dolfin team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

On Thu, Sep 02, 2010 at 02:32:09PM +0200, Anders Logg wrote:
> On Thu, Sep 02, 2010 at 05:55:27PM +0530, Harish Narayanan wrote:
> > On 9/2/10 5:54 PM, Anders Logg wrote:
> > > On Thu, Sep 02, 2010 at 05:51:11PM +0530, Harish Narayanan wrote:
> > >> On 9/2/10 5:49 PM, Garth N. Wells wrote:
> > >>>
> > >>>
> > >>> On 02/09/10 13:17, Anders Logg wrote:
> > >>>> I'm running into problems on Karmic. Ig has CMake version 2.6 and we
> > >>>> require 2.8 when building DOLFIN.
> > >>>>
> > >>>> Is there a reason for using 2.8 or can we change to 2.6.
> > >>>
> > >>> Main reason not to go back to 2,6 is the extra testing.
> > >>
> > >> I asked about this earlier, and it wasn't clear to me why this specific
> > >> version of CMake was needed.
> > >>>
> > >>>> The other
> > >>>> option would be to drop Karmic support.
> > >>>>
> > >>>
> > >>> Dorsal can download and build CMake.
> > >>
> > >> Yes, this is an option too. Dorsal can build CMake.
> > >
> > > Is it already supported so we just need to turn it on in
> > > karmic.packge?
> >
> > Yes.
> >
> > > I see the cmake.package file now, will try that.
> >
> > Just make sure the package points to recent cmake since I haven't done
> > updates of that sort in a few months.
> >
> > Harish
>
> Looks like it's 2.8.2. That should be enough.

Karmic seems to be working now after I added

  Boost_USE_MULTITHREADED=ON

in karmic.platform and also modifying CMakeLists.txt in DOLFIN to look
for this variable.

This is probably not the best solution. I still don't understand why
we don't have a separate FindBoost.cmake where we can do a bit more
checking.

The problem now is that Karmic works only for the unstable build, not
the stable build. So I guess it should remain in contributed.

--
Anders