Because definitions are not expressions, I think they want to start with a keyword.
On Thu, Mar 5, 2009 at 12:56 PM, Geoffrey Irving <[email protected]> wrote: > On Thu, Mar 5, 2009 at 9:49 AM, Jonathan S. Shapiro <[email protected]> wrote: >> On Thu, Mar 5, 2009 at 12:40 PM, Geoffrey Irving <[email protected]> wrote: >>> What would be the syntax for appending a definition to the current scope? >> >> It's mainly a keyword distinction, I think. In BitC, for example, all >> type definitions append to current top level scope (restricted to top >> level). DEFINE appends to current scope, and LET introduces and >> encloses a nested scope. >> >> The difference is mainly one of keyword selection (perhaps "def" vs. >> "let"). I'm hopefully that we can normalize the two forms a bit as we >> re-syntax. > > Actually, do we need a keyword for adding a variable to the current > scope at all? I.e., what about > > x = 3; > > The obvious problem with this is that it overlaps with assignment, > which might make it a non-starter. That said, if I had to make I > choice I'd prefer to make assignment more verbose (and more visually > obvious) than creating a new variable. > > Geoffrey > _______________________________________________ > bitc-dev mailing list > [email protected] > http://www.coyotos.org/mailman/listinfo/bitc-dev > > _______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
