On 30/04/2013 11:00 AM, Max Cantor wrote:
I just realized that what I proposed earlier is pretty much exactly what
predicate functions were in the days of typestate (sorry, I'm a bit new
here). I completely understand and agree with the decision to remove
typestate which does seem quite complex, however, predicate functions
have a lot of value beyond typestate.
So, my questions are, did Predicate Functions themselves introduce a
high degree of complexity? If not, would the community be open to
bringing them back? If so, I'd be willing to put some work into doing
that.
Yes, there was a lot of complexity. See the other message I just posted
on the matter. Short version: figuring out where to split the space of
functions is hard, any split will be disagreeable, it has a lot of
complexity-cost, and will cause library bifurcation anyways, and
probably not achieve the many benefits you're imagining.
-Graydon
_______________________________________________
Rust-dev mailing list
[email protected]
https://mail.mozilla.org/listinfo/rust-dev
