[Grace-core] comments from Philippe & UBC

[James Noble]

So Shahad(sic?) was right: Kim & I are emailing each other while sitting next to each other:

> It is from the outer scope.  I gather we still have not agreed as to whether or not we allow items to be redefined from the outer scope.  I assume that if we do then the inner definition would scope out the outer one.  I believe Andrew was against having a "final" annotation, but something like that might be handy for dialects.

sure. in any event you'd have to say "is override" or "is shadowing" or something
and if you take of the safety catch and point the gun at your feet...

> I assume that is part of a method definition.  If so, then it is because it is not evaluated until the method request.  By the way, we are dropping the "return"s.

which is what I said.
You can't always get rid of "returns" (unfortunately).
We could go to shift/reset, but I think that would easily undermine any claim we had to be a simple language.   (and no, calling them "Setjmp/longjmp" won't help :-)

> I never liked this because it is contrary to all our other naming rules.  Personally I'd make people write their own setter methods, as in many cases it is the wrong thing to have setters.  (I can live with the getters as they actually follow our naming rules.)

yes I know, but I still think that cure is worse than the disease: you have to access your variables somehow!

J