IRC channel logs

<PurpleSym>Urgh, and now most of my imported Node packages don't build any more. This is *really* frustrating.

<zimoun>civodul presentation at KTH about Guix: https://www.pdc.kth.se/about/events/seminar-reproducible-software-environments-with-gnu-guix-1.1249710

<zimoun>arf, it’s canceled. )-:

<civodul>already! :-)

<rekado>PurpleSym: you may need the old node-build-system that I added back to guix-science

<PurpleSym>rekado: Yeah, probably. I'm not going down that rabbit hole of vendoring a build system and will instead just pin a working Guix commit. There's more recently broken dependencies that I cannot deal with right now.

<PurpleSym>(Beyond node, I mean.)

<rekado>the changes to the build system were very minor

<rekado>I think we should modify the node-build-system to restore building with an older version of node

<rekado>(thus removing the need for this almost exact copy of the node build system)

<zimoun>PurpleSym: is wip-haskell, wip-haskell-updates or wip-haskell-updates or non the branch of the Haskell team?

<zimoun>rekado: you seem the last to have pushed to wip-haskell-updates and wip-haskell-updates-2 back on 2020. Do you remember the aim of these branches?

<zimoun>Does it make sense to cut them?

<zimoun>ACTION zimoun pushed https://hpc.guix.info/blog/2023/05/reproducible-research-hackathon-let-redo

<rekado>zimoun: it’s fine to remove these branches. I have no intention of touching Haskell any time soon.

<rekado>wip-haskell-updates-2 and wip-haskell-updates were both about reducing the closure of pandoc

<rekado>the results of these branches have already made it into the master branch a long time ago

<zimoun>rekado: Thanks. I do not feel confident to cut them… I am going to send an email to guix-devel in case someone depends on it (I do not think so :-))

<rekado>I’ll delete them, no need for an email

<zimoun>ok, thanks.

<rekado>uhm… they have already been deleted

<rekado>maybe I did that a week or so ago when I cleaned up my old branhces

<rekado>*branches

<zimoun>ah, that’s probably me that have not run --prune when fetching in the mean time. In all cases, thanks. :-)

<rekado>the wip-r branch still has that commit to switch to Bioconductor git: https://git.savannah.gnu.org/cgit/guix.git/commit/?h=wip-r

<rekado>no idea how to move forward there

<rekado>in the interest of long-term reproducibility we should move to git, but I’m annoyed by the added friction

<zimoun>well, I do not know. Considering how Bioconductor manage their repository, indeed I do not know if it is worth the effort. Especially when the tarballs should be now ingested by SWH and I am working to also ingest the past. The question is thus more or less Disarchive.

<zimoun>Using Git, it would remove Disarchive from the dance.

<zimoun>But it adds complexity with Bioconductor tags and co.

IRC channel logs

2023-05-16.log