Ola Skavhaug wrote: > On Thu, Oct 22, 2009 at 10:35 PM, Anders Logg <l...@simula.no> wrote: >> On Thu, Oct 22, 2009 at 09:47:39PM +0200, Ola Skavhaug wrote: >>> I see that it is possible to write to the options database in SLEPc >>> and PETSc by sending a couple of strings to a function. Should we add >>> a loophole in DOLFIN to get access to this database? The issue is >>> particularly relevant from Python. >>> >>> Cheers, >> How about subclassing Parameters to get a new class PETScParameters >> and then let that class call the appropriate PETSc function whenever a >> parameter is updated. Then we can add this as a nested parameter set >> for the global DOLFIN parameters: >> >> parameters["petsc"]["foo"] = bar: > > That's probably the way to go with it. > > Someone (tm) should with it down on launchpad so I can do it in a > couple of weeks, when I have time. >
Done https://blueprints.launchpad.net/dolfin/+spec/petsc-slepc-parameters Garth > Ola > >> -- >> Anders >> >> -----BEGIN PGP SIGNATURE----- >> Version: GnuPG v1.4.9 (GNU/Linux) >> >> iEYEARECAAYFAkrgwfcACgkQTuwUCDsYZdF1swCfResOXzC96Hu59CRyKQxp7X67 >> tfwAmwarpOhS/GRqTz0cE1RLzLCpnHoe >> =WN8l >> -----END PGP SIGNATURE----- >> >> _______________________________________________ >> DOLFIN-dev mailing list >> DOLFIN-dev@fenics.org >> http://www.fenics.org/mailman/listinfo/dolfin-dev >> >> > > > -- Dr Garth N Wells Department of Engineering University of Cambridge Trumpington Street Cambridge CB2 1PZ United Kingdom tel. +44 1223 3 32743 e-mail gn...@cam.ac.uk http://www.eng.cam.ac.uk/~gnw20/ _______________________________________________ DOLFIN-dev mailing list DOLFIN-dev@fenics.org http://www.fenics.org/mailman/listinfo/dolfin-dev