Re: [ggt-devel] gmtl-config error

[Patrick H. <pa...@13...>]

On Wed, 2004-09-22 at 16:26 -0500, Allen Bierbaum wrote:
> I have fixed this bug in the build.  I was able to simplify the=20
> dependency tracking for gmtl-config by just using an SCons Value() node=20
> to track the value of all items that are substituted into gmtl-config.=20
> The long and short of what this means is that if anything changes that=20
> is used for the creation of gmtl-config, the file will be re-created.

Nice.  Thanks.

 -Patrick

> Patrick Hartling wrote:
> > On Mon, 2004-09-20 at 16:50 -0500, Daniel E. Shipton wrote:
> >=20
> >>I install gmtl by
> >>scons
> >>scons install prefix=3D$HOME/tmp/gmtl-0.4
> >=20
> >=20
> > I could be wrong, but I don't think that this is a valid way to install
> > GMTL as the SCons build is currently written.  The gmtl-config script i=
s
> > generated as part of the first call to scons, and since there is no
> > prefix set, it uses a default value.  When I install GMTL, I just go
> > straight for the install step.  AFAIK, SCons ensures that everything is
> > built before installing.  If you're not building PyGMTL, there is
> > nothing to build aside from gmtl-config.  Of course, PyGMTL doesn't get
> > installed by SCons anyway, so there is a different sort of problem
> > there.
> >=20
> >  -Patrick
> >=20
> >=20
> >>the generated gmtl-config script output=20
> >>/usr/local/include for the CXXFLAGS
> >>which is obviously wrong.
> >=20
> >=20
> >=20
>=20
>=20
--=20
Patrick L. Hartling                     | Research Assistant, VRAC
http://www.137.org/patrick/             | 2274 Howe Hall Room 2624
PGP: http://tinyurl.com/2oum9           | http://www.vrac.iastate.edu/