announcements

Jacob Emcken 2026-02-24T15:04:12.656809Z

https://github.com/jacobemcken/aws-simple-sign 2.3.0 is available A library that can generate presigned URLs for S3 and sign HTTP requests for AWS (Babashka compatible). Added: • Allow overriding response headers Thanks to @will.semple3 for the suggestion

❤️ 1
🎉 6
2026-02-24T18:55:34.368829Z

https://cljdoc.org/d/io.github.noahtheduke/splint/1.23.1/doc/home: a linter focused on style and code shape splint v1.23.0 & v1.23.1 New Rules • lint/empty-loop-in-fn: Prefer (defn foo [] (... (recur))) over (defn foo [] (loop [] (... (recur)))). • style/prefer-var-dispatch: Use a var for a multimethod's dispatch function: (defmulti example #'dispatch-fn). Only for CLJ, disabled by default. • style/defmulti-arglists: Always include :arglists in the metadata of a defmulti: (defmulti example {:arglists '([obj])} #'dispatch-fn). Relevant changes: • lint/redundant-call: Make sure the function isn't aliased from another namespace. lint/defmethod-names: Skip methods with unquoted dispatch vals. (For example, (defmethod foo ~some-val ...).) • Fix some bugs in the clj-kondo hook for noahtheduke.splint.rules/defrule. • Add hook for noahtheduke.splint.pattern/pattern to ease usage in other libraries. • Deprecate lint/try-splicing, extend lint/body-unquote-splicing to cover try.

2
🎉 11
borkdude 2026-02-24T18:56:52.557689Z

I wouldn't recommend this for CLJS > style/prefer-var-dispatch: Use a var for a multimethod's dispatch function: (defmulti example #'dispatch-fn).

👍 1
2026-02-24T18:57:54.488939Z

oh? i don't know anything about vars in cljs

borkdude 2026-02-24T18:58:36.133519Z

it bloats the final artifact with unnecessary info

2026-02-24T19:00:00.108709Z

interesting

2026-02-24T19:00:13.889389Z

i'll cut a point release to limit this to CLJ

borkdude 2026-02-24T19:00:28.169489Z

in CLJS you should warn against it ;)

😂 1
2026-02-24T19:00:45.515739Z

lol i'll leave that to you

borkdude 2026-02-24T19:00:52.380799Z

people are going to change this in .cljc files without being aware of this

2026-02-24T19:08:17.463969Z

sure, good points

borkdude 2026-02-24T19:08:43.926779Z

it's not disastrous but not exactly a recommendation in general

2026-02-24T19:09:24.069729Z

yeah, i'll narrow it to clj and leave it off by default

2026-02-24T19:10:14.733819Z

i've been thinking about a counterpoint rule for performance, about vars. i'm not sure how it would look, so i haven't written it yet, but i know recommending vars is recommending slower code

borkdude 2026-02-24T19:10:44.713479Z

have you benchmarked multimethods on vars vs the deref-ed function?

borkdude 2026-02-24T19:10:58.244419Z

I've tested it informally but I didn't see any difference