On Dec 18, 2009, at 8:34 AM, Ken Dyck wrote:
> On Friday, December 18, 2009 11:12 AM, Daniel Dunbar wrote:
>>
>> I haven't really been following this work, but is there a
>> strong motivation for implementing
>> is{Zero,One,Positive,Negative} instead of just implementing
>> comparison against an integer type?
>
> Mainly, it protects against accidentally comparing quantities in
> character units with those in bit units. But it also leads to more
> succinct tests than, for example, "size == CharUnits::fromRaw(0)".
>
> I'd be fine with removing those predicates, if there's a general feeling
> that they are superfluous. But I think to add a comparison against a raw
> integer would weaken the main purpose of the type.
Agreed.
Apologies for not reviewing this; I just missed it in the backlog. It looks
fine to me.
John.
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
