Search code examples
clojurenamespacesclojurescriptread-eval-print-loopshadow-cljs

Why does CIDER's REPL some times start as "shadow.user>" and in others it starts as "cljs.user>"?

I am using Clojure, ClojureScript, shadow-cljs, Emacs, Cider, and Lein to develop a web app project.

This is the usual workflow: inside a project and for ClojureScript projects, execute cider-jack-in-cljs, choose shadow-cljs, then shadow for REPL type, and app for build option.

Sometimes, I start a project and see as the top-level in the REPL:

shadow.user>

In other times, I see:

cljs.user>

I believe the names indicate the namespace been used.

But, why are different start-ups happening? What do the different names imply (beyond the namespace)?

My expectation was that I would have the same namespace on every startup of the REPL.

Solution

  • shadow.user> is generated after executing cider-jack-in-clj. And cljs.user> is generated after executing cider-jack-in-cljs.

    The s at the end of the command's name is subtle but crucial. I can't say much about the implications.

    Since the commands are similar, I believe that I accidentally executed cider-jack-in-clj instead of cider-jack-in-cljs. This generated different namespaces for the REPL and a wrong impression that the REPL namespace would vary under same conditions. It only varies if the command changes!