<clever>
go back to the project edit page, what branch is it pulling spec.json from?
<dhess>
clever: I get a 403 for that URL
<clever>
only an admin can open it
<clever>
youll need to open the same page on your own hydra
<dhess>
yeah sorry I figured that out
<dhess>
ok ok ok
<dhess>
I know what's going on
<dhess>
ugh this would have taken AGES to find
<dhess>
I renamed the default branch for our Hydra jobspect project from master -> main :)
<dhess>
clever: THANK YOU. :)
<clever>
so hydra was parsing the old spec.json
<dhess>
yep
<dhess>
your hints led me in the right direction
<dhess>
thanks so much, one of these days I'm going to buy you a beer :)
<dhess>
or 6
<clever>
heh, i dont drink beer
<dhess>
I'll buy you 6 of whatever you like
<dhess>
and there it goes
<dhess>
sir, I love you.
<dhess>
clever++
<{^_^}>
clever's karma got increased to 536
<samueldr>
dhess: next nixcon that clever's there, bring him pepsi (diet?) from your country :)
<dhess>
will do. I was actually going to see him at NixCon 2018, but ironically I just missed it because I was moving the UK permanently a week or 2 after it happened :)
<samueldr>
wow, moving the whole UK?
<clever>
samueldr: i never touch diet either
<dhess>
moving *to* :)
<samueldr>
I didn't recall if it was max or something else than "pure unadulterated" :)
zarel has quit [Ping timeout: 260 seconds]
zarel has joined #nixos-dev
justanotheruser has quit [Ping timeout: 246 seconds]
<jonringer>
is anyone from infra in here?
<jonringer>
@grahamc or @aminechikhaoui
<jonringer>
We need your help to finalize the 20.09 release
<worldofpeace>
gchristensen:
<worldofpeace>
aminechikhaoui:
<worldofpeace>
lol, anyways. this is like being tardy to ur funeral
<jonringer>
lol
<worldofpeace>
the irc bot needs like a lol counter
<worldofpeace>
how many times I said lol in a public channel a day, with milestones and stats
<worldofpeace>
<worldofpeace "lol, anyways. this is like being"> this carries over the tradition of releasing 20.03. which was dead in the EST night
<worldofpeace>
#reasons #choices
<samueldr>
(if there's anything I can directly help with, ask)
<worldofpeace>
I think it's just stableBranch = true; on hydra
<worldofpeace>
it seems like, somehow we confused reading our own docs to think "oh is this done at a certain point before, or does it have to have commits when hydra is green with the according stuff committed"
<worldofpeace>
turns out there's no real cascading effect IIRC and it could've been done many hours ago without an issue
<samueldr>
yeah, at any moment in time you want the result to show in the release name that it's ready IIRC
<worldofpeace>
yep, no "beta"
<drakonis>
this one should be called 20.11
<drakonis>
at this rate
<worldofpeace>
nooo we have few days 🤣
<samueldr>
no, since it's named after the branch-out tag
<worldofpeace>
worst case first thing in our morning tomorrow
<worldofpeace>
oh wait, it is?
<worldofpeace>
reads code again
<samueldr>
the "no" was about naming it anything else than 20.09
<worldofpeace>
oh lol aminechikhaoui said he just got back in ✨
<worldofpeace>
I will say, FTR, I'm a good developer and claim to know what I'm doing... for now. Once I move into the night I will slowly become incoherent
<abathur>
nixoween
<drakonis>
i'm just pulling your leg
<worldofpeace>
drakonis: I appreciated it 😸
<worldofpeace>
okay, 20.09 jobset has an eval scheduled
<worldofpeace>
I believe just waitin time now
<samueldr>
you got someone to stableBranch it?
<samueldr>
nice
<samueldr>
(yes you did)
<worldofpeace>
yep, amine just got back a bit ago
<worldofpeace>
we also have Jon with ability to modify jobset configurations ✨
<worldofpeace>
so this step won't require pinging someone to click a button
<worldofpeace>
we both agree that sleeping isn't a part of the release manager schedule. today was "eat release ❌sleep"
<worldofpeace>
aminechikhaoui: yep, it all seems to happen naturally for me if I've mentioned that before
<danderson>
question not related to the release (yay release!!!) : where might I find the list of packages that hydra builds for the cache? Or the list of excluded packages?
