On Fri, Oct 09, 2009 at 09:19:03AM +0200, Johannes Ring wrote: > On Fri, Oct 9, 2009 at 8:53 AM, Anders Logg <l...@simula.no> wrote: > > On Fri, Oct 09, 2009 at 08:49:35AM +0200, Johan Hake wrote: > >> On Friday 09 October 2009 08:44:43 Johannes Ring wrote: > >> > On Fri, Oct 9, 2009 at 8:41 AM, Anders Logg <l...@simula.no> wrote: > >> > > In addition to a couple of timeouts, the buildbot is currently failing > >> > > on this demo: > >> > > > >> > > demo/function/nonmatching-interpolation/cpp > >> > > > >> > > *** Error: DOLFIN has been compiled without GTS, intersection detection > >> > > not available. > >> > > > >> > > Does anyone remember how we used to handle this? Did we have an #ifdef > >> > > in the demo or how was it handled? > >> > > >> > Yes, #ifdef. See demo/mesh/intersection/cpp/main.cpp. > >> > >> But why is not the python demo complaining? We have some has_gts in the > >> mesh/intersection demo too, but none in function/nonmatching-interpolation, > > > > Yes, strange. Can someone who hasn't GTS check why? Otherwise, I can > > check it later but I'll need to rebuild without GTS. > > I ran the python demo on the mac-osx slave. I'm not sure what to look > for but this is the output I get: > > fenicssl...@amira-imac:python$ DOLFIN_NOPLOT=1 python demo.py > Ignoring plot. Unset DOLFIN_NOPLOT to enable plotting. > Ignoring plot. Unset DOLFIN_NOPLOT to enable plotting. > 0.47717658682 > 0.475242291949 > fenicssl...@amira-imac:python$
Strange that it works. > Actually, the mac-osx slave do have GTS installed but it was discarded > because of a too strict version requirement in dolfin/scons.cfg. I > have fixed this now. The winxp-mingw32 slave is the only one without > GTS now. It's good to have one slave without GTS for testing. -- Anders
signature.asc
Description: Digital signature
_______________________________________________ DOLFIN-dev mailing list DOLFIN-dev@fenics.org http://www.fenics.org/mailman/listinfo/dolfin-dev