Clojurians Log v2
Clojure programming
Channels
Apps
scittle
vemv 2024-06-23T20:34:06.304749Z
A cider user is seeing a nil classpath for the scittle runtime. https://github.com/clojure-emacs/cider/issues/3716#issuecomment-2185315270 Does that ring a bell? We had seen (and you gracefully fixed) a similar thing for another runtime (bb probably)
borkdude 2024-06-23T20:44:26.170969Z
scittle doesn't have a classpath, as far as I know
borkdude 2024-06-23T20:44:42.893909Z
scittle doesn't even assume a file system
vemv 2024-06-23T20:53:51.604579Z
Would it make sense if https://github.com/babashka/scittle/blob/691ed3a31a9fd7e794cb5bf302f4b693ed002c56/bb.edn#L29 exposed as a classpath the set of dirs that can be possibly served? We'd be fine if the classpath was nil when running scittle in other ways
borkdude 2024-06-23T21:03:23.745389Z
scittle's nrepl thing is a generic tool that doesn't know anything about what directories are served (those are served by an app server in a different process which sci.nrepl doesn't know anything about)
borkdude 2024-06-23T21:04:04.541889Z
so I don't think it makes sense to think in terms of classpath here maybe
borkdude 2024-06-23T21:05:05.704039Z
if it's useful I could add an option to sci.nrepl so you can optionally pass a classpath which it then sends back
borkdude 2024-06-23T21:05:18.300319Z
but I'm not sure if that makes sense
borkdude 2024-06-23T21:05:23.890739Z
checking back tomorrow
vemv 2024-06-23T21:08:00.927259Z
Thanks much for the hammocking. I'll post a thing in the GH thread, feel free to follow
🙏 1
borkdude 2024-06-23T21:09:14.488289Z
I'll subscribe, thanks