IRC channel logs
2021-09-23.log
back to list of logs
<civodul>everything Bellard does is so impressive <efraim>anyone know if its possible to see conda build recipies? <civodul>efraim: there's a link in the update on conda-forge in the blog post <civodul>my understanding is that before conda-forge, things were decentralized: you could have recipes that you'd keep for yourself <civodul>rekado_: i can spend a bit of time answering a couple of points on birdsite if you want <rekado_>I’m out for a couple of appointments today and wouldn’t be able to do that even if I knew how to write :) <civodul>though if you don't know how to write, then i don't even know how to type :-) <rekado_>hah, i actually don’t know how to type these days! I finally got around to setting up the keyboardio variant of the Atreus keyboard with home row modifiers… what a ride! *civodul is on a basic Dell keyboard with us-intl layout, not very hacker-looking <rekado_>my wrists are a bit sensitive ever since I switched to using these laptop keyboards. I have the kinesis advantage2 from work, but I don’t have enough space on my desk, so I’m trying to get used to this small atreus instead. <civodul>rekado_: done, but do chime in as you see fit! <PurpleSym>rekado_: I see a few packages in bioinformatics.scm use pandoc-citeproc. Do you know whether they manually invoke citeproc or just use it via pandoc? I’m going to remove pandoc-citeproc, since it’s deprecated. <rekado_>PurpleSym: at least PiGx explicitly looks for pandoc-citeproc. That’s because pandoc complained if it wasn’t there. But I don’t think PiGx explicitly *calls* pandoc-citeproc. <zimoun>And is Guix somehow a Monad? ;-) <civodul>zimoun: this paper looks interesting <civodul>how do you always find those nice references? <dstolfa>zimoun: spivak has other pretty good introductory books and even recorded lectures <dstolfa>re guix being a monad: maybe, in some obscure undefined k-category (: <zimoun>civodul: I procrastinate too much ;-) <zimoun>not always the opinion of partner :-) <rekado_>today I’ll try to figure out what to do about my rserver patches. <zimoun>dstolfa: yeah for sure. :-) Although I think this functional hygiene Guix provide is under represented and for me it has been a blow-up. Maybe my «monad is burrito» fallacy. ;-) <rekado_>there’s a bug: when rserver runs as root it won’t actually use the target user’s session to determine the R version; it uses the root user’s configuration (which likely doesn’t even exist). <dstolfa>zimoun: i understand the temptation to think of these things in a structured way, i'm just more of a formal person that would like a, well, formal definition of things without any handwaving. e.g. a claim "guix is a monad" has to define the monad fully, and ideally find the adjunction that corresponds to it (with the definition of the categories) <dstolfa>i see too many things claim they are something only to fall apart when i try to use them as such <rekado_>PurpleSym: the rstudio-server package uses ghc-pandoc instead of pandoc. <rekado_>this leads to a really large closure. <rekado_>should work to use the static pandoc package instead. <nckx>dstolfa: As someone with Haskell on the perennial ‘things I want to try when I have the time’ list (which I never do), your ‘notable example’ comment intrigues me. Care to elaborate? How does it fall short of what? <rekado_>PurpleSym: I’m trying to run rserver from a git checkout, built within a Guix environment <rekado_>it keeps complaining about a missing database table <rekado_>it creates the database just fine, but it doesn’t seem to contain everything <rekado_>needed to copy the database migration file to the “db” directory relative to the current working directory