<clever>
also, using a git rev would have been better
<gchristensen>
asymmetric: most people use the "tofu" method
<clever>
the problem, is that the filename is based on the rev and hash
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<clever>
and both network and protocol have v1.4.0.tar.gz
<gchristensen>
"trust of first use": put in the wrong hash, have nix tell them what the rightone is
<asymmetric>
gchristensen: that's how i got myself in this situation though
<gchristensen>
oh
<clever>
if you used a rev, the filenames wouldnt have matched, and it would have missed
<clever>
asymmetric: that "wrong" hash was right for another file
<gchristensen>
oh, yeah
<gchristensen>
change / break the hash :)
<clever>
it has to be a hash that doesnt match any file you have downloaded
<MP2E>
that's what I usually do. I use "nix-hash --type sha256 --base32 Gimmick.jpg", gimmick.jpg being an image of my github avatar (any random file works that isn't going to be an input to a nixpkgs derivation)
<MP2E>
though that still has a chance of colliding i suppose
<MP2E>
still, unlikely!
<asymmetric>
doesn't nix check that the hash belongs to the right repo?
<clever>
nix doesnt understand git
<clever>
it just has a name and hash
<gchristensen>
asymmetric: nix doesn't care where it came from as long as the hash was correct
<clever>
which form a storepath
<asymmetric>
ah, good point
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 1 new commit to release-17.09: https://git.io/vbyt7
alexteves has quit [(Quit: My Mac Mini has gone to sleep. ZZZzzz…)]
nuncanada has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] samvher opened pull request #32925: idris-modules: clarify usage of with-packages (master...patch-2) https://git.io/vbyuj
NixOS_GitHub has left #nixos []
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
<samae>
hey, any reason why using grep in runCommand would systematically fail?
otherjrolfs has joined #nixos
jrolfs has joined #nixos
ma27 has quit [(Ping timeout: 252 seconds)]
jrolfs has quit [(Ping timeout: 248 seconds)]
otherjrolfs has quit [(Ping timeout: 248 seconds)]
jensens has joined #nixos
asuryawanshi has joined #nixos
asuryawa_ has joined #nixos
asuryawanshi has quit [(Read error: Connection reset by peer)]
otherjrolfs has joined #nixos
jrolfs has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
rogue_koder has quit [(Quit: Konversation terminated!)]
otherjrolfs has quit [(Ping timeout: 248 seconds)]
jrolfs has quit [(Ping timeout: 252 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] markuskowa opened pull request #32926: nrsc5: init at 20171129 (master...nrsc5) https://git.io/vbygO
NixOS_GitHub has left #nixos []
otherjrolfs has joined #nixos
jrolfs has joined #nixos
nuncanada has quit [(Ping timeout: 272 seconds)]
jmeredith has quit [(Quit: Connection closed for inactivity)]
<ij>
If something's in my nix-channel --list, should it also appear in $NIX_PATH after profile.d/nix.sh evaluation?
cmcdragonkai has quit [(Quit: WeeChat 1.9.1)]
pac has quit [(Ping timeout: 246 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to staging: https://git.io/vbygp
<NixOS_GitHub>
nixpkgs/staging c303047 Vladimír Čunát: fetchurl on Darwin: gssSupport = false (see 9b54a00160)...
NixOS_GitHub has left #nixos []
civodul has joined #nixos
<ij>
samae, How does it fail exactly? You've posted no debug info.
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
b has quit [(Ping timeout: 260 seconds)]
otherjrolfs has quit [(Ping timeout: 264 seconds)]
jrolfs has quit [(Ping timeout: 268 seconds)]
seanparsons has quit [(Ping timeout: 272 seconds)]
laggy_wifi_ has quit [(Remote host closed the connection)]
laggy_wifi_ has joined #nixos
seanparsons has joined #nixos
laggy_wifi_ has quit [(Remote host closed the connection)]
laggy_wifi_ has joined #nixos
<Wizek>
Would someone who can read nix (hopefully better than I do currently) be able to tell me how I could override a non-haskell dependency going through reflex-platform's project development workflow?
otherjrolfs has quit [(Ping timeout: 260 seconds)]
jrolfs has quit [(Ping timeout: 264 seconds)]
proteusguy has quit [(Remote host closed the connection)]
warrickmacmillan has quit [(Quit: warrickmacmillan)]
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<i-am-the-slime>
Probably easier if I just install a VM with ubuntu on it :D
ylwghst has quit [(Ping timeout: 252 seconds)]
otherjrolfs has quit [(Ping timeout: 240 seconds)]
jrolfs has quit [(Ping timeout: 240 seconds)]
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
ssmike has quit [(Ping timeout: 240 seconds)]
<samae>
Is there any way to get the list of IPs assigned to the machines, in the case of virtualbox dep, at build configurations time?
<samae>
in nixops ^^^
Lisanna has joined #nixos
<samae>
I mean, I know an /etc/hosts is generated for every machine
<samae>
But I really need to be able to add "aliases" (and I don't want to set up a full DNS service)
iyzsong has joined #nixos
<Lisanna>
hey, maybe I'm just misunderstanding something here... but I'm trying to get nixpkgs copied somewhere as part of a derivation. but cp ${<nixpkgs>} ... isn't copying nixpkgs, instead it's copying a bunch of stuff that would be in a /nix/store (derivations and products). I thought <nixpkgs> evaluates to a location containing the channel files (default.nix, etc.)?
<tilpner>
Lisanna - ${/any/path} copies that path into the store
jtojnar has quit [(Read error: Connection reset by peer)]
jtojnar has joined #nixos
<Lisanna>
tilpner: yeah, I figured as much, but I would like it to always reliably be whatever nixpkgs is set to in the NIX_PATH of the derivation itself
<tilpner>
Lisanna - So ${<nixpkgs>} first copies <nixpkgs> into the store, then calls cp /nix/store/.../ ...
<Lisanna>
right, so <nixpkgs> goes into the store as /nix/store/foo, and then cp is run on /nix/store/foo, so shouldn't that still just copy ./default.nix etc.?
<Lisanna>
not sure why I'm ending up with *the entire store* ):
<tilpner>
I'll need commands and output to answer that properly
<Lisanna>
huh, okay
<tilpner>
It's not clear what you mean by entire store
<tilpner>
It should copy more than just default.nix
<tilpner>
But it should copy all of /nix/store into /nix/store/...
<tilpner>
*But it shouldn't copy all of /nix/store into /nix/store/...
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
<Lisanna>
okay, so if I put a builtins.toString in front of it, it resolves to /nix/var/nix/profiles/per-user/root/channels/nixos/nixpkgs/
<Lisanna>
which, when I ls, has the Nix expressions in it that I'm expecting
<Lisanna>
but, it's not copied into the /nix store, since it's just a string now :/
<Lisanna>
so, those commands are failing now because it's giving an absolute path on my system, which is outside the sandbox (let alone the virtual machine), so it can't access it like that
<tilpner>
Ahh, that's what you're trying to do. Can you revert to "${<nixpkgs>}", and describe what's bad about that, again?
<Lisanna>
sure
<tilpner>
If I understood your correctly, you want it to copy to the store, and then cp /nix/store/... $mountPoint/nixpkgs, which this should do
otherjrolfs has quit [(Ping timeout: 256 seconds)]
<Lisanna>
when I revert it to ${<nixpkgs>}, the result I get is:
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] markus1189 opened pull request #32932: bepasty: Add `checkInput` for tests (master...bepasty-fix) https://git.io/vbyMA
NixOS_GitHub has left #nixos []
<Lisanna>
"/nix/store/k8g2ma0sg58mzfjr1my8hys8gm374yqv-nixpkgs" gets printed (From the echo)
<Lisanna>
then, the ls prints: "02xjvflg03d9wlzdw1ig54rgwa32jwxq-findutils-4.6.0\n0hwra7yv9rdmg32slmk463a3l25zp36g-pcre-8.41\n0ns4r436sp18fgjf3njp7965j44z327y-paxctl-0.9..."
<tilpner>
Okay, just to check: ls -la $(nix-instantiate --find-file nixpkgs) returns the expected default.nix?
<Lisanna>
just on my regular shell?
<tilpner>
Yes
davidak has joined #nixos
dan_b` has joined #nixos
laggy_wifi_ has quit [(Remote host closed the connection)]
<Lisanna>
the result is "/nix/var/nix/profiles/per-user/root/channels/nixos/nixpkgs -> ."
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] andir opened pull request #32936: gopass: init at 1.6.6 (master...add-gopass) https://git.io/vbyDR
NixOS_GitHub has left #nixos []
leat has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
<Lisanna>
'/nix/var/nix/profiles/per-user/root/channels/nixos/nixpkgs' is what I expect ${<nixpkgs>} to put into the store for me
<Lisanna>
and it's what ${builtins.toString <nixpkgs>} evaluates to
laggy_wifi_ has joined #nixos
jrolfs has joined #nixos
<Lisanna>
maybe ${builtins.toPath (builtins.toString <nixpkgs>)} would work
<Lisanna>
but that looks silly
laggy_wifi_ has quit [(Remote host closed the connection)]
jrolfs has quit [(Ping timeout: 240 seconds)]
laggy_wifi_ has joined #nixos
<Lisanna>
and here's the weird part... ls ${<nixpkgs>} has extra stuff in it that is *not* in nixpkgs, like, it has custom stuff that I've built
<Lisanna>
it's like ${<nixpkgs>} is instead evaluating to /nix/store
<tilpner>
But you already tested against that: "/nix/store/k8g2ma0sg58mzfjr1my8hys8gm374yqv-nixpkgs" gets printed (From the echo)
slyfox has quit [(Ping timeout: 268 seconds)]
<Lisanna>
right, /nix/store/k8g2ma0sg58mzfjr1my8hys8gm374yqv-nixpkgs contains thigns that only /nix/store can contain
<tilpner>
I guess I'll have to try this locally
<Lisanna>
stuff from nixpkgs and stuff I've build with nix-build
slyfox has joined #nixos
<Lisanna>
toPath toString <nixpkgs> doesn't work... I guess there's a difference between path values and "path values" (it's not being copied into the store ):)
laggy_wifi_ has quit [(Remote host closed the connection)]
laggy_wifi_ has joined #nixos
<tilpner>
Lisanna - What's a valid value for "partition"?
otherjrolfs has joined #nixos
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
<Lisanna>
so, that function boots an existing disk image that you pass in as srcImg
<Lisanna>
partition is which partition of srcImg you want to mount
<tilpner>
Actually, nevermind, we can probably reduce this
<Lisanna>
yeah
<Lisanna>
shouldn't require runInLinuxVM to reproduce
Mic92 has quit [(Read error: Connection reset by peer)]
<Lisanna>
just open nix-repl
<Lisanna>
':l <nixpkgs>'
<Lisanna>
'a = "${<nixpkgs>}"'
<Lisanna>
'a'
<Lisanna>
then take the result, go to your regular shell, and run "ls" on it
<Lisanna>
in my case, it prints the entire universe and then some
<gchristensen>
PlasmaStar: you mentioned something w.r.t. censorship, what did you mean?
simukis has joined #nixos
<tilpner>
Lisanna - I put this into lisanna.nix: with import <nixpkgs> {}; writeScript "foo" ''cp -a ${<nixpkgs>} /tmp/nixpkgs''
<Lisanna>
I'd just start with an echo first..
<Lisanna>
or an ls
<tilpner>
Lisanna - Then $(nib /tmp/lisanna.nix) and ls /tmp/nixpkgs shows the content of my <nixpkgs>
otherjrolfs has quit [(Ping timeout: 240 seconds)]
<Lisanna>
as in, default.nix etc.?
<tilpner>
Yes
<Lisanna>
:|
<tilpner>
What happens for you?
<Lisanna>
I get the same results you do
<Lisanna>
wtf
<Lisanna>
try the nix-repl thing
<tilpner>
As a last resort, try replacing ${<nixpkgs>} with ${path}
<Lisanna>
yeah.... trying hard to avoid that
i-am-the-slime has quit [(Ping timeout: 256 seconds)]
erictapen has quit [(Ping timeout: 260 seconds)]
<tilpner>
Lisanna - Is it that bad? I mean literally path, this is not a placeholder
<Lisanna>
err, what is ${path} ?
pie__ has quit [(Remote host closed the connection)]
<Lisanna>
oh... is it a special symbol?
pie__ has joined #nixos
<tilpner>
That path provided by nixpkgs
<tilpner>
*+is
<Lisanna>
nix-repl> "${path}"
<Lisanna>
this hangs
<tilpner>
Give it a few seconds
i-am-the-slime has joined #nixos
<Lisanna>
okay, it finished
<Lisanna>
okay, will that always be nixpkgs?
<tilpner>
It will always be nixpkgs. It will not always be <nixpkgs>
<tilpner>
That's an advantage IMO
<Lisanna>
huh, okay
<Lisanna>
yeah, I guess so
<Lisanna>
...where does "path" come from though?
<tilpner>
In my setup, find $(nix-instantiate --find-file nixpkgs) | wc -l => 4
mizu_no_oto has quit [(Quit: ["Textual IRC Client: www.textualapp.com"])]
jrolfs has quit [(Ping timeout: 256 seconds)]
otherjrolfs has quit [(Ping timeout: 272 seconds)]
Mic92 has joined #nixos
<Lisanna>
tilpner: thanks for the tip about pkgs.path, that seems to be exactly what I needed.
<tilpner>
Well, it's not quite the same, but it should work for the traditional channel setup
<Lisanna>
so, you said that it's vanilla nixpkgs, not *my* nixpkgs?
<Lisanna>
so, one potential thing I need to avoid is having a nixpkgs installed on the disk image that's different from the one being used to build the derivation that's running the runInLinuxVM in the first place
<Lisanna>
actually, that shouldn't matter
erictapen has joined #nixos
erictapen has quit [(Remote host closed the connection)]
erictapen has joined #nixos
<Lisanna>
since users on the disk image won't be issuing nix-build on products of my derivation... I always use direct paths to things I've installed in the disk image /nix/store. Adding nixpkgs is more of a formality at this point, it won't be used for anything
<Lisanna>
but, nix-shell requires .nix-defexpr to exist, and nix-channel needs a channel before it creates that dir, so have to have one :/
sary has joined #nixos
ssmike has joined #nixos
the has quit [(Remote host closed the connection)]
pie__ has quit [(Ping timeout: 250 seconds)]
<Lisanna>
okay, dumb question: how do I then proceed to add that $mountpoint/nixpkgs as a channel? it says it needs to be a URL, and /nixpkgs isn't a URL.
FRidh has quit [(Quit: Konversation terminated!)]
<Lisanna>
the man page isn't too helpful :/
FRidh has joined #nixos
<tilpner>
Are you sure nix-shell uses defexpr? I was under the impression only nix-env needs that
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
<tilpner>
And you don't *need* to use channels at all, I'm Fine™ with just NIX_PATH
<Lisanna>
oh, I'm dumb
<Lisanna>
I was trying to import <nixpkgs> in a nix expression on the disk image
<Lisanna>
so it was trying to find nixpkgs
<srhb>
The hydra manual doesn't mention anything about creating channels. Is there a nice example somewhere?
<Lisanna>
ugh, this is tricky
<Lisanna>
I need buildEnv, but buildEnv is a functor... I can't just copy it into the image's /nix/store and have it be available...
<Lisanna>
srhb: are you trying to create a channel with hydra specifically?
<tilpner>
Lisanna - If you want to save image space (about 110M), you can store nixpkgs in compressed form
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<tilpner>
But I'm not entirely sure what you're doing, so...
