Re: [sigc] failure building from CVS

[Murray Cumming <murrayc murrayc com>]

On Tue, 2006-04-25 at 08:30 -0500, Jonathon Jongsma wrote:
> On 4/25/06, Murray Cumming <murrayc murrayc com> wrote:
> > On Mon, 2006-04-24 at 18:22 -0500, Jonathon Jongsma wrote:
> > > On 4/24/06, Murray Cumming <murrayc murrayc com> wrote:
> > > > On Sun, 2006-04-23 at 21:17 -0500, Jonathon Jongsma wrote:
> > > > > I'm getting the following error (see below) when building libsigc++2
> > > > > from CVS with jhbuild.  I never used to get this error, and I haven't
> > > > > been able to figure out what it's trying to tell me about
> > > > > nested-name-specifiers.  I recently upgraded my computer to Ubuntu
> > > > > Dapper (pre-release) which has g++ version 4.0.3.  Breezy had 4.0.1,
> > > > > so it's possible that the new compiler is more strict about something,
> > > > > but I haven't been able to track down how to fix it.  Any ideas?
> > > >
> > > > Try a completely fresh checkout.
> > >
> > > I just deleted the libsigc++2 directory and re-ran and I get exactly
> > > the same error.
> >
> > That's very odd. I am also using Ubuntu Dapper (g++ version 4.0.3). I
> > thought that this had something to do with incorrect configure test
> > results that needed a new configure run, but I don't see any #ifdefs
> > near that line. Maybe I was thinking of a similar compilation error.
> >
> > Just in case, here is my sigc++config.h file, for comparison.
> 
> That looks identical to mine.

Is your Dapper completely up-to-date? Maybe it was something that was
fixed in the meantime. In case the g++ version numbers are different,
here's mine:

murrayc ubuntumurrayc:~/cvs/gnome216/branches/HEAD$ g++ --version
g++ (GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)
Copyright (C) 2006 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is
NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
PURPOSE.

> > I'm not sure what the error is trying to tell us.
> >
> 
> I'm not sure either.  I spent a couple hours trying to figure it out
> before I sent the original email, but I obviously wasn't successful. 
> I found a couple references to other projects fixing that error, and
> in each case, the fix was something like the following:
> 
> (inside a class or struct)
> typedef some_type X::other_type
> 
> changed to
> typedef some_type other_type
> 
> For example, here's one such fix:
> http://openscenegraph.org/viewcvs/include/osg/State?r1=1.72&r2=1.73
> But that doesn't seem to really fit with the error we're seeing (as
> far as I can tell).  But maybe it gives you some ideas...
> 
> I also just noticed another reference to this exact problem on the
> libsigc-list from december:
> http://mail.gnome.org/archives/libsigc-list/2005-December/msg00000.html
> That problem is also with g++ 4.03 on Debian, so perhaps it does have
> something to do with compiler version.

-- 
Murray Cumming
murrayc murrayc com
www.murrayc.com
www.openismus.com