#haskell.nix on 2020-10-09

2020-09-01 01:44 angerman changed the topic of #haskell.nix to: https://input-output-hk.github.io/haskell.nix - alternative haskell infrastructure for nix; logs at https://logs.nix.samueldr.com/haskell.nix

00:18 proofofkeags_ has quit [Ping timeout: 265 seconds]

00:18 proofofkeags has quit [Ping timeout: 265 seconds]

02:00 acarrico has quit [Ping timeout: 264 seconds]

02:07 proofofkeags_ has joined #haskell.nix

02:07 proofofkeags has joined #haskell.nix

03:15 proofofkeags has quit [Ping timeout: 260 seconds]

03:15 proofofkeags_ has quit [Ping timeout: 260 seconds]

07:31 fendor has joined #haskell.nix

08:32 o1lo01ol1o has joined #haskell.nix

08:39 o1lo01ol1o has quit [Ping timeout: 256 seconds]

08:41 o1lo01ol_ has joined #haskell.nix

08:50 o1lo01ol_ has quit [Read error: Connection reset by peer]

08:50 o1lo01ol1o has joined #haskell.nix

09:13 o1lo01ol1o has quit [Ping timeout: 260 seconds]

09:14 o1lo01ol1o has joined #haskell.nix

09:18 o1lo01ol1o has quit [Ping timeout: 240 seconds]

09:24 __monty__ has joined #haskell.nix

09:26 o1lo01ol1o has joined #haskell.nix

09:32 o1lo01ol1o has quit [Ping timeout: 272 seconds]

09:38 o1lo01ol1o has joined #haskell.nix

09:50 o1lo01ol1o has quit [Remote host closed the connection]

09:50 o1lo01ol1o has joined #haskell.nix

10:42 o1lo01ol1o has quit [Remote host closed the connection]

10:44 o1lo01ol1o has joined #haskell.nix

10:51 o1lo01ol1o has quit [Remote host closed the connection]

11:00 o1lo01ol1o has joined #haskell.nix

11:00 o1lo01ol1o has quit [Remote host closed the connection]

11:00 o1lo01ol_ has joined #haskell.nix

11:02 jonge[m] has quit [Quit: killed]

11:02 domenkozar[m] has quit [Quit: killed]

11:02 ptival[m] has quit [Quit: killed]

11:02 siraben has quit [Quit: killed]

11:02 michaelpj has quit [Quit: killed]

11:02 Ericson2314 has quit [Quit: killed]

11:11 jonge[m] has joined #haskell.nix

11:18 <__monty__> I keep getting traces mentioning *-calculateSha and *-generateMaterialized programs that don't exist in my store. Should they be? And if they shouldn't why am I getting these traces anyway?

11:30 o1lo01ol_ has quit [Ping timeout: 240 seconds]

11:32 o1lo01ol1o has joined #haskell.nix

11:33 acarrico has joined #haskell.nix

11:34 julm has quit [*.net *.split]

11:34 joshmeredith has quit [*.net *.split]

11:35 julm has joined #haskell.nix

11:40 joshmeredith has joined #haskell.nix

11:41 hekkaidekapus} has joined #haskell.nix

11:43 hekkaidekapus{ has quit [Ping timeout: 240 seconds]

11:53 siraben has joined #haskell.nix

11:53 michaelpj has joined #haskell.nix

11:53 ptival[m] has joined #haskell.nix

11:53 Ericson2314 has joined #haskell.nix

11:53 domenkozar[m] has joined #haskell.nix

11:59 acarrico has quit [Ping timeout: 260 seconds]

12:04 <terrorjack> How to add a non-Haskell executable dependency? e.g. one of my packages require nixpkgs.nodejs-14_x to build, but I don't find something like buildInputs in the documentation

12:12 acarrico has joined #haskell.nix

12:24 <ocharles> __monty__: https://github.com/input-output-hk/haskell.nix/issues/781?

12:29 lemmih has quit [Ping timeout: 272 seconds]

13:08 <__monty__> Ah, thank you, ocharles. Is a pretty confusing message though.

13:11 <ocharles> Sure, it's a bug

13:11 <ocharles> They tend to be confusing :D

14:12 fendor_ has joined #haskell.nix

14:14 fendor has quit [Ping timeout: 258 seconds]

14:15 <__monty__> Hmm, does haskell.nix never use stackage GHCs? I'm ending up compiling a GHC for a stackage lts.

14:23 <ocharles> haskell.nix applies some patches to GHC

14:24 <__monty__> Which combined with GHC 8.6.3 probably not being cached means I need to build it?

14:25 <__monty__> Awkwardly phrased. If the GHC of the LTS *was* more common, like 8.6.5 it would probably be cached?

14:39 <angerman> __monty__: we'll only build the latest release per each minor branch. And no, we'll never use stackage provided ghcs; they won't work nicely in our nix setup anyway. We'll however build the same ghc version as stackage advertises.

14:39 <angerman> Keeping a *large* set of ghcs around is quite hard on CI. I wish we had a good story for that.

14:42 <__monty__> Hmm, when you say the GHC stackage lists, do you mean latest LTS or something? Because most GHC versions have at one point been in a stack LTS.

14:43 <angerman> each stackage snapshot lists one ghc.

14:43 <angerman> so we'll build that listed ghc (by version).

14:43 o1lo01ol1o has quit [Remote host closed the connection]

14:44 <__monty__> So I shouldn't end up building a GHC for stack projects?

14:44 <angerman> If it happens to be the latest 8.6.5, 8.8.2, 8.10.2, if I remember the current version right, you'll get a cache hit. If you need 8.6.3, you sadly won't.

14:45 <angerman> Once 8.10.3 is released, we'll only build 8.10.3 in CI, and 8.10.2 won't see cache hits anymore.

14:45 <__monty__> But 8.6.3 is in LTS-13.10 and 13.11.

14:47 <angerman> right, so you'll get the proper derivation for it, but we'll only have the latest ghc release versions cached. It's a limitation on how much we can build in CI.

14:48 <angerman> supporting more ghc versions would be nice, but we just don't have the ressources for that right now.

16:03 proofofkeags_ has joined #haskell.nix

16:03 proofofkeags has joined #haskell.nix

17:11 fendor_ is now known as fendor

22:23 __monty__ has quit [Quit: leaving]

22:47 fendor has quit [Remote host closed the connection]