<danderson>
I'd like to send a PR to enable building mingw cross-compilers, because they're generally useful for windows-from-linux development, and right now they're not getting built so my nix-shell takes 1h to update whenever I bump channels
<samueldr>
I don't know that there is a _list_ of excluded packages
<samueldr>
but for all that's included it's stuff that's in all-packages.nix, and tests, and their dependents
<danderson>
I guess put another way, pkgsCross.mingwW64.buildPackages.gcc empirically doesn't have cached builds, and I'd like cached builds to exist
<samueldr>
in your particular case it's not something you'd see listed at any point yeah :)
<danderson>
but I have no idea what knobs to twiddle to make that happen
<samueldr>
looking under 5 seconds, doesn't look like there is a precedent for such a thing yet
<danderson>
it looks like some other cross packages do get cached as a side-effect of being pulled in by top-level items.
<danderson>
so, that's the closest to a precedent - some cross compilers do get built and cached
justanotheruser has joined #nixos-dev
<samueldr>
yes
<danderson>
not exactly a ringing endorsement, admittedly :)
<samueldr>
I meant I couldn't see in under 5 secs any pkgsCross member being specifically asked for
<samueldr>
pkgsCross for aarch64-linux gets built mostly because of Mobile NixOS :)
<danderson>
yeah, and various ARM and AVR compilers for embedded-ish projects
<samueldr>
maybe open an issue or discourse post and ask for the usual suspects about cross, and maybe eelco, if there would be unforeseen issues, and how they would handle it?
<danderson>
sure. I was going to create a PR adding mingw packages and see what happens, but I can start with an issue :)
<samueldr>
I *guess* it'd look like adding something to the release.nix used by nixpkgs
<samueldr>
as aliases under all-packages.nix?
<danderson>
I figured I could just add `mingw = pkgsCross.mingwW64.buildPackages.gcc;` to all-packages.nix
<samueldr>
I don't think that's the way to go
<samueldr>
that's not a real package really
<danderson>
then if you want a windows cross-compiler, you nix-env -iA nixpkgs.mingw and voila
<danderson>
it's not?
<samueldr>
well
<samueldr>
yes, but I don't think that's the way to make it correct
<danderson>
okay, I'll file an issue. Who would the usual cross-compiling suspects be, if you happen to know?
<samueldr>
I don't think having aliases to pkgsCross in all-packages is desirable
<aminechikhaoui>
probably stdenv for every pkgsX variant shouldn't be too controversial, I also thought I'd find pkgsMusl.stdenv the other day built but seems we don't do that
<samueldr>
but yes, having them built I think is desirable
<{^_^}>
#101814 (by danderson, 27 seconds ago, open): Build and cache cross-compile toolchains in cache.nixos.org
<danderson>
and while I was doing that, my double gcc build finished \o/
<danderson>
back to work :)
andi- has quit [Remote host closed the connection]
andi- has joined #nixos-dev
orivej has joined #nixos-dev
Reventlov has quit [Read error: Connection reset by peer]
rnhmjoj has quit [Changing host]
rnhmjoj has joined #nixos-dev
rnhmjoj has joined #nixos-dev
bgamari has quit [Ping timeout: 260 seconds]
bgamari has joined #nixos-dev
<etu>
ryantm: Thanks to rapid releases and that I submitted a merge request of this package, nixpkgs master is now out of date by one minor version: https://repology.org/project/php:phpstan/versions This is a good testcase :)
jonringer has quit [Ping timeout: 260 seconds]
lopsided98 has quit [Quit: No Ping reply in 180 seconds.]
AlwaysLivid has joined #nixos-dev
lopsided98 has joined #nixos-dev
<cole-h>
worldofpeace++ jonringer++
<{^_^}>
jonringer's karma got increased to 8
<{^_^}>
worldofpeace's karma got increased to 245.00000000000003
<etu>
{^_^}++
<{^_^}>
{^_^}'s karma got increased to 221
<etu>
Nice
alp has joined #nixos-dev
cransom has quit [Ping timeout: 240 seconds]
cransom has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
saschagrunert has joined #nixos-dev
<NinjaTrappeur>
Are there any plans to backport builtins.fetchTree to the Nix 2.x lineage?
vcunat has joined #nixos-dev
alp has quit [Ping timeout: 272 seconds]
alp has joined #nixos-dev
__monty__ has joined #nixos-dev
teto has joined #nixos-dev
cole-h has quit [Ping timeout: 260 seconds]
dhess has left #nixos-dev ["ERC (IRC client for Emacs 27.1)"]
orivej has joined #nixos-dev
tilpner has quit [Remote host closed the connection]
tilpner has joined #nixos-dev
avn has joined #nixos-dev
averell has quit [Quit: .]
saschagrunert has quit [Remote host closed the connection]
saschagrunert has joined #nixos-dev
averell has joined #nixos-dev
alp has quit [Ping timeout: 272 seconds]
vcunat has quit [Ping timeout: 272 seconds]
saschagrunert has quit [Remote host closed the connection]
saschagrunert has joined #nixos-dev
alp has joined #nixos-dev
alp has quit [Ping timeout: 272 seconds]
vcunat has joined #nixos-dev
<gchristensen>
hey worldofpeace I can help later today!
