<tilpner>
greglearns___ - In your position, I would just give in and install rustup :/
<copumpkin>
the rust stuff in nixpkgs is still pretty young, unfortunately
<greglearns___>
but, with rustup, I can't use nixOps... (I think)
magnetophon1 has quit [(Ping timeout: 240 seconds)]
acertain has quit [(Read error: Connection reset by peer)]
acertain has joined #nixos
nil has quit [(Ping timeout: 260 seconds)]
ison111 has joined #nixos
erictapen has quit [(Ping timeout: 240 seconds)]
jsgrant_ has quit [(Quit: Peace Peeps. o/ If you need me asap, message me at msg[(at)]jsgrant.io & I'll try to get back to you within 24-36 hours.)]
<greglearns___>
tilpner how were you using rust nightly in your projects?
<tilpner>
greglearns___ - I was using rustNightlyBin from nixpkgs. I migrated to stable, so I didn't need that anymore. When I tried to make the old .nix files work again, I hit the same error
<greglearns___>
stable nix? or stable rust?
<tilpner>
Stable Rust
<greglearns___>
have you done rust web dev? any recommendations for what works with stable/
<tilpner>
Oh, it's been years... I was using Iron, but I didn't even know what Nix was back then, and Rust was pre-1.0
ertes has joined #nixos
<greglearns___>
ok, thanks (as a single tear rolls down my cheek)
ryanartecona has joined #nixos
xadi has joined #nixos
yegortimoshenko has joined #nixos
<greglearns___>
tilpner I really appreciate your help today. You got me unstuck!
<tilpner>
What? I thought we ended up right where we started? :/
ryanartecona has quit [(Ping timeout: 240 seconds)]
ryanartecona has joined #nixos
<dalaing>
any recommendations around a vps for a NixOS machine?
<greglearns___>
well, we did, BUT now I know that it's not going to work to fight that tide... before, I felt like I was missing something. NOW, I'm feeling like something between NixOS and Rust's implementation is just broken and so it's not worth fighting it for now... so, I can peacefully move on to using stable Rust on Nix and sleep well tonight.
<tilpner>
dalaing - I'm running mine on Vultr
<tilpner>
greglearns___ - Oh, okay. That's a very optimistic view :D
<greglearns___>
*laughing* OR, the peaceful resignation that comes when one is fully beaten :-)
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Ericson2314 opened pull request #27368: cc-wrapper: Unify and improve dynamic linker flag logic (staging...cc-wrapper-interpreter) https://git.io/vQQPo
NixOS_GitHub has left #nixos []
<dalaing>
tilpner: thanks for that
mith[m] has joined #nixos
<dalaing>
Hetzner looks to have some pretty decent pricing, trying to work out how much pain I'd have to go through to set things up there
<dalaing>
I guess in that case I could just follow the NixOps manual and see how I go
<tilpner>
dalaing - In comparison to other VPS providers, Vultr allows uploading an ISO, making the "Install NixOS from a recovery image" trick unnecessary
<dalaing>
I'm not sure if I'm reading the prices wrong, or if I'm not making an apples-to-apples comparison, but Hetzner looks much cheaper for similar specs
<dalaing>
ie $35/month vs $120/month for similar specs
<tilpner>
That's much larger than my server is. Are you looking at Hetzners dedicated or virtual servers?
<gchristensen>
dalaing: I use packet.net a lot
sigmundv_ has quit [(Remote host closed the connection)]
<dalaing>
tilpner: I've looked at both, and the price ratio seems fairly stable
<mpcsh>
dalaing: try vultr.com
<tilpner>
dalaing - If you don't have any experience with Hetzner, maybe ask around (though I've mostly heard good things about them). Some issues aren't immediately apparent
<dalaing>
Hetzner seems about 4x cheaper (or more) than vultr or packet.net
<dalaing>
maybe I should drink more coffee and read the fine print more carefully
<tilpner>
(Like "Oops, we rebooted your stuff, hope you don't mind")
georgew has joined #nixos
<tilpner>
If you care a lot about, get both and do benchmarks :)
<mpcsh>
tilpner: my vultr instance is $2.50/month, hetzner looks more expensive??
greglearns_ has quit [(Quit: Leaving)]
* tilpner
-> dalaing
<dalaing>
I'm look at 4 cores / 16 GB of RAM and up, maybe that's where the disconnect is
lambdael has quit [(Quit: WeeChat 1.7.1)]
<mpcsh>
dalaing: ahh, in that case I'm not sure what the best option is
* dalaing
wipes forehead
lambdael has joined #nixos
Rotaerk has joined #nixos
<dalaing>
I thought I was missing something more crucial
ryanartecona has quit [(Ping timeout: 246 seconds)]
<dalaing>
vultr and packet.net both look pretty great for the smaller instances
<gchristensen>
even for big instances it is well pricede
<dalaing>
that's what lead to the question - I'm wondering if I'm reading something drastically wrong
<gchristensen>
dalaing: hetzner might be older, or might not have 2x bonded gigabit nics
<gchristensen>
copumpkin: 17.03 ok?
<copumpkin>
sure
ryanartecona has joined #nixos
greglearns___ has quit [(Client Quit)]
<gchristensen>
commented
hyphon81 has joined #nixos
<copumpkin>
gchristensen: thanks! I fixed it to actually produce output
<copumpkin>
so you don't have to dinstinguish "good failure" from bad failure :P
<gchristensen>
works for me
<copumpkin>
cool, thanks
<copumpkin>
but fails on darwin, right?
<copumpkin>
:D
<copumpkin>
it fails on mine, that's good enough for me
ryanartecona has quit [(Quit: ryanartecona)]
<yegortimoshenko>
copumpkin: going to test your 10.13 nix build in several hours
xadi has joined #nixos
mudri has quit [(Ping timeout: 240 seconds)]
magnetophon1 has joined #nixos
sigmundv_ has joined #nixos
xadi has quit [(Remote host closed the connection)]
<dalaing>
gchristensen: seems like the hetzner servers I'm looking at have a dedicated 1Gb uplink - which is probably fine for us
<gchristensen>
ok
<dalaing>
Australian internet softens the impact of that for us I guess :)
<gchristensen>
ouch
xadi has joined #nixos
sigmundv_ has quit [(Ping timeout: 255 seconds)]
vaibhavsagar has joined #nixos
zeus_ has quit [(Remote host closed the connection)]
zeus_ has joined #nixos
Supersonic112 has quit [(Disconnected by services)]
Supersonic112_ has joined #nixos
Supersonic112_ is now known as Supersonic112
zeus_ has quit [(Ping timeout: 276 seconds)]
justelex has joined #nixos
Wizek_ has quit [(Ping timeout: 255 seconds)]
justelex_ has quit [(Ping timeout: 260 seconds)]
slack1256 has joined #nixos
<mpcsh>
hey all, I'm using lightdm and bspwm, but whenever I switch to a tty (e.g. ctrl+alt+f2), bspwm quits, and I can't find a way to start lightdm again. what gives?
eacameron has joined #nixos
yegortimoshenko has quit [(Ping timeout: 248 seconds)]
lambdamu has joined #nixos
yegortimoshenko has joined #nixos
yegortimoshenko has left #nixos []
<mpcsh>
*headdesk* but I didn't think to try systemctl start display-manager....
davidak has quit [(Ping timeout: 248 seconds)]
lambdamu_ has quit [(Ping timeout: 246 seconds)]
zeus_ has joined #nixos
vaibhavsagar has quit [(Remote host closed the connection)]
vaibhavsagar has joined #nixos
eacameron has quit [(Remote host closed the connection)]
xadi has quit [(Quit: Leaving.)]
s33se_ has joined #nixos
<Infinisil>
mpcsh: Does it actually quit? It's on the 7th terminal
davidak has joined #nixos
<Infinisil>
Ctrl-Alt-F7
<Infinisil>
Ohh you solved it, never mind
<Infinisil>
But it shouldn't terminate just by doing that
s33se has quit [(Ping timeout: 240 seconds)]
mbrgm has quit [(Ping timeout: 240 seconds)]
<mpcsh>
Infinisil: so you're right, it doesn't terminate, but I still can't bring it back up - Ctrl-Alt-F7 (or any F key above 6) does nothing
<mpcsh>
(but ctrl-alt-f{1-6} work fine)
mbrgm has joined #nixos
Jackneillll has joined #nixos
<Infinisil>
mpcsh: Weird, no idea
Jackneilll has quit [(Ping timeout: 240 seconds)]
<mpcsh>
different note, does anyone know how to enable command-not-found?
<vaibhavsagar>
hi, how do I check the latest version of a PyPI package for download?
<vaibhavsagar>
I can't browse pypi.io/packages/source
<vaibhavsagar>
and doing the obvious thing (nix-prefetch-url mirror://pypi/b/bash_kernel/0.6.tar.gz) doesn't work
<mpcsh>
tilpner: that did it, thanks!!
<Infinisil>
tilpner: Weird that doesn't work for me, option isn't there
<tilpner>
You two might be on different versions. It's also not yet listed in the web options search
<Infinisil>
Hmm, I'm using the 17.03 channel
<Infinisil>
and I can see it via grepping
darlan has joined #nixos
davidak has quit [(Quit: Leaving.)]
<radvendii>
what is command-not-found? It sounds useful...
paperd has joined #nixos
<Infinisil>
radvendii: I think instead of complaining that a command wasn't found which is available in this package, it just directly installs the package and runs the command
<radvendii>
O.o
<radvendii>
that's crazy talk
<radvendii>
does it do it in a nix-shell?
danl1240 has quit [(Quit: My iMac has gone to sleep. ZZZzzz…)]
<radvendii>
like is it like running `nix-shell -p thing --run "thing"`, or like installing it with nix-env?
<Infinisil>
I don't know, the option isn't available for me so I can't checkt
danl1240 has joined #nixos
<Infinisil>
Wait I just had a look at unstable, seems to work there, not in 17.03 though
<Infinisil>
Happy 1.5b! Even though nobody's here right now
<paperd>
Hello
<joehh>
is there any reason why "Preprocessing library xxx" when running cabal build would start taking minutes to run with cpu sitting at 100%?
<joehh>
using nixos and ghc 8.0.
<joehh>
2
<paperd>
how can I fix error: Typesafe Activator was removed in 2017-05-08 as the actual package reaches end of life.
<Infinisil>
paperd: Seems like a dead package to me, why do you need it?
<Infinisil>
You need to get an old version back somehow if you really need it
<paperd>
Infinisil: I don't need it
spinus has quit [(Ping timeout: 248 seconds)]
<paperd>
when I do nix-build -f ~/.nixpkgs -iA hello it throws that error
orivej has joined #nixos
<paperd>
or when I try and build anything from nixpkgs
<Infinisil>
paperd: I assume ~/.nixpkgs is your local checkout?
<paperd>
no, it's master from github
<radvendii>
paperd: are you sure? why do you have it under .nixpkgs? that's where config.nix should live
<paperd>
I haven't edited config.nix
bhipple has joined #nixos
<paperd>
I don't want to see it when I do an ls in my terminal
<radvendii>
oh hm
<radvendii>
I'm just wondering if that's the problem, but probably not
<paperd>
I cloned to ~/nixpkgs and it throws the same error
<radvendii>
okay, not that then
<paperd>
It links to a github thread, which I've read, but doesn't really tell me anything.
<Infinisil>
paperd: Didn't you mean nix-env instead of nix-build?
<Infinisil>
nix-env -f ~/nixpkgs -iA hello
<Infinisil>
Works for me
<bhipple>
Hi guys, is there a way to get the transitive closure of build-time dependencies of a pkg? I'm trying: nix-store -q --requisites /nix/store/4akxixlbhg18lvm978lgvzf9xfjm148j-fzf-0.16.4 for example, but it doesn't seem to work
<bhipple>
In the above case, it just shows glibc, bash, and tzdata
<Infinisil>
paperd: Or `cd ~/nixpkgs; nix-build -A hello`
<paperd>
Infinisil: yes, that works for me, but I also use nix-build to see if things will build
<bhipple>
but I can see from the pkg that ncurses and the whole golang ecosystem is a build input
<Infinisil>
bhipple: Maybe you need to work with the .drv in the store instead
<bhipple>
Infinisil: aha, 417 lines of output. Looks like that did it, thanks!
<Infinisil>
paperd: According to my nix-build, there is no such -f flag
<radvendii>
paperd: nor is there a -i flag...
<Infinisil>
You are most certainly referring to nix-env -f ~/.nixpkgs -iA hello
<radvendii>
I feel like there should maybe be a more consistent flag-scheme for nix-* functions
<paperd>
Infinisil: I must be, thank you.
<Infinisil>
radvendii: It's pretty consistent if you ask me, what's not?
<radvendii>
like why isn't there a -f option for nix-build?
<Infinisil>
Ah
<radvendii>
I don't actually use these commands that much, so I don't know
<tilpner>
nixUnstable has $ nix build -f ...
<Infinisil>
I feel like nix-build is more for development purposes, in which case you should really be in the correct directory already, not sure though
<Infinisil>
tilpner: Oh nice
<tilpner>
(But no -a)
<Infinisil>
`man nix`
<Infinisil>
No manual entry for nix
<Infinisil>
:/
<Infinisil>
Okay it's unstable, i shouldn't complain about missing manpages
<paperd>
So I've just been typing the wrong command for like months... that is hilarious.