<Lisanna>
So, I do actually need nixpkgs on the disk image, so I can use buildEnv
<Lisanna>
so, I guess I can just use a raw NIX_PATH
<Lisanna>
and set that manually
<Lisanna>
it's too bad that nix-channel apparently doesn't support using local files though
<nikivi>
I assume it being a sandbox build enviroment is something you want to have on a server rather than a personal OS
FRidh has joined #nixos
<Lisanna>
tbh I don't use Nix on MacOS for a lot, since I spend most of my time on a linux box, but I do use it for a few things, like when I get frustrated at the way Mac packages software...
<Lisanna>
and I just want a simple program and don't want to spend the time to fix my broken homebrew installation :)
<nikivi>
You ssh to linux from mac you mean?
<gchristensen>
nikivi: I use it mostly because nix-shell is magic
<Lisanna>
nikivi: yeah
<Lisanna>
tilpner: no dice. error: unable to download ‘file:///nixpkgs/MANIFEST’: Couldn't read a file:// file (37)
<nikivi>
gchristensen That's the thing. I wish to see this magic first hand instead of reading about it.
<gchristensen>
try it :) it works great on mac
<nikivi>
But I am worried to mess some things as I am already running a package manager with many packages
<Lisanna>
nikivi: Nix doesn't interfere with your other package managers
<Lisanna>
that's one of its selling points :)
<gchristensen>
nix will not interfere with brew
<Lisanna>
it's fully isolated
<gchristensen>
see: reasons you want sandboxing even on your laptop, and you don't even know it :)
mbrgm has quit [(Quit: ZNC 1.6.5+deb1 - http://znc.in)]
<nikivi>
Also it seems nix is really tailored to systems that have more than 1 user
<nikivi>
But on mac I just have 1 user and never switch
laggy_wifi_ has quit [(Remote host closed the connection)]
<gchristensen>
it isn't, the "multi-user" terminology you see is a bad name
<Lisanna>
nikivi: not really, I have many single user systems I use Nix on
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<gchristensen>
Nix's design philosophy just happens to make many user systems much easier
<srhb>
gchristensen: Hm, thanks.. It's not obvious to me how these integrate into the hydra channels tab, but I'll try to figure it out :)
<gchristensen>
oh the channels tab
<gchristensen>
I have no idea
<gchristensen>
:$
<srhb>
Ah... :-P
pie__ has joined #nixos
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<srhb>
I should have been clearer.
spietz has quit [(Ping timeout: 252 seconds)]
{^_^} has quit [(Remote host closed the connection)]
{^_^} has joined #nixos
{^_^} has quit [(Changing host)]
{^_^} has joined #nixos
vidbina has joined #nixos
<infinisil>
Hmm, can somebody merge this maybe? I'm not able to make it user+group configurable and there's little reason for it. https://github.com/NixOS/nixpkgs/pull/30518 cc gchristensen Dezgeg peterhoeg
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] CMCDragonkai opened pull request #32939: nodePackages.dat init at 13.9.2 (master...dat) https://git.io/vby7H
NixOS_GitHub has left #nixos []
pie__ has quit [(Ping timeout: 256 seconds)]
<infinisil>
Has been left standing for a while now and the service has been really useful for me and has never caused problems
{-_-} has joined #nixos
<{-_-}>
{^_^}
jrolfs has quit [(Ping timeout: 240 seconds)]
otherjrolfs has quit [(Ping timeout: 260 seconds)]
<{^_^}>
{-_-}
ma27 has joined #nixos
{o_o} has joined #nixos
<{o_o}>
{-_-}
{^_^} has quit [(Remote host closed the connection)]
<NixOS_GitHub>
[nixpkgs] CMCDragonkai opened pull request #32940: nodePackages: Added instructions to README about packages with binary addons (master...nodepackagesreadme) https://git.io/vby5E
NixOS_GitHub has left #nixos []
<hyper_ch>
infinisil: is your "s" key broken?
<infinisil>
Heh, because zetup?
<hyper_ch>
yes
<hyper_ch>
and ho do you send it to a remote pool?
<grahamc>
Nothing like a forced reboot by nixops and a fat finger to spoil a morning
<hyper_ch>
how
<infinisil>
Well I thought I'd keep the scheme of znapzend, the setup binary is called znapzendzetup
jensens has quit [(Remote host closed the connection)]
<hyper_ch>
ah ok :)
gchristensen has quit [(Ping timeout: 240 seconds)]
<infinisil>
and pandoc's markdown is probably the best intergrated in pandoc, so it's gonna be easy to generate correct other formats (pdf/html/man)
<srhb>
gchristensen: Ouch
<roberth>
The primary goal of reStructuredText is to define a markup syntax for use in Python docstrings and other documentation domains
<roberth>
I'm not much of a snake person though
<gchristensen>
roberth: you really stepped in it, asking questions about docs ;)
{^_^} has joined #nixos
{^_^} has quit [(Changing host)]
{^_^} has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
<roberth>
haha yeah, I guess I was trying to find the word pythonista but that's a good point ;)
<infinisil>
Well, just having written a big module yesterday and not knowing how to write the options descriptions with proper xml attributes, I'm really for everything except xml (I know this discussion is about Nix comments, not option descriptions but still)
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
<gchristensen>
yeah I guess it is unreasonable to learn a few small bits of simple XML
<infinisil>
gchristensen: how do you do a man page entry in docbook xml?
<gchristensen>
jtojnar: or is an operator which works when fetching an attr from an attrset
<gchristensen>
jtojnar: {}.foo or "bar" -> returns bar, {foo = "hi";}.foo or "bar" -> returns "hi"
<jtojnar>
oh
<gchristensen>
I've been punished too many times by the vague parsing of markdown, the differently applied rules between engines, and the lack of decent semantics
FRidh has quit [(Quit: Konversation terminated!)]
jrolfs has quit [(Ping timeout: 264 seconds)]
FRidh has joined #nixos
otherjrolfs has quit [(Ping timeout: 268 seconds)]
<NixOS_GitHub>
[nixpkgs] svanderburg closed pull request #32940: nodePackages: Added instructions to README about packages with binary addons (master...nodepackagesreadme) https://git.io/vby5E
NixOS_GitHub has left #nixos []
otherjrolfs has quit [(Ping timeout: 248 seconds)]
jrolfs has quit [(Ping timeout: 260 seconds)]
<joepie91>
Lisanna: probably doesn't help that there's no language spec :P
<gchristensen>
it is very difficult to get right though, because it is all based on regexes
<joepie91>
gchristensen: it's probably theoretically fixable based on my past experiences with textmate bundles
<joepie91>
it'll just be a lot of hacky experimentation
<yegortimoshenko>
gchristensen: i agree, i think markdown is as much of a format as csv. i use org-mode whenever i can, but XML docs are also fine
<joepie91>
because there's basically zero debugging tools for this stuff
proteusguy has joined #nixos
pkill9 has joined #nixos
<joepie91>
might look into it when bored
<gchristensen>
(ノ◕ヮ◕)ノ*:・゚✧ yegortimoshenko
<gchristensen>
it is unpopular to like docbook-based docs in the nix community :D
FRidh has quit [(Remote host closed the connection)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 6 new commits to master: https://git.io/vbyAa
<NixOS_GitHub>
nixpkgs/master 5b5e98b Lorenzo Manacorda: libbitcoin: 2.11.0 -> 3.4.0
<NixOS_GitHub>
nixpkgs/master 4fea55e Lorenzo Manacorda: libbitcoin-protocol: init at 3.4.0
<NixOS_GitHub>
nixpkgs/master 86ee454 Lorenzo Manacorda: libbitcoin-client: 2.2.0 -> 3.4.0
NixOS_GitHub has left #nixos []
FRidh has joined #nixos
<yegortimoshenko>
gchristensen: i don't see why XML gets so much hate when HTML is relatively tolerated
<roberth>
gchristensen: interesting experiment, but it involves a lot of mechanics that make it hard to use everywhere
<gchristensen>
IMO? fear, and it is a bit tricky to get the editing environment setup, and the error messages can be unparseable
<roberth>
documentation should be easy to create
<gchristensen>
roberth: I agree, I was trying to avoid actually embedding docbook direcly in strings, I wouldn't try that again
<sphalerite>
I can't seem to build Rust stuff using cargo anymore, the cc wrapper (I presume) complains about an impure path being used in the build
<sphalerite>
Is there a way I can disable that check? It used to work
<Lisanna>
yegortimoshenko: yeah, the correct attitude is to hate both ^^
<joepie91>
yegortimoshenko: depends on context. XHTML was poorly received compared to HTML due to strictness and unforgiving failure modes, XML is generally poorly received as a serialization format because it's unnecessarily complex and doesn't directly map to a native data type in all languages (unlike JSON, which does), and many XML-based things are poorly received because they tend to over-specify things and require a lot of work to specify or handle common
<joepie91>
cases even if you never use the corner cases
<yegortimoshenko>
Lisanna: i disagree, i think HTML/XML is a very good idea, i would only be happier if it used s-expressions, but syntax is tolerable
davidak has joined #nixos
<joepie91>
yegortimoshenko: most of it, in the end, comes down to people feeling that XML-y things are unnecessarily complex for their usecase
<joepie91>
which, in many cases, is a valid concern
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<yegortimoshenko>
joepie91: XML is good for representing trees, it's just people tried to represent everything in XML
<yegortimoshenko>
sometimes map
<infinisil>
it's just annoying to write in a lot of editors, and I won't switch editors just so i can write docs easier
<yegortimoshenko>
sometimes maps are a better idea
<joepie91>
yegortimoshenko: trees of what, though? you can represent trees using basically anything that can represent objects/maps and lists, including JSON
<Lisanna>
there's at least one other way to express that that I'm forgetting :)
<yegortimoshenko>
joepie91: XML is of course better than JSON for representing trees, it properly supports metadata (attributes) on nodes, and representing trees in JSON is very verbose
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej closed pull request #32922: libsoundio: Fix build on darwin (master...feature/libsoundio-darwin) https://git.io/vbyWM
NixOS_GitHub has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 2 new commits to master: https://git.io/vbyxW
<NixOS_GitHub>
nixpkgs/master 644a404 Mario Rodas: libsoundio: Fix build on darwin
<joepie91>
yegortimoshenko: but this assumes that you *need* to specify metadata on nodes, and that it's a usecase where that's easier to deal with than objects that have a `children` attribute with an array of children as its value, which is usually not the case
<joepie91>
so no, XML is not universally 'better than JSON for representing trees'
<joepie91>
it depends entirely on the usecase
<Lisanna>
FRidh: I have no idea what on earth is in that paste, but thanks for the bookmark :)
<dan_b`>
I'm thinking of buying a bike shed, can anyone tell me what the best colour is, please?
<joepie91>
Lisanna: looks like strace output :P
<joepie91>
dan_b`: purple, clearly :)
otherjrolfs has quit [(Ping timeout: 252 seconds)]
<Lisanna>
so I've been thinking of switching to emacs from vim... pros/cons?
<joepie91>
anyway, not really bikeshedding here, just explaining why people dislike XML
<gchristensen>
dan_b`: docbook-flavoured
jrolfs has quit [(Ping timeout: 272 seconds)]
<gchristensen>
Lisanna: #emacs has a good sales pitch
<Lisanna>
gchristensen: I was joking :)
<gchristensen>
Lisanna: they are too
<yegortimoshenko>
gchristensen: hehehe
<Lisanna>
:)
<gchristensen>
their sales pitch is "We aren't gonna lie. Emacs sucks. Some of us tolerate it, but we can't tell you if YOU'LL be able to. Try it and make up your own mind."
acarrico has joined #nixos
<etu>
Lisanna: I ride the emacs boat. But personally I couldn't care less what others use as long it works for them. But feel free to ask me questions if you have any.
<Lisanna>
bad joke... I felt the XML and JSON thing was a vim vs. emacs thing
<joepie91>
yegortimoshenko: anyway, to complete what I was saying above because I got distracted halfway through: the problem with having a 'native' distinction between attributes and children in your serialization format, is that most languages have no way to represent this as data structures; most languages rely entirely on some variant of 'maps and lists' for dealing with serialized data... meaning that even if you *do* have a native distinction like in XML,
<joepie91>
the code that the developer writes is going to have to access a .children property or something like that anyway, because that's the only way most languages can represent that distinction
<joepie91>
so the effective representation that the developer works with is going to be pretty much the same as how you'd represent the tree in JSON anyway
<joepie91>
and that's too costly a translation step for many developers
<joepie91>
(mental translation step, that is)
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
ma27 has quit [(Ping timeout: 252 seconds)]
<yegortimoshenko>
joepie91: that i agree with, but XML is supposed to be parsed into custom data structures, and not in lists/maps like json.
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<joepie91>
sure, but 1) that's not how people usually handle deserialized data in reality, and 2) in the cases where they do, they often end up having to write some sort of logic of configuration anyway that has to work around this lack of distinction in their language of choice
<joepie91>
specifying mappings from XML to custom data structures suffers from the exact same problem of having no obvious language-native distinction between children and attributes
<joepie91>
(on top of that, a lot of developers work with languages where there's not such a strong focus on 'custom data structures'-)
<gchristensen>
let's take this to #nixos-debates-xml
<yegortimoshenko>
joepie91: i agree, it's problematic.
jrolfs has quit [(Ping timeout: 240 seconds)]
* joepie91
has to leave for a bit anyway
otherjrolfs has quit [(Ping timeout: 240 seconds)]
orivej has quit [(Ping timeout: 260 seconds)]
szicari has joined #nixos
ma27 has joined #nixos
stranger__ has joined #nixos
<dan_b`>
dezgeg: "it might be that cross stripping is not being done (or implemented incorrectly)"
<stranger__>
hey, is it possible to print meta information about package?
<dan_b`>
this turns out to have been the case
ssmike has quit [(Remote host closed the connection)]
<dan_b`>
at least, as of master when I started this branch
ssmike has joined #nixos
mizu_no_oto has joined #nixos
<dan_b`>
it's runningf 'strip' not 'mipsel-unknown-linux-gnu-strip' or whatever
FRidh has quit [(Quit: Konversation terminated!)]
<dan_b`>
(and then it's suppressing the error messages, but I understand the reasoning there)
<roberth>
stranger__: use nix-repl or nix-instantiate --eval --strict -E '(import <nixpkgs> {}).hello.meta'
FRidh has joined #nixos
<stranger__>
roberth: thank you!
<roberth>
stranger__: you're welcome
otherjrolfs has joined #nixos
jrolfs has joined #nixos
mizu_no_oto has quit [(Client Quit)]
NightTrain has joined #nixos
stranger__ is now known as stranger
<woffs>
nixos on sparc, anyone? ;)
freusque has joined #nixos
stranger is now known as Guest56528
<Guest56528>
woffs: not earlier than nixos on russian "elbrus" i suppose :D
hellrazo1 has quit [(Ping timeout: 248 seconds)]
Guest56528 is now known as stranger_
otherjrolfs has quit [(Ping timeout: 260 seconds)]
jrolfs has quit [(Ping timeout: 240 seconds)]
stranger_ is now known as stranger__
stranger__ is now known as stranger___
NightTrain has quit [(Quit: leaving)]
NightTrain has joined #nixos
ma27 has quit [(Ping timeout: 252 seconds)]
FRidh has quit [(Quit: Konversation terminated!)]
sary has quit [(Ping timeout: 248 seconds)]
FRidh has joined #nixos
mizu_no_oto has joined #nixos
iyzsong has quit [(Ping timeout: 252 seconds)]
jrolfs has joined #nixos
otherjrolfs has joined #nixos
berce has joined #nixos
<berce>
ping jtojnar
<PlasmaStar>
gchristensen: You banning on certain words, then immediately regretting it. lol
<gchristensen>
oh, PlasmaStar, it is because someone tested to see if the bot worked :) I wasn't trying to ban people, just spammers.
