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
johnny101 has joined #nixos-dev
<samueldr> clever: did you receive the notification for the review request for #65212 ?
<{^_^}> https://github.com/NixOS/nixpkgs/pull/65212 (by thefloweringash, 21 hours ago, open): nixos-test-driver: allow configuration of net frontend and backend
<samueldr> you're not forced to review it, but I feel like you're knowledgeable in the matter :)
layus has quit [Quit: ZNC 1.7.3 - https://znc.in]
<clever> samueldr: i dont always keep an eye on my github notifications
layus has joined #nixos-dev
<samueldr> right, I'll keep that in mind if it's ever important
<clever> *looks*
andi- has quit [Read error: Connection reset by peer]
andi- has joined #nixos-dev
<clever> samueldr: ive got 2 main ipxe examples you could look over if you want
<samueldr> if you have further tests to validate the PR that would be lovely
<clever> samueldr: this spins up an entire nixos router, with full ipxe netboot support, and a nixos ramdisk booting
<clever> samueldr: this loads ipxe (packaged as a linux kernel) into qemu, to test the ipxe scripts
<clever> samueldr: the romfile= stuff from the PR you linked, is qemu emulating a dhcp+tfto server, that does basically the same thing as netboot_server.nix
<clever> so you dont need a seperate VM
Guanin has quit [Remote host closed the connection]
<clever> samueldr: most of the changes there look good, main thing i can think of is to check the bittorrrent tests
<clever> [clever@system76:~/nixpkgs]$ nix-build nixos/release.nix -A tests.bittorrent.x86_64-linux
<clever> running...
<clever> samueldr: it fails
<samueldr> can you document the failure on the PR?
<samueldr> (and thank you)
<clever> samueldr: done
<samueldr> clever++
<{^_^}> clever's karma got increased to 158
<thefloweringash> hmm, passes for me
<clever> thefloweringash: from the same git rev?
<thefloweringash> yep
<clever> fun, non-deterministic failures!
<clever> client1: running command: test -e /tmp/test.tar.bz2
<clever> client1: exit status 1
<clever> thefloweringash: mine just loops this forevr
<thefloweringash> I'll pastebin my log
<clever> the difference is at about line 2662
<clever> /tmp/test.tar.bz2 eventualy shows up for you, but not me
<clever> with an uptime of about 40 seconds
<clever> thefloweringash: can you add your results to the PR also?
<thefloweringash> sure, done
Drakonis has quit [Quit: WeeChat 2.4]
Jackneill has joined #nixos-dev
aszlig has quit [Quit: Kerneling down for reboot NOW.]
aszlig has joined #nixos-dev
FRidh has joined #nixos-dev
joko has joined #nixos-dev
<joko> Hey, has anyone access to wendy.ewi.tudelft.nl which is used by Hydra? Some builds are timing out there
magnetophon has quit [Ping timeout: 244 seconds]
magnetophon has joined #nixos-dev
cransom has quit [Quit: WeeChat 2.4]
Jackneill has quit [Ping timeout: 246 seconds]
Jackneill has joined #nixos-dev
Jackneill has quit [Ping timeout: 244 seconds]
Jackneill has joined #nixos-dev
worldofpeace_ has joined #nixos-dev
worldofpeace_ has quit [Quit: worldofpeace_]
aria has joined #nixos-dev
orivej has quit [Ping timeout: 272 seconds]
johanot has joined #nixos-dev
orivej has joined #nixos-dev
orivej has quit [Ping timeout: 248 seconds]
<tilpner> Oh, this is just great. The GitHub tag has different tests than what's on PyPi
<tilpner> Which should I prefer?
<Taneb> I want to say PyPi, it's better to go for releases than source repos
<tilpner> That's what I thought too, so I tried pypi first
<gchristensen> on the other hand, nix typically wants to go from sources not from what is on pypi. ... but perhaps that is true everywhere but pypi :P
orivej has joined #nixos-dev
__Sander__ has joined #nixos-dev
<adisbladis> gchristensen: Most python stuff in nixpkgs is fetched from pypi
<adisbladis> tilpner: Sometimes in nixpkgs we use fetchFromGitHub precisely to get tests
<adisbladis> They are often omitted from the sdist tarballs on pypi
<tilpner> adisbladis: This was relevant for #65262
<{^_^}> https://github.com/NixOS/nixpkgs/pull/65262 (by tilpner, 1 hour ago, open): python3Packages.PyGithub: 1.36 -> 1.43.8
<tilpner> I ended up disabling the tests entirely in one package, and partially in another :(
<tilpner> But in this case, Pypi had more tests than GH
<FRidh> that's not very ....common.
johnny101m has quit [Remote host closed the connection]
johanot has quit [Quit: WeeChat 2.4]
magnetophon has quit [Ping timeout: 248 seconds]
Drakonis has joined #nixos-dev
magnetophon has joined #nixos-dev
<gchristensen> samueldr: :D
primeos has quit [Quit: WeeChat 2.4]
__Sander__ has quit [Quit: Konversation terminated!]
primeos has joined #nixos-dev
<joko> nixpkgs.tarball seems to failing for nixos-unstable-small: https://hydra.nixos.org/job/nixos/unstable-small/tested#tabs-constituents
<joko> Seems like the package/release name changed, "tarball" is following nixpkgs
<joko> Looking for what changed
<samueldr> though I don't think that's the issue, exit code 137 (sigkill)
<joko> samueldr: could anyone restart it then? Although github seems dead currently
<samueldr> I'm saying it must not be the issue causing the failure as https://hydra.nixos.org/build/96729091#tabs-summary also uses nixpkgs-* for the release name
<samueldr> joko: restarted
<joko> samueldr: tyvm
<samueldr> earliest 19.03 eval; tarball seems to always have been named nixpkgs-* https://hydra.nixos.org/build/89535556
<samueldr> (after all, this is AFAIUI the nixpkgs tarball for that nixos [channel] eval)
orivej has quit [Ping timeout: 245 seconds]
<joko> samueldr: it seems to be built, thanks once again!
Drakonis has quit [Quit: WeeChat 2.4]
Drakonis has joined #nixos-dev
drakonis1 has joined #nixos-dev
drakonis_ has quit [Ping timeout: 258 seconds]
Drakonis has quit [Ping timeout: 252 seconds]
Drakonis has joined #nixos-dev
drakonis1 has quit [Ping timeout: 244 seconds]
Drakonis has quit [Ping timeout: 250 seconds]
webster23 has joined #nixos-dev
__monty__ has joined #nixos-dev
<pierron> hum … I don't know where to find my windows manager stderr … so I went into strace-ing it while reproducing the error and looking for write(2, …)
pierron is now known as nbp
orivej has joined #nixos-dev
Jackneill has quit [Remote host closed the connection]
orivej has quit [Ping timeout: 258 seconds]
tilpner has quit [Quit: WeeChat 2.4]
<gchristensen> it would be cool for https://r13y.com/ to provide a link back to a github issue, but I can't really do a "search or create new" workflow
tilpner has joined #nixos-dev
Drakonis has joined #nixos-dev
<infinisil> (Reminder that 19.09 is coming up soon and there is no new release manager yet afaik)
tilpner has quit [Ping timeout: 245 seconds]
<gchristensen> disasm
<infinisil> Ah cool, milestone can be updated then <https://github.com/NixOS/nixpkgs/milestone/16>, and the topic as well
tilpner has joined #nixos-dev
phreedom_ has joined #nixos-dev
<samueldr> reminder that 20.03 is coming next and there's no new release manager yet :)
<gchristensen> definitely
phreedom has quit [Ping timeout: 260 seconds]
<worldofpeace> hmmmm......
* gchristensen waggles eyebrows at worldofpeace
<worldofpeace> I must admit it's something I've been, looking into... just thinking of a good timing
<infinisil> worldofpeace: It's kinda hard to predict what you'll be doing in half a year
<infinisil> Well for me at least :)
<worldofpeace> infinisil: exactly
<infinisil> Who knows, maybe we'll become Guix people in that time!
<worldofpeace> lol, I don't think that's likely though
<worldofpeace> I just know if I get an opportunity in that time frame I'd have to drop everything and just do that
<infinisil> Maybe yeah
<samueldr> tbf, with disasm we spoke of it at last nixcon, he already showed interest and I told him to wait until closer to the release to reaffirm, same reasons: no way to know what will happen in close to a year
<gchristensen> it is a lot of work and a big commitment, for sure
<samueldr> so I figure the RMs for 19.09 can do the same, and get back to you later
<gchristensen> I feel we need to find a way to delegate more of the work
<worldofpeace> gchristensen: totally, thinking the same
<infinisil> Maybe have a set of ~5 "release-helpers"
<gchristensen> we would not want to distribute the responsibility though. important to have one (or two) people at the top.
magnetophon has quit [Ping timeout: 246 seconds]
<worldofpeace> do you mean "at the top" or "at most" gchristensen
<gchristensen> at the top
<samueldr> maybe we need a better way to show "ready to help" for good people around the release
<gchristensen> a person needs to be responsible andfinal in charge for decisions. the rest ofthe people are tohelp
<samueldr> so the RMs can somehow get their attention of needed
<gchristensen> samueldr: that sounds good
<gchristensen> mayb ewe could make a team which is very easy to join as a way to ping "willing and able"
<infinisil> I personally like the idea of having a small team of dedicated release helpers
<gchristensen> metoo
<infinisil> Because otherwise people aren't really feeling any dedication
<infinisil> Or so
<worldofpeace> I'd say the people "ready to help" are just the ones who are able to be active at the current time
<gchristensen> not so loose, I think
<worldofpeace> and that rotates because of changing circumstances
<gchristensen> I think there is a number of people who are largely interested in helping a lot
<samueldr> I'm thinking people like srhb who have been really helpful in getting things going along last two releases in the ZHF thread
<worldofpeace> i keep a list :)
<gchristensen> worldofpeace: OH then I should just shut up and let you be the subject matter expert :P
<infinisil> Idea: Split the release management up into smaller sections. E.g. a NixOS module section, one for docs, one for python packages, ...
<worldofpeace> gchristensen: oh nope, I don't shut anyone up. experts gotta learn too :)
<worldofpeace> but on topic with the ZHF thread, it works because it's a sign of a collective effort going on where people can figure how they can insert themselves
<worldofpeace> infinisil: fedora had this for python packages https://fedoraproject.org/wiki/Changes/F31_Mass_Python_2_Package_Removal ...
<gchristensen> exactly, worldofpeace
<gchristensen> it is a clear way to participate, with a direct need
<infinisil> worldofpeace: Hehe, "To prevent massive breakage, ...". We don't have that problem as much with nix because people can just use an older nixpkgs version to get back whatever they need :)
Drakonis has quit [Ping timeout: 248 seconds]
* gchristensen hopes the big nixops PR will be in 19.09
<worldofpeace> link gchristensen ?
<infinisil> Ah, the big nixops PR we all know about which is completely obvious to everybody
<gchristensen> exactly
<gchristensen> the one which deletes all the code
<infinisil> Ohh I was actually thinking of that one
<{^_^}> nixops#1179 (by johnalotoski, 1 week ago, open): Nixops Core, Plugin Restructuring PR
<infinisil> Haven't seen this new one though *subscribes*
<gchristensen> iohk took on my initial PR :)
Drakonis has joined #nixos-dev
<gchristensen> note this PR is useful because it makes the process of porting to py3 more piecemeal
<infinisil> Hehe, would be kind of ironic if nixpkgs removed python 2 support with nixops still using it
magnetophon has joined #nixos-dev
drakonis_ has joined #nixos-dev
drakonis1 has joined #nixos-dev
<infinisil> Deprecating python 2 would actually be a nice goal for 19.09
Drakonis has quit [Ping timeout: 276 seconds]
<gchristensen> we absolutely should
<gchristensen> as it'll "expire" half way through
<infinisil> Although, EOL is January 2020. So it would make sense to keep it in for at least 19.09
<infinisil> But at least defaulting to python 3 would be nice for 19.09
drakonis_ has quit [Ping timeout: 264 seconds]
<gchristensen> typically I think if it will EOL during the release, we should drop it before the release
Jackneill has joined #nixos-dev
<samueldr> how much can we remove from /proc from the sandbox for nix-build, without breaking everything?
<gchristensen> uhh...?
<samueldr> for reproducibility, if the /proc/cpuinfo is not available, less bad things could happen
<gchristensen> oh cool
<samueldr> or if it was somehow overlaid with a static "nix86_64" cpu or "nixaarch64" or something
<samueldr> this fails in two ways
<samueldr> (1) computer without SSE [though unlikely now] would build differently than one with
<samueldr> (2) cross-compilation!
Drakonis has joined #nixos-dev
drakonis_ has joined #nixos-dev
drakonis_ has quit [Client Quit]
__monty__ has quit [Quit: leaving]
Jackneill has quit [Remote host closed the connection]
<worldofpeace> lol I still don't know how to merge #61876. probably just going to go for it
<{^_^}> https://github.com/NixOS/nixpkgs/pull/61876 (by worldofpeace, 8 weeks ago, open): gdk-pixbuf: rename from gdk_pixbuf
webster23 has quit [Remote host closed the connection]
webster23 has joined #nixos-dev
webster23 has quit [Ping timeout: 264 seconds]
magnetophon has quit [Ping timeout: 272 seconds]
magnetophon has joined #nixos-dev