nixos/modules/fluxbase.nix
<ixxie> then there is no need to repeat pkgs for every element
rmeis has joined #nixos
eacameron has joined #nixos
<madonius> hmm i see
<madonius> thanks
ttuegel has quit [(Remote host closed the connection)]
<ixxie> madonius: finally, the language overview is found in the Nix manual (for basic packaging https://nixos.org/nix/manual/#ch-simple-expression but in more generality https://nixos.org/nix/manual/#ch-expression-language )
<madonius> yeah i saw that
<ixxie> the nixpills are also cool http://lethalman.blogspot.fi/search/label/nixpills
<LnL> this is also pretty good for the basics https://medium.com/@MrJamesFisher/nix-by-example-a0063a1a4c55#.djbc0mguk
<madonius> but to be honest not very beginner friendly /o\
<madonius> LnL: thnx
<ixxie> I feel your pain madonius
<madonius> I think for now I have a LOT to read up
<LnL> that post is the most beginner friendly resource I know of
rmeis has quit [(Ping timeout: 258 seconds)]
reinzelmann has joined #nixos
ixxie has quit [(Ping timeout: 276 seconds)]
pi3r has joined #nixos
Jackneill_ has quit [(Ping timeout: 276 seconds)]
mizu_no_oto has joined #nixos
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
nckx has quit [(Quit: ☭ + 🐧 + GNU Guix: https://gnu.org/s/guix)]
josePhoenix has joined #nixos
<josePhoenix> Hi all
ambro718 has joined #nixos
<ambro718> is there still no "nixos-rebuild boot" equivalent in nixops?
<NixOS_GitHub> [nixpkgs] pradeepchhetri opened pull request #22237: vault: 0.6.3 -> 0.6.4 (master...vault) https://git.io/vDUff
<josePhoenix> I am following the example in http://nixos.org/nixos/manual/index.html#sec-user-management of adding a key to "openssh.authorizedKeys.keys" for a user, but the key does not appear in ~/.ssh/authorized_keys after nixos-rebuild switch
<ambro718> josePhoenix: just try is the login with that key works :)
<josePhoenix> It doesn't :(
<josePhoenix> I see "debug1: Offering ECDSA public key: /Users/jdl/.ssh/id_ecdsa" from my side but then get a password prompt anyway
<ambro718> the docs for this option say "A list of verbatim OpenSSH public keys that should be added to the user's authorized keys. The keys are added to a file that the SSH daemon reads in addition to the the user's authorized_keys file. You can combine the keys and keyFiles options."
<ambro718> so it seems expected to not see them in ~/.ssh/authorized_keys. Don't know why it doesn't work for you though.
<josePhoenix> That's interesting
<josePhoenix> But yeah, still doesn't appear to be working
<ambro718> josePhoenix: does adding the key to ~/.ssh/authorized_keys work?
<Yaniel> just to make sure your openssh.authorizedKeys.keys looks like ["ssh-rsa blahblahblah foo@bar"] ?
<josePhoenix> Adding to ~/.ssh/authorized_keys works
<josePhoenix> And yes Yaniel, more or less
<josePhoenix> key type is ecdsa-sha2-nistp256 and foo@bar is the user@host where the key was generated
AveryC has joined #nixos
<ambro718> from the code I think it will convert the config into /etc/sshd/authorized_keys.d - maybe examine that
<ambro718> * ssh not sshd
<josePhoenix> Hmm
<NixOS_GitHub> [nixpkgs] offlinehacker pushed 2 new commits to master: https://git.io/vDUfB
<NixOS_GitHub> nixpkgs/master c0fd124 Pradeep Chhetri: vault: 0.6.3 -> 0.6.4
<NixOS_GitHub> nixpkgs/master 8a29566 Jaka Hudoklin: Merge pull request #22237 from pradeepchhetri/vault...
