On Tue, 2010-06-15 at 01:58 -0700, Brad Roberts wrote: > On 6/15/2010 1:41 AM, dsource.org wrote: > > phobos commit, revision 1646 > > > > > > user: kyllingstad > > > > msg: > > Made std.string.icmp() a template function and fixed issue 3386, > > "to!bool(string) is not implemented" > > > > http://www.dsource.org/projects/phobos/changeset/1646 > > Should there be a template constraint on there to limit C1 and C2 to char > types? > _______________________________________________ > phobos mailing list > [email protected] > http://lists.puremagic.com/mailman/listinfo/phobos
I don't know, I was wondering about that too. I just copied cmp()'s signature, which also doesn't have any constraints. Since there is only one (i)cmp template, a template constraint isn't needed for disambiguation. So the only reason to add one is to improve documentation and error messages, both of which are arguable. -Lars _______________________________________________ phobos mailing list [email protected] http://lists.puremagic.com/mailman/listinfo/phobos
