[ODE] test_