<josePhoenix> Oh, what the heck
<josePhoenix> It was word-wrapped in 'nano'
<josePhoenix> In configuration.nix
<josePhoenix> that would do it
<ambro718> nano -w FTW :)
<ambro718> I never use nano, only nano -w
<josePhoenix> Also, does users.mutableUsers = false; mean passwords get reset to unusable on rebuild?
<josePhoenix> ambro718: I think Ubuntu might alias nano to nano -w? I'm used to not having to specify it :)
<ambro718> josePhoenix: quite possibly
<ambro718> josePhoenix: about the first question - I think it would result in all user accounts including passwords being reset to whatever the configuration specifies
<ambro718> "If you set users.mutableUsers to false, then the contents of /etc/passwd and /etc/group will be congruent to your NixOS configuration. For instance, if you remove a user from users.extraUsers and run nixos-rebuild, the user account will cease to exist. Also, imperative commands for managing users and groups, such as useradd, are no longer available."
<josePhoenix> Makes sense
<ambro718> I suppose you should simply understand it as that /etc/passwd and /etc/shadow are generated from nixos configuration and cannot be modified outside of nixos-rebuild
<josePhoenix> I guess if I have key-based auth and passwordless sudo set up, I don't need a password at all >_>
Jackneill_ has joined #nixos
<ambro718> lack of good "secrets" support in nix/nixos is a known issue :)
<ambro718> including any password in the nixos configuration would make it readable to any local user, as the nix store is publicly readable
sellout- has joined #nixos
nckx has joined #nixos
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
gilligan_ has quit [(Ping timeout: 255 seconds)]
glines has joined #nixos
<NixOS_GitHub> [nixpkgs] pradeepchhetri opened pull request #22239: certbot: 0.9.3 -> 0.10.2 (master...certbot) https://git.io/vDUJL
<swistak35> I'm using NAT through networking.nat nixos options. If my sourcePort is 1234, is there any way to limit that only requests from one IP can use this port?
Shou has joined #nixos
<swistak35> In NAT there's no drop-all rule in the end, so by default everything is accepted to the NATed ports, and no `extraCommands` or something like that to control it
<swistak35> Any way to do it without just specifying all NAT rules manually?
<swistak35> by manually I mean not using networking.nat, but put everything into network.firewall.extraCommands instead
slyfox has quit [(Read error: Connection reset by peer)]
slyfox has joined #nixos
alx741_ has joined #nixos
Shou has quit [(Ping timeout: 255 seconds)]
alx741 has quit [(Ping timeout: 264 seconds)]
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
eacameron has quit [(Read error: Connection reset by peer)]
eacameron has joined #nixos
tarinaky has quit [(Ping timeout: 258 seconds)]
<cransom> the module doesn't monkey with forwarding rules so you'd need to do it manually.
gilligan has joined #nixos
<NixOS_GitHub> [nixpkgs] pradeepchhetri opened pull request #22240: openconnect: 7.06 -> 7.08 (master...openconnect) https://git.io/vDUUy
athan has joined #nixos
dbmikus has joined #nixos
dtzWill has quit [(Quit: leaving)]
reinzelmann has quit [(Ping timeout: 255 seconds)]
goibhniu has joined #nixos
<LnL> I don't understand why this doesn't work https://gist.github.com/LnL7/3227f4c8800fe7d7d170156a38822b05
<LnL> The option value `containers.foo.systemd.services.bar.serviceConfig' in `<unknown-file>' is not a attribute set of systemd options.
