On Thu, Jul 26, 2012 at 11:08 AM, Satish Balay <[email protected]> wrote:

On Thu, 26 Jul 2012, Jed Brown wrote:

...

On Thu, Jul 26, 2012 at 10:41 AM, Satish Balay <[email protected]> wrote:

...

The naming/clasification is not perfect.

Looks like Barry tried to fix this - by adding a target tests_DATAFILESPATH - but now it has both C and fortran examples in it? [i.e TESTEXAMPLES_FORTRAN_NOCOMPLEX and TESTEXAMPLES_C_NOCOMPLEX]

http://petsc.cs.iit.edu/petsc/petsc-dev/rev/c37110e4b450

Perhaps this stuff needs an overhaul..

I also think it should be overhauled so that examples can be tagged when they are declared. I think that Python declarative syntax is easy, but it would be good to parse the existing makefiles to avoid needing a manual conversion.

Is everyone okay with the test suite depending on Python?

I think python is fine - as long as the run targets still remain in the makefile.

I am fine with having run targets, AS LONG as they are not shell. Make them Python. Make can run python just as well as shell/ Matt

Are you saying each 'run' target in the makefile will be tagged - and the python script will parse this tag?

Satish

-- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener