IRC channel logs
2021-05-18.log
back to list of logs
<PurpleSym>zimoun: The incoming traffic to guix-science from cnrs.fr are probably you guys? :) <zimoun>PurpleSym: hey! yeah probably. Cool! How do you see the traffic ? <PurpleSym>(Not sure if you can see them as an org member.) <zimoun>oh cool, I did know this feature <civodul>this reproducibility workshop we're having is pretty exciting! <civodul>there were ~80 people yesterday, 65 now <zimoun>efraim: I have seen you pushed Julia packages, cool! Do you use Julia ? <efraim>I figured starting with the Julia dependencies first would be easiest, then I'd wrap the whole thing in a container <efraim>I have about another 70 packages so far <zimoun>efraim: ah cool! You have already packages this 70 packages or WIP? <efraim>some are WIP, probably more than half just need some clean-up work before upstreaming <efraim>here's what I had at the beginning of the day <efraim>of course that time it didn't paste :/ <zimoun>oh neat, you have julia-documenter. I have been too lazy to package it. :-) <zimoun>do you it by “hand“? Or do you have kind-of “importer”? <zimoun>because I am trying to have a Julia importer and piouf! boring… :-) <efraim>all by hand. I search for julia-<packagename> or PackageName.jl and normally get lucky <efraim>oh yeah, documenter had some cyclical dependencies, I must've figured something out but i'll need to trace through it again to delete the unneeded ones <zimoun>yeah me too, but the dependencies sometimes lead you really far from the initial package. ;-) <efraim>sortingalgorithms and statsbase sorted itself when I packaged sortingalgorithms@1 <zimoun>yeah, and I do not know how Pkg manage that. <efraim>I'm sure they skip the tests, that's how I made it work in the end <efraim>or precompiled packages to break it at some point <efraim>do you find the julia tests run slowly? <efraim>they all seem to be single threaded <zimoun>yeah, the test suites are really slow. Always using only one core, at least each time I am checking <civodul>i pushed edits to the 1.3.0 announcement for hpc.guix.info <civodul>should we go ahead and push right now? <zimoun>I note you do not like my «It extends and simplify the use of [`guix -repl`]» ;-) <civodul>zimoun: yes, i wasn't sure what it meant or what the message was <civodul>let's push it tomorrow at a better time :-) <zimoun>civodul: let synchronize the push with a message on guix-science. :-)