[racket] typed racket and new racket languages?

From: Richard Lawrence (richard.lawrence at berkeley.edu)
Date: Thu Jun 16 21:17:14 EDT 2011

Hi Raoul,

> to what degree can one leverage typed racket for new languages in the
> racket ecosystem? 

(You mean for *creating* new languages, right?)  I'm actually in the
middle of doing this right now.

> tho i read that even if there are ways to use it, there be dragons in
> terms of getting the new language just right wrt type enforcement.

Where did you read this?  On the whole, I have found the type system
much more helpful than dragon-like.  I have run into a couple of issues
that I haven't had time/energy to understand/debug yet, especially with
using syntax objects.  I mostly chalk this up to my own ignorance,
though.  At the moment I am working around it with a very thin layer of
plain Racket atop my TR library.

One caveat: I haven't implemented a new language in plain Racket before,
so I can't honestly give an opinion on whether TR makes the task
relatively easier or harder.  But I can say that TR has the only type
system I've ever seen whose usefulness-to-learning-effort ratio has
worked out favorably for me.


Posted on the users mailing list.