Re: singular tests

[Burcin Erocal]

Hi Francois,

On Sun, 19 Aug 2012 23:20:45 +1200
François Bissey <francoi...@canterbury.ac.nz> wrote:

> With Thomas we are looking at singular a bit moe carefully.
> It appears to us that the tests are completely broken.
> The list of source files for factory is full of files that are not
> there in CMakeList.txt.
> I note that this is factory again.
> Does anyone really care about the tests in singular?

Yes, we care about the tests. But the current test suite will not run
without errors on all machines. It doesn't support distinguishing
results between 32/64 bit architectures for instance.

If you are improving the Singular ebuilds, can I suggest that you take
a look at the upcoming version instead of the current releases? The
spielwiese branch on github has a new autotools based build system
which should be far better than the current stable tree. It should also
allow building omalloc and factory separately.

[1] https://github.com/Singular/Sources/

I haven't tried the tests on Spielwiese recently. Since it's in
development, there are some known failures. I will see if we can
specify a subset that passes to help include tests in the ebuild.
Extending this subset later would be trivial once the ebuild knows how
to drive the tests.


Thank you for your help.

Best,
Burcin

[Alexander Dreyer]

Hi Fran�ois
> thanks for the pointer. My own take is that in 3.1.5 tests are just broken
> and won't even build. Unless I follow the wrong recipe for testing.
Maybe: switching to CMake was never completed, autotools will be used
instead. So you refer to the directory 'tests'? It's not in use any more
and outdated. spielwiese takes a different direction.

The actual tests for recent Singular are the highlevel tests in Tst.

perl regress.cmd -s ../Singular/Singular Short/*lst Long/*lst

(and even more like VeryLong/*lst).

Best regards,
Alexander


--
Dr. rer. nat. Dipl.-Math. Alexander Dreyer

Abteilung "Systemanalyse, Prognose und Regelung"
Fraunhofer Institut f�r Techno- und Wirtschaftsmathematik (ITWM)
Fraunhofer-Platz 1
67663 Kaiserslautern

Telefon +49 (0) 631-31600-4318
Fax +49 (0) 631-31600-5318
E-Mail alexande...@itwm.fraunhofer.de
Internet http://www.itwm.fraunhofer.de/sys/dreyer.html