Re: [sigc] [Fwd: libsigc++ 1.9.13 fails on Solaris 9 sparc]

[Murray Cumming <murrayc murrayc com>]

On Fri, 2004-02-13 at 13:14, Martin Schulze wrote:
> Am 2004.02.11 20:13 schrieb(en) Murray Cumming:
> > Here is the first result of a Forte build that I know of.
> 
> Great! Do we finally have access to a Solaris machine?!

No, I asked on #sun in irc, and someone decided to help by testing the
build.

> Reading the log ...
> 
> > automake: sigc++/Makefile.am: not supported: source file `functors/
> > slot.cc' is
> > in subdirectory
> > automake: sigc++/Makefile.am: not supported: source file
> > `adaptors/lambda/lambda.cc' is in subdirectory
> > sigc++/Makefile.am:54: invalid variable
> > `nobase_library_include_HEADERS'
> 
> libsigc++2 requires automake 1.6 or later.

Strange. No tarball should require any automake when doing a configure.
Maybe he did not use a tarball. I'm CCing him.

> 
> > CC -g -i -o .libs/test_trackable test_trackable.o  -L/opt/gnome-2.4/ 
> > lib
> > ../sigc++/.libs/libsigc-2.0.so -R/opt/gnome-2.4/lib
> > Undefined			first referenced
> >  symbol  			    in file
> > sigc::slot_base&sigc::slot_base::operator=(const sigc::slot_base&)
> > test_trackable.o
> 
> This would be because of the automake problems: functors/slot.cc is
> not built and not linked into libsigc-2.0.so
> 
> Do we have a chance to use automake 1.6 on the solaris machine?
> Alternatively, we could test the latest tarball?!

I think that sf was down when he tried. Damien, could you please try
with the latest tarball from here? We really appreciate it.
http://sourceforge.net/project/showfiles.php?group_id=1970&package_id=76644

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