[racket-dev] Single-flonum-ness not preserved in racket/math functions

From: Vincent St-Amour (stamourv at ccs.neu.edu)
Date: Tue Jun 5 17:57:24 EDT 2012

At Tue, 05 Jun 2012 15:45:09 -0600,
Neil Toronto wrote:
> Vincent, is there a quick way for me to test whether the types I give 
> the new functions are sound?

I recently added redex-based random testing for TR's float types. It's
in collects/tests/typed-racket/tr-random-testing.rkt.

Currently, it only tests flonum (i.e. double precision) operations and
types, but it should be easy to expand to single-flonums.


Posted on the dev mailing list.