<etu>
ryantm: Hmm, that attribute name "php-phpstan-0.12.49" is expected to be broken
<ryantm>
etu: I messed around a bit with my setup and now it is finding it okay. I'm not sure what's different though. I think it's likely that it will figure out how to update these.
<etu>
ryantm: Ok, I'll refrain from updating this specific package so we'll have it as a testcase :)
<aminechikhaoui>
It's not really urgent but should I ping other people for https://github.com/NixOS/nixpkgs/pull/101189, I just don't want the PR to bitrot. I'm not sure who's reviewing musl related changes these days
<{^_^}>
#101189 (by AmineChikhaoui, 6 days ago, open): gcc9: add patches required for target musl
<ryantm>
etu sounds good. I'll keep monitoring it.
saschagrunert has quit [Remote host closed the connection]
alp has joined #nixos-dev
saschagrunert has joined #nixos-dev
justanotheruser has quit [Ping timeout: 265 seconds]
justanotheruser has joined #nixos-dev
<worldofpeace>
gchristensen: amine was able to do stableBranch yesterday
<worldofpeace>
though u could still help
<worldofpeace>
the channel update is taking FOREVER
<gchristensen>
is anyone using recursive nix? cc domenkozar[m] maybe... I'm wondering if it works for you on 20.09
vcunat has quit [Ping timeout: 272 seconds]
pinpox is now known as yawe
yawe is now known as pinpox
pinpox is now known as yawe
yawe is now known as pinpox
saschagrunert has quit [Quit: Leaving]
<utsl>
@aminechikhaoui I'm also working on trying to make pkgsMusl (and pkgsStatic) work better. I'd be happy to collaborate.
rajivr has quit [Quit: Connection closed for inactivity]
<jtojnar>
worldofpeace: I am not sure how feasible is backporting individual fontconfig commits
<jtojnar>
I think switching to master, testing it on unstable for a while, and then backporting it as a whole would be relatively safe
<jtojnar>
since we might not notice regressions immediately
<jtojnar>
or to minimize risk, just cherry-pick mclasen’s patches to unstable
<jtojnar>
so we can test them there
vcunat has joined #nixos-dev
<samueldr>
hi, hopefully someone that is somewhat on authoritative on Hydra things can review and get this fast-tracked so it can be put in place before the next time we get a weird failure: https://github.com/NixOS/hydra/pull/825
<{^_^}>
hydra#825 (by samueldr, 1 day ago, open): Fix unhelpful error messages in aggregate jobs.
alp has quit [Ping timeout: 272 seconds]
<worldofpeace>
Jan Tojnar: I'm not sure what u mean with "not sure how feasible backporting individual fontconfig commits" if the suggestion is to cherry-pick mclassen's patches? At least, that's what I thought to do, since it minimizes the change surface
<jtojnar>
worldofpeace: I meant with regards to QAing the changes
<worldofpeace>
Jan Tojnar: oh yeah, I think the latter is the way to go then. fontconfig seems to have issues that could be really really annoying for us if introduced into a stable release. I think there might be other bugfixes that would be nice... not sure then to just use master
<worldofpeace>
lol, I've disagreed and agreed with my self
<jtojnar>
the main issue is time, though
<jtojnar>
#101813 sounds bad enough and going through unstable will take several weeks of builds and testing at least
justanotheruser has quit [Ping timeout: 246 seconds]
nschoe has joined #nixos-dev
orivej has quit [Ping timeout: 265 seconds]
alp has joined #nixos-dev
teto has joined #nixos-dev
<gchristensen>
it looks like nix-build on 20.09's nixUnstable doesn't print log messages to the screen
<gchristensen>
anyone seeing this?
dstzd_ is now known as dstzd
dstzd has joined #nixos-dev
dstzd has quit [Changing host]
tokudan has quit [Remote host closed the connection]
justanotheruser has joined #nixos-dev
gchristensen has quit [*.net *.split]
emilazy has quit [*.net *.split]
emilazy has joined #nixos-dev
tokudan has joined #nixos-dev
gchristensen has joined #nixos-dev
tokudan has quit [Remote host closed the connection]
tokudan has joined #nixos-dev
justanotheruser has quit [Ping timeout: 260 seconds]
justanotheruser has joined #nixos-dev
nschoe has quit [Remote host closed the connection]
fadenb has quit [Quit: killed]
<jonringer>
should I make a post to reddit.com/r/linux about the 20.09 release?
