Clojurians Log v2

I ❤️ lumo, and really believe in the approach, especially the bit about embracing the host platform and tools. I've recently been trying to get an nrepl integration working - looks promising based on https://github.com/djblue/nrepl-cljs. I'm also interested in the idea of a solid browser development flow. Surely others are too? It would make me very sad if this project were to die.

I don't think Lumo is "dying" anytime soon. I don't expect new features, but that also depends on you and anyone willing to make a contribution. I expect Antonio or Andrea to keep lumo in sync with latest clojure/script and node releases. There's still great untapped potential with expanding lumo, but competing with jvm based dev tools is a lost battle in terms of speed, not an excuse not to push for self-hosted clojurescript compiler research and experimentations.

> I expect Antonio or Andrea to keep lumo in sync with latest clojure/script and node releases. Don’t.

expect was a wrong word here, but I haven't followed up on recent events. Tough still maintaining and fixing it on nixpkgs. And using lumo ofc.

To be clear: Lumo is not going to get any updates. I’m not interested in spending any more time working on it or reviewing PRs

Ok, that's clear enough, so I'll fork it.

Happy to add you as a collaborator

Do mean specifically competing with the likes of figwheel and shadow, lein, deps? Why do you think that is?

many reasons for it, but I'm not a compiler researcher, but the jvm runtime is just lightyears ahead in almost every regard, cpu, multithreading and memory. But the libraries and their stability in java is enough for library maintainers to build powerful tools on solid foundation than the wilderness of npm packages. But node or any native js based environment will probably catch up big part of this difference in next decade.

Agreed re: jvm vs node (at least for now), but I guess I always envisaged something like lumo also providing a killer cljs web-browser experience (like shadow-cljs, but without deps.edn, JVM and so on).

sooner or later, it will arrive

It does seem rather obvious that it will, but given what's going on in the rest of the cljs community around shadow, figwheel, etc, and the trouble lumo is in, I'm questioning my logic here. Lumo seems like a good basis for building such a stack. I'm no compiler researcher either, but my smell-ometer goes crazy when my projects have two dependency descriptors (in fact, with shadow-cljs and cursive, I have three!).

I wonder what the Clojure(script) leads think about all this? It's hard to imagine that they think ultimately the JVM has a roll in the cljs landscape (beyond tactical tooling solutions in the short term).

"Donating" the project to clj-commons could also work, but that still has the problem of finding someone who wants and is able to maintain it