<dtz>
no thoughts == could be trivial, could be impossible, who can say
<dtz>
:D
<dtz>
haha okay
<dtz>
aw MAN worked my way towards using latest hydra + latest Nix and aw man things are smoother on my poor server ^_^
<dtz>
\o.
<dtz>
\o/
<dtz>
*two arms
<Sonarpulse>
dtz: nice!
<dtz>
I... it's kinda ridiculous. haha :D.
<dtz>
had to check many times that my jobs are running and actually being uploaded
<dtz>
lol
<Sonarpulse>
I've heard from a coworker that moving to newer hydra is really hard
<dtz>
well I have that relatively solved, but it was somewhat problematic / "new" to tackle using latest Nix with it for uninteresting reasons
<Sonarpulse>
hmmm
<Sonarpulse>
also haskellPackages.hello built on linux-to-darwin branch
<Sonarpulse>
which *should* be the same haskell changes as master
<Sonarpulse>
but trying it there now
<Sonarpulse>
(macOS in both cases)
<dtz>
oooooo re:linux-to-darwin
<dtz>
awesome stuff, I poked at that a while back but didn't have time to "bring to home"
<dtz>
time and perhaps experience :D. But anyway that's great
<dtz>
i'm sure darwin users would be very happy about that , if nothing else we can test w/o darwin builders :P
<dtz>
(although borg seems blessed with surprisingly equal-ish darwin build-power
<dtz>
)
<Sonarpulse>
dtz: well i'm sort of tip-toeing around actual LLVM cross so cfar
<dtz>
:D
<dtz>
one day we'll just build clang for each platform (bootstrap?) and everything else will be interacting with that platforms clang-based universal driver
<dtz>
:D
<dtz>
now to figure out how to best move my hydra logs to S3 (existing logs)
<dtz>
esp since `log-servers` was removed on the path to 1.12/2
<dtz>
(honestly I like the idea of fetching via store URI better anyway but still I have to script the movement it seems hehe)
<dtz>
probably NBD in the morning :)
<dtz>
downside of my latest deployment: hydra-queue-runner prints "upload progress" lol
<dtz>
only really a problem b/c I have little faith in the robustness of journalctl
<dtz>
i'll probably just bash script something :3 l ol
<Sonarpulse>
dtz: good luck!
<Sonarpulse>
yeah you'll have to tell me about your long term llvm plans more
<Sonarpulse>
dtz: build on master
<dtz>
! Well good to hear, hmm. Can you build through pandoc? (Not just a personal need, bunch of things use it fwiw). And what rev are you testing exactly? I'll use that when investigating tomorrow :)
<dtz>
Thanks for giving it a go, sorry for possible false alarm. Hmm.
jtojnar has quit [Read error: Connection reset by peer]
jtojnar has joined #nixos-dev
<srhb>
Has anyone explored passing information to the builders in a different way than environment variables? I'm starting to believe it's a fundamentally problematic concept due to the differences in environment sizes across platforms, and we're taking up more and more space.
jtojnar has quit [Ping timeout: 245 seconds]
jtojnar has joined #nixos-dev
jtojnar has quit [Read error: Connection reset by peer]
jtojnar has joined #nixos-dev
<niksnut>
srhb: Nix 2.0 supports passing attributes via a JSON file
<srhb>
niksnut: That could certainly help...
<srhb>
I wonder if we could abstract the generic builders enough to make it easy to _never_ put anything in the environment.
<srhb>
$(getattr foo) vs $foo or something
<Dezgeg>
non-exported environment variables shouldn't have a size limit
<srhb>
Dezgeg: How do you mean?
<Dezgeg>
if they are passed in JSON you can just source a script that sets them as local variables in bash, not something that's exported
<srhb>
wait, that doesn't affect arg_max?
<Dezgeg>
no
<srhb>
o_o
<srhb>
TIL.
<Dezgeg>
why would it, they're not being passed to child processes (and thus, the kernel) :P
<srhb>
That makes sense of course. I hadn't thought of that.
pxc1 has joined #nixos-dev
drakonis has joined #nixos-dev
pxc1 has quit [Ping timeout: 256 seconds]
Jackneilll has quit [Quit: Leaving]
Jackneill has joined #nixos-dev
jtojnar has quit [Ping timeout: 252 seconds]
jtojnar has joined #nixos-dev
<sphalerite>
Dezgeg: but then they aren't really environment variables are they?
<Dezgeg>
that was the plan
jtojnar has quit [Read error: Connection reset by peer]
jtojnar has joined #nixos-dev
jtojnar_ has joined #nixos-dev
jtojnar has quit [Ping timeout: 252 seconds]
jtojnar_ is now known as jtojnar
__Sander__ has joined #nixos-dev
jtojnar_ has joined #nixos-dev
jtojnar has quit [Read error: Connection reset by peer]
jtojnar_ has quit [Ping timeout: 240 seconds]
jtojnar has joined #nixos-dev
orivej has quit [Ping timeout: 248 seconds]
drakonis has quit [Remote host closed the connection]
jtojnar_ has joined #nixos-dev
jtojnar has quit [Read error: Connection reset by peer]
drakonis has joined #nixos-dev
<Sonarpulse>
peti: trying to redeam myself with the little one :)
jtojnar has joined #nixos-dev
drakonis has quit [Read error: error:1408F10B:SSL routines:ssl3_get_record:wrong version number]
drakonis has joined #nixos-dev
jtojnar_ has quit [Ping timeout: 260 seconds]
Lisanna has joined #nixos-dev
<peti>
Sonarpulse: No worries. Mistakes happen during extensive refactoring. In hindsight, I should have gone through haskell-updates to test the PR before merging.
<Sonarpulse>
peti :) ironically normally i test on linux more than darwin, but this time it was linux not darwin with the issue
<Sonarpulse>
I couldn't find it in the logs, but I think dtz reported seeing an issue with the --framework-depends too
<Sonarpulse>
so perhaps that flag is newer than I thought
<Sonarpulse>
(I know latest cabal-install recognizes it, and 1.24's ChangeLog has the cabal file version if it being added)
__Sander__ has quit [Quit: Konversation terminated!]
Lisanna has quit [Quit: Lisanna]
Lisanna has joined #nixos-dev
Lisanna has quit [Client Quit]
Lisanna has joined #nixos-dev
Lisanna has quit [Client Quit]
Lisanna has joined #nixos-dev
Lisanna has quit [Client Quit]
Lisanna has joined #nixos-dev
Lisanna has quit [Client Quit]
Lisanna has joined #nixos-dev
Lisanna has quit [Client Quit]
Lisanna has joined #nixos-dev
Lisanna has quit [Remote host closed the connection]
lassulus has quit [Quit: WeeChat 2.0]
lassulus has joined #nixos-dev
lassulus has quit [Client Quit]
lassulus has joined #nixos-dev
drakonis has quit [Remote host closed the connection]
FRidh has joined #nixos-dev
FRidh has quit [Ping timeout: 245 seconds]
Lisanna has joined #nixos-dev
Lisanna has quit [Remote host closed the connection]
jtojnar has quit [Remote host closed the connection]
<Sonarpulse>
like it's from the compiler's perspective not library's perspective
<dtz>
sorry, but :(
<dtz>
haha
<Sonarpulse>
I half want to add proper bootstrapping support to meson
<Sonarpulse>
and then pull some coup to convince gcc and llvm to use it
<Sonarpulse>
meson is the better cmake
<Sonarpulse>
it understands simple cross
<Sonarpulse>
it doesn't understand bootstrapping though
<Sonarpulse>
dtz: I dunno if Ralith is ever on IRC
<Sonarpulse>
but I pushed a comomit
<Sonarpulse>
and rebased
<Sonarpulse>
such that clang is built with *0* libraries *gasp*
<Sonarpulse>
now just need to show it where compile-rt and libgcc is
<Sonarpulse>
we can then remove libgcc from libc
<Sonarpulse>
something like that
<sphalerite>
Sonarpulse: yes Ralith is on IRC quite a bit (via matrix so they'll see whatever the matrix bridge doesn't eat), but doesn't seem to be in here, only in #nixos
<Sonarpulse>
sphalerite: thanks
<Sonarpulse>
dtz: it's just amazing how bad all this stuff can be...