Hi @U05V9JANJQK! Currently, the poly tool only supports backend code, but it's on our list to also support ClojureScript together with https://github.com/thheller/shadow-cljs (which will also support cljs for the backend). You can read more about it https://github.com/polyfy/polylith/discussions/301 and follow or participate in the discussion.

Ah, that's good news. Still I am curious if it's a good idea to pair a web framework/template with Poly in principle or if Poly works best when rolling your own stack so as not to conflict with some opinions that a framework might have?

I've refactored a kit app to use polylith. It essentially boiled down to starting a polylith app from scratch and adding the libraries that Kit uses to the mix and re-creating some of the Kit boilerplate (esp. around the repl-driven development). Note: I wasn't using "kit modules" but directly using the libraries that Kit depends on, like ring or next.jdbc or migratus, etc) Bonus: I have a lot more knowledge now of how Kit works and how to build a a web app from components and libraries, not so dependent on Kit anymore. And I've swapped out ring/etc with Pedestal.

My frontend uses htmx so I haven't had to deal with cljs

Would you be interested in open sourcing your Kit/Polylith edition?

@U05V9JANJQK - I cannot publicly release the source of this particular project, however I have a (personal) project that I was considering releasing at some point. It might need a bit of cleanup first but uses a similar architecture (though an earlier iteration) - polylith + pedestal + next.jdbc + htmx I just realized that I rewrote the whole thing in Common Lisp, so that's not the personal project where I used poly+pedestal+htmx... Anyway if I find/do something that can be open sourced I definitely will!