@stbgz This is exactly what I’ve been waiting for!

But if I were to wish for one thing, it would be a library that given a project.clj (or some other description of dependencies), gives a graph of dependencies.

Now, I see that you would need to download stuff to be able to make the whole dependency graph, so maybe the fn that figures out the deps-graph could take another function which given a url, produced the artifact. I guess that’s what you’re doing with the os namespace in huckleberry

The reason I’m going on about this is that planck has it’s own http-implementation, so it’d be neat if we with planck could do something like (huck/resolve deps planck.http/get)

I think we'll ultimately have something like Boot/Lein for self-hosted ClojureScript

Huckleberry seems to be a step in that direction

@anmonteiro I guess calvin is the boot/lein thing here

right but it doesn't currently have tasks to show the dep graph as you say

and it won't work with Planck if it uses Lumo namespaces

@anmonteiro but I guess my point is that since a lot of io is undefined for cljs as a platform, io needs to be pluggable

yeah, which is why I'm still uncertain how to proceed

I've spoken briefly with Mike to unify IO into something based on protocols that both Planck/Lumo can implement downstream

So have I, but that was before Lumo

But I guess then you’ll end up in trying to define an IO subsystem/protocol for cljs’s to adhere to.

But with Lumo this is becoming interesting again.

But, it would be cool if I could basically do whatevercljsrepl -e '(http/get “”)’

or even whatevercljsrepl -e '(slurp “”)’

Yeah we need to figure out io for the repls so we are not tied to a single platform

That being said Calvin just learned to start Planck repls