reinzelmann has joined #nixos
hotfuzz_ has quit [(Quit: WeeChat 1.6)]
<NixOS_GitHub> [nixpkgs] domenkozar pushed 1 new commit to master: https://git.io/vDUT3
<NixOS_GitHub> nixpkgs/master 4c22b95 Domen Kožar: Add a script to get failures for hydra eval /cc @globin
<LnL> I have a nixops deployment with a service that basically does the same
justan0theruser has joined #nixos
michas has joined #nixos
justanotheruser has quit [(Ping timeout: 248 seconds)]
hotfuzz has joined #nixos
dbmikus has quit [(Ping timeout: 255 seconds)]
dbmikus has joined #nixos
michalrus has quit [(Quit: leaving)]
michalrus has joined #nixos
rly has quit [(Ping timeout: 248 seconds)]
ryantrinkle has quit [(Ping timeout: 240 seconds)]
dbmikus has quit [(Ping timeout: 276 seconds)]
hiratara has quit [(Ping timeout: 255 seconds)]
bennofs1 has joined #nixos
hiratara has joined #nixos
hotfuzz has quit [(Quit: WeeChat 1.6)]
reinzelmann has quit [(Quit: Leaving)]
hotfuzz has joined #nixos
<nhooyr> I use envirnoment.systemPackages to install everything I use on the command line as root and my normal user. Should I instead be installing these things with `nix-env`? Also, is there a way to declaratively install them for each user?
_rht has quit [(Quit: Connection closed for inactivity)]
nhooyr has quit [(Remote host closed the connection)]
nhooyr has joined #nixos
nhooyr has quit [(Remote host closed the connection)]
nhooyr has joined #nixos
alpounet has quit [(Read error: Connection reset by peer)]
alpounet has joined #nixos
<NixOS_GitHub> [nixpkgs] phryneas closed pull request #22168: rambox: 0.4.5 -> 0.5.2 (master...master) https://git.io/vMhXw
<Yaniel> nhooyr: the declarative way to install stuff as a user is to write a ~/.nixpkgs/config.nix
<nhooyr> ok thanks
<Yaniel> which defines stuff in packageOverrides
<Yaniel> I have a package called all there, which is basically `with pkgs; buildEnv { all packages I want }`
<Yaniel> then I install that with `nix-env -iA nixos.all`
athan has quit [(Remote host closed the connection)]
mkoenig has joined #nixos
<nhooyr> ah ok thanks
hiratara has quit [(Quit: ZNC - http://znc.in)]
ryantrinkle has joined #nixos
hiratara has joined #nixos
<tnks> has anyone made integrated calls to cabal2nix into their Haskell builds?
<tnks> I know there's callHackage, but that's different.
<tnks> the way I was going to do it would put source in /nix/store, so I guess maybe not ideal?
goibhniu has quit [(Ping timeout: 245 seconds)]
<justan0theruser> trying to install tensorflow, getting "protobuf-3.0.0-beta-2 not supported for interpreter python3.5m", how do I fix this? https://paste.debian.net/plain/910852
<NixOS_GitHub> [nixops] Nadrieril opened pull request #596: Fix #499 (master...fix-499) https://git.io/vDULP
ttuegel has joined #nixos
<NixOS_GitHub> [nixpkgs] corngood opened pull request #22241: WIP: amdgpu-pro: 16.50 -> 16.60 (master...amdgpu-update) https://git.io/vDUL1
<ttuegel> I'm trying to use the NixOS installer to rescue a broken NixOS installation. I see that `nixos-install --chroot` will allow me to run a command in the chroot, but trying to build anything results in "error: cloning builder process: Operation not permitted". Does anyone know what to do about this?
<madonius> i am no expert myself, but you did check the filesystem premissions, right?
kampfschlaefer has quit [(Ping timeout: 256 seconds)]
<ttuegel> madonius: You mean the mount permissions? Yes, everything is rw.
<madonius> thought so
<ttuegel> I just need to change the /etc/fstab, but it seems to be impossible :'(
<swistak35> ttuegel: Can't you use the fact that all nixos generations are in GRUB? And choose last working generation from it?
rmeis has joined #nixos
djames has joined #nixos
<ttuegel> swistak35: No, I'm reinstalling after upgrading my motherboard. The partition labels changed, grub is gone, and there are no working generations.
<djames> I'm a relatively new NixOS user -- most interested in the package management side at the moment. I know how to install things but haven't yet figured out how to inspect what got installed and why. What would you suggest as next to read or try?
<ttuegel> Aha! The magic option in `--option build-use-sandbox false`!