<disasm>
I should know Monday if I'm presenting at DevOpsDays Philly on reproducible deployments using nixos... If it doesn't work out, I'm going to go for it and see if my boss will let me go to NixCon in Munich :)
<bhipple>
Mostly go, ncurses, openssl, and all the stdenv stuff
<Infinisil>
bhipple: Damn, using IPFS for that, pretty sweet
<Infinisil>
I've never even thought of that, but it's pretty much perfect for pastebinning stuff
<simpson>
Infinisil: Generalization: Systems which let you assign names to data are perfect matches for processes which take names and operate on data.
<Infinisil>
Oh, now it's a different error, i removed ipfs from systemPackages
<Infinisil>
still permission denied, but on a different file, I may need to chmod again
<clever>
what does "type ipfs" say?
<Infinisil>
Nope, lock permission denied again
<Infinisil>
/run/current-system/sw/bin/ipfs
<bhipple>
`groups` shows you're in IPFS?
<Infinisil>
bhipple: Yup
<Infinisil>
Well it doesn't really matter much, I'm going to dig into it later, unless you really want to solve this right now
<bhipple>
Weird. Good luck
<clever>
ive only used it with the data in home
<Infinisil>
How can i view the numbers instead of names of groups/users?
<bhipple>
`getent passwd`
<Infinisil>
And for files?
georges-duperon has quit [(Ping timeout: 255 seconds)]
<clever>
Infinisil: run "id"
indi_ has quit [(Remote host closed the connection)]
<clever>
and stat for a file/dir
<pop>
A version of a program I want to use is on the master branch of nixpkgs. When I run `nixos-rebuild switch` it does not pull this version of the application. How can I use the master version of applications?
indi_ has joined #nixos
indi_ has quit [(Remote host closed the connection)]
indi_ has joined #nixos
<Infinisil>
Hmm, nope, it's all 261 for ipfs, am definitely in group, maybe the lock file is taken by something else
<pop>
Oops, I just found the --upgrade flag. I'm pretty sure that's what I want.
<vaibhavsagar>
pop: clone nixpkgs, and then `nix-env -f <path/to/nixpkgs> -i <application>
<clever>
Infinisil: can you pastebin the output of ls -ltrhR /var/lib/ipfs
<vaibhavsagar>
in case your package isn't in the latest 17.03
<pop>
vaibhavsagar: Where can I find out what's in "the latest 17.03"?
<pop>
Just the online nixos packages website thingie?
<Infinisil>
Yup, that was it, works now, using 8081 for the webserver now
<Infinisil>
God damnit
<Infinisil>
So many ports, but everybody wants 8080
<simpson>
8080 is just the traditional running-as-user HTTP port.
hyphon81 has quit [(Remote host closed the connection)]
<simpson>
IPFS really should request a port.
<Infinisil>
yeah
mkoenig_ has quit [(Ping timeout: 248 seconds)]
<bhipple>
Is there a standard tool for requesting ports on a host?
<clever>
i just pick a random number between 1024 and 65535 when i start a project
<Infinisil>
bhipple: I think it's more of a "internet organizations declares that port xxx belongs to IPFS and shouldn't be used for anything else"
<clever>
yeah, that has to be done for ipfs
dtzWill has joined #nixos
<dash>
yeah it's IANA
<Infinisil>
Hmm, but this whole port idea is kinda outdated imo
<simpson>
bhipple: I mean in the sense of getting a port number from IANA.
mkoenig has joined #nixos
<simpson>
Infinisil: Well, yes, but only because TCP's getting a little outdated.
<Infinisil>
How about declaring port 0 as the port that says where services are located
eacameron has joined #nixos
<dash>
Infinisil: yes, the future will be all HTTP and the OS will give every process an IPv6 address that can only listen or connect on 80 and 443
<Infinisil>
E.g. you send a message to port 0, asking for the io.ipfs service, you get back port xxx
<dash>
Infinisil: DNS has SRV records for this
<Infinisil>
But DNS is too high of a level, DNS is already above the port abstraction
<Infinisil>
The solution I just said sounds actually pretty good
<dash>
Infinisil: shrug, it's plausible to use DNS lookups to lookup ports since it's already used to lookup IPs
<simpson>
Infinisil: Why can't your message just be "Hi, can I have the data associated with this IPFS hash?" Why must it be anything more?
<clever>
dash: i recently switched one of my services to SRV, but now i'm finding new fun bugs
<clever>
dash: one users router, doesnt use the record type (a, aaaa, srv) as a key in the cache
<dash>
clever: yes, approximately nobody uses it
<clever>
dash: so it can return a records when the client asked for srv
<clever>
dash: it has even returned aaaa records in response to an a query, with no ipv6 setup
<dash>
clever: classy
<Infinisil>
simpson: Well ipfs is separated between server and client so you can run it on an actual server, so it needs a port
<dash>
clever: someone should visit his house and accidentally break it
<dash>
Infinisil: only if you use TCP or UDP
<simpson>
Infinisil: I'm asking, *why* do you need ports? Because of TCP/IP, right? Is there any other reason?
<clever>
dash: and googles 8.8.8.8 keeps randomly claiming the domain doesnt exist
<clever>
simpson: what about rpcbind?
<simpson>
clever: NFS is one of those things that we will not need.
eacameron has quit [(Ping timeout: 260 seconds)]
<clever>
i think rpcbind just lets you lookup the port behind a named service
<Infinisil>
simpson: Hmm, I mean what's the alternative to using ports?
<simpson>
Yeah, but literally only NFS uses it~
<dash>
Infinisil: a networking system that doesn't use ports
<Infinisil>
dash: Example?
<simpson>
Infinisil: One alternative is named services. The OSI model has a layer for it. The idea is that, both locally and remotely, you don't actually care which routing port's needed, just whether it's hitting HTTP/FTP/Gopher/etc.
<dash>
Infinisil: ethernet, IP, etc etc don't use port numbers, only TCP and UDP do
<simpson>
I knew a guy who wrote a daemon to do this in userspace.
<dash>
Infinisil: so, something else could be used (over top of any of these) that doesn't use port numbers
<Infinisil>
dash: Ah right, but our internet infrastructure would need to be changed
<dash>
Infinisil: Not necessarily.
<Infinisil>
Every router uses ports for routing I think
<Infinisil>
And NAT's
<dash>
Infinisil: I recommend Van Jacobson's talk "A New Way To Look At Networking" about this
<dash>
Infinisil: he talks about how the Internet got started over top of the original computer network
<simpson>
Infinisil: Huh? No, it's really 100% TCP and UDP. IP routers don't care about the ports, except when doing NAT or DPI, which are to both be shunned as rude.
<dash>
Infinisil: and about how the process could be repeated
<Infinisil>
dash: Sounds interesting
<Infinisil>
simpson: Well I guess I'll have to refresh my internet knowledge some more heh
<Infinisil>
even with `sudo echo hi`, how has this anything to do with samba?
<clever>
i think that module is to let you use a samba password via pam
<Infinisil>
Ohh
<Infinisil>
I enabled password sync, I'll turn that off for now then
<Infinisil>
Yup that was it, thanks!
ryanartecona has joined #nixos
<Infinisil>
When I was using macOS I was really annoyed at the constant stream of error messages that appeared in the logs
<Infinisil>
Not having any of that in linux
indi_ has quit [(Remote host closed the connection)]
indi_ has joined #nixos
<copumpkin>
file issues please!
<copumpkin>
we have limited people working on this stuff on darwin but we read issues
<Infinisil>
copumpkin: The constant error stream had nothing to do with nix
<Infinisil>
Or do you mean filing issues in general?
<copumpkin>
well, the build instructions (which presumably lead to errors on macOS) are still nixpkgs's responsibility
<copumpkin>
in many cases clang 4 is much stricter and warning-happy than gcc
<copumpkin>
but we can still suppress things that are nonsense
<copumpkin>
anyway yeah, issues in general are good :)
<copumpkin>
with any luck we'll pick up more darwin contributors and they'll want stuff to do :D
<Infinisil>
No I really mean it has nothing to do with nix, back when I was using macOS I had so many warnings and errors in the system console, that's what annoyed me, haven't even used nix at that time
<copumpkin>
oh I see what you mean
<copumpkin>
yeah, there's a ton of junk in there
<Infinisil>
If I would still enjoy using macOS I would totally want to help make nix better there
<disasm>
Infinisil: ditto... I haven't rebooted back into osx in over a month.
<copumpkin>
boo, we need to make linux less appealing :P
<copumpkin>
(tbc I still use it on all my servers, obviously)
<Infinisil>
I have just recently, I'll have to move over the rest of the data before I can nuke osx from the main disk and give linux the full 480GB
<disasm>
I probably will next week though to test out this install script gchristensen is working on :)
<Infinisil>
I have just installed a fresh OSX on an external disk, I could test it there as well
<disasm>
that's a good point, I do have a macbook air in the laptop I originally used for testing the nixos waters before doing it on my main laptop. I could reinstall osx there for nix darwin testing.
<Infinisil>
And since my node/js knowledge is about 0/10, that's how far I can get
eacameron has quit [(Ping timeout: 248 seconds)]
<Infinisil>
Should I open an issue/PR for that?
<copumpkin>
yeah probably :)
<disasm>
Infinisil: I would, and just mention WIP. Someone like me might try cloning it down and getting it working :)
magnetophon1 has quit [(Ping timeout: 246 seconds)]
ebzzry has quit [(Ping timeout: 240 seconds)]
<Infinisil>
And why the hell is this ./generate.sh script taking a lifetime
<Infinisil>
Damn node and its hundred thousands of packages for every single thing
ebzzry has joined #nixos
<copumpkin>
wanna add two numbers? there's a library for that
<Infinisil>
What was that 10 line package named again that everyone used but then the world ended as it disappeared or something? Right aling something
<Infinisil>
It doesn't even look at the file's contents, only extension. What the fuck is the use of this package, if you require a specific extension, then check for that instead
<Infinisil>
Sorry I'm going off topic, but sheesh
ryanartecona has quit [(Quit: ryanartecona)]
newhoggy_ has joined #nixos
indi_ has quit [(Remote host closed the connection)]
indi_ has joined #nixos
indi_ has quit [(Remote host closed the connection)]
<sphalerite[m]>
I certainly wouldn't expect it to look at the contents considering that it has "path" in its name
newhoggy has quit [(Ping timeout: 276 seconds)]
paperd has quit [(Quit: WeeChat 1.7.1)]
<Infinisil>
sphalerite[m]: Fair enough
<Infinisil>
but still
<disasm>
sphalerite[m]: I agree, but I probably would anticipate it checks if a binary is in your path by the name (which it obviously doesn't do)
hamishmack has quit [(Quit: hamishmack)]
<sphalerite[m]>
You mean an executable?
<sphalerite[m]>
And I think this sort of thing is good.
newhoggy_ has quit [(Remote host closed the connection)]
<Infinisil>
great, apparently I'm building V8 from source..
hamishmack has joined #nixos
<disasm>
sphalerite[m]: I mean if I wrote something called is-binary-path, I would expect it to take a string binary and return true/false if that command is in your current path.
hamishmack has quit [(Client Quit)]
<sphalerite[m]>
These tiny 3-line pieces of code whose intentions aren't so obvious are so frequently duplicated
<disasm>
I wouldn't expect it to look at the extension and determine if it's a binary or not, lol :)
<Infinisil>
disasm: You're thinking too much linux :P
<sphalerite[m]>
I agree that filename extensions are generally a terrible idea
<sphalerite[m]>
Or rather reliance on them
slack1256 has quit [(Ping timeout: 240 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] hhm0 opened pull request #27373: hebcal: init at 4.13 (master...hebcal) https://git.io/vQQAQ
NixOS_GitHub has left #nixos []
newhoggy has joined #nixos
<sphalerite[m]>
disasm: but it could be on your path and not be a binary. This is the case for many many programs on nixos!
<Ralith>
I am a big fan of not accessing any files you don't already know the format of
justbeingglad has joined #nixos
justbeingglad has left #nixos []
<Infinisil>
Ralith: Well as long as it's not executed it's fine by me
<Ralith>
bonus points for already knowing the exact contents!
<Infinisil>
Stupid me didn't know what `exec` did a few weeks ago, don't run it on random files that are not bash, does no good
<sphalerite>
works just fine on anything executable :)
<Infinisil>
sphalerite[m]: Hmm... I'm afraid of testing it
<sphalerite>
(that is, both "executable bit set" and "can actually be executed on the current machine")
<sphalerite>
why?
<Infinisil>
Had some weird behaviour, aka window jumbling for a microsecond, then the terminal window got closed
<Infinisil>
not sure what I ran it on
<sphalerite>
I mean, there's barely ever a reason to use exec in an interactive shell unless you're execing another interactive shell to replace the previous one (arguably a neater way than source ~/.whatevershellrc)
<sphalerite>
Window jumbling..?
