Re: [Darwin] Patch c++/15428

[Matt Austern <austern@apple.com>]

On May 24, 2004, at 12:03 PM, Jason Merrill wrote:

> The whole business about not setting the explicit bit is a kludge to 
> allow
> explicit instantiations of a class when not all of the members have 
> been
> defined.  It would probably be better to handle that directly in
> do_type_instantiation, and just always set the explicit bit.

I agree that that would  be cleaner.  It would also be consistent with
the direction that I think the standards committee is moving in.

I'll make sure that doesn't break the build (i.e. that nothing in 
libstdc++
or libjava depends on the old behavior), and, if it looks OK, I'll 
submit a
patch to make that change.

			--Matt