On Thu, 2016-03-24 at 11:12 -0700, Van Snyder wrote:
> On Thu, 2016-03-24 at 08:20 -0600, Dan Nagle wrote:
> > Hi,
> >
> > > On Mar 24, 2016, at 07:04 , Bill Long <[log in to unmask]> wrote:
> >
> > > And, it is perhaps mentioning that any change that affects the typing system (apart from trivial syntax enhancements like type(integer)) can involve non-trivial implementation issues.
> >
> > So if someone wants to propose enums, first tell us what
> > you want the enums to do. If we can find a good set of use-cases,
> > then we’ll ask the standards committee’s Data subgroup to propose
> > the specifications.
> >
> > Are logical (i/o) units to become enums?
> > How ‘bout kind parameters per type?
> > Array subscripts? Character indexes?
> >
> > What else ya got? ;-)
>
> Exception identifiers. But then, we've backed away from block
> structured exception handlers repeatedly.
>
> Array bounds and subscripts using ordered enumeration types at first
> blush are appealing, because bounds checking is done at compile time.
> But what happens if COLORS(RED:BLUE) is an actual argument? Are the
> bounds of the actual argument still 1:42?
----------------^^^^^^
OOPS! DUMMY argument!
>
> How about Vipul's example? Isn't that a use case?
>
> > --
> >
> > Cheers!
> > Dan Nagle
|