gchristensen changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html | 18.09 release managers: vcunat and samueldr | https://logs.nix.samueldr.com/nixos-dev
Synthetica has quit [Quit: Connection closed for inactivity]
lopsided98 has quit [Quit: Disconnected]
lopsided98 has joined #nixos-dev
drakonis has quit [Remote host closed the connection]
pie__ has joined #nixos-dev
pie___ has joined #nixos-dev
pie__ has quit [Ping timeout: 246 seconds]
jtojnar has quit [Ping timeout: 250 seconds]
pie___ has quit [Ping timeout: 250 seconds]
<ekleog> thank you all for having looked into this!
lassulus_ has joined #nixos-dev
lassulus has quit [Ping timeout: 246 seconds]
lassulus_ is now known as lassulus
orivej has quit [Ping timeout: 245 seconds]
worldofpeace has quit [Quit: worldofpeace]
xeji has joined #nixos-dev
xeji has quit [Client Quit]
xeji has joined #nixos-dev
xeji has quit [Quit: WeeChat 2.3]
xeji has joined #nixos-dev
xeji has quit [Client Quit]
jtojnar has joined #nixos-dev
Jackneill has joined #nixos-dev
MichaelRaskin has quit [Quit: MichaelRaskin]
Synthetica has joined #nixos-dev
orivej has joined #nixos-dev
Cale has joined #nixos-dev
init_6 has joined #nixos-dev
xeji has joined #nixos-dev
xeji has quit [Client Quit]
<domenkozar> is there anyone that is interested to guide us through GSOC2019? :)
genesis has joined #nixos-dev
lokado has joined #nixos-dev
xeji has joined #nixos-dev
pie___ has joined #nixos-dev
layus has quit [Ping timeout: 245 seconds]
layus has joined #nixos-dev
init_6 has quit []
orivej has quit [Ping timeout: 268 seconds]
pie___ has quit [Ping timeout: 246 seconds]
<LnL> hmm the installer doesn't create a 'default' profile anymore?
Jackneill has quit [Remote host closed the connection]
orivej has joined #nixos-dev
lokado has quit [Quit: WeeChat 2.3]
pie__ has joined #nixos-dev
<aristid> samueldr: you're samuel dionne-riel?
<aristid> samueldr: i was extremely surprised to see nixos use a non-long-term kernel, and it seems you made the change. i don't really mind, but the fact that the commit message doesn't explain why such a huge change was made does bother me.
<aristid> on kernel.org, 4.19 is only called "stable", not "longterm". is it really long-term anyways?
<samueldr> aristid: this is under the longterm maintenance
<aristid> samueldr: so the kernel.org mainpage is wrong/misleading?
<samueldr> misleading, it's still stable I guess :)
<aristid> oh it's both "stable" _and_ "longterm" but they only call it one of the two? :D
<samueldr> I don't know for sure, but I guess so
<aristid> samueldr: still, would've been good if they commit message had said "4.14 -> 4.19 because 4.19 is Greg KH's new long-term supported kernel"
<samueldr> :/ in the body of the commit I could have, yeah, though I did handle that through the PR #52863
<{^_^}> https://github.com/NixOS/nixpkgs/pull/52863 (by samueldr, 1 day ago, merged): Switch default linux package to the new LTS, 4.19
<samueldr> I assumed people interested in the kernel, default versions and LTS would know which release was LTS. maybe my assumption was wrong
<samueldr> the 4.19 release announcement re-confirms it's an LTS release https://lkml.org/lkml/2018/10/22/184
<samueldr> though I think there was a separate announcement for when it was designated as LTS
<domenkozar> which was before release :)
<domenkozar> we used to switch to LTS after a few bugfixes
<domenkozar> I think waiting on another minor kernel is a good idea for all kernel modules to catch up
<domenkozar> as we did this time.
<samueldr> yeah, I was pleasantly surprised to see no breakage, once the two failing modules were updated
<samueldr> (nixos tests failures excluded :/)
<samueldr> oh! channel updated, finally :)
<domenkozar> it did?
<samueldr> nixos-unstable apparently did, according to howoldis, not nixpkgs-unstable
<domenkozar> yeah pandoc failed since haskell pkgs got bumped
<domenkozar> but it was fixed so soon
<samueldr> (I'll ask again next year as there's lack of activity) I think the nixos:unstable-aarch64 jobset is good to be disabled; its check interval is already disabled anyway, and it's tracking another older branch rather than master... unless there's still some use for it?
