[racket-dev] [plt] Push #29023: master branch updated

From: Greg Hendershott (greghendershott at gmail.com)
Date: Tue Jul 15 19:48:54 EDT 2014

> I just don't think the additional line in that error message is very
> helpful, and it's already a long and scary error message.

Not that anyone asked for my opinion, but I agree.

Similarly, I wouldn't find it helpful if rackunit failure messages
added a caveat, "assuming the unit test isn't buggy".

I imagine most people already consider the provenance of the contract
or test, and prioritize. If it's in a standard or "respect-worthy"
library, _possibly_ it's buggy but more likely the other code is.
Whereas if I wrote the contract or test, it's as suspect as that which
it claims to validate.

Have many people not approached it that way, and it's caused problems?

Posted on the dev mailing list.