I was starting to worry that protorepl was abandoned. I'm still very new to using RDD. Do you switch to the namespace of a function you're working on for that namespace? It seems like the general guidance is to do RDD in a namespace that isn't part of application code. I see people create a file and define a user namespace a lot. I'm still unclear as to how to develop/modify functions via RDD that exist in a defined namespace.

I use "Rich Comment Forms" -- code inside

(comment
  <scratch code goes here>
  ,)

Rich Hickey uses that approach -- you'll see RCFs at the end of several Clojure files in the core libraries -- and Stu Halloway and others also advocate using comment forms to hold scratch code as you're exploring the solution.

Never type into the REPL directly: always write code and eval it -- that's it in a nutshell.

Since you're working in a source file, nearly all Clojure integrations for editors take care of evaluating the code in the correct namespace automatically, so you never need to worry about that.

BTW, the Chlorine package for Atom doesn't even have a REPL "input" area: so it forces you to always write code and eval it. But you'll normally start a REPL yourself from a terminal, with a Socket REPL running, and you can still type into that if you really feel the urge to just try out an expression without putting it in source code.

I have a couple of videos on YouTube showing my Chlorine/Atom workflow https://www.youtube.com/channel/UC8GD-smsEvNyRd3EZ8oeSrA -- in those videos I'm also using Cognitect's REBL data browser but I've switched to Reveal now so I really need to do some updated videos.

Thank you! This is most helpful. Going to try chlorine and the rich comment workflow right now. I had already gotten used to writing forms and sending them to the repl with protorepl which was quite a revelation.

I need to brush up on the history of all these different repls at some point.

The #chlorine channel nearly always has some folks who can help you, if you get stuck.