IRC channel logs
2025-12-12.log
back to list of logs
<ieure>I understand this is likely to change, so I'm mostly farting in the wind. <rlb>(In clj that's (:require [something :refer :all]).) <ieure>I don't love everything about Clojure, but it got a lot of things right, this included. <ieure>Every other Lisp should steal how Clojure does hash-maps. <rlb>Lot of things I like, but yeah, not perfect. <ieure>rlb, Yeah. Lots of developer experience optimizations at the micro level, seemingly not all that much thought put into big picture stuff like struct types. I'm sure you've seen the flowchart. <ArneBab>ieure: I’d also prefer this do be different, but I think it comes from old C and a ton of code would break if we changed that. So the path forward is teaching that as best practices. <rlb>It's also true that for a good while when clojure started, :refer :all was conventional, so it started off with the same effective default, and then changed style. <ArneBab>(the last "that" refers to #:select or only ☺) <ieure>ArneBab, Yeah, it'd break substantally all existing Guile code. <dsmith>rlb, Unfortunately, dash doesn't do $'...' <rlb>I think I checked, but not sure if newer versions can... <dsmith>Not in trixie: echo $'123' -> $123 <rlb>I think maybe I looked upstream, but I might also be misremembering. <dsmith>ACTION tries to use as few bashisms as possible <rlb>That's actually posix now. <dsmith>Yes. Was just looking at $'' a few weeks ago. <rlb>Oh, right, you just meant in general. <rlb>Up to a point, lately I tend to just require bash for anything "complicated enough", in particular for arrays/hashes and/or printf %q with respect to making it easier to get quoting/escaping and safe interpolation right. <rlb>e.g. "${some_array[@]}", etc. <rlb>But I also stick to sh when I need to -- e.g. sometimes for *bsd portability etc. <rlb>Merged lightening main to pick up ekaitz copyright updates, and added the new manpages. <lechner>Hi, bytevector-slice is great when splitting large bytevectors. Is there also a way to merge two adjacent bytevectors? <rlb>Not without creating a new one afaik, i.e. via bytevector-append.