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
hekkaidekapus_ is now known as hekkaidekapus
la-jesystani has joined #haskell.nix
<la-jesystani> howdy! really liking the look of this, just reading through the guide and was wondering how easy it is to get this working with nix flakes?
la-jesystani has quit [Ping timeout: 260 seconds]
la-jesystani has joined #haskell.nix
<julm> la-jesystani: it should work with flake.nix by now, eg. https://github.com/J-mie6/ParsleyHaskell/issues/5#issue-690751843
<julm> note however that all-hies will not work with flake.nix and is about to be deprecated in favor of haskell-language-server
__monty__ has joined #haskell.nix
la-jesystani has quit [Ping timeout: 265 seconds]
fendor has joined #haskell.nix
dhess has joined #haskell.nix
<dhess> Hi, I'm finally biting the materialization bullet :) When I see something like this:
<dhess> Should I set index-state in my nix expressions to "2020-10-26T00:00:00Z" or "2020-10-25T23:13:25Z" ?
<dhess> Next question: the `plan-sha256` and `materialized` hints I'm getting from the build look like this:
<dhess> and not like the messages shown in the documentation here: https://input-output-hk.github.io/haskell.nix/tutorials/materialization/
<dhess> but for example, regarding this hint from my build:
<dhess> trace: To make this a fixed-output derivation but not materialized, set `plan-sha256` to the output of /nix/store/mvzg21xz426b7pbdj7rsdy4y5q1h40vn-calculateSha
<dhess>
<dhess> that path /nix/store/mvzg21xz426b7pbdj7rsdy4y5q1h40vn-calculateSha doesn't exist
<dhess> help? :)
<__monty__> Since cabal has to fall back I believe you should use the older index-state. Though it'll probably just keep falling back if you use the newer one.
<__monty__> Those binaries don't exist until you've finished building the expression afaiui.
<dhess> building which expression?
<__monty__> Whatever expression you're building when you get those warnings.
<dhess> oh no, the expression has been built
<dhess> the path still doesn't exist.
<__monty__> I think those dogs are just showing their age a little, re the different traces.
<__monty__> Hmm, does the generateMaterialized path exist?
<dhess> __monty__: no, that doesn't exist after the build, either.
<__monty__> Have you checked all the paths? (Starting to have to guess, I'm out of my depth.)
<dhess> not sure what you mean by that
<__monty__> There's many calculateSha paths in your output, did you check all of them or just the first ones? I'm thinking *maybe* you got the deps from the cache and those binaries don't come along with them.
<dhess> Ahh, I see. None of them exist.
<__monty__> I guess we'll wait for someone to drop a knowledge bomb : )
<dhess> :)
<dhess> so then I'm left wondering how the maintainers generate their materializations
<dhess> known issue since June 13 :()
<dhess> :(
<__monty__> The other issue does have instructions for how to materialize though.
<dhess> at least one person was not able to make that work
<dhess> anyway I'll try that after I'm done reading all these issus
<dhess> issues
<__monty__> I think arian was talking about the original issue, i.e., they don't exist after just building the expression.
<angerman> dhess: that that -calculateSha path doesn't exist is a bit strange.
dutchie has left #haskell.nix ["WeeChat 2.9"]
<michaelpj> angerman: it's the same old issue
<michaelpj> I should just get rid of the helpful messages :( and go back to telling people how to build the script manually
<michaelpj> it's quite annoying, but I don't see any way to reliably ensure that the script exists
<angerman> gah. I need to spend more time with haskell.nix again :-/
<__monty__> Maybe add a note to the helpful message to a wiki/doc/issue in case it doesn't exist?
<dhess> michaelpj: what's your recommended solution for now, then?
<dhess> And how do you all materialize outputs for haskell.nix if the calculators don't get built?
fendor has quit [Ping timeout: 272 seconds]
<michaelpj> dhess: if you build with `checkMaterialization=true` they will get built, otherwise the `updateMaterialization` script can be found as an attribute on the project: https://github.com/input-output-hk/plutus/blob/master/nix/dev.nix#L17
<dhess> oh I did not know about `updateMaterialization`!
<dhess> oh my so this can all be automated?
<dhess> well, partially anyway :)
acarrico has joined #haskell.nix
dutchie has joined #haskell.nix
<michaelpj> dhess: yes, in principle it should be all nice, it's just getting the information to the right place!
dutchie has left #haskell.nix ["WeeChat 2.9"]
acarrico has quit [Ping timeout: 258 seconds]
acarrico has joined #haskell.nix
acarrico has quit [Ping timeout: 240 seconds]
dutchie has joined #haskell.nix
fendor has joined #haskell.nix
proofofkeags has joined #haskell.nix
acarrico has joined #haskell.nix
la-jesystani has joined #haskell.nix
joshmeredith has joined #haskell.nix
ocharles has joined #haskell.nix
hekkaidekapus_ has joined #haskell.nix
hekkaidekapus has quit [Ping timeout: 240 seconds]
<dhess> michaelpj: still around?
dutchie has left #haskell.nix ["WeeChat 2.9"]
__monty__ has quit [Quit: leaving]
hekkaidekapus_ is now known as hekkaidekapus
dhess has quit [Remote host closed the connection]
dhess has joined #haskell.nix