dolfin team mailing list archive

[Andre Massing <massing@xxxxxxxxx>]

Den 19.01.11 16.04, skrev Johannes Ring:
> On Wed, Jan 19, 2011 at 3:57 PM, Andre Massing<massing@xxxxxxxxx>  wrote:
> > Den 19.01.11 15.46, skrev Johannes Ring:
> > > On Wed, Jan 19, 2011 at 3:44 PM, Andre Massing<massing@xxxxxxxxx>    wrote:
> > > > Den 19.01.11 15.44, skrev Johannes Ring:
> > > > > On Wed, Jan 19, 2011 at 3:38 PM, Andre Massing<massing@xxxxxxxxx>
> > > > >   wrote:
> > > > > > Hi!
> > > > > >
> > > > > > I just wanted to test the newest version of dolfin (since for some
> > > > > > reason
> > > > > > the assembly of functional has slowed down on my machine by a huge
> > > > > > factor!)
> > > > > > and realized that DOLFIN cmake (via dorsal) took my python27 instead of
> > > > > > python26 and therefore messes up my installation since ufl, ffc were
> > > > > > installed as python26 variants.
> > > > > > Any idea what happened? Were there recently changes in the cmake
> > > > > > concerning
> > > > > > python finding procedure?
> > > > > Try to give these option to CMake:
> > > > >
> > > > >    -DPYTHON_EXECUTABLE:FILEPATH=/opt/local/bin/python2.6
> > > > >    -DPYTHON_LIBRARY:FILEPATH=/opt/local/lib/libpython2.6.dylib
> > > > >
> > > > > Johannes
> > > > And via dorsal?
> > > Add them both to CONFOPTS.
> > Tusen takk :)! That works. Just for curiosity, why/when did that behaviour
> > changed?
> I guess when you updated MacPorts recently and Python 2.7 was
> installed? ;-)
Educated guess ;)? Yes, your are completely right. Interesting that not 
python26 is picked-up although it is the default, wondering how cmake 
decides which one to take if several python installation exists, which 
is quite usual.
> I had similar problems on the Mac buildbot after doing
> "port selfupdate" and "port upgrade outdated". I didn't ask for Python
> 2.7 to be installed but suddenly there it was. I struggled for several
> hours to get things running again.
Same here, it was caused by some gnome doc tools, which requires 
py27-libxml2.
Andre

> Johannes

--
André Massing       		email: massing@xxxxxxxxx
Ph.D. student			mobile: +47 469 57 401
Simula Research Laboratory
NO-1367 Fornebu, Norway