Re: locally changing global variables

[Judd Storrs]

I think I misunderstood slightly. Shouldn't local scoping of these
parameters be the default behavior? It seems like the way things are
breaks the concept of encapsulation pretty badly and I don't see
what's gained. It seems to me that getting these to scope and inherit
correctly would be the better solution and none of this syntax would
be needed.

What's the use case for not unwind_protect-ing these things?
Environment initialization functions? I've always somewhat wondered
why all of these settings weren't just part of some copy-on-write
struct/object that is passed around silently through function calls.


--judd