<gchristensen>
yea!
__monty__ has quit [Quit: leaving]
<worldofpeace>
lol, how did fedora also release today
<supersandro2000>
because it is a big conspiracy
<supersandro2000>
Nix and Fedora are planning the big coupe
<supersandro2000>
Fnixora is coming
<jonringer>
well, nixpkgs can be installed on anything that hash a linux kernel. So you could say that any distro is a potential "host"
<garbas>
you mean "FixOS" :)
<jonringer>
has*
<jonringer>
garbas++
<{^_^}>
garbas's karma got increased to 29
red[evilred] has joined #nixos-dev
<red[evilred]>
in the firefox nixpkg meta - there is a key: badPlatforms
<red[evilred]>
can someone help me understand whhat that means and how/why it's used?
<jonringer>
it's respected by hydra
<supersandro2000>
NüchtOS
<jonringer>
essentially prevent hydra from trying to build the package on those platforms
<supersandro2000>
*NüschtOS
<jonringer>
hydra being our CI tool which builds packages and uploads them to the binary cache
<red[evilred]>
Bingo - so thaht would explain why I'm able to build it locally
justanotheruser has quit [Ping timeout: 272 seconds]
<red[evilred]>
but hydra didn't build a cache?
<jonringer>
correct
<jonringer>
either it's really expensive to do so, or some other reason
<red[evilred]>
Well, it's definitely expensive
<red[evilred]>
it's firefox
<red[evilred]>
it's exceedingly expensive
<jonringer>
I'm assuming you're talking about badPlatforms = lib.platforms.darwin;
<red[evilred]>
yes
<jonringer>
there's probably some issue where the hydra builds were just taking forever, and it was cut for resource reasons
<jonringer>
darwin compute is significantly more expensive than linux compute
<jonringer>
legal darwin compute*
<red[evilred]>
Gotcha
<red[evilred]>
it's a shame I can't upload my build (once it's done)
<jonringer>
problem with build outputs is that they are binary packages at that point, and you will probably want them to be built on managed hardware to avoid malware side attacks
<red[evilred]>
righht
<red[evilred]>
hmm
<jonringer>
also, nowadays most browsers will "self-update"
<red[evilred]>
so I want to upload a thing that has darwin as a valid target
<red[evilred]>
let me rephrase that
<red[evilred]>
I want to submit a package to nixpkgs, but unfortunately it has firefox as a dependency
<red[evilred]>
I guess that means that my build on darwin in hydra will just refuse to evaluate?
<garbas>
jonringer: red[evilred]: probably it would be nice to separate this 2 reasons and maybe add expensivePlatforms (or some better name)
<red[evilred]>
I guess as we move towards a new flakey-world... maybe we can build those expensive poackages once in 10 changes or something
<red[evilred]>
so the flake can refer to the cache
<jonringer>
from hydras perspective, it makes sense (build packages for `platforms`, but not for `badPlatforms`)
<garbas>
oh sorry there is broken flag that tell that the package can not be built
<jonringer>
another common scenario is `platforms = platforms.linux;`, `badPlatforms = platforms.aarch64`
justanotheruser has joined #nixos-dev
<supersandro2000>
and emulating aarch64 on amd64 is quite slow
<red[evilred]>
ugh, afte all that - it failed on Unknown option: --disable-xcode-checks
<jonringer>
that's probably why it was disabled. Not an easy problem to solve. I would just recommend installing it outside of nixpkgs
<jonringer>
sometimes the fight isn't worth it
<jonringer>
or move over to linux! :)
<supersandro2000>
DSL - Darwin Subsystem for Linux
<red[evilred]>
Well, I'm trying to port EyeWitness to OSX
<red[evilred]>
(it already works on linux) :-)
<red[evilred]>
brb - grabbing a snack while that compiles
<supersandro2000>
See you tomorrow
<supersandro2000>
red[evilred]: or Firefox is only built every X days
<supersandro2000>
or when it is necessary
<supersandro2000>
wait thats almost flakes
justanotheruser has quit [Ping timeout: 260 seconds]