<domenkozar> hmm that gets me thinking
<domenkozar> we split up nixpkgs into different channels per platform
<domenkozar> maybe same should be done for nixos?
<domenkozar> anyway I've disabled it to reflect current reality
<samueldr> domenkozar: graham was thinking about removing that distinction
<samueldr> and in fact, other than 18.09-aarch64, of which there won't be an equivalent for 19.03, the only split is with darwin, right?
<domenkozar> yeah
<domenkozar> so I did the whole darwin split because people complained that they don't want linux channel to hang if there's no darwin maintainer available
<domenkozar> which makes sense also for aarch, since one needs different hardware
<samueldr> imho, one major distinction, but not enough by itself, that aarch64 is less problematic than darwin in this regard, is that there's no licensing issue; you can (even though it's going to be a pain) run aarch64 emulated legally
<samueldr> and then, a big chunk of the compatibility and test results will go in pair with x86_64; when we ran aarch64 as supported, the only tests that failed were the predictable names, and AFAIUI it's because the (virtualised) hardware differs, thus the names differ
<samueldr> oops, got a bit unrelated in the initial span
<samueldr> ah, and as I'm not a nixops user, I forgot about how it gets annoying for nixops users when the channels aren't in sync
<samueldr> though you're 100% right that it needs maintainers :/
<samueldr> though I'm not sure that splitting channels are the solution when there's no maintainers
pie__ has quit [Remote host closed the connection]
pie__ has joined #nixos-dev
orivej has quit [Ping timeout: 245 seconds]
orivej has joined #nixos-dev
pepesza has quit [Quit: ZNC 1.6.3+deb1ubuntu0.1 - http://znc.in]
orivej has quit [Ping timeout: 272 seconds]
<domenkozar> samueldr: well I think there is
<domenkozar> since one doesn't block the others
<domenkozar> so if there is a maintainer that comes around every week or so that's fine
<domenkozar> but it makes no sense to block 1k people due to 20 needing a completely different thing
<samueldr> I was mainly pointing out how there were talks about going the other way, and I'm not sure what the appropriate solution is
<samueldr> I see how both solutions have good and bad points; having a common channel is as good as having separate channels
<samueldr> (here considering software like nixops expect to use one revision, possibly from a channel, to deploy to a mixed set of platforms)
<domenkozar> I guess we could achieve both, but we'd need to change infrastructure
<samueldr> right, things need to change, I think there's agreement there, though the solution needs thought
<domenkozar> I think we should encourage people to pin nixpkgs
<domenkozar> and then they can pick nixpkgs revision that includes all platforms
<domenkozar> I know it's not the smoothest
<samueldr> tooling would need to exist to allow users to know whether it includes all platforms, and whatever they need is also in the cache
<domenkozar> nixops deploy --build-only --dry-run
<domenkozar> I mean passing channel doesn't guarantee much beyond tests
<domenkozar> so it could easily happen that aarch64 build for kodi failed,but not for linux
<samueldr> right, still it's something that
<domenkozar> so if we could get hydra to do interval builds always from the same offset in the hour
<domenkozar> the only thing they'd have to do is find inverval for which all platforms passed
<samueldr> -- that's important here; you wouldn't want to continue if tests fails :/
<samueldr> "interval builds always from the same offset in the hour"?
<domenkozar> in other words, we'd have to align for both evaluations to start at the very same time
<domenkozar> (note that this would also help reducing evaluation time)
orivej has joined #nixos-dev
sir_guy_carleton has joined #nixos-dev
pie__ has quit [Ping timeout: 252 seconds]
orivej has quit [Ping timeout: 250 seconds]
orivej has joined #nixos-dev
pie__ has joined #nixos-dev
orivej has quit [Ping timeout: 250 seconds]
orivej has joined #nixos-dev
sir_guy_carleton has quit [Quit: WeeChat 2.2]
pepesza has joined #nixos-dev
hedning_ has joined #nixos-dev
hedning_ is now known as hedning
pie__ has quit [Ping timeout: 250 seconds]
pie__ has joined #nixos-dev