<PlasmaStar>
Ok :)
<gchristensen>
(and the bot did indeed work, lol)
jrolfs has quit [(Ping timeout: 260 seconds)]
otherjrolfs has quit [(Ping timeout: 272 seconds)]
otherjrolfs has quit [(Ping timeout: 248 seconds)]
mizu_no_oto has joined #nixos
nevermind has joined #nixos
nevermind is now known as Guest93517
<FRidh>
Seems to be something with the systemd unit since starting the daemon manually allows users to build
i-am-the-slime has joined #nixos
otherjrolfs has joined #nixos
i-am-the-slime has quit [(Ping timeout: 240 seconds)]
otherjrolfs has quit [(Ping timeout: 268 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] svanderburg pushed 1 new commit to master: https://git.io/vbSkE
<NixOS_GitHub>
nixpkgs/master 065acc4 Sander van der Burg: jsdoc: init at 3.5.5
NixOS_GitHub has left #nixos []
FRidh has quit [(Quit: Konversation terminated!)]
alexteves has joined #nixos
pkill9 has quit [(Ping timeout: 246 seconds)]
FRidh has joined #nixos
<dan_b`>
given a default.nix file, is there any command that will tell me which of its dependencies (recursively) have already been built and which ones will need to be built?
<dan_b`>
without actually buildng it, I mean
<dan_b`>
oh, cock. my git bisect failed. i forget that gnu hello has a 'make check' target that doesn't work when xbuilding
jrolfs has joined #nixos
otherjrolfs has joined #nixos
i-am-the-slime has joined #nixos
pkill9 has joined #nixos
jrolfs has quit [(Ping timeout: 252 seconds)]
otherjrolfs has quit [(Ping timeout: 252 seconds)]
leat has joined #nixos
davidak has quit [(Remote host closed the connection)]
davidak has joined #nixos
mizu_no_oto has quit [(Quit: Computer has gone to sleep.)]
ssmike has quit [(Remote host closed the connection)]
ssmike has joined #nixos
hke has quit [(Ping timeout: 248 seconds)]
otherjrolfs has quit [(Ping timeout: 264 seconds)]
jrolfs has quit [(Ping timeout: 272 seconds)]
Guest93517 has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
nevermin_ has joined #nixos
hke has joined #nixos
hellrazo1 has quit [(Ping timeout: 252 seconds)]
otherjrolfs has joined #nixos
i-am-the-slime has quit [(Ping timeout: 240 seconds)]
i-am-the-slime has joined #nixos
drakonis has joined #nixos
otherjrolfs has quit [(Ping timeout: 240 seconds)]
ssmike has quit [(Remote host closed the connection)]
ssmike has joined #nixos
<pkill9>
does libreoffice work for anyone else?
chreekat has joined #nixos
<woffs>
pkill9: I remember having it opened some days ago on 17.09
<pkill9>
nevermind, got it to work
<samueldr>
libreoffice-5.3.6.1 does for me
FRidh has quit [(Quit: Konversation terminated!)]
<pkill9>
i copy system libs over to a directory and use patchelf to make them look in my system paths, then set LD_LIBRARY_PATH to the directory containing those patched libs
FRidh has joined #nixos
<pkill9>
seems to work so far
jrolfs has joined #nixos
otherjrolfs has joined #nixos
ssmike has quit [(Remote host closed the connection)]
ssmike has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] PierreR opened pull request #32942: HaskellPackages.language-puppet: enable x86_64 build again (master...master) https://git.io/vbSOU
NixOS_GitHub has left #nixos []
<chreekat>
The nixpkgs manual has a section on how to build haskell packages with profiling enabled. It creates a duplicate of haskellPackages with an overridden mkDerivation. How would I create a copy of haskellPackages that only overrides mkDerivation for a single package?
dan_b has quit [(Ping timeout: 248 seconds)]
<chreekat>
There's so many 'self's and 'super's and other repeated words in the examples that it's hard to make heads or tails of it
<srhb>
chreekat: The example works by overriding mkDerivation to enable profiling for all its arguments
<srhb>
chreekat: Instead, override just the haskell package you want to enable profiling for
butchery has quit [(Quit: leaving)]
otherjrolfs has quit [(Ping timeout: 240 seconds)]
jrolfs has quit [(Ping timeout: 240 seconds)]
ssmike has quit [(Remote host closed the connection)]
ssmike has joined #nixos
<chreekat>
I think I need to reread section 7.1...
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<srhb>
chreekat: What do you really want to do? Have a separate set with some libraries profiled or override haskellPackages itself with some libraries being profiled?
i-am-the-slime has quit [(Quit: Konversation terminated!)]
ottidmes has joined #nixos
NightTrain has quit [(Quit: leaving)]
<chreekat>
srhb: great question. I just tried installing brittany from unstable, but one of its deps (multistate) is failing its tests atm. I thought I'd just try setting noCheck on that package and see where I end up
NightTrain has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
FRidh has quit [(Client Quit)]
FRidh has joined #nixos
otherjrolfs has quit [(Ping timeout: 248 seconds)]
<srhb>
chreekat: self is the finished packageset (the fixpoint), super is the packageset *before* this overlay, selfHS and superHS likewise but just for haskellPackages, oldAttrs are the attributes of multistate before this overlay
zzamboni has quit [(Quit: Leaving.)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] zimbatm closed pull request #32285: Make less known wayland compositors usable (master...wayland-wm) https://git.io/vbOVW
NixOS_GitHub has left #nixos []
<srhb>
chreekat: Does that help?
<chreekat>
yes, thanks! "self is the fixpoint" in particular. Let me give it a spin - I need to make sure that multistate is really the problem; there might be two
hotfuzz has joined #nixos
<srhb>
You can add another override in the same definition of haskellPackages if you want. :)
* chreekat
nods
ThatDocsLady has quit [(Quit: Leaving)]
FRidh has quit [(Remote host closed the connection)]
hotfuzz_ has quit [(Ping timeout: 252 seconds)]
FRidh has joined #nixos
<chreekat>
bummer, looks like the test module is still being built even without doCheck. I can think of various reasons for that. I think I'll just fall back to building brittany locally
<chreekat>
/manually
drakonis has quit [(Ping timeout: 252 seconds)]
<srhb>
Oh, it fails building, not running?
<chreekat>
yeah, which is a point I probably should have mentioned
<chreekat>
heh
<hexagoxel>
chreekat: btw, how does the multistate test fail?
<chreekat>
hexagoxel: i misspoke. I think it's the monad-memo
<chreekat>
package
<hexagoxel>
oh well, i have no control over that :)
<chreekat>
:)
<Fare>
Hi. Given a NixOS partition mounted from Ubuntu, what is the "right" way to chroot into it?
<clever_>
Fare: there ix a nixos-install --chroot, but its only present on another nixos
clever_ is now known as clever
clever has quit [(Changing host)]
clever has joined #nixos
<Fare>
clever, yes, and trying to run the nixos-install that is already in the chroot doesn't obviously work; I'm stuck where it tries to call nix-instantiate.
<Fare>
But I'm trying to define PATH, etc.
endformationage has joined #nixos
<clever>
Fare: you need a tmpfs at /run and a symlink from /run/current-system to the latest build of nixos in the store
<clever>
then /run/current-system/sw/bin/ in $PATH
<Fare>
/run is there already, no tmpfs
lonokhov has quit [(Quit: leaving)]
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<Fare>
I'm trying to bootstrap NixOS from a machine that cannot boot NixOS, yet, at least not in a fully functional way
civodul has quit [(Remote host closed the connection)]
nevermin_ has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
dan_b has joined #nixos
<chreekat>
srhb: cool - i had a followup question, but learning about 'extend' let me answer it on my own. I wanted to try creating a *copy* of haskellPackages rather than overriding it 'in place'. I managed to do that by setting '~/.config/nixpkgs/config.nix'.packageOverrides.myHSCopy = self.haskellPackages.extend (...)
<chreekat>
(well, packageOverrides is a function, but hopefully you see what I mean)
<srhb>
chreekat: With the overlay it'd be as simple as renaming the lefthand side
<srhb>
ie myHaskellPackages instead of haskellPackages
jrolfs has quit [(Ping timeout: 240 seconds)]
otherjrolfs has quit [(Ping timeout: 252 seconds)]
<srhb>
(It's not that different from the packageOverrides version, but overlays seem nicer)
<chreekat>
oh! ok. That's why I described my soln: in case there was a nicer way of doing it
<chreekat>
thanks again
<srhb>
:)
stranger___ has quit [(Quit: Connection closed for inactivity)]
orivej has joined #nixos
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
jb55 has quit [(Ping timeout: 264 seconds)]
ssmike has quit [(Remote host closed the connection)]
ssmike has quit [(Remote host closed the connection)]
otherjrolfs has quit [(Ping timeout: 256 seconds)]
jrolfs has quit [(Ping timeout: 272 seconds)]
<pmeunier>
Hi! I get the following error when trying to deploy to GCE with NixOps: ERROR: google_compute_engine.test_compat (unittest.loader.ModuleImportFailure)
<pmeunier>
ImportError: No module named mock
<pmeunier>
Any idea?
FRidh has quit [(Remote host closed the connection)]
<pmeunier>
(I'm using the Git version of nixpkgs and nixops)
<srhb>
pmeunier: Which git versions?
FRidh has joined #nixos
FRidh has quit [(Client Quit)]
FRidh has joined #nixos
<pmeunier>
srhb: from the git repository (not sure I understand the question)
<srhb>
pmeunier: which branch and commit :)
<srhb>
There are multiple branches.
<pmeunier>
master, latest
<pmeunier>
(for both)
<srhb>
Brave!
<pmeunier>
well, I need features that are super recent in nixpkgs (such as buildRustCrate)
jcb__ has quit [(Quit: jcb__)]
<srhb>
pmeunier: Perhaps it needs to be added to the pythonPath of nixops?
<iqubic>
What is the url for unstable?
<srhb>
(I'm not very good at Python, so take with a grain of salt)
nevermind has joined #nixos
<iqubic>
And then, how do I tell nix that I want to take stack from unstable?
nevermind is now known as Guest19244
<srhb>
iqubic: Try searching for nixos channels, you should find it quite quickly :)
frontsideair has joined #nixos
<srhb>
iqubic: The name you give the channel will be it's namespace in nix-env commands
<srhb>
iqubic: As in, nix-env -iA nixos-unstable.stack
<frontsideair>
Hi everyone, I have a question. I have installed yarn, a node package while node 6.x was installed. Now I want it to use node 9.x but I don't know how to tell nix-env that. I have removed both node 6.x and yarn but since my env depends on them they aren't garbage collected.
<srhb>
frontsideair: How does your env depend on them?
<frontsideair>
My older env, I mean.
<srhb>
frontsideair: If they're not installed in your current env, they will be garbage collceted.
Guest19244 has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<iqubic>
So I added the channel nixos-unstable to my system
asuryawanshi has quit [(Ping timeout: 272 seconds)]
<iqubic>
However, trying to add nixos-unstable.stack to my list of packages in configuration.nix fails.
<iqubic>
I get told that nixos-unstable is an invalid variable.
otherjrolfs has quit [(Ping timeout: 268 seconds)]
<chreekat>
I'm trying to use 'nix-shell "<nixpkgs>" -A haskellPackages.brittany.env', and it's telling me it can't find the attribute 'brittany'. It works with lens, however. Any pointers?
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
lebel has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] vbgl closed pull request #32758: ocsigen-i18n: do not strip (master...ocsigen-i18n) https://git.io/vbKA6
NixOS_GitHub has left #nixos []
NightTrain has joined #nixos
<lebel>
there's no way right now to have a mdadm daemon running and emailing if there's a raid failure right now on NixOS?
raynold has joined #nixos
<srhb>
chreekat: Why are you trying that though?
<hyper_ch>
mdadm should autosend message to a designated account if there's a failure
<srhb>
I'm not sure why it wouldn't be working. Do you have an override on brittany?
asymmetric has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 1 new commit to master: https://git.io/vbS87
<chreekat>
srhb: I was using the example straight of the manual where it says, "If you need such an environment because you’d like to compile a Hackage package outside of Nix — i.e. because you’re hacking on the latest version from Git..." becasue that's exactly what I'm doing. :)
<iqubic>
no, I'm not using nix-env
<srhb>
iqubic: I linked you a guide to make it available in configuration.nix
<srhb>
chreekat: Aha, OK! You're doing it right!
<chreekat>
I changed a config option from 30 to 0 and it caused a ghc panic - thought I would look at the exact version of things to try to duplicate
<srhb>
chreekat: An "impossible" panic?
<chreekat>
yeah
<srhb>
eek.
<srhb>
But I think you must have a brittany override somewhere that's ruining the env attribute somehow.
asuryawanshi has joined #nixos
jrolfs has quit [(Ping timeout: 252 seconds)]
otherjrolfs has quit [(Ping timeout: 252 seconds)]
<chreekat>
first of all, what does "<nixpkgs>" mean again? Is that going to be pulling from a certain channel? I have two
<srhb>
It grabs the nixpkgs path from $NIX_PATH
<srhb>
try echo $NIX_PATH
<srhb>
Generally any <foo> grabs the value of foo from $NIX_PATH
iqubic has quit [(Ping timeout: 250 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 1 new commit to master: https://git.io/vbS4p
<NixOS_GitHub>
nixpkgs/master a04ae2c Orivej Desh: ocamlPackages.curses: no longer broken
NixOS_GitHub has left #nixos []
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
<chreekat>
Ok. in that case it looks like it's using nixos (per-user/root/channels/nixos)
<srhb>
That is indeed the default on nixos.
<chreekat>
I did make that overlay earlier where I changed the configure flags for monad-memo. But I (hoped I) created a copy of haskellPackages where I did that
galbolle has joined #nixos
<srhb>
Could you disable it and check if you can evaluate brittany.env then?
<chreekat>
oh, brittany may not even BE in nixos stable
<srhb>
Aha!
<chreekat>
So I suppose I need to modify the nix-shell invocation to use my unstable channel
<srhb>
You can just point it at that path instead of <nixpkgs>
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<galbolle>
Hi all, I'm wondering how I can override an emacs package from ~/.config/nixpkgs/config.nix: I'd like to build my emacs using emacsWithPackages, but have it use an overridden version of tuareg
<cmiles74>
I'm working on a systemd service/module that wants to add symlinks to a directory in /etc when it's running. Is there way to have Nix create a symlink during module deployment to satisfy this requirement?
<NixOS_GitHub>
[nixpkgs] FRidh pushed 2 new commits to master: https://git.io/vbSRG
<NixOS_GitHub>
nixpkgs/master 276d60e Markus Hauck: odpdown: add `checkInputs`...
<NixOS_GitHub>
nixpkgs/master b446161 Frederik Rietdijk: Merge pull request #32931 from markus1189/odpdown-fix...
NixOS_GitHub has left #nixos []
<lebel>
srhb: thanks, I'll try that.
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] FRidh pushed 2 new commits to master: https://git.io/vbSR8
<NixOS_GitHub>
nixpkgs/master e1dbe8e Markus Hauck: bepasty: Add `checkInput` for tests...
<NixOS_GitHub>
nixpkgs/master 005628a Frederik Rietdijk: Merge pull request #32932 from markus1189/bepasty-fix...
NixOS_GitHub has left #nixos []
<cmiles74>
This is kind of icky, I see what it's doing. When snapd installs a snap, it tries to create a symlink for systemd to mount that snap during boot.
m0rphism has quit [(Quit: WeeChat 1.9.1)]
otherjrolfs has quit [(Ping timeout: 248 seconds)]
jrolfs has quit [(Ping timeout: 260 seconds)]
<yegortimoshenko>
cmiles74: sorry i never got back to you with snapd
<cmiles74>
yegortimoshenko: That is totally okay. I spent more time on it and I've made some decent progress. :-)
<cmiles74>
yegortimoshenko: At this point, I have it building and creating a package; I have a module added to my config that I can invoke. I'm chasing down runtime issues at this point. :-)
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] P-E-Meunier opened pull request #32944: Fixing tests for Google Compute Engine (master...fix-gce) https://git.io/vbS0W
otherjrolfs has quit [(Ping timeout: 252 seconds)]
MichaelRaskin has joined #nixos
jb55 has joined #nixos
jrolfs has joined #nixos
mizu_no_oto has joined #nixos
jrolfs has quit [(Ping timeout: 252 seconds)]
<hexagoxel>
chreekat: if you succeeded, is it worth adding some instructions to the brittany readme?
<srhb>
I believe it builds with the extra argument to memo, so that should just go in the packageset config.
<srhb>
(ie. please PR :-))
<chreekat>
hexagoxel: I don't think it's brittany-specific. The problem is not nix-specific - monad-memo is failing to build in a certain configuration
warrickmacmillan has joined #nixos
frankpf has joined #nixos
<hexagoxel>
ok
pkill9 has quit [(Ping timeout: 248 seconds)]
<chreekat>
except hang on. how does nixpkgs choose versions of haskell packages? I was vaguely under the impression that it tracks stackage, but perhaps not, eh? Does it do any work to check if packages work together?
<srhb>
chreekat: I believe hackage-packages.nix is loosely based on some LTS snapshot.
<chreekat>
that file says it's autogenerated, but doesn't provide a clue as to how :)
<srhb>
hackage2nix I think?
<chreekat>
sounds promising
<srhb>
Or maybe stackage2nix
<srhb>
Unsure.
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] jwiegley pushed 1 new commit to master: https://git.io/vbSzK
<NixOS_GitHub>
nixpkgs/master 2134543 John Wiegley: coqPackages.contribs: new set with all packages from coq-contribs
NixOS_GitHub has left #nixos []
otherjrolfs has joined #nixos
<srhb>
There-s also a make-package-set.nix
<srhb>
In the same directory
<srhb>
brb
<srhb>
(hopefully...)
srhb has quit [(Quit: Quit)]
jrolfs has joined #nixos
ssmike has quit [(Remote host closed the connection)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] srhb closed pull request #32944: Fixing tests for Google Compute Engine (master...fix-gce) https://git.io/vbS0W
NixOS_GitHub has left #nixos []
srhb has joined #nixos
asymmetric has quit [(Ping timeout: 265 seconds)]
<srhb>
Success.
yegortimoshenko has quit [(Ping timeout: 252 seconds)]
<pkill9>
if anyone is looking to exercise their nix-package-making abilities, then i'd like this to be installed with Nix https://github.com/drawpile/Drawpile
<pkill9>
I would do this myself but I know nothing of building/packaging with Nix ;(
<tnks>
Is the flood of crazy over?
<tnks>
I guess so... sorry for the question, I was still in scrollback.
MercurialAlchemi has quit [(Ping timeout: 265 seconds)]
ssmike has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] vbgl opened pull request #32946: ocaml: adds an argument to enable flambda (master...ocaml-flambda) https://git.io/vbSaL
NixOS_GitHub has left #nixos []
taktoa has quit [(Ping timeout: 252 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 pushed 2 new commits to master: https://git.io/vbSaC
<NixOS_GitHub>
nixpkgs/master cfd4299 Robert Schütz: abcm2ps: 8.13.17 -> 8.13.18
<srhb>
My current workflow consists of xrandr --output X --off --output Y --auto && xrandr --output X --auto --primary
<srhb>
Because... Hmm.
capisce has quit [(Ping timeout: 268 seconds)]
<clever>
srhb: lol
<clever>
the xfce gui is great
pie_ has quit [(Ping timeout: 252 seconds)]
<srhb>
Yeah, I should probably just give in. :P
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] jb55 opened pull request #32948: python-bitcoinlib: init at 0.9.0 (master...python-bitcoinlib) https://git.io/vbSKf
NixOS_GitHub has left #nixos []
<jb55>
autorandr works great
capisce has joined #nixos
Neo-- has quit [(Ping timeout: 252 seconds)]
ethan is now known as ethanwinn
<hyper_ch>
kde tames randr
<jb55>
desktop env... never...
<jb55>
just need xmonad and some glue
<samueldr>
gchristensen: if you'd prefer not to learn xrandr, there's arandr that's pretty standalone, and can save xrandr incantations as scripts
<jb55>
one day clicking on stuff will actually open them...
telent has joined #nixos
<jeaye>
xrandr, in my experience is the most no-nonsense way of handling monitors.
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
<jeaye>
Something will be wrong on my wife's machine, and the UI doesn't show what or provide much info at all. xrandr doesn't lie and it makes things pretty simple, at least for my uses.
<gchristensen>
samueldr: nice!
<gchristensen>
samueldr: I don't think I have enough room in my brain for xrandr :$
<telent>
Gnu hello installs its binary in $out with perms 0555
<jb55>
shut everything down
<telent>
Which seems to mean that strip wont strip it
tmaekawa has joined #nixos
<telent>
Should I fix this by changing the package or would it be better to patch stripDir for all packages?
<telent>
(I don't know if there are other places not being stripped because the stderr of strip is swallowed silently)
tmaekawa has quit [(Client Quit)]
frankpf has quit [(Ping timeout: 252 seconds)]
ilyaigpetrov has quit [(Quit: Connection closed for inactivity)]
alexteves has quit [(Quit: My Mac Mini has gone to sleep. ZZZzzz…)]
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<gchristensen>
telent: all executables are forced to be 0555 with nix
otherjrolfs has quit [(Ping timeout: 252 seconds)]
jb55 has quit [(Ping timeout: 240 seconds)]
<acowley>
I'm trying to package a perl utility, but when I run the script it complains it can't find its own library module. I generated the defintion with nix-generate-from-cpan, and put the definition in perlPackageOverrides.
<acowley>
Can anyone nudge me in the right direction?
<acowley>
(The utility is FileRename, btw)
<acowley>
Oh, and I tried to run the program in a nix-shell -p perlPackages.FileRename
<acowley>
The store path does have a lib directory with what I guess is the module, but @INC doesn't include it, I guess
otherjrolfs has quit [(Ping timeout: 272 seconds)]
FRidh has quit [(Quit: Konversation terminated!)]
ssmike has quit [(Remote host closed the connection)]
FRidh has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
hmensch2 has quit [(Quit: Lost terminal)]
<cmiles74>
Is it possible to specify a runtime dependency? I have squashfsTools as a propagatedBuildInput, but it's still not available when my app runs (my package is compiling from source).
<acowley>
If I set PERL5LIB I can get the script to work, so perhaps my package could do that with its own out directory. Though just dumping that into shellHook did not work.
<acowley>
cmiles74: I think buildInputs should be available at runtime, too
<cmiles74>
acowley: I wonder if the resulting binary doesn't directly use mksquashfs and that's why it can't find it.
d4rk_ has joined #nixos
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<d4rk_>
afternoon everyone -- Quick question here : I'm trying to rebuild but keep getting "The option `services.gitlab.secrets.secret' is used but not defined." Fair deal. But how do I define it exactly? (I'm fairly new to NixOS ^^)
jmeredith has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
<srhb>
d4rk_: You set it in your configuration.nix
FRidh has joined #nixos
ma27 has quit [(Quit: WeeChat 2.0)]
FRidh has quit [(Client Quit)]
<srhb>
d4rk_: As in services.gitlab.secrets.secret = "sortasecret";
FRidh has joined #nixos
<srhb>
d4rk_: The error message may be a bit opaque. What's happening is that you've enabled the module (either directly through enable = true or through some other service) and the module itself utilizes other options that you haven't set yet, so it fails.
jrolfs has quit [(Ping timeout: 268 seconds)]
statusbot has quit [(Ping timeout: 276 seconds)]
otherjrolfs has quit [(Ping timeout: 260 seconds)]
<srhb>
d4rk_: You'll probably bump into that a few times until you have all the required options in place. :)
ma27 has joined #nixos
<d4rk_>
Mkayyyy. I gotchya. We'll see how it goes :)
<acowley>
meanwhile, I regenerated the 2k files I needed to rename while trying to package the rename utility, so now I don't actually need it.
<acowley>
Solid work there.
Rick[29SMPittsbu has joined #nixos
page has quit [(Quit: leaving)]
otherjrolfs has joined #nixos
jrolfs has joined #nixos
<d4rk_>
@srhb: thanks. Just read the gitlab.nix I wasnt sure where I needed to define it. the configuration.nix makes so much sense. I'll set everything up now :)
otherjrolfs has quit [(Ping timeout: 252 seconds)]
frankpf has quit [(Ping timeout: 252 seconds)]
FRidh has quit [(Remote host closed the connection)]
jrolfs has quit [(Ping timeout: 265 seconds)]
FRidh has joined #nixos
telent has quit [(Quit: telent)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Ericson2314 opened pull request #32953: buildPythonPackage: Make setup hook part of nativeBuildInputs (staging...python-dep-fix) https://git.io/vbSDs
<NixOS_GitHub>
[nixpkgs] vcunat opened pull request #32954: glibc: maintenance update and support RHEL 6 -like kernels (staging...p/rhel6) https://git.io/vbSys
NixOS_GitHub has left #nixos []
jrolfs has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
nevermind has joined #nixos
nevermind is now known as Guest56663
verite has joined #nixos
jrolfs has quit [(Ping timeout: 260 seconds)]
otherjrolfs has quit [(Ping timeout: 260 seconds)]
dpren has joined #nixos
civodul has joined #nixos
jb55 has quit [(Ping timeout: 252 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Ericson2314 closed pull request #32953: buildPythonPackage: Make setup hook part of nativeBuildInputs (staging...python-dep-fix) https://git.io/vbSDs
NixOS_GitHub has left #nixos []
Guest56663 has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
otherjrolfs has joined #nixos
ma27 has quit [(Ping timeout: 272 seconds)]
nevermin_ has joined #nixos
FRidh has quit [(Quit: Konversation terminated!)]
otherjrolfs has quit [(Ping timeout: 252 seconds)]
FRidh has joined #nixos
bennofs has joined #nixos
Lisanna has quit [(Ping timeout: 260 seconds)]
<ethanwinn>
Is this still the best approach for installing Inkscape plugins: http://git.io/vT9KZ ?
<Ralith>
ethanwinn: if it were me I'd refactor that package to take a list of plugins as an argument
bennofs has quit [(Remote host closed the connection)]
nevermin_ has quit [(Ping timeout: 265 seconds)]
dan_b has joined #nixos
ottidmes has quit [(Quit: WeeChat 1.9.1)]
jrolfs has joined #nixos
jrolfs has quit [(Ping timeout: 252 seconds)]
<ethanwinn>
Ralith: makes sense, similar to weechat.nix
<ethanwinn>
s/weechat.nix/weechat's default.nix/
FRidh has quit [(Quit: Konversation terminated!)]
yegortimoshenko has joined #nixos
FRidh has joined #nixos
FRidh has quit [(Client Quit)]
FRidh has joined #nixos
<Ralith>
yeah, and many others
zzamboni has quit [(Quit: Leaving.)]
<cmiles74>
acowley: I got it working with makeWrapper and mangling the path that way. :-P
<ethanwinn>
thanks Ralith, I'll take a stab at it.
otherjrolfs has joined #nixos
jrolfs has joined #nixos
rihards has quit [(Quit: rihards)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] vcunat pushed 1 new commit to staging: https://git.io/vbSQC
<NixOS_GitHub>
nixpkgs/staging a7d4fb8 Vladimír Čunát: Merge branch 'master' into staging...
NixOS_GitHub has left #nixos []
otherjrolfs has quit [(Ping timeout: 240 seconds)]
jrolfs has quit [(Ping timeout: 252 seconds)]
hiratara has quit [(Ping timeout: 265 seconds)]
hiratara has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] pbogdan opened pull request #32956: [WIP/RFC] update firefox on 17.09 (release-17.09...update-firefox-17.09) https://git.io/vbSQF
NixOS_GitHub has left #nixos []
Itkovian has joined #nixos
zzamboni has joined #nixos
<cmiles74>
snapd wants to add systemd targets (for snap mounts) everytime a new snap is installed (at runtime). Is there a clever way to make a director in /etc writeable, so snap can do this? In this case it's /etc/systemd/system/multi-user.target.wants... The guys in the snappy channel say the path is hard coded into snap, changing it would require patching the source. I can do that, but would rather not.
FRidh has quit [(Quit: Konversation terminated!)]
FRidh has joined #nixos
FRidh has quit [(Client Quit)]
FRidh has joined #nixos
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
srhb has quit [(Quit: Quit)]
asuryawanshi has quit [(Ping timeout: 252 seconds)]
jrolfs has joined #nixos
otherjrolfs has joined #nixos
<dtzWill>
cmiles74: symlink to a directory snapd controls, maybe?
<dtzWill>
oh but it's not a directory or even a file that really truly belongs to snapd. hmm.
<dtzWill>
snapd should emit nix code and installing snaps will be a new system generation, updating the /etc/ file. :P
otherjrolfs has quit [(Ping timeout: 240 seconds)]
jrolfs has quit [(Ping timeout: 252 seconds)]
<Sonarpulse>
where do I get dig?
<Sonarpulse>
it says bind
<Sonarpulse>
that's not true
<sphalerite>
Sonarpulse: dnsutils iirc
<Sonarpulse>
thanks sphalerite
<dtzWill>
Sonarpulse: bind.dnsutils ?
<dtzWill>
ah
<dtzWill>
bummer command-not-found reports wrong attribute though
<Sonarpulse>
those are the same hmm
d4rk_ has quit [(Quit: Page closed)]
<cmiles74>
dtzWill: Yeah, it's getting messier and messier. I was hoping for something that maybe would make it easy to bind_mount or something inside a package expression to try and fake it out.
<vcunat>
Sonarpulse: same? I don't get what you mean.
mr_sm1th has quit [(Quit: No Ping reply in 180 seconds.)]
mr_sm1th has joined #nixos
phreedom has quit [(Ping timeout: 248 seconds)]
alhariel has quit [(Remote host closed the connection)]
FRidh has quit [(Remote host closed the connection)]
FRidh has joined #nixos
spietz has quit [(Quit: WeeChat 1.9.1)]
davidak has quit [(Ping timeout: 248 seconds)]
galbolle has quit [(Quit: galbolle)]
jrolfs has quit [(Ping timeout: 256 seconds)]
otherjrolfs has quit [(Ping timeout: 272 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] dtzWill opened pull request #32957: gnuplot_full: init with support for lua/tikz/tex (master...feature/gnuplot-extra-terminals) https://git.io/vbSjm
NixOS_GitHub has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] orivej pushed 5 new commits to master: https://git.io/vbSjG
<NixOS_GitHub>
nixpkgs/master b513fc7 Bas van Dijk: python.pkgs.voluptuous: init at 0.10.5...