<Infinisil>
can't remember exactly, eh whatever, I'll report when I encounter it again
<sphalerite>
well what exec does is replace the current shell with another process
<sphalerite>
so yeah, the terminal window will get closed when that other process exits
<Infinisil>
sphalerite[m]: Oh that would make sense
<Infinisil>
Does anyone by chance know how long it takes to compile V8? I'd like to know if I should wait or just go to bed
<sphalerite>
that's why it's used in most wrappers, so that you don't have a shell process hanging around when it's not necessary anymore
<Infinisil>
sphalerite[m]: Yeah I just used it recently to avoid a double fork
<sphalerite>
Just go to bed. Even if it finishes as soon as you've gone, sleep is good for you ;)
<Infinisil>
True
<Infinisil>
I'll enjoy the 100% CPU sound while sleeping :P
<Infinisil>
oh i think it just finished
acertain has quit [(Read error: Connection reset by peer)]
acertain has joined #nixos
<srhb>
I'm puzzled about this failure: https://travis-ci.org/NixOS/nixpkgs/jobs/253390686 -- it doesn't seem to have anything to do with my commit. I'm wondering if it's the same reason as the nox-review fail I had locally, that also didn't relate to my fix (as far as I could see)
<srhb>
Should I do anything, or just wait?
vaibhavsagar has joined #nixos
<sphalerite>
srhb: I'd go with ignore. People often just ignore travis anyway afaik
<NixOS_GitHub>
[nixpkgs] vbgl pushed 1 new commit to master: https://git.io/vQQhz
<NixOS_GitHub>
nixpkgs/master 53a50df Vincent Laporte: ocamlPackages.camomile: fix build for OCaml 4.05
NixOS_GitHub has left #nixos []
magnetophon has quit [(Ping timeout: 255 seconds)]
yee has joined #nixos
magnetophon has joined #nixos
yee has quit [(Client Quit)]
<joehh>
sphalerite: nice - thanks for that - I'll give it a go
<LnL>
joehh: adding the nix profile.d line to your bashrc instead of profile should fix that
zeus_ has quit [(Remote host closed the connection)]
<LnL>
bash won't load .profile with ssh host -- cmd
<joepie91>
today's work is going to be inventorying all my miscellaneous systems and migrating everything to a NixOS host one by one... wish me luck :)
magnetophon has quit [(Ping timeout: 246 seconds)]
<avn>
joepie91: will you use something like nix-in-place, or just plain reinstall?
<joepie91>
avn: NixOps + KVM VPSes (ie. QEMU) that have NixOS proper installed on them, moving services between systems one by one
<joepie91>
avn: then the emptied systems will either be thrown out or reinstalled
<joepie91>
I was intending to consolidate some VPSes anyway
<joepie91>
so this is a fine moment to do so
magnetophon1 has joined #nixos
<joepie91>
avn: some of the 'source systems' are OpenVZ VPSes so that's going to be a fun exercise to get to work
<joepie91>
I may use nixos-infect for some KVM VPSes where I can't provide custom ISOs and the provider won't add NixOS on request
<joepie91>
so yeah, it's all going to be some degree of experimentation
<joepie91>
:P
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 pushed 2 new commits to master: https://git.io/vQQjB
<NixOS_GitHub>
nixpkgs/master 85070e1 Silvan Mosberger: fast-cli: init at 1.0.0
<NixOS_GitHub>
nixpkgs/master ab87dc5 Jörg Thalheim: fast-cli: add reference to all-packages.nix...
<NixOS_GitHub>
[nixpkgs] wizzup opened pull request #27378: django-extensions : init at 1.8.1 (master...add-python-django-extensions) https://git.io/vQ7qh
NixOS_GitHub has left #nixos []
orivej has quit [(Remote host closed the connection)]
<ixxie>
say I want to define dotfiles for my user account in my configuration.nix, so that they are subsequently placed in my home directory; is there a nice way to do this?
<tilpner>
(It works fine here, but please don't blame if it eats your dotfiles)
<tilpner>
*blame me
<tilpner>
ixxie - Instead of those /cfg paths, you could use writeText to define the content inline
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
hellrazor has quit [(Quit: WeeChat 1.9)]
<ixxie>
tilpner: thanks, thats perfect :) just out of curiousity - the way you use this is have your dotfiles in cfg folder with your nixos configuration and have it build them into home?
slack1256 has quit [(Remote host closed the connection)]
<tilpner>
ixxie - Exactly. All my configuration is gathered in /cfg, and /home contains only symlinks to dotfiles, not the actual files
<ixxie>
adisbladis[m]: I recall this home manager thing, but it seems like overkill for this
<ixxie>
I also am a bit confused by it and it would take me a while to figure it out xD
<tilpner>
There's a slight variation to this. You can remove the quote from the paths, and nix will copy the files into the nix store
<adisbladis[m]>
ixxie: I agree it might be :) I just thought it was worth considering at least
roberth has joined #nixos
<tilpner>
That would mean any changes to your dotfiles are only activated on nixos-rebuild
<ertes-w>
roberth: just set it in your shell profile or session startup script
<roberth>
Wake up, Gravious...
<adisbladis[m]>
A federated communication protocol, think XMPP (Jabber) but more modern
<Gravious>
roberth, :)
<ixxie>
adisbladis[m]: kinda optimized for nix and not nixos eh?
<roberth>
ertes-w, thanks. Doesn't feel very Nixy, but seems reasonable
ThatDocsLady has joined #nixos
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
<adisbladis[m]>
ixxie: No? Currently nixos does not come with a declarative way to install applications in your user profile and manage that
<adisbladis[m]>
I like to keep my systemPackages fairly small and install most stuff as my user
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
<ixxie>
adisbladis[m]: I see.... my user doesn't even have a nixpkgs yet xD
schoppenhauer has quit [(Ping timeout: 260 seconds)]
mudri has joined #nixos
<ertes-w>
roberth: i don't think there are per-user shell profiles in configuration.nix, and even if there were i wouldn't use them, because you would have to be root to change your own NIX_PATH
wak-work has joined #nixos
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
hyphon81 has joined #nixos
MoreTea has joined #nixos
Thra11 has quit [(Ping timeout: 255 seconds)]
<roberth>
ertes-w, you could always override it in your shell. I suppose the default NIX_PATH could be more configurable, or even be a bit dynamic, but that might cause some surprises... .profile is the user's entrypoint, so it makes sense to keep it simple and just maintain all user configuration from there using the user's favorite tools (manual editing, or something like what adisbladis incidentally just posted :)
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
arianvp2 has quit [(Client Quit)]
Wizek_ has joined #nixos
arianvp2 has joined #nixos
<ertes-w>
i have an s6-based setup there, but currently it's not nixified (even though i do actually have all the necessary nix expressions to do it)
<ertes-w>
however, it doesn't include dot-files and is mainly about services… personally i don't see too much value in nixifying dot-files
<gchristensen>
domenkozar: ping
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nix] edolstra pushed 2 new commits to master: https://git.io/vQ7sV
<NixOS_GitHub>
nix/master 0681f8c Eelco Dolstra: Shut up a memory leak warning
<NixOS_GitHub>
nix/master 38374a9 Eelco Dolstra: Tarball job: Include libseccomp on Linux only
NixOS_GitHub has left #nixos []
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nix] edolstra closed pull request #1459: Build tarball per system (master...darwin-tarball) https://git.io/vQy9M
NixOS_GitHub has left #nixos []
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
mpcsh has quit [(Quit: THE NUMERICONS! THEY'RE ATTACKING!)]
nslqqq has quit [(Ping timeout: 260 seconds)]
mpcsh has joined #nixos
<domenkozar>
gchristensen: pong
<roberth>
ertes-w, I didn't hear of s6 before. What are you using it for?
<roberth>
ertes-w, I didn't see much value in nixifying dot files, but have to do some remote development now, so it seems a bit more compelling. I guess I need to bit the bullet.
<gchristensen>
niksnut: we found some unanticipated breakage in how the darwin installer works. I'm hoping to get domenkozar's feedback on my fixes before merging, but I think we'll want to make another nix release soon, to fix people's automation :(
<adisbladis[m]>
roberth: I think it's nice in case I get a new laptop or the disk breaks down
<niksnut>
gchristensen: ok
m0rphism1 has joined #nixos
m0rphism has quit [(Ping timeout: 255 seconds)]
<gchristensen>
domenkozar: we can rename that variable too ... :)
<domenkozar>
gchristensen: reviewing
<gchristensen>
domenkozar: note at the end I have a test case where I starve the TTY of input, and another test case where it ensures things in /nix/store are allowed to exist before installing
<domenkozar>
gchristensen: this will still trigger lots of textt (actually more than before) when not asking questions right?
<domenkozar>
it feels a bit scary to see so much text and it's polluting the CI logs
<domenkozar>
could we also then have a quiet mode?
<gchristensen>
domenkozar: there are 2 files there
<clever>
domenkozar: and i have found uses for that, a failed build leaves the $out intact, and then i can investigate logs it left there
<clever>
domenkozar: the above also helps greatly in fixing reference cycles in multiple output derivations
<clever>
nix-shell also wont delete the outputs, so you can shell into the failing derivation, and then grep for $out in $lib
<domenkozar>
nice
<domenkozar>
gchristensen: nice :)
<domenkozar>
lgtm
<gchristensen>
domenkozar: as-is? or the further-reduced version?
<clever>
i also saw a trick in the github issues many months ago, where you first nix-build the kernel, then nix-shell into the kernel, so you can use its deps and $out to build modules
<domenkozar>
the quiet file
<gchristensen>
cool, I _can_ add another env var to reduce the output further, but I'm running out of time this AM to test it.
arianvp2 has joined #nixos
<LnL>
domenkozar: I think you need do do something similar to hydra's store_url = file:///tmp/nix-cache and cache that
<gchristensen>
domenkozar: I can PR what I have, now, if you think it is ready
<clever>
LnL: oh, and then nix-push what you want to save into that, so travis can tar it back up
<domenkozar>
gchristensen: if you have implemented the quiet file, then I think it's good enough (with 200 lines)
<gchristensen>
yep already in that branch you reviewed :)
simukis has joined #nixos
<gchristensen>
domenkozar: is the environment variable name too silly?
<domenkozar>
yes :D
<domenkozar>
it should be self-descriptive
<gchristensen>
:)
<gchristensen>
ALLOW_USER_TO_EXPERIENCE_PROBLEMS?
<domenkozar>
ALLOW_PREEXISTING_INSTALLATION
<LnL>
clever: yes, I think 1.12 can do that automatically with NIX_REMOTE=local?root=/tmp/nix-cache
<clever>
LnL: oh, interesting, not sure what that will do exactly, will need to play with it
pie_ has joined #nixos
<clever>
it sounds like all store actions will go to that dir, and not /nix/store/
<clever>
i believe the 1st example treats a store as a cache, and the 2nd example lets you act on a store that isnt at /nix/store yet (chroot'ing for you)
<clever>
LnL: yeah, this sets up a store that isnt mounted at the right spot yet, and the paths in the above cmd are wrong
<gchristensen>
domenkozar, niksnut doing some final automatic testing but it should be good to go. I'll ping when the tests finish.
<clever>
and you can then either copy it to /nix/store, or use option substituters to implement a copy-closure thing
<domenkozar>
kk
magnetophon1 has joined #nixos
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
<LnL>
clever: yeah, I thought there was something that would use the local store to build but write the output in a different path
Thra11 has joined #nixos
<clever>
LnL: i think --option substituters could be used to treat /nix/store as a binary cache, when doing a build that targets /tmp/nix-cache/nix/store/
<LnL>
yes something like that
<LnL>
but I have not played with the local options yes do I might be wrong
<clever>
same, i have some notes from questions i asked before, but havent done much
* LnL
wants a 1.12 release :)
MercurialAlchemi has quit [(Read error: No route to host)]
mexisme2 has joined #nixos
<gchristensen>
niksnut, domenkozar: it works -- ready for 1.11.13
<clever>
mpickering: unquote the path and it should work
<mpickering>
perfect thanks
<chominist[m]>
clever (IRC): so how do the stores interact with one another?
eacameron has quit [(Ping timeout: 260 seconds)]
<clever>
chominist[m]: when using NIX_REMOTE=local?real= ?
<chominist[m]>
I read the manuals, but didn't understand how the stores in nix/store interacts with /
<chominist[m]>
(Or mb I missed something)
<gchristensen>
you only have one store, at /nix/store
Thra11 has quit [(Ping timeout: 260 seconds)]
<chominist[m]>
Because I found the xkb there in the xkeyboard_config store, but I wasn't sure why /etc/X11 didn't have a symlink to it
<chominist[m]>
*something else besides store
<MoreTea>
chominist[m], when possible, we do not store configuration settings in /etc, but somewhere in /nix/store. Typically, we point to that config file in the systemd unit.
<chominist[m]>
MoreTea (IRC): then how can you get localectl to find the xkb folder?
arianvp2 has quit [(Quit: arianvp2)]
<chominist[m]>
And more importantly, how can I do more complex xkb settings (through configuration.nix or otherwise)
<MoreTea>
this is the generic case, I admit to have no clue about how exactly locatectl / xkb work in this case.
orivej has quit [(Ping timeout: 248 seconds)]
orivej has joined #nixos
<chominist[m]>
I have completely misunderstood how stores work
<chominist[m]>
Not that the xkb thing is solved, but it's great to undo a blindspot
<gchristensen>
oh?
arianvp2 has joined #nixos
grburst has joined #nixos
xadi has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nix] edolstra pushed 1 new commit to master: https://git.io/vQ74Q
<NixOS_GitHub>
nix/master 112ff78 Eelco Dolstra: nix: Show help when no arguments are given...
NixOS_GitHub has left #nixos []
Thra11 has joined #nixos
<grburst>
hey, if i declare a meta package in ~/.config/nixpkgs/config.nix that includes a package providing a services. is there to declare in the meta package also enable the services services.foo.bar.enable like i do globally in configuration.nix?
<grburst>
is there a way to*
<domenkozar>
gchristensen: after 3 evals, no success
<grburst>
another question ^^: could not find where i can lookup nixpkgs functions, e.g. i came across writeTextFile but could not find any documentation about it
<grburst>
are these function somewhere documented?
<joepie91>
does anybody happen to have any previously undocumented notes on how to run NixOS in an OpenVZ guest (without access to the host, ie. an OpenVZ VPS from a provider)? my next project is to figure that out, and any upfront notes would be helpful :)
<joepie91>
cc clever in particular since you tend to have notes about weird stuff :P
<srhb>
I'm working on upversioning kubernetes services to 1.7. It's pretty complex, and I'm not sure I can get all the tests to work on my own. Is it okay to (eventually, probably after the weekend) create a PR with a note not to merge this but just take it up for further discussion? Or should I keep it as an issue?
<gchristensen>
srhb: yeah, put "WIP" in the title
<srhb>
gchristensen: Thank you.
<srhb>
I see there's some prior work already by LnL
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
<grburst>
gchristensen okay thanks - but a documentation / list of available nixpkgs options and functions and so forth are not available (as far as you know)?
<MoreTea>
@grburst, they're not documented properly.
<gchristensen>
grburst: either incomplete or doesn't exist, and is a regularly talked about problem :(
<MoreTea>
I had a PR to fix this somewhere, by abusing __functor
<gchristensen>
omg
<MoreTea>
but not sure what the performance overhead of that would be.
<MoreTea>
gchristensen, you helped me with that ;)
<ikwildrpepper>
domenkozar: just my opinion here wrt that specific issue. except for some strong opinion and a bit annoying way of expressing him/herself, I do not think the comments are really abusive
<ikwildrpepper>
(the initial comment, I meant)
<gchristensen>
the problem is it is the exact same person
<ikwildrpepper>
gchristensen: could be
<ikwildrpepper>
github should be able to check that, I imagine?
<MoreTea>
"Are you afraid of him?" after a warning is not a very nice thing to say.
takle has quit [(Remote host closed the connection)]
takle has joined #nixos
<MoreTea>
Which approach would you all prefer, to document nixpkgs's lib functions (and your own helper functions)
<domenkozar>
ikwildrpepper: there are many insults in the post
<domenkozar>
ou would get a lot more respect if you presented facts and arguments instead of trying to import Eastern-European practices to a mostly Western online society.
<ikwildrpepper>
domenkozar: yeah, I was talking about the first comment, should have been clearer about that
<gchristensen>
the biggest complaint I have about the first message "Why wasn't this merged on the very same day the PR was opened?" is "because we're volunteers"
<gchristensen>
among other things :)
<ikwildrpepper>
gchristensen: yeah, agree that would have been a good response to the question
<ikwildrpepper>
or any other answer that would explain it
<gchristensen>
fair
<domenkozar>
you assume it would help to go down that path arguing
<domenkozar>
I disagree
<gchristensen>
and we should have done that, but we were perhaps prematurely assuming the user was 0xabab, and we were right
<gchristensen>
...imo :)
<ikwildrpepper>
I do think that we should also apply restraint in responses that could be inflamatory. e.g. I don't really see insults in the initial comment that user made
takle has quit [(Ping timeout: 240 seconds)]
<gchristensen>
ikwildrpepper++
<gchristensen>
I agree
<ikwildrpepper>
the later post obviously is abusive
<ikwildrpepper>
but you could argue that that might be triggered by domen's comment. still not a good reason to post such an abusive comment in reply.
<domenkozar>
there were implicit insults in the first comment
<domenkozar>
carefully read
<domenkozar>
> It appears that everyone except one is in favor of this. The obvious solution seems to be that this person puts a big ifdef block in the code he is running on his machine and let's the rest of the world enjoy their 256GB RAM servers with TBs of storage, or their 8GB RAM phones.
<srhb>
I think everyone agreed they're there. But not whether they were sufficient.
<ikwildrpepper>
domenkozar: just sharing my opinion ;)
<globin>
Yes I wouldn't really call it abusive, but it's definitely passive agressive
<ikwildrpepper>
domenkozar: I agree though that this is likely that other dude, and the 0x dude was definitely abusive before
<gchristensen>
IMO, I don't think domenkozar's comment was wrong, but I think being more restrained in the first reply would have been an improvement, since he was going to reply like that pretty much no matter what you said.
<srhb>
And I'm in favour of tossing the personar after just the first anyway. Keeping bad people out of nice communities is a necessity.
<gchristensen>
but whatever, what is done is done
<gchristensen>
and they showed their colors
* srhb
nods
<ikwildrpepper>
gchristensen: yeah, I don't think the outcome would have been different, if this is actually 0x....
<Gravious>
i sympathise though, it can be really difficult to work on hard on something and see it go unnoticed
<gchristensen>
Gravious: this person doesn't actually produce anything but angry issue comments
<gchristensen>
^ slight hyperbole
kiloreux has joined #nixos
<gchristensen>
^ but mostly true
kiloreux_ has joined #nixos
k0001 has joined #nixos
<nh2>
domenkozar: I would not block anybody on github, it's generally counterproductive. Thumbsdown emoji and ignoring is typically better
<ToxicFrog>
It says something about how long this has been an issue that I had already referenced the corresponding bugs in nixos-homepage and then completely forgot about them
<gchristensen>
I strongly disagree
<ToxicFrog>
I should just figure out how to hack on the website and fix it already
<ToxicFrog>
Even if those fixes never get merged I can at least run it locally
arianvp2 has quit [(Read error: Connection reset by peer)]
<nh2>
gchristensen: with whom, with me?
<gchristensen>
yes
arianvp2 has joined #nixos
<domenkozar>
nh2: this guy is a big pain, it's not worth the effort
<Gravious>
gchristensen, i say this because i just walked into another channel and vented a bit about closing a pr that was a lot of work and had been open for 3 months without a single comment
<ToxicFrog>
...apparently at some point I wrote a fix for this and then never created a PR for it
<ToxicFrog>
wow
<gchristensen>
Gravious: yeah that sucks
<ToxicFrog>
I must have gotten interrupted by toddler or something
<domenkozar>
but we can wait until things escalate
<domenkozar>
I'd rather ban him before we lose another contributor
<Gravious>
gchristensen, a little, but when you talk to people you just realise how incredibly busy they are, it's just hard to appreciate that from silence
<gchristensen>
^
<domenkozar>
even if he *sometimes* is constructive
<domenkozar>
what counts is people, not fixing/commenting
romildo has joined #nixos
<gchristensen>
Gravious: yeah I know, I definitely understand the author's frustration. the person who we're discussion regarding that PR isn't the author though, just a perpetually unhappy issue troll
<niksnut>
domenkozar: I think he's 0xABAB too
<domenkozar>
anyway we can wait if you'd like
<domenkozar>
as long as your eyes are open :)
grburst has quit [(Ping timeout: 240 seconds)]
<nh2>
domenkozar gchristensen: communities that block negative voices are unhealthy; healthy communities can handle such people. In the Haskell community we also recently had a case where some people blocked other voices on github and it reflected very badly on them
<domenkozar>
nh2: that's a different case, they are far more complicated in terms of politics
<domenkozar>
we don't even want to go there
<gchristensen>
nh2: I guess you're welcome to manage your communities that way, but I'm not interested in being part of a community where it is okay for people to send personal attacks and threats to other contributors over github issues and email.
<domenkozar>
in Haskell there are two groups disagreeing what should be banned based on political views
<domenkozar>
I hope we never get to that point
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor opened pull request #27384: vim_configurable: add ".../after" plugin dirs to runtimepath (master...vim-plugin-after) https://git.io/vQ7ud
NixOS_GitHub has left #nixos []
<__Sander__>
we can also just ignore what he says
<domenkozar>
it's not that simple, he is driving people away
<shlevy>
+1 for blocking/banning
<domenkozar>
ignoring is not enough
<domenkozar>
it works for some trolls, but not all
<__Sander__>
this guy is not secretely terry a. davis or something? :D
<shlevy>
Successful communities (in person or online), even ones generally in favor of welcoming outsiders, absolutely have mechanisms for turning away toxic elements. A private community isn't the same thing as a government and different standards should apply.
<gchristensen>
nh2: there are a few on github, and a few in my inbox
arianvp2 has quit [(Read error: Connection reset by peer)]
<domenkozar>
but style of comments and timing aligns, not that correlation is causation though.
<domenkozar>
not implying*
<gchristensen>
I am :P
<joepie91>
nh2: there is a point where you need to consider what you prioritize more: 1) active contributors not burning out or 2) giving everybody a chance
orivej has quit [(Remote host closed the connection)]
<joepie91>
while I dislike banning in general, there are some cases where somebody just doesn't leave any other option
<joepie91>
other than letting them drag down the community by being unnecessarily hostile
<nh2>
domenkozar gchristensen joepie91: this comment, too, I would recommend to handle by ignoring. Somebody whines on the Internet, so what? IMO that's not a good enough reason to implement censorship. Calrama's comment at https://github.com/NixOS/nixpkgs/issues/27258#issuecomment-313974449 shows that other nix contributors can identify idiots and not get discouraged by them.
<gchristensen>
nh2: okay
romildo has quit [(Quit: Leaving)]
<domenkozar>
I'm also very much for giving people a chance and I do belive they change, but if this guy created another user to continue his behavior, we're just not there
<joepie91>
nh2: 'ignoring' is a misnomer. these kind of comments cannot be ignored, they can only be not-responded-to; but that does not mean that it does not affect the mental state of contributors.
<joepie91>
nh2: and getting discouraged by hostile behaviour is a long-term process; it's not something you can pin down on any one particular comment.
<joepie91>
nh2: so yes, the first attempt or even attempts (plural) should be to communicate with a problematic user and try to get them to change their behaviour towards something more constructive. but if somebody is actively hostile towards that, and is producing nothing but noise, then at some point the chances run out.
<gchristensen>
this is a great talk about how open source that is relevant to this conversation, and I recommend people watch it :) https://www.youtube.com/watch?v=PlagRkD4e8s
<domenkozar>
nh2: I do agree that we shouldn't randomly ban people though, I think so far in all the history of NixOS it was 2 users
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 closed pull request #27160: make-autostartitem: Use the right path for storing the desktop files (master...makeautostartitem_fix) https://git.io/vQVt8
NixOS_GitHub has left #nixos []
<nh2>
joepie91: Calrama's mental state seems totally fine; he has identified whether he should care about that user's comment or not
<joepie91>
nh2: you're missing my point. you're still trying to infer consequences from a single comment, about a single person.
<joepie91>
nh2: whether generally disruptive and hostile people are kept around, is a key factor in whether contributors will consider the community a welcoming environment to contribute to. if there are enough incidents where somebody is hostile without impunity - and that can be hostility towards *anybody* on public display - then people will stop contributing, because the cost of a contribution is too high to them.
<joepie91>
er
<joepie91>
with impunity *
<joepie91>
this is not something you can infer from a single post, or a single person's response, or any one single bit of data. but the moment somebody starts riling other people up with little to no constructive value, that's the first warning sign that there's a situation you need to deal with. that doesn't mean you need to immediately resort to bans, but you can't just wave it away as "just ignore it" either
<joepie91>
lots of small hostilities will eventually create an overall hostile environment, and people will leave it (often without telling you).
<gchristensen>
we gave him a whole year before we banned him
newhoggy has joined #nixos
<nh2>
domenkozar: I agree this is problematic but I still think that the better approach is to publicly show that most people disagree with "0xABAB"'s opinion on the matter, not to take their ability to state their opinion away. And instead tell the contributor to ignore that user and continue their good work.
<joepie91>
nh2: so really, the metric you want to use is "was this person acting in good faith?" -- even people with good intentions will occasionally make hostile remarks without intending it. but if somebody is clearly showing bad faith and refusing to change their behaviour, then it's time to remove that person from the conversation.
<gchristensen>
all I can say is "ok", nh2. I hear your opinion.
<joepie91>
nh2: the problem isn't the opinion, though, it's the presentation.
<joepie91>
it's unnecessarily hostile.
<nh2>
joepie91: if communities banned "unnecessarily hostile" things then censorship would rule the world, especially in those countries currently known for free speech
<nh2>
also, looking at that specific post, I see no big problem with somebody coming and say "the quality of this software is shit". The correct answer to that is "OK, we don't care about that or your opinion on it for packaging pursposes"
<joepie91>
nh2: no, it would not./
<gchristensen>
free speech isn't a right in open source projects. you can say what you want, but a project owner isn't required to let you stay
<joepie91>
and I'm really not sure how you're drawing a line from "expecting people not to be unnecessarily hostile" to "censorship".
<domenkozar>
nh2: it's hard to say to each new contributor: hey, we're a welcoming community, but ignore these 5 people
<domenkozar>
because that's what blocking a person does without instructing people
<joepie91>
nh2: btw, emphasis on "unnecessarily".
arianvp2 has quit [(Ping timeout: 260 seconds)]
<gchristensen>
ok I'm going to leave this conversation now, good luck
arianvp2 has joined #nixos
<LnL>
yeah, was just ignoring him but I've seen some contributes get frustrated by his reviews because they don't know
<joepie91>
nh2: in the end, the primary purpose of an open-source project is to maximize the contributions and quality thereof, not to provide a platform for arbitrary speech presented in arbitrary ways. therefore, if those two goals conflict, then prioritizing contributions over arbitrary speech is the correct thing to do.
<joepie91>
and if somebody is harassing constructive contributors... yeah, well.
<nh2>
gchristensen: sure, I didn't suggest there was a right in anything, and surely the one with moderator perms has the eventual decision power. But what I'm saying is that it's a good thing to have, not a right, and it's healthy for a community to be able to deal with idiots on the Internet.
gchristensen has left #nixos ["WeeChat 1.7.1"]
<joepie91>
nh2: but the reason for existence of an open-source project isn't to give people practice in dealing with idiots on the internet.
<joepie91>
forums do a pretty good job of that already.
cpennington has joined #nixos
ixxie has quit [(Ping timeout: 240 seconds)]
<nh2>
joepie91: I agree with you on what the purpose is, but I certainly put more trust (and thus willingness to contribute) into communities that can handle nasty people without pulling out the ban hammer.
ixxie has joined #nixos
<joepie91>
nh2: and like I said, a banhammer shouldn't be the first option, and it isn't the first option here. but there comes a point where that is the only option left.
<joepie91>
and at that point, it's a completely valid measure.
<domenkozar>
> What can work is to limit and work to repair the damage that psychopaths do. Like bullies at school, psychopaths do not suffer from depression. It is family, friends, and colleagues who pay the cost. Once you see the process behind the psychopaths' impact craters, you can intervene.
<domenkozar>
> Intervention is not simple. To deal with people who have spent their lives charming, manipulating, and bullying others is by default impossible. If you try to warn a group, you will find yourself blamed. If you try to warn individuals, you will find them turning against you. You must move slowly, carefully, and with the right knowledge.
<joepie91>
nh2: to be clear, I'm all for targeted criticism, including blunt criticism (hell, I give that out myself plenty often), but that is a completely separate matter from the way that criticism is presented - if it involves unnecessary personal attacks, hostility, etc., then it's completely valid to say "no" to that.
<nh2>
joepie91: OK. Also note I'm arguing purely from the standpoint of what I've seen in the two posts linked. If there are more that I don't know about then surely I'm not in the picture and banning might be appropriate. Only thing I'm saying is that what I've seen *so far* seems to be below the threshold of what an OSS community should be able to deal with
<ToxicFrog>
nh2: the thing is, some "nasty people" cannot be handled without "pulling out the ban hammer", either because they consider themselves so unassailably in the right that attempts to change their behaviour just result in them doubling down, or because being an asshole is an end in itself for them and they have never been acting in good faith.
<joepie91>
nh2: what an OSS community "should be able to deal with" is contributions and bug reports in good faith. for anything beyond that, there's absolutely no obligation to deal with it in any way other than to show somebody the door.
<ToxicFrog>
A community that has never banned anyone is either lucky enough never to have encountered these people, or is dominated by them.
nwspk has quit [(Ping timeout: 255 seconds)]
<domenkozar>
nh2: we have emails from that users that strongly suggest it's a cause not worth fighting for
<ToxicFrog>
In the latter case, because once you get a critical mass of assholes they end up driving out most other contributors.
<joepie91>
(btw, a bug report is a contribution in and of itself, I'm just naming it separately because often people don't perceive it as such and I want to make sure I'm not misunderstood :P)
<nh2>
domenkozar: just to be clear, I'm not fighting for that user, I'm already as convinced as you are that the person is an idiot. I'm solely argueing for the approach to deal with that kind of idiocity
nwspk has joined #nixos
erictapen has quit [(Ping timeout: 260 seconds)]
<domenkozar>
nh2: ah ok :)
mbrock has joined #nixos
<mpickering>
I am trying to override a package in the haskell package set but it seems the override is not applied somehow when using ghcWithPackages
<mpickering>
I see my mistakes, I was using haskellPackages rather than the right package set
<nh2>
domenkozar joepie91: Also I do see the problem that new contributors may not be able to easily identify these types of people, and that they *can* be discouraged if they can't do so. Github definitely doesn't make that easy. On the other flame locations on the Internet (e.g. Reddit) you can at least click on a user and quickly get an opinion on what they do and what their usual tone and responses to them are. So that's a point supporting your a
<joepie91>
nh2: message got cut off after "supporting your a"
<nh2>
approach
<joepie91>
right
justelex has joined #nixos
<joepie91>
even with that kind of feature, though, there will be a chunk of contributors that will be put off by hostile responses and that'll just quietly leave and never return, without ever bothering to look into the history of somebody
<nh2>
joepie91: OK but has some reasonable level of nastiness resistance. Otherwise I'd we'd be clicking on spam all day, and get depressed by the angry kids throwing us swear words in online games. I don't think people are _that_ soft
takle has joined #nixos
ndrei has joined #nixos
<joepie91>
nh2: don't make the mistake of assuming that your own nastiness-resistance is representative of that of others :)
<joepie91>
a good friend of mine is brilliant from a technical perspective, but he completely avoids online social venues (including generally not contributing to open-source projects) because he doesn't want to have to deal with the crap that almost inevitably gets thrown around
<joepie91>
these are the people you will never hear of because they never come around to your place in the first place
<joepie91>
so your view of how nastiness-resistant people are, is biased by the people who can survive in whatever environment you've created
<ToxicFrog>
nh2: ^
<joepie91>
that's not to say that we must all be perfectly polite all the time, because sometimes there are just good reasons to eg. be blunt (think on-point criticism), but you shouldn't assume "oh people can deal with the nasty stuff anyway" if it isn't necessary to have that nasty stuff in the first place
<nh2>
domenkozar: so yeah, I guess my plea is: if you decide to ban whatever users, please make sure to put a thorough justification somewhere so that it's clear to observing third-parties (who may not have the seen the whole picture yet) what the whole picture is, so that they can themselves consider it well-justified and do not think that the nix community pulls out the ban hammer too easily (stating your earlier mentioned statistics of 2 banned us
arianvp2 has quit [(Read error: Connection reset by peer)]
arianvp2 has joined #nixos
<joepie91>
that might be a good idea, in particular to prevent misrepresentation on the part of banned users acting in bad faith with a grudge :P
gchristensen has joined #nixos
<nh2>
joepie91: yes
<gchristensen>
I'd be okay with their emails to me being published
<joachifm>
gchristensen: people are sending you nasty emails? wow ...
<nh2>
gchristensen: that's a good idea, I think that sends a strong signal of "it's pointless to try and harrass our community members privately, because we will expose such behaviour publicly and even more people will think you are an idiot then"
<niksnut>
hehe: "The only reason nixos.org package search exists, is because they want to spy on you."
<gchristensen>
niksnut: where did that get posted?
<LnL>
ibrahims: ~/.nix-defexpr/etc/profile.d/nix.sh should set that for you
* mbrock
wishes he could get NixOS to work on Raspberry Pi 3
obadz has quit [(Ping timeout: 240 seconds)]
obadz has joined #nixos
bennofs has joined #nixos
dbmikus has joined #nixos
pie_ has quit [(Remote host closed the connection)]
pie_ has joined #nixos
<ibrahims>
LNL: i had a weird fish shell setup, somehow it didn't export that variable. thanks for the point. you rule!
<nh2>
how does nix decide which paths to download and which ones to build?
<LnL>
ibrahims: ah yes, I don't think the installer work for fish out of the box
<nh2>
I have an issue where it puts a package into "these paths will be fetched" but then fails to download because that path doesn't exist on cache.nixos.org
<nh2>
so I wonder how it knows ahead of time what can be fetched
<mbrock>
ah, yeah, I just realized I used another .img
<mbrock>
well, I used the .img from that URL you posted, but the GitHub thread links to another one, called `sd-image-aarch64-linux.img'
<disasm>
domenkozar: I've met a few people like 0xABAB/butterfyla in my life. Some of them are really bright (usually on the aspergers spectrum), others are just plain trolls that like to complain about everything and do nothing... Regardless of which type, if they can't conform to the community rules, both can be toxic.
ericsagnes has quit [(Ping timeout: 255 seconds)]
<copumpkin>
(I'd try to keep "medical diagnoses" out of any conversations like these)
<ixxie>
mbrock: well, dezgeg is around here sometimes, so if you still get stuck after trying the other thing they may be able to help out
yegortimoshenko has joined #nixos
<disasm>
copumpkin: not suggesting he has a medical diagnosis, just that I know people that have been diagnosed that act like that.
<gchristensen>
ok let's move on from that particular part
<mbrock>
thanks. getting NixOS on the RPi seems kind of crucial for being able to maintain a RPi-based widget kind of thing
<yegortimoshenko>
is it possible to use nix-build to build one result symlink with multiple packages in it?
<gchristensen>
yegortimoshenko: yep, buildEnv
<yegortimoshenko>
gchristensen: thanks!
magnetophon1 has joined #nixos
nh2 has quit [(Quit: Leaving.)]
xadi has quit [(Ping timeout: 240 seconds)]
arianvp2 has quit [(Quit: arianvp2)]
phinxy has joined #nixos
* ToxicFrog
catches up on #17126, wait what
ericsagnes has joined #nixos
cpennington has quit [(Ping timeout: 268 seconds)]
nslqqq has quit [(Ping timeout: 246 seconds)]
ixxie has quit [(Quit: Lost terminal)]
xadi has joined #nixos
<copumpkin>
niksnut: I now have a couple of shady things I want to do with scopedImport
<globin>
copumpkin, domenkozar, niksnut: how are we continuing to handle him? just ignore him?
<copumpkin>
globin: I don't know. If he's had a warning before (I haven't been tracking all his conversations) I'd probably ban, but otherwise stand by the current warning and see where it goes from there
<gchristensen>
niksnut: can you ban them?
<copumpkin>
niksnut: I'm starting to think a more useful primitive is something like "deepScopedImport" might be easier to implement (with caching) and cover 90% of use cases (and all of mine)
<shlevy>
copumpkin: ooh say more
<shlevy>
I've been annoyed at the limits of scopedImport since it was added
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] ttuegel pushed 2 new commits to master: https://git.io/vQ7Pr
<NixOS_GitHub>
nixpkgs/master 640137d Thomas Tuegel: dropbox: 29.4.20 -> 30.4.22
<NixOS_GitHub>
nixpkgs/master 8c9273a Thomas Tuegel: Merge pull request #27382 from ttuegel/update/dropbox...
NixOS_GitHub has left #nixos []
<copumpkin>
so one of the first combinators I wrote over scopedImport was a recursive thing that replaces a few functions all the way down an import chain
<copumpkin>
we can already do that with scopedImport
<gchristensen>
no
<shlevy>
gchristensen: scopedImport has existed in nixpkgs for quite some time :D
<copumpkin>
but it's a bit messy and scopedImport is slow
<shlevy>
s/nixpkgs/nix
<gchristensen>
I know ... but I didn't realize scopedImport could override builtins :o
<gchristensen>
ok time to start a code obfuscation contest
<copumpkin>
anyway, I'm using it for the deterministic nix evaluation stuff
<clever>
gchristensen: it can also override NIX_PATH
<gchristensen>
clever: that is what I knew it ctould do
<copumpkin>
but I'm also thinking it'll be a crucial ingredient in the internal monorepo conversation
<copumpkin>
that I was having with catern yesterday
<joepie91>
copumpkin: monorepo conversation?
<copumpkin>
yeah, how best to manage an internal big repo of code using Nix
<joepie91>
ah
<joepie91>
my intuitive response would be "don't have an internal big repo of code" :P
<gchristensen>
rm -rf nixpkgs
<copumpkin>
even if you don't, there's still issues with managing your own code with nix
<copumpkin>
it's almost easier to do FOSS with it than your own projects
<copumpkin>
I mean manage other FOSS than your own
<copumpkin>
we don't have a good solution to deuplicating the .nix files inside the project and the ones in your central repository of packages
eacameron has joined #nixos
<copumpkin>
so you can either manage releases and bump hashes all the time
<copumpkin>
and do import-from-derivation
<copumpkin>
or you can have a monorepo-ish setup
<copumpkin>
both of them are kinda painful
<copumpkin>
or at least not as smooth as I think they should be
jgeerds has quit [(Quit: Leaving)]
nslqqq has joined #nixos
<copumpkin>
the monorepo setup has the advantage that you can use local source references (e.g., src = ./src; in your mkDerivation calls) but the disadvantage of making creating a channel a real PITA
<copumpkin>
unless you want to bundle all your source code into the channel
newhoggy has quit [(Remote host closed the connection)]
<copumpkin>
anyway, will write more thoughts later, should probably do some work
<gchristensen>
last job we went the route of per-project repos and it was hell, we had to use AOSP's `repo` to manage it all, and it was pure, concentrated pain
<copumpkin>
anyway, I believe scopedImport can make this all a lot better ;)
<copumpkin>
yup
<gchristensen>
and our bill for github was a small fortune, despite each repo being small
radvendii has joined #nixos
<domenkozar>
copumpkin: my 2c: I suggest not having further discussions with butterflya
<gchristensen>
ToxicFrog: ^
<Eisfreak7>
Is there a nice sandboxing solution for nixos? The ability to determine exact dependency graphs should make for a nice sandboxing feature, shouldn't it?
<copumpkin>
domenkozar: yeah, I'm not going to continue
<copumpkin>
Eisfreak7: at runtime? not really
<copumpkin>
Eisfreak7: at build time definitely and it already exists
<domenkozar>
Eisfreak7: you can use systemd sandboxing
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor pushed 1 new commit to master: https://git.io/vQ71K
<Eisfreak7>
Hm yes I was talking about runtime. I was thinking something like firejail integration: I can enable firejail globally / on a per application basis an nix generates "firejail wrappers" that automatically call the binaries with firejail and an appropriate config
<joepie91>
+1 to ban, clearly showing no intent to change
<joepie91>
but doubling down instead
<gchristensen>
I mean, it is literally the same person who is already banned, not sure why we need to vote
<joepie91>
also, fwiw, domenkozar, that first link suggests they're Dutch -- the :') emoticon is particularly commonly used in NL due to a Dutch forum (Fok!)
<joepie91>
in case that may be a useful data point
<joepie91>
(I rarely see it used by non-Dutch people)
<copumpkin>
gchristensen: they are? 0xABAB is but butterflya still seems able to post
<domenkozar>
this guy is so advanced, he has 1 year old github accounts for backup
<domenkozar>
I mean, seriously
<gchristensen>
I know but it is ban evasion
newhoggy has joined #nixos
<gchristensen>
I don't think I've seen such blatant ban evasion since I +o'd ##php like 10 years ago
<copumpkin>
anyway, no objection from me
<shlevy>
+1 to banning, -1 to voting. I trust our org admins to make these calls and I don't trust the random selection of irc users who have opinions on this :D
<gchristensen>
shlevy++
<domenkozar>
well I wouldn't say random, but I agree
<shlevy>
If we want to ensure community feedback, transparency around decisionmaking and a clear forum for surfacing issues with the way decisions are made may make sense. Crowdsourcing decisions as they happen, not so much :D
<copumpkin>
I don't actually know who has the power to ban :) is it just niksnut and ikwildrpepper?
<copumpkin>
time to bring up the usual CoC discussion ;)
<domenkozar>
I can as well
* copumpkin
runs
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] globin pushed 1 new commit to openssl-1.1: https://git.io/vQ7Mp
<NixOS_GitHub>
nixpkgs/openssl-1.1 6aa2c14 Robin Gloster: opendkim: add patch for openssl 1.1
NixOS_GitHub has left #nixos []
<globin>
domenkozar: +1
<domenkozar>
banned
sibi_ has quit [(Quit: Connection closed for inactivity)]
yegortimoshenko has quit [(Read error: Connection reset by peer)]
<__Sander__>
makes me wonder when the next dubious person shows up
<__Sander__>
that happens to be him :)
<__Sander__>
for a strange reason, I think this guy is enjoying what he is doing
<gchristensen>
thank you, domenkozar
* ToxicFrog
gnaws irritably on packages.tt
<ToxicFrog>
Why is Javascript.
<bennofs>
hmm, what was the problem with 0xABAB? I mean, it looks like most of his review comments were pretty bad / useless / wrong, but it doesn't look to me as if he did it on purpose
<niksnut>
domenkozar: thanks
<__Sander__>
and by responding to all his provocations, we encourage him to keep going
<copumpkin>
bennofs: pretty active disparagement in some of them
<niksnut>
he has been banned from various other projects as well
<copumpkin>
bennofs: usually if someone disagreed with the review, it went downhill quickly
<gchristensen>
domenkozar: oh did you see 1.11.13 finished? :)
<bennofs>
work-stealing scheduler :)
<domenkozar>
gchristensen: I need to run now, hopefully can test in a few hours
<gchristensen>
dang!
<gchristensen>
maybe niksnut can release anyway, it is already an improvement
<bennofs>
gchristensen: do a few hours make such a big difference ?
<srhb>
bennofs: It's exciting! :-)
orivej has quit [(Ping timeout: 248 seconds)]
newhoggy has quit [(Ping timeout: 240 seconds)]
<gchristensen>
bennofs: depends how many people are failing to use it, but probably not too big
<ToxicFrog>
why does jquery not have an escape_text function? Why doesn't *javascript*?
<ToxicFrog>
AFAICT the "best" way to do this is to create a temporary <div> node, stuff the text into it, extract the HTML, then throw the node away
<niksnut>
doing multiple releases per day looks so brown paper bag :p
<domenkozar>
yeah I want to test this first :)
<domenkozar>
otherwise we'll end up with Nix 1.11.99
<domenkozar>
<3 gchristensen
<copumpkin>
Nix 1.11.-1
<domenkozar>
1.11.-256
<copumpkin>
8 bits per version component ,and it's signed
<__Sander__>
1.11.111
<__Sander__>
that would be an awesome version number :D
<joepie91>
ToxicFrog: innerText
<ToxicFrog>
joepie91: yeah, but again, that requires a DOM node
<joepie91>
ToxicFrog: ...sure? but why would you need to escape HTML if not for use in a DOM node?
<gchristensen>
domenkozar: :P
<catern>
wow scopedImport seems real neat
newhoggy has joined #nixos
<ToxicFrog>
joepie91: because I want to do other stuff with it before inserting it into the node.
<gchristensen>
__Sander__: as long as we don't get to that python dependency which was asymptotically approaching 1, ie: 1.11.999999
<joepie91>
ToxicFrog: what kind of stuff specifically?
<LnL>
oh god, it still didn't bump it's version?
<__Sander__>
haha :D
<gchristensen>
domenkozar: I made this image for you, to encourage the ban, but I was too slow: http://i.imgur.com/LoBRVlG.png (the joke is the phrase on the show is "Move That Bus")
<catern>
copumpkin: I don't know that much Nix language idioms yet so I don't know to what extent I can participate on my own merits in the monorepo conversation :)
<ToxicFrog>
joepie91: what I'm actually working on here is the nixos package search license display. A license is either a string (needs to be htmlescaped and displayed), an object (needs to have a name extracted and escaped, then optionally wrapped in <a> if a url is also present), or an array of strings and/or objects (map as above, then interpose ', ')
<ToxicFrog>
I wonder if it would be easier to just emit a TextNode for the string versions, text or <a> for the object versions, and use .map().interpose($("<div/>").text(", ")) to staple everything together for the last version
newhoggy has quit [(Ping timeout: 240 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] qknight closed pull request #27276: postfix: complete remake of postfix service (master...postfix_redone) https://git.io/vQPmZ
NixOS_GitHub has left #nixos []
<joepie91>
ToxicFrog: I can see this being an issue in case 3 with hyperlinks - however, I've never run into that kind of issue since I wouldn't string-concatenate things together, but rather represent each license as a separate DOM node (probably a <ul> with <li>s), using CSS to affect the display to be a single comma-separated line
<joepie91>
that is, take the semantic approach
<clever>
joepie91: there is also document.createTextNode i believe
<joepie91>
(in which case this issue doesn't exist since you can just assign the text to the hyperlink nodes)
<ToxicFrog>
I don't know enough about CSS to do that. I'm just trying to improve the current display, which omits license info for the vast majority of packages.
<clever>
joepie91: you could also make a single ul, containing multiple text nodes
<joepie91>
(could make it snazzier with margin etc. but meh)
newhoggy has joined #nixos
<joepie91>
clever: items within a <ul> must be <li>s
<clever>
var ul = document.createElement("ul"); ul.appendChild(document.createTextNode("foo")); ul.appendChild(document.createTextNode(", ")); ....
<ToxicFrog>
Hmm. What's javascript for `apply`?
<joepie91>
yeah, that is invalid :)
<clever>
within the li i mean
<joepie91>
don't really see the value of that though
<clever>
var li = document.createElement("li"); ul.appendChild(document.createTextNode("foo")); ul.appendChild(document.createTextNode(", ")); ....
<joepie91>
ToxicFrog: can you rephrase that question? :P
<ToxicFrog>
apply(f, [1,2,3]) is eqv to f(1,2,3)
<ToxicFrog>
It unpacks an ordered collection into the arguments of a function.
arianvp2 has joined #nixos
<joepie91>
right
<joepie91>
in JS that'd be f.apply(null, [1, 2, 3])
<joepie91>
the first argument being the `this` context
<joepie91>
(ie. to affect the value of `this` within the function call's scope)
xadi has quit [(Remote host closed the connection)]
<joepie91>
ToxicFrog: in ES6 there's also f(...[1, 2, 3]) but I doubt that you're using Babel + a bundler, so that would probably not be a viable option
<joepie91>
ToxicFrog: you should rarely need .apply in JS, though
<ToxicFrog>
Oh, it's a method on the function itself. That makes sense.
<ToxicFrog>
I probably don't need it here, but I got sidetracked
<joepie91>
right :)
mudri has joined #nixos
MoreTea has joined #nixos
newhoggy has quit [(Ping timeout: 258 seconds)]
xadi has joined #nixos
hellrazor has joined #nixos
<joepie91>
ToxicFrog: also, bit of advice when working with JS: never use constructors or 'classes'
<joepie91>
you don't need them, and they will just make things more difficult
<joepie91>
just use object literals, return them from a factory function where necessary, etc.
<MoreTea>
gchristensen, domenkozar did you have a location for the multi-user osx installer?
hyphon81 has quit [(Remote host closed the connection)]
<MoreTea>
one can run at least one virtualized VM right?
phinxy has quit [(Read error: Connection reset by peer)]
ted_ has quit [(Client Quit)]
erictapen has joined #nixos
<ToxicFrog>
joepie91: will do, thanks
<gchristensen>
MoreTea: 1 apple machine allows you 1 running copy of OSX on that machine
Itkovian has quit [(Quit: My MacBook has gone to sleep. ZZZzzz…)]
<MoreTea>
so, if we'd only use them for integration tests, we should still be OK, right?
<ToxicFrog>
I had this working fine for text, but then someone in review pointed out that it will not properly handle license names that require escaping, and now there's spiders everywhere.
<gchristensen>
MoreTea: that is where I get to "It'd be a bit of a waste to have a whole, capable Apple machine not running full-bore on other x86_64-darwin Hydra jobs."
<MoreTea>
They should not exclusively build tests?
<joepie91>
ToxicFrog: this sounds about right for programming :)
<joepie91>
(in general)
<MoreTea>
dinner, brb.
contrapumpkin has quit [(Ping timeout: 260 seconds)]
muyfine has joined #nixos
radvendii has quit [(Ping timeout: 260 seconds)]
<pietranera>
Hi I would like to install terraform 0.9.11 with nix-env, but an older version is provided by nixpkgs (NisOS stable). The terraform derivation is just a call to buildGoPackage, and I don't know how I can create an overlay where the version and sha256 is overridden.
<copumpkin>
pietranera: you could also just pick up nixos-unstable which has a newer version
Infinisil has joined #nixos
<copumpkin>
that doesn't have to update your entire system
<copumpkin>
you can mix and match packages
newhoggy has joined #nixos
<pietranera>
copumpkin thanks
<pietranera>
I wanted to see whether the overlay mechanism worked
contrapumpkin has joined #nixos
erictapen has quit [(Ping timeout: 240 seconds)]
<pietranera>
I will try and add the unstable channel then.
arianvp2 has quit [(Read error: Connection reset by peer)]
<pietranera>
as I didn't find a way to override the arguments of buildGoPackage
arianvp2 has joined #nixos
<pietranera>
(still happy to take suggestions though)
Mateon2 has joined #nixos
<copumpkin>
TBC, the override can probably work too, but I haven't looked
magnetophon1 has joined #nixos
Mateon1 has quit [(Ping timeout: 248 seconds)]
Mateon2 is now known as Mateon1
k0001 has quit [(Remote host closed the connection)]
k0001 has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor opened pull request #27387: nixos/libvirt: prevent OVMF path from being garbage collected (master...libvirt-fix-ovmf-gc) https://git.io/vQ75H
<vaibhavsagar>
actually even if it's a nixops deploy, you should be able to ssh into it and run nix-rebuild
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nix] grahamc opened pull request #1466: Multi-user installer for Darwin (master edition) (master...nix-master-darwin-install) https://git.io/vQ7x5
NixOS_GitHub has left #nixos []
__Sander__ has quit [(Quit: Konversation terminated!)]
<himmAllRight>
Does anyone know how I can install the needed Latex binaries in nixOS?
yegortimoshenko has joined #nixos
07IAAHG20 has joined #nixos
<07IAAHG20>
[nix] edolstra closed pull request #1466: Multi-user installer for Darwin (master edition) (master...nix-master-darwin-install) https://git.io/vQ7x5
07IAAHG20 has left #nixos []
17SABIOYV has joined #nixos
<17SABIOYV>
[nix] edolstra pushed 18 new commits to master: https://git.io/vQ7pL
<17SABIOYV>
nix/master a0ad8ba Graham Christensen: Shellcheck the existing installer
<17SABIOYV>
nix/master fb40d73 Graham Christensen: Switch to a fancy multi-user installer on Darwin
<17SABIOYV>
nix/master 2b5ab03 Graham Christensen: multi-user install: move the profile in to the nix etc/profiles.d output
17SABIOYV has left #nixos []
<gchristensen>
niksnut: so fast :D
newhoggy has joined #nixos
m0rphism1 has quit [(Ping timeout: 248 seconds)]
erictapen has joined #nixos
MoreTea has quit [(Ping timeout: 260 seconds)]
m0rphism1 has joined #nixos
k0001 has quit [(Remote host closed the connection)]
<vaibhavsagar>
himmAllRight: nix-env -qaP 'tex.*'
<vaibhavsagar>
to see the packages available to you
k0001 has joined #nixos
<vaibhavsagar>
texlive looks like the one you want
<himmAllRight>
vaibhavsagar: Thanks. Will I be able to connect to it through emacs (spacemacs really)
<vaibhavsagar>
I don't see why not
newhoggy has quit [(Ping timeout: 240 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nix] edolstra pushed 4 new commits to master: https://git.io/vQ7pd
<NixOS_GitHub>
nix/master fdc9da0 Eelco Dolstra: Avoid a call to derivationFromPath()...
NixOS_GitHub has left #nixos []
Filystyn has quit [(Remote host closed the connection)]
<himmAllRight>
vaibhavsagar: Okay. I'm trying the texlive-combined-full-2016 package. I had tried texmaker but that and emacs couldn't find the text binary
nh2 has joined #nixos
Filystyn has joined #nixos
Filystyn has quit [(Changing host)]
Filystyn has joined #nixos
Filystyn has quit [(Remote host closed the connection)]
erictapen has quit [(Ping timeout: 240 seconds)]
mudri has joined #nixos
orivej has joined #nixos
newhoggy has joined #nixos
Filystyn has joined #nixos
Filystyn has quit [(Changing host)]
Filystyn has joined #nixos
yegortimoshenko has quit [(Ping timeout: 268 seconds)]
<Filystyn>
clever added line in grube after hitting e -> init=/bin/sh
<Filystyn>
after this F10 and...
<Filystyn>
nothign happened
<Filystyn>
loged me normaly
vaibhavsagar has quit [(Ping timeout: 246 seconds)]
newhoggy has quit [(Ping timeout: 248 seconds)]
ambro718 has joined #nixos
newhoggy has joined #nixos
zeus_ has joined #nixos
justanotheruser has joined #nixos
newhoggy has quit [(Ping timeout: 255 seconds)]
Filystyn has quit [(Remote host closed the connection)]
justanotheruser has quit [(Client Quit)]
justan0theruser has quit [(Ping timeout: 240 seconds)]
justanotheruser has joined #nixos
lambdael has joined #nixos
<pbogdan>
anyone here on 17.03, using chromium, and wiling to do me a small favour? if possible I would like to see the output of chrome://sandbox/
newhoggy has joined #nixos
dfranke has quit [(Quit: Lost terminal)]
<gchristensen>
pbogdan: suid sandbox: no, namespace: yes, pid: yes, network: yes, seccomp-bpf: yes, seccomp-bpf tsync yes, yama lsm enforcing no
<gchristensen>
"you are adequately sandboxed"
pietranera has quit [(Quit: Leaving.)]
newhoggy has quit [(Ping timeout: 255 seconds)]
newhoggy has joined #nixos
<pbogdan>
gchristensen: thanks! does it say "PID name spaces" or "PID namespaces" for you, or was that exact output?
<gchristensen>
it was quickly retyped output :) it says "PID namespaces"
<gchristensen>
and the "Namespace: yes" one was "Namespace Sandbox: yes" ... I can copy-paste exactly what I have if you'd like :P
phinxy has joined #nixos
<pbogdan>
no, that's fine, was particularly interested in the pid one
<greglearns>
... I'm still trying to solve Rust on NixOS, so that I can deploy Rocket.rs web apps to AWS using NixOps....
<greglearns>
yesterday, tilpner and I tried to get a Nix-expression working, but... I couldn't get it working. Today, I'm exploring an alternate path...
<greglearns>
is it possible to compile a Rust app and then just have it copied to a NixOS machine (via NixOps)?
<gchristensen>
sure
k0001 has quit [(Ping timeout: 240 seconds)]
ibrahims has quit [(Ping timeout: 240 seconds)]
proteusguy has joined #nixos
riclima has joined #nixos
riclima has quit [(Client Quit)]
newhoggy has joined #nixos
[0x4A6F] has quit [(Ping timeout: 240 seconds)]
<Infinisil>
I'm just thinking about what nixpkgs should become over time
<Infinisil>
There are going to be more and more packages
<Infinisil>
And having all of that in a single git tree, not so sure about that
<Infinisil>
I'm finding lots and lots of packages which could be added to nixpkgs, but I don't want to clutter it, but that's somehow really what one should do today
newhoggy has quit [(Ping timeout: 268 seconds)]
<dash>
Infinisil: what problems do you predict from making nixpkgs bigger?
m0rphism1 has quit [(Quit: WeeChat 1.8)]
newhoggy has joined #nixos
<gchristensen>
does anyone remember who came up with the docker tools that created a layer per dependency?
<LnL>
that's a thing?
<gchristensen>
yeah
<gchristensen>
I mean it isn't, in nixpkgs
<gchristensen>
but it is a thing in their nixpkgs :)
<LnL>
interresting
radvendii has quit [(Ping timeout: 240 seconds)]
<catern>
aaaaaaaaaaaaaaaaaaaaaaaargh
<catern>
nix-dev is moving
<dash>
gchristensen: exarkun did something like that
<catern>
this is bad
<gchristensen>
why?
<gchristensen>
thank you dash
<LnL>
are layers cachable like that regardless of their base image?
<catern>
because afaik gmane still isn't accepting new mailing lists to translate to NNTP
<LnL>
yeah I thought so, not the image itself but that's not as big of a deal
nix-gsc-io`bot has quit [(Client Quit)]
newhoggy has quit [(Ping timeout: 240 seconds)]
<gchristensen>
bash: ls: command not found ... looks like I forgot coreutils
<LnL>
yeah or busybox
newhoggy has joined #nixos
<Infinisil>
dash: I'm not sure what problems this could bring actually, I'll just add whatever I feel like, we can deal with problmes when they arise
proteusguy has quit [(Ping timeout: 246 seconds)]
Mateon2 has left #nixos []
hotfuzz_ has joined #nixos
hotfuzz has quit [(Ping timeout: 240 seconds)]
newhoggy has quit [(Ping timeout: 260 seconds)]
mudri has quit [(Ping timeout: 240 seconds)]
proteusguy has joined #nixos
newhoggy has joined #nixos
<gchristensen>
" runAsRoot is a bash script that will run as root in an environment that overlays the existing layers of the base image with the new resulting layer, including the previously copied contents derivation. This can be similarly seen as RUN ... in a Dockerfile. " this is the most confusing sentence
newhoggy has quit [(Ping timeout: 240 seconds)]
ambro718 has quit [(Ping timeout: 255 seconds)]
newhoggy has joined #nixos
<greglearns>
has anyone compiled Rust with MUSL and then copied that to a NixOS machine as a binary?
<catern>
gchristensen: no, that won't work... gmane is pretty much dead to the world except for existing lists that are already being archived :(
newhoggy has quit [(Ping timeout: 248 seconds)]
newhoggy has joined #nixos
newhoggy has quit [(Ping timeout: 240 seconds)]
ChongLi has joined #nixos
hotfuzz_ is now known as hotfuzz
tokudan has joined #nixos
cpennington has quit [(Remote host closed the connection)]
newhoggy has joined #nixos
Supersonic112 has quit [(Ping timeout: 240 seconds)]
<schoppenhauer>
extend it inside some nix expression for, for example, some nixOps module
<schoppenhauer>
I mean for some nixOps vm specification
newhoggy has quit [(Ping timeout: 240 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor opened pull request #27389: [backport] OVMF: separate output for ovmf binaries (release-17.03...ovmf-backport) https://git.io/vQ53L
NixOS_GitHub has left #nixos []
<joachifm>
schoppenhauer: you can add your module to imports, e.g., { config, ... }: { imports = [ ./my-module.nix ]; config = ...; }
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] pbogdan opened pull request #27390: chrome/chromium: beta and dev fixes for release-17.03. (release-17.03...chrome-17.03) https://git.io/vQ53j
NixOS_GitHub has left #nixos []
newhoggy has joined #nixos
ertes has joined #nixos
athan has joined #nixos
newhoggy has quit [(Ping timeout: 240 seconds)]
<schoppenhauer>
joachifm: where?
<clever>
schoppenhauer: you can add an imports list to any nixos module, to extend the module list
<clever>
schoppenhauer: configuration.nix is another module
<clever>
schoppenhauer: nixops also takes a list of modules when you create a deployment
takle has quit [(Remote host closed the connection)]
radvendii has joined #nixos
takle has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor pushed 1 new commit to master: https://git.io/vQ5GN
<clever>
ison111: and the code behind services.http.enable sets wantedBy for you
MoreTea has joined #nixos
<clever>
ison111: to undo that, you would need to do systemd.service.httpd.wantedBy = lib.mkForce [];
<ison111>
So do I keep the rest of my configuration under services.httpd and just remove the "enable" line and put it under systemd instead? Or do I move my entire configuration under systemd.services?
<clever>
keep the existing config, including the enable line
<clever>
and add the line i gave above
<ison111>
alright, thanks
<clever>
if you remove enable, then the systemd service wont even be created
<ison111>
clever: not sure if I did it right. I set systemd.service.httpd.enable = true; and also false. When I had it set to true it seemed to still start up the services automatically, when I set it to false it prevented it didn't automatically start them, but now when I try to manually start them it says it's "masked".
newhoggy has quit [(Ping timeout: 240 seconds)]
fetch_ has joined #nixos
fetch_ has left #nixos []
<clever>
ison111: if you set enabled = false;, then nixos wont even create the service
<clever>
so there is nothing to start
<clever>
you must set enable = true; and add the line i gave above
<gchristensen>
fetchurl: it is curious to me that you regularly hang out in here, but as a different nick each time (abspath cryptsetup dns_ fetch_ fetchurl foobardns grayen guacamole_ inf-rec mongodb-native nixnode nixos-install nomachine phpfpm qmake robomongo subl3 xcb xcursor)
<clever>
ison111: yeah, that enable must be set to true
<clever>
ison111: and you must also add systemd.service.httpd.wantedBy = lib.mkForce [];
jtojnar has joined #nixos
<catern>
gchristensen: heh, good observation, that's cute
<catern>
must be all the things they are looking for help with :)
<catern>
or just different pieces of software
<fetchurl>
Why is it that Nix when a different hash has been found than expected and you correct this, that it will download the file again, while it should still have the previously downloaded file available? It should be able to check for the already downloaded in the store, right? Am I misunderstanding something?
newhoggy has quit [(Ping timeout: 255 seconds)]
<LnL>
fetchurl downloads things to the store, since the hash was not correct the build failed and the output path is incorrect
eacameron has quit [(Remote host closed the connection)]
<gchristensen>
catern: you are absolutely right! cat .weechat/logs/irc.freenode.#nixos.weechatlog | awk '$3 ~ /abspath|cryptsetup|dns_|fetch_|fetchurl|foobardns|grayen|guacamole_|inf-rec|mongodb-native|nixnode|nixos-install|nomachine|phpfpm|qmake|robomongo|subl3|xcb|xcursor/ {print $0}'
<LnL>
lol
radvendii has joined #nixos
contrapumpkin has quit [(Ping timeout: 260 seconds)]
<LnL>
all my generated python packages are green \o/
<fetchurl>
gchristensen: I did not know it tracked me like that, although I did notice that my IP was mentioned with my nickname, I just took nicknames of things I was working on, or figuring out at that time
MercurialAlchemi has quit [(Ping timeout: 240 seconds)]
<gchristensen>
fetchurl: that is cool :) I didn't grep by your IP, but your "user" name: d9672d13@gateway/web/freenode/ip.217.103.45.19 "d96.."
<gchristensen>
which I guess is a hash of your IP
<LnL>
and no more evaluation errors :D
<fetchurl>
gchristensen: I really have to try weechat some time, I just keep using webchat.freenode.net, because it good enough for my use
<gchristensen>
fair! no worries :) just was a surprising curiosity :D
greglearns has joined #nixos
<fetchurl>
I was suprised someone noticed!
oida_ has joined #nixos
jgertm has joined #nixos
<LnL>
yeah, the only one I've noticed is "so" he must have a good mention filter :p
<clever>
LnL: that one always messed with my tab-complete, and my mention filter only triggers if my name is at the very start of a line
<fetchurl>
But regarding my question, why has the choice been made to do it like that, I would have expected the source to be saved seperately from the build, because now I have had cases were it had to download gigabytes instead of just once a few hundred MB
<MoreTea>
fetchurl, you can use `nix-prefetch-url $URL`
<Infinisil>
Can I get internet in a --pure nix-shell?
<MoreTea>
it will download, add to the store, and print out the SHA
<LnL>
fetchurl: nix doesn't really know about what fetchurl is doing inside the build
<Infinisil>
Hmm
<Infinisil>
Okay it's the packages fault I think, never mind then
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor pushed 1 new commit to release-17.03: https://git.io/vQ5Ro
<NixOS_GitHub>
nixpkgs/release-17.03 1436cc4 Joachim Fasting: OVMF: separate output for ovmf binaries...
NixOS_GitHub has left #nixos []
<fetchurl>
MoreTea: Yeah, but you need to specify the correct URL and name right? I figure that, e.g. in my case downloading Factorio the name would be: /nix/store/gxrik76yn1kyi6j64mj5v3nbd1d7cznk-factorio_alpha_x64-0.15.30.tar.xz but then I would still have to reverse engineer the URL from the Nix sources, since I dod not see it mentioned in the output anywhere
<pbogdan>
how is trust of a binary cache established? is it a case of "I trust this entity" or is there something in nix that would enforce that contents of said cache can't be tampered with?
newhoggy has quit [(Ping timeout: 260 seconds)]
<fetchurl>
LnL: Do you mean it (fetchurl) can differ depending on its inputs, but if it stores it in the store, should this not matter, only that the name and hash match should matter right?
<MoreTea>
fetchurl, yes, but the command nix-prefetch-url takes the URL, and outputs the hash.
<LnL>
yes you could implement an alternative for fetchurl that uses ipfs or whatever that generates the exact same output
<LnL>
changing the url of fetchurl also doesn't change the output since the hash didn't change
jtojnar has quit [(Read error: Connection reset by peer)]
jtojnar has joined #nixos
<ToxicFrog>
joepie91: so I ended up using CSS as you suggested in the end and I hate CSS now.
<ToxicFrog>
But it works.
<joepie91>
ToxicFrog: if you hate CSS, that means you've learned something about styling :)
<joepie91>
(don't worry, most people do)
<joepie91>
turns out that styling content is a Really Hard Problem and the quirks that CSS has don't help with that
<fetchurl>
I will test nix-prefix-url, but from what I remember, it will still redownload the file, since the name did not match, so I download it, get the correct hash, fill in the correct hash, but then it redownloads, because the filename was different, but maybe I remembered wrong
<gchristensen>
you can pass --name
<gchristensen>
IIRC
<fetchurl>
yes, but then you need to know the name, reverse engineering it from the Nix source, or stopping the build process the moment it mentions it
<clever>
i usualy just increment a random number in the hash
<clever>
so its invalid, and wont match up with the old file
<clever>
then let nix-build fail
<fetchurl>
clever: That is exactly what I do, but then it downloads it twice, hence my initial question why it does that
<joepie91>
clever: 'parametric module' pattern, for maintaining certain scopes across modules/files
<clever>
fetchurl: there is also a new thing, nix-prefetch-url '<nixpkgs>' -A hello.src i think it was
<clever>
joepie91: that will download it, and then rename its entry in the store, to match the real hash
<clever>
so when you run the nix, it wont do a 2nd download
spinus has quit [(Ping timeout: 248 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] dezgeg pushed 1 new commit to staging: https://git.io/vQ5ES
<NixOS_GitHub>
nixpkgs/staging 5ff18c4 Tuomas Tynkkynen: libuv: Disable yet another test...
