Re: [sigc] libsigc++ & Forte: template operator()
- From: Murray Cumming <murrayc murrayc com>
- To: Martin Schulze <martin-ml hippogriff de>
- Cc: libsigc++ list <libsigc-list gnome org>
- Subject: Re: [sigc] libsigc++ & Forte: template operator()
- Date: Mon, 22 Mar 2004 00:53:44 +0100
On Sun, 2004-03-21 at 19:56, Martin Schulze wrote:
> Am 2004.03.20 17:38 schrieb(en) Murray Cumming:
> > On Sat, 2004-03-20 at 14:30, Martin Schulze wrote:
> > > > My SUN Forte C++ 5.3 and 5.4 still get stuck on
> > > >
> > > > CC -DHAVE_CONFIG_H -I.. -I.. -g -c signal.cc -KPIC -DPIC -o
> > > > .libs/signal.o
> > > > "../sigc++/functors/functor_trait.h", line 48: Error: Function
> > > > templates
> > > > may not have default template parameters.
> > >
> > > This is utter nonsense.
> >
> > Well, it does look like an unusual technique.
>
> We are using it in several places. It allows us to distinguish
> between classes that do or don't inherit from functor_base /
> adaptor_base / lambda_base transparently.
Hmm, I wonder if other uses succeed. Can you hack up a quick test case?
> I'm not sure how to hack
> around this compiler bug and whether it's worth the effort. I suggest
> that we first decide what to do about the ".template operator()<...>()"
> problem.
I don't think that problem exists anymore, because I added the
configure-time test.
I think you should do a tarball release so we can have it tested on SUN
Forte C++ 5.5.
--
Murray Cumming
www.murrayc.com
murrayc murrayc com
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]