[maehwasu]

As someone who uses Clojure a lot:

It would be a lot better for language adoption if people would write more blog posts showing best practices for tooling, and less vague stuff about how awesome the language is.

For example, the startup time thing is completely irrelevant if you use Emacs/VSCode with a standalone REPL that you connect to.

It's non-trivial to set this up (more because of lack of consolidated info rather than time it takes), but it can easily be standardized across machines/envs (I Dockerize all projects in a simple way to allow team members to get started up with REPLs easily).

[dragandj]

I totally agree, and that's why I am going to write a book about that:

https://twitter.com/draganrocks/status/1226811229362147331

It has to wait until I wrap up the books that are currently in progress [1], but I expect to start this summer.

[1] https://aiprobook.com

[maehwasu]

Awesome.

To put my own money when my mouth is, here is the bash script to start up a nREPL server in Docker: https://pastebin.com/PPmfDPyA

Here is the Dockerfile: https://pastebin.com/ymcUFdYT

And here is a sample deps.edn with the cider alias: https://pastebin.com/2a7vSFD7

To run this, you'd just put them all in one directory, run cider-up.sh, and then connect to the REPL with Cider in EMACS, or in Calva for VSCode with "Jack in to or Connect to REPL Server" and then "localhost" "4444"

[reilly3000]

Random fan here: you DO rock. Thanks for your incredible work on making ML a first-class citizen in Clojure!