[kamaal]

I think from the discussions and other threads what I infer is that we need a total re-look at the concept of an editor, Your way actually looks good. Except that we need to retain the 'Emacs infrastructure'.

We need to really keep Lisp as the extension language.

I think what we need is Light Table with something like this : http://www.kickstarter.com/projects/568774734/emacsy-an-embe...

You have to develop Light Table ON the Emacs OS!

[ibdknox]

Lisp is the extension language :) Everything so far is Clojure + ClojureScript with the language backends written in the language they support (python in python and so on)

[stcredzero]

How about Stallman's idea for GUILE (Scheme) uber-alles. Do source-to-source translation of approximations of the supported languages to Clojure. This would allow everyone access to almost everything in the tool.

[lucian1900]

That effort hasn't even materialised for Scheme+Elisp for Emacs.

The whole approach is guaranteed to not produce a perfectly compliant of any other language.

[stcredzero]

Yes, that is the price. That's the price for native access to the entire codebase of the underlying editor.

So long as it's close enough, you don't need perfect compliance.

[lucian1900]

I disagree. If the support isn't perfect, you might as well not bother and help people learn your native language.

[stcredzero]

I think the popularity of CoffeeScript is a datapoint against your idea. I bet a lot of pythonistas use it because it's less of a shift than Javascript. Also, in such a system, if people are curious enough, they will learn the native implementation language as a result of the system being so constructed.