Hacker News Reader

Comment by markokocic | Hacker News Reader

I remember reading a while ago about an initiative to switch Emacs elisp implementation to use Guile backend. I haven't been watching that closely.

Does anyone have more information about the status of that project/initiative?

View on HN

It's dead. It never really got legs; similar projects are proposed pretty much monthly on emacs-devel only to die quietly.

Yeah, it's fighting with "Let's reimplement Emacs in Common Lisp" for most often tried, yet ultimately unsuccessful project. I think there was a GSoC project recently, but haven't heard anything about its results.

There is Deuce [1], a project to port Emacs to Clojure. Specifically, the goal is to port the C core entirely to Clojure and replace many redundancies with JVM primitives; also, source-to-source transformation from Elisp to Clojure allows the editor to interpret Elisp on the fly.

[1] https://github.com/hraberg/deuce

The JVM seems like a really silly target for this. OH BOY! An editor that won't behave correctly in ANY environment!

You might have more experience in this area than I, but is this not the point of the JVM?

By targeting the JVM, you are giving up on a native user experience. Makes perfect sense when the user interface of your software is programmatic; insanity when you're trying to make software to be used by humans. Clojure is a good lisp, but hardly reason enough to toss all UI convention for. C + Guile makes about a zillion times more sense.

I disagree; you are not tied to Swing or any Java UI toolkit merely by using Clojure on the JVM. You can just as easily integrate with a GTK, Cocoa, and whatever Windows uses.

You'll notice how the current Emacs is weird on all three platforms by default (at least they switched the scrollbar position). C/Guile isn't a panacea for a UI frontend, they're entirely orthogonal.

Reply on news.ycombinator.com