NixOS_GitHub has left #nixos []
<joepie91>
clever: mishighlight :)
<clever>
oops
<joepie91>
(nevertheless an important tidbit of knowledge)
<fetchurl>
clever: Thanks, I will look into that!
brodul has joined #nixos
<clever>
with normal fixed-output downloads in nix, you tell nix what the hash of $out will be ahead of time, and nix computes $out based on that hash, and ignores all build inputs
<clever>
and nix will enforce the data matching that hash
<clever>
with the new nix-prefetch-url -A, it will rename the output after the download is finished, so it always lands in a path matching its hash, even if the hash isnt what nix was expecting
<manveru>
oooh
<manveru>
how new?
<clever>
not sure, try it and see what happens
<Infinisil>
I have this package I'm adding, it's basically just a bunch of useful bash scripts, but each one of them has dependencies such as {curl,wget,fetch}, {vlc,mvp,mplayer}, and more. Should I just not make any of these included? Or should I include everything?
newhoggy has joined #nixos
<clever>
[clever@amd-nixos:~]$ nix-prefetch-url '<nixpkgs>' -A hello.src
<fetchurl>
Is it documented somewhere, googling it does not really help
<clever>
works on nixos-unstable
<manveru>
well, i don't see it in the manpage, that's why i'm asking :)
<clever>
fetchurl: its in the changelogs for nix
<fetchurl>
ah
jtojnar has quit [(Ping timeout: 260 seconds)]
Mateon2 has joined #nixos
<fetchurl>
Its there since 1.11, nice
newhoggy has quit [(Ping timeout: 240 seconds)]
spinus has joined #nixos
takle has joined #nixos
yegortimoshenko has joined #nixos
jtojnar has joined #nixos
Ivanych has quit [(Ping timeout: 248 seconds)]
takle has quit [(Ping timeout: 255 seconds)]
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor pushed 2 new commits to master: https://git.io/vQ5zj
<NixOS_GitHub>
nixpkgs/master 292827b Bjørn Forsman: nixos/libvirt: modify xml with xmlstarlet...
<NixOS_GitHub>
nixpkgs/master b8e109d Bjørn Forsman: nixos/libvirt: prevent OVMF path from being garbage collected...
NixOS_GitHub has left #nixos []
newhoggy has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor closed pull request #27387: nixos/libvirt: prevent OVMF path from being garbage collected (master...libvirt-fix-ovmf-gc) https://git.io/vQ75H
NixOS_GitHub has left #nixos []
yegortimoshenko has quit [(Ping timeout: 258 seconds)]
newhoggy has quit [(Ping timeout: 248 seconds)]
newhoggy has joined #nixos
takle has joined #nixos
yegortimoshenko has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] bjornfor pushed 2 new commits to release-17.03: https://git.io/vQ52W
<NixOS_GitHub>
nixpkgs/release-17.03 cf9bb65 Bjørn Forsman: nixos/libvirt: modify xml with xmlstarlet...
<NixOS_GitHub>
nixpkgs/release-17.03 7b390ba Bjørn Forsman: nixos/libvirt: prevent OVMF path from being garbage collected...
NixOS_GitHub has left #nixos []
newhoggy has quit [(Ping timeout: 260 seconds)]
<Infinisil>
How can I have a nix-shell with a package I just added to my local nixpkgs tree?
<Infinisil>
This works somehow, I'm having some problems with it though:nix-shell -E 'with import ./. {}; runCommand "h" {buildInputs = [ mypacket ];} ""' --pure
<gchristensen>
Infinisil: well because a directory isn't a function maybe
<nh2>
I'm having a nixops problem: When deploying from a NixOS machine (as root), I get the error "/nix/store/pbhxc864gqach922mgl4v03d9dff93h4-grub-config.xml:1: parser error : Document is empty". And indeed that file is empty. But the same path is not empty on my Ubuntu machine, which is weird because files with same hashes should have same contents, right?
<nh2>
I guess a "grub-config.xml" file should never be empty, right?
<Infinisil>
gchristensen: Huh?
<Infinisil>
I'm calling this in the nixpkgs checkout
<Infinisil>
Hold on, I'll demonstrate the problme
<gchristensen>
sorry, I was confused by the import ./.
<nh2>
clever: it happens when I run `nix-shell -p nixops` as root and use the nixops provided by that environment. danbst found that problem originally (cc aszlig)
DutchWolfie has joined #nixos
DutchWolfie has quit [(Changing host)]
DutchWolfie has joined #nixos
newhoggy has quit [(Ping timeout: 260 seconds)]
<clever>
nh2: ah, so it sounds like one of the machines had already "built" a copy of that file, and it didnt have to copy it over from the good machine
<clever>
nh2: and the build "passed" even with a fatal error
<fetchurl>
clever: After fixing a few things in my configuration, including some annoying unclear infinite recursion, it does work, but not for factorio, since does not use fetchurl, so it fails on missing a urls attribute, but a nice thing to know about for other packages
<clever>
nh2: using nix-store --delete, you should be able to remove the failed build, and then try it again
<clever>
fetchurl: yeah, factorio uses a pretty weird and unique fetching system
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 pushed 1 new commit to master: https://git.io/vQ5rM
<NixOS_GitHub>
nixpkgs/master 5bee13f Jörg Thalheim: restbed: fix license and evaluation...
NixOS_GitHub has left #nixos []
newhoggy has joined #nixos
DutchWolfie has quit [(Quit: Konversation terminated!)]
magnetophon1 has quit [(Ping timeout: 260 seconds)]
newhoggy has quit [(Ping timeout: 260 seconds)]
erictapen has joined #nixos
tokudan has quit [(Quit: Leaving)]
ixxie has joined #nixos
<ToxicFrog>
After getting license info showing, I think I now see how to implement support for nonfree packages, too.
<ToxicFrog>
Shouldn't be too hard to have an "include nonfree y/n" toggle and put a little ⚠ next to them or something.
<gchristensen>
ToxicFrog: cool! :D
arianvp2 has quit [(Ping timeout: 260 seconds)]
seanparsons has quit [(Read error: Connection reset by peer)]
newhoggy has joined #nixos
seanparsons has joined #nixos
<radvendii>
If I add a nixos module, how do I test it?
<ToxicFrog>
Broken is proving a bit harder because if I enable includeBroken in the package list generator, it doesn't build _at all_
<ToxicFrog>
(but I care about that less, so I'll probably just implement nonfree support for now and leave broken for someone else to deal with~)
<gchristensen>
ToxicFrog: not great advertising broken stuff anyway
romildo has quit [(Quit: Leaving)]
<gchristensen>
domenkozar: any luck testing?
<ToxicFrog>
gchristensen: no, but as a default-off option it could be useful
<gchristensen>
meh, I wouldn't bother :P
<ToxicFrog>
After all, someone asking "does nixos have package X" might respond differently to "no" and "yes, but it doesn't build right now"
<LnL>
my idea at the time was that a temporary broken package wouldn't disappear from the website
<gchristensen>
but what is the practical difference?
<LnL>
but might be better to just leave it out
erictapen has quit [(Ping timeout: 248 seconds)]
<ToxicFrog>
gchristensen: if they're in a patching mood, it might prompt them to look at the derivation and submit a patch :)
<ToxicFrog>
But like I said, I will leave this to other people to implement, since while I think it would be a good thing I don't actually care that much
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Mic92 pushed 1 new commit to master: https://git.io/vQ5Kb
<NixOS_GitHub>
nixpkgs/master 3ef073e Jörg Thalheim: nodePackages.js-beautify: init at 1.6.14...
NixOS_GitHub has left #nixos []
<gchristensen>
cool
newhoggy has quit [(Ping timeout: 248 seconds)]
<gchristensen>
a improvement is better than no improvement
<ToxicFrog>
Whereas not showing nonfree packages has bitten me several times, and I'm far from the only one
<nh2>
clever: I'm not copying form the good machine, I just had the good machine for comparison. I'm comparing the deployment "good machine -> 3 hosts deployed with nixops" with "bad machine -> 3 hosts deployed with nixops"
<sphalerite[m]>
sorry to bring this up again, but FWIW I agree with 0xABAB's banning. I contribute to NixOS because I enjoy using it and feel that my contributions are valued. 0xABAB's shitting all over stuff did not contribute to my enjoyment!
<gchristensen>
thank you :)
kiloreux has quit [(Remote host closed the connection)]
kiloreux_ has quit [(Remote host closed the connection)]
<sphalerite[m]>
I do agree with most of their technical opinions but I think their value is more than offset by their unpleasantness.
<sphalerite[m]>
Also, I've got myself a new baby laptop (chromebook) and was looking to install nixos on it. Has anyone here used nixos with coreboot?
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] Ericson2314 closed pull request #27368: cc-wrapper: Unify and improve dynamic linker flag logic (staging...cc-wrapper-interpreter) https://git.io/vQQPo
NixOS_GitHub has left #nixos []
newhoggy has quit [(Ping timeout: 240 seconds)]
schoppenhauer has quit [(Ping timeout: 260 seconds)]
yegortimoshenko has quit [(Remote host closed the connection)]
<sphalerite[m]>
Considering installing debian+nix instead also, because there are convenient step-by-step instructions for debian and I'm quite confident that I'll manage to install nix from there :p
<simpson>
sphalerite[m]: I'm afraid that you've chosen poorly as far as customizeable hardware; Crouton is a thing.
<sphalerite>
simpson: I chose it because libreboot will run on it
<simpson>
sphalerite: I'm just saying that, despite the appearance of being general-purpose hardware, they really aren't.
<sphalerite[m]>
What's missing from them?
newhoggy has joined #nixos
<simpson>
Horsepower, mostly. Not much RAM, mediocre GPU/CPU packages, that sort of thing.
mexisme has quit [(Ping timeout: 240 seconds)]
<sphalerite[m]>
right, that doesn't surprise me in the slightest
mexisme has joined #nixos
<simpson>
I used to be on the team that makes the software that runs on Chromebooks. We did *not* develop on them. They were solely deployment and testing devices. I did plop Crouton on a couple of mine, and it works, but it's also pretty unfun. The fanciest I got with a Chromebook was getting a few emulators running on it many years ago, and it wasn't happy doing that either.
<gchristensen>
I love the idea of a chomebook though
<sphalerite[m]>
Main reasons I got it are A, so I have something that will run libreboot just because that's cool, and B, for taking to the US. I don't want to take my normal driver laptop with me
hiratara has quit [(Ping timeout: 246 seconds)]
Sonarpulse has joined #nixos
newhoggy has quit [(Ping timeout: 240 seconds)]
<radvendii>
how often is the unstable channel updated
hiratara has joined #nixos
<radvendii>
and how do i check when the last update was?
<sphalerite[m]>
simpson: I'm expecting to use it for a little python program, some light browsing, and as a thin client to access beefier machines via SSH
erictapen has quit [(Ping timeout: 260 seconds)]
<sphalerite[m]>
surely those are reasonable expectations?
<simpson>
sphalerite[m]: Oh, totally, don't come to the USA with hardware that you value.
<simpson>
100% agree. Carry shitty laptop with you instead.
newhoggy has joined #nixos
<gchristensen>
radvendii: https://channels.nix.gsc.io/graph.html is usually correct, but sometimes a bug in the nix channel update process causes it to say a channel update happened, however in reality the channel update failed
ng0 has quit [(Quit: Alexa, when is the end of world?)]
<Infinisil>
When I added a package, shouldn'te I just be able to use outputs = [ "out" "man" ] and it adds those manpages to the list of manpages when installing?
<Infinisil>
Installing it with nix-env -f path/to/nixpkgs -iA bashSnippets
newhoggy has quit [(Ping timeout: 260 seconds)]
<Infinisil>
But i can't use `man weather` or `man bash-snippets
erictapen has joined #nixos
newhoggy has joined #nixos
erictapen has quit [(Ping timeout: 240 seconds)]
newhoggy has quit [(Ping timeout: 248 seconds)]
newhoggy has joined #nixos
NixOS_GitHub has joined #nixos
<NixOS_GitHub>
[nixpkgs] hamishmack opened pull request #27392: pango: temp fix for font size on Retina displays (master...master) https://git.io/vQ55i
NixOS_GitHub has left #nixos []
<samae>
hey, am I the only one that has a: “manual-combined.xml fails to validate” when building the following derivation /nix/store/jzdvny020lx4j09ic1jl1hfcf0yvfh2i-nixos-manual-combined.drv
newhoggy has quit [(Ping timeout: 260 seconds)]
tikhon has quit [(Ping timeout: 240 seconds)]
newhoggy has joined #nixos
Neo-- has quit [(Ping timeout: 276 seconds)]
<samae>
but of course, I'm using unstable-small, and that kind of breaking can be expected
<samae>
I'll try it later on
darlan has joined #nixos
Neo-- has joined #nixos
ambro718 has quit [(Quit: Konversation terminated!)]