related to the discussion just above, I wonder how are people using babashka and cider together. If you are working on multiple scripts, do you start multiple bb nrepl servers or just use the same one? And you have a different cider repl for each or you normally share that as well?

it depends. bb supports the concept of a project and classpath

but there is no default classpath (yet). you have to set it manually like in bb.edn :

{:paths ["."]} ;; add local dir to classpath

ah ok yeah, maybe to be on the safe side you can start a nrepl-server for each project

sure, that also works

just a bit more tricky with the ports I guess, you have to find the next free port to use

but should not be that hard

you can start an nrepl server on port 0

and this will select a free port. you will have to parse that port from the output then to connect

ah nice that's cool

we could make bb write an .nrepl-port file but so far I've hesitated if we should do this since there might also be other Clojure REPLs doing that

but in #nbb that is how it works already since @viesti just decided to do this in his implementation

so perhaps #babashka should also just do it now ;)

ah yeah, it's likely that were there is a bb script there is also a clojure project in the same project

exactly

I guess it helps the jack-in process. we could decide to not overwrite existing nrepl-port files or so. it's a tricky problem

Is there a way for babashka.curl to not automatically follow redirects? Currently --location seems to be hard coded.