Re: separate compilation

[Thomas Wicklund]

Jason R. McVetta writes:
 > I don't see why the compile time static type checking system should
 > prevent separate compilation, either.  A pre-compiled binary
 > representation of a class would only need provide the class's name and
 > it's type signatures; those things would alone satisfy type-checking.

Separate compilation was a victim of insufficient resources from what
I've gleaned in the many years I've followed Sather.

If I remember correctly there are also some technical reasons why it's
difficult, but couldn't say what they are.