<djames> I've used apt, homebrew, etc... looking for something analogous to `brew info <package>`
rmeis has quit [(Ping timeout: 255 seconds)]
sitedyno has quit [(Ping timeout: 248 seconds)]
teh__ has quit [(Ping timeout: 260 seconds)]
mudri has quit [(Quit: WeeChat 1.6)]
mudri has joined #nixos
<NixOS_GitHub> [nixpkgs] kquick opened pull request #22243: Use Brick 0.17 because 0.16 is not compatible with vty_5_15 specified… (master...master) https://git.io/vDUt7
jgeerds has joined #nixos
bennofs has quit [(Ping timeout: 276 seconds)]
sitedyno has joined #nixos
rardiol has joined #nixos
michas has quit [(Ping timeout: 258 seconds)]
ttuegel has quit [(Remote host closed the connection)]
michas has joined #nixos
<NixOS_GitHub> [nixpkgs] jagajaga pushed 2 new commits to master: https://git.io/vDUqm
<NixOS_GitHub> nixpkgs/master 56ed9f8 Pradeep Chhetri: openconnect: 7.06 -> 7.08
<NixOS_GitHub> nixpkgs/master d948e55 Arseniy Seroka: Merge pull request #22240 from pradeepchhetri/openconnect...
eacameron has quit [(Remote host closed the connection)]
jgertm has quit [(Ping timeout: 252 seconds)]
<Ralith> djames: perhaps review the OPERATION --QUERY section of `man nix-store`?
eacameron has joined #nixos
gpyh has joined #nixos
<gpyh> hi
<djames> Ralith: thanks -- I didn't see an answer there. Is there a corresponding command to "brew info <package>" ... e.g. human-written text that gives guidance on a package. For example, postgresql has various things that a package manager might or might not do
<djames> My two cents is that I would rather not have to dig too far into nix in order to use it effectively
<djames> of course, I haven't defined "too far". but the "brew info _" abstraction is simple and intuitive. looking for its analogue, if any, in nix
<rardiol> is there some way to use the steam fhs/environment without running the steam command?
<Ralith> I have no idea what `brew info` does
<djames> Ralith : I just told you
athan has joined #nixos
<corngood> Are you talking about dependencies etc?
<djames> the first google hit also explains it
<gpyh> djames: nix is declarative; it's not a problem to have to dig into it
<djames> corngood if that was for me, yes dependencies are part of it
<djames> however, a human-written chunk of text that gives recommendations what to do post install are useful. that's why brew does it
<corngood> The metadata in nix packages is pretty limited. It's like, name, maintainer, URL, license pretty much
<djames> gpyh what does that mean? I love declarative languages. I'm simply saying that I've spent over 15 minutes searching for a basic mapping from brew to nix. I'm not alone.
<gpyh> nix does have the concept of dependency
<gpyh> doesn't*
<gpyh> It's quite unfriendly yes
<gpyh> and there's a lot of room for improvement
<gpyh> but I don't believe in the solution you propose
<bennofs1> djames: using nix without understanding the Nix language and reading parts of nixpkgs is almost impossible at the current state
<djames> gpyh what solution am I proposing?
<gpyh> generating cli help from metadata
<djames> I was just asking if something like 'brew info' existed. if not, that's ok. corngood talked about the metadata so it makes it sound like probably not
<gpyh> oh sorry
<gpyh> then no, there isn't a direct equivalent
<djames> so, next, I'll look into nix expressions and the package repository on git to see what the installs do exactly
<corngood> I suggest pulling up the nix source for a package you're interested in.
<gpyh> definitely clone it
<djames> gpyh thanks, just asking what I thought was a sensible question :)
panostimos has quit [(Ping timeout: 240 seconds)]
<corngood> You'll understand what I mean about the metadata
Jackneill_ has quit [(Remote host closed the connection)]
<gpyh> greping through nixpkgs is my 1. source of documentation
<corngood> And if the info isn't in there it would belong in the manual I believe
<gpyh> it was frustating at first but every package is its special snowflake when things get hard
<gpyh> and when things aren't, there's usually nothing to document
ryantrinkle has left #nixos []
<gpyh> +1 corngood, and on the website you also have documentation for everything that's configurable for nixos
<gpyh> I mean every config attribute
<corngood> I think nix-env tends to give people the impression that nix is more like other package managers than it actually is when you're using it declaratively
<gpyh> I am seeking assistance to setup NAT on my machine; I've spent a couple of hours already to no avail
<djames> thanks for the context!~
<corngood> Though in this case having metadata easily available and some link between packages and docs is certainly not an unreasonable request
<djames> so in the case of postgresql (and other services) should I generally guess/hope/think that they will setup services? (in macOS? in linux?) is there a doc page on service starting w.r.t nix?
<djames> e.g. startup services
<gpyh> are you on nixos djames?
<djames> gpyh macOS just using nix as a package manager. is that crazy?
<gpyh> no, but you can't use services (AFAIK)
<djames> ah, good to know
<LnL> corngood: you can only get it in json/xml
<gpyh> if you're on nixos, you have the concept of modules, some of which being service; and for those the documentation is excellent
<LnL> corngood: nix-env -qa -f '<nixpkgs>' -A hello --meta --xml
<LnL> djames: I have a project that brings a bunch of nixos stuff to darwin
<corngood> I can't say I would use a human readable version, but I can see how someone might expect it to be there
<LnL> yeah, --meta without the xml or json flag doesn't show anything at the moment
<gchristensen> how's it going, domenkozar?
<LnL> gchristensen: did the gc help?
<NixOS_GitHub> [nixpkgs] dezgeg pushed 2 new commits to master: https://git.io/vDUmY
<NixOS_GitHub> nixpkgs/master d1d8ed2 Tuomas Tynkkynen: stdenv: Aarch64 bootstrap: Update bootstrap tarballs to hydra-built ones...
<NixOS_GitHub> nixpkgs/master 4d059f6 Tuomas Tynkkynen: spidermonkey_17: Add one more patch for Aarch64
<NixOS_GitHub> [nixpkgs] joachifm pushed 1 new commit to master: https://git.io/vDUmG
<NixOS_GitHub> nixpkgs/master ba1aeb8 Joachim Fasting: felix: 2.0.5 -> 5.6.1, fix build...
<NixOS_GitHub> [nixpkgs] joachifm pushed 2 new commits to master: https://git.io/vDUml
<NixOS_GitHub> nixpkgs/master 46c0da1 Michael Peyton Jones: arbtt: multi-user.target does not exist in user systemd
<NixOS_GitHub> nixpkgs/master ac1e65c Joachim F: Merge pull request #22230 from michaelpj/services/arbtt-fix-wanted-by...
pi3r has quit [(Quit: Nettalk6 - www.ntalk.de)]
<gchristensen> LnL: I think so! :)
<LnL> allright, so using autogc on the master should fix this
<NixOS_GitHub> [nixpkgs] AndersonTorres opened pull request #22244: Intercal: a little fix (master...intercal) https://git.io/vDUmu
<NixOS_GitHub> [nixpkgs] ttuegel opened pull request #22245: dropbox: use `paxmark m` to fix Dropbox when grsecurity enabled (master...dropbox-grsecurity) https://git.io/vDUm2
Dezgeg has quit [(Ping timeout: 240 seconds)]
Dezgeg has joined #nixos
flyx has quit [(Excess Flood)]
flyx has joined #nixos
eacameron has quit [(Remote host closed the connection)]
eacameron has joined #nixos
mudri has quit [(Quit: WeeChat 1.6)]
mudri has joined #nixos
<LnL> gchristensen: going to take another look at the jobset stuff tomorrow
<LnL> hopefully I can get something working to improve the diff
JagaJaga has quit [(Ping timeout: 255 seconds)]
<gchristensen> cool :)
xwvvvvwx has joined #nixos
mkoenig has quit [(Read error: Connection reset by peer)]