sphalerite changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | NixOS 19.03 released! https://discourse.nixos.org/t/nixos-19-03-release/2652 | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html https://r13y.com | 19.03 RMs: samueldr,sphalerite | https://logs.nix.samueldr.com/nixos-dev
alp has joined #nixos-dev
alp has quit [Ping timeout: 245 seconds]
phreedom has quit [Quit: No Ping reply in 180 seconds.]
phreedom has joined #nixos-dev
phreedom has quit [Remote host closed the connection]
phreedom has joined #nixos-dev
orivej has quit [Ping timeout: 272 seconds]
MichaelRaskin has quit [Quit: MichaelRaskin]
Moredread[m] has joined #nixos-dev
phreedom has quit [Quit: No Ping reply in 180 seconds.]
phreedom has joined #nixos-dev
<teto> I created a nixos test for multipath TCP but ultimately, I would like the VMs to be multihomed. Any example in the test framework ? what would be the best way to achieve this ?
Jackneill has joined #nixos-dev
Jackneilll has joined #nixos-dev
Jackneill has quit [Ping timeout: 248 seconds]
<niksnut> IIRC it's possible to assign multiple interfaces to the VMs (see the nat tests for example)
<clever> the bittorrent test also has nat, because its testing the upnp features too
<teto> I don't think nat machines are multihomed, (the router has several interfaces but not the other nodes). I wonder how the network is created, seems like there is some DHCP server or /etc/hosts going on since I don't configure anything and the network just works
<teto> seems like I will find answers in build-vms.nix :)
alp has joined #nixos-dev
<teto> (answering my question: config.virtualisation.vlans is responsible for the tests multihoming)
<yorick> Shados: it wouldn't really work with nixos tests I think
orivej has joined #nixos-dev
pie_ has quit [Read error: Connection reset by peer]
pie__ has joined #nixos-dev
johanot has joined #nixos-dev
drakonis1 has quit [Quit: WeeChat 2.4]
drakonis_ has joined #nixos-dev
Drakonis has quit [Ping timeout: 245 seconds]
Drakonis has joined #nixos-dev
drakonis_ has quit [Read error: Connection reset by peer]
Drakonis has quit [Read error: Connection reset by peer]
Drakonis has joined #nixos-dev
<domenkozar[m]> do nixos options compose between two modules?
<domenkozar[m]> never tried that :D
<clever> domenkozar[m]: the merge function on the type controls that
<clever> domenkozar[m]: https://github.com/NixOS/nixpkgs/blob/master/lib/types.nix#L196-L214 and the merge func on 205
<domenkozar[m]> that's for config
<domenkozar[m]> I'd like to merge two options
<domenkozar[m]> so actually option definition
__Sander__ has joined #nixos-dev
<arianvp> Oh good question
<arianvp> No clue
<arianvp> If it is then the acme module could be a bit simpler. It's currently spread over nginx, lighttpd and acme modules
<arianvp> Whilst would be nicer if it all was in one place
<arianvp> Let me know if you find out :D
<gchristensen> I think yes
orivej has quit [Ping timeout: 246 seconds]
<clever> domenkozar[m]: ahhh, i have added custom options to fileSystems.<name>
<clever> that seamlessly adds 3 new options to all filesystems and swap devices
<clever> but it doesnt really deal with conflicting types on a single option
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 248 seconds]
vdemeest has joined #nixos-dev
<domenkozar[m]> yeah I have the same option that is defined at two places
<domenkozar[m]> type, description at one place
<domenkozar[m]> and default at another
orivej has joined #nixos-dev
<clever> domenkozar[m]: defaults dont have to be under options, due to mkDefault
<domenkozar[m]> yeah but that doesn't show up in the manual
vdemeest has quit [Quit: vdemeest]
<clever> domenkozar[m]: ahhh, yeah
__Sander__ has quit [Ping timeout: 272 seconds]
<domenkozar[m]> what kind of naming would be best for eval and runtime paths? nixops is so terrible in that regard
__Sander__ has joined #nixos-dev
<domenkozar[m]> best seems `BlablaFile` and BlablaRuntimeFile`
<gchristensen> what are you trying to communicate?
<domenkozar[m]> that one path is referenced at eval time
<domenkozar[m]> another one at runtime
<gchristensen> ah
<gchristensen> like a "keys" file
<domenkozar[m]> nixops currently overloads path vs file
<domenkozar[m]> path = file || directory
<domenkozar[m]> then nixops messes that up
<domenkozar[m]> and says path is reference, file is "read time"
<domenkozar[m]> while it should convey when are these files read
<domenkozar[m]> so best I could come up is runtime vs evaltime
<gchristensen> another property is the runtime one is "host location"
<gchristensen> this is a tricky thing
<gchristensen> what is the use case?
<domenkozar[m]> need to reference to both paths in Nix for secrets
<domenkozar[m]> service.foo.optionPath = ./file.json
<domenkozar[m]> service.foo.optionRuntimePath = "/var/lib/foo/file.json"
<domenkozar[m]> path vs file in Nix doesn't help much :D
<gchristensen> dhess and I were dreaming about an abstraction layer where you wouldn't need to come up with this sort of clear terminology
<gchristensen> but I can't find my notes about it
<domenkozar[m]> well one way is to define base directory for secrets
<domenkozar[m]> so runtime path becomes internal implementation
<domenkozar[m]> but that's only when you design the software :)
<domenkozar[m]> gchristensen: would love to read your notes at some point :)
<gchristensen> I'll try and dig 'em up
<gchristensen> it was neat, would support things like secrets being fulfilled by not just deployment.keys, but vault/kms, and also rotated credentials
<domenkozar[m]> keep teasing :P
domenkozar[m] has quit [Changing host]
domenkozar[m] has joined #nixos-dev
__Sander__ has quit [Ping timeout: 245 seconds]
alpounet has joined #nixos-dev
alp has quit [Ping timeout: 252 seconds]
<domenkozar[m]> ok can't merge options
<gchristensen> I do merge options
<domenkozar[m]> eh
<gchristensen> hehe
<gchristensen> hum
<domenkozar[m]> wish it said why they colllide
<gchristensen> that would be real nice
<domenkozar[m]> would be great if we can get weekly out this week for nixcon CFP
<gchristensen> shipit?
<domenkozar[m]> :D
coconnor has quit [Ping timeout: 245 seconds]
coconnor has joined #nixos-dev
<yorick> gchristensen: so, what's your packet build setup?
johanot has quit [Quit: WeeChat 2.4]
<arianvp> Lots of machines
drakonis_ has joined #nixos-dev
drakonis_ has quit [Read error: Connection reset by peer]
drakonis1 has joined #nixos-dev
Drakonis has quit [Ping timeout: 268 seconds]
alpounet has quit [Ping timeout: 252 seconds]
drakonis1 has quit [Ping timeout: 276 seconds]
alpounet has joined #nixos-dev
marek has joined #nixos-dev
marek has quit [Changing host]
<arianvp> Keep finding and squashing more bugs in the ACME module :)
<arianvp> the PR is starting to grow a bit big now though...
<gchristensen> maybe should merge small improvments one at a time if possible?
<arianvp> gchristensen:I found a nice hack for the oneshot testing btw
<gchristensen> oh?
<arianvp> it's hacky but it works
<arianvp> :D
<gchristensen> oh dear!
<gchristensen> maybe write up some docs in the code? :P
v0|d has joined #nixos-dev
<adisbladis> domenkozar[m]: It's looking likely that we'll release the tickets in the next couple of days too so we could have a weekly nixcon edition? :)
<samueldr> eek, tickets already?
drakonis1 has joined #nixos-dev
orivej has quit [Ping timeout: 245 seconds]
drakonis1 has quit [Ping timeout: 252 seconds]
<adisbladis> samueldr: Why not?
<samueldr> not why not, just surprised
<samueldr> and more about the excited definition of surprised than the unexpected
<adisbladis> =)
Jackneilll has quit [Remote host closed the connection]
orivej has joined #nixos-dev
Jackneill has joined #nixos-dev
Jackneill has quit [Remote host closed the connection]
justanotheruser has quit [Ping timeout: 258 seconds]
<worldofpeace> hey, anyone know how ta handle hackpage-packages.nix when you've renamed a package and you need to regenerate it?
<infinisil> worldofpeace: It's done automatically once a day on the haskell-updates branch
<infinisil> worldofpeace: You can only make changes to configuration-hackage2nix.nix
<worldofpeace> infinisil: cool so It should be fine to merge a change to staging that would break eval because of this?
<infinisil> A change to hackage-packages.nix?
<infinisil> What kinda change?
<{^_^}> https://github.com/NixOS/nixpkgs/pull/61876 (by worldofpeace, 5 weeks ago, open): gdk-pixbuf: rename from gdk_pixbuf
<infinisil> worldofpeace: Yeah I don't think it's a good idea to change hackage-packages.nix for that
<infinisil> The problem would just reappear next time it regenerates
<worldofpeace> So would I have to merge the change to haskell-updates and haskell-updates into staging after it regenerates?
<infinisil> Can't fix it in nixpkgs
<infinisil> I'll post a comment with that in the PR
<worldofpeace> Understood infinisil. Now It's just fuzzy how that'll propagate down
<infinisil> worldofpeace: Make PR, PR gets merged, peti updates hackage-packages.nix with the master version of hackage2nix on the haskell-updates branch, after some time when most the new updated packages seem to be working, he'll merge haskell-updates into master
<worldofpeace> infinisil: Ok, so that pr will have to wait until that happens
<infinisil> worldofpeace: Hm yeah, there wasn't an alias for it previously?
<worldofpeace> infinisil: I think its always been gdk_pixbuf. That change will add an alias though
<worldofpeace> (is double checking)
<infinisil> Yeah then the cabal2nix PR will have to wait
<worldofpeace> lol, the process to rename one package
<infinisil> Hehe yeah
<samueldr> isn't there an eval done with aliases disabled?
<samueldr> if so, it would cause issues
<worldofpeace> yep, that's why in part it has to wait if I'm understanding correctly (and why the commit manually changing hackage is there so I can check eval properly)
v0|d has quit [Ping timeout: 248 seconds]
justanotheruser has joined #nixos-dev
justanotheruser has quit [Excess Flood]