Pink Gorilla ClojureScript kernel dependencies

How to run it:

  ; run once to install shadow-cljs
  install.sh 

  ; build bundles:
  ./build-small.sh
  ./build-gorilla.sh
  ./builld-mariacloud.sh

  ; run the webserver that provides the bundles
  ./run-web.sh

What happens

• Bundles are created in /out/small /out/gorilla /out/mariacloud
• Each BUndle has:
• index.transit.json - this file is used by shadow-cljs loader.
• ana/*.transit.json - analysis files that are used by the self hosted clojurescript anlyzer
• js/*.js - compiled namespaces from npm and maven that will be read by the shdow-clj loader
• src/. - source files (can be cljs or js) of the bundled namespaces.

Advantages

• CLJS Kernel has no dependencies on the libraries that it exectutes

Further Work:

• compile all namespaces for a dependency without specifying namespaces
• dont provide bundles, but serve on a per namespace basis
• allow multiple versions numbers for each library.
• dynamic on the fly compilation of dependencies via api