Another implicit typename bug

[Mark Mitchell]

>>>>> "Jason" == Jason Merrill <jason@cygnus.com> writes:

>>>>> Mark Mitchell <mark@markmitchell.com> writes:

    >> Here's a rather nasty implicit typename bug, which I checked in
    >> as g++.pt/typename13.C.

    Jason> I don't see this as "rather nasty"; code that changes a
    Jason> name from type to function between a primary template and a
    Jason> specialization should be taken out and shot, even if it is
    Jason> well-formed.

Sorry; I was unclear.  I meant that it would be rather nasty to fix;
I'm not sure quite what it would take to have that (well-formed, IMO)
code compile correctly in the presence of the implicit typename
extension.  Ideally (again, IMO), ANSI/ISO C++ would be a (proper)
subset of the language accepted by g++ (without any arguments).
That's not presently true for lots of reasons, of course.

-- 
Mark Mitchell 			mark@markmitchell.com
Mark Mitchell Consulting	http://www.markmitchell.com