Michiel Borkent
Michiel Borkent
How would this affect performance? Are there any other future limitations that might come with switching?
> looks like if its perf critical, raw sqlite is to be used otherwise for smaller things the perf tradeoff is worth the cgo drop. I do not follow this...
Maybe a new project is the safest and sanest way to go, but on the other hand I also don't feel like maintaining multiple sqlite projects. Perhaps we can wait...
Btw are you also aware of this one? https://changelog.com/podcast/626. It came onto my radar today in my podcast feed, haven't listened to it.
Went with musl support for now. We'll keep this PR as a reference for an alternative pod if things don't work out.
Sounds good to me, what about you @oxalorg ?
Good idea!
@dbohdan This is probably a bug. E.g. this works: ``` $ bb -e '(-> (babashka.process/pipeline (babashka.process/pb "echo hello") (babashka.process/pb "cat")) last deref :out slurp)' "hello\n" ``` but the `:in` option...
@yuhan0 For some reason, via the reader macro path, what should have been a MapEntry is a vector in the backing map. If we change this: ``` clojure (valAt [this...
A repro of the problem: ``` clojure user=> (defn foo [v] (clojure.lang.MapEntry. (first v) (second v))) #'user/foo user=> (foo [1 2]) [1 2] user=> (type (foo [1 2])) clojure.lang.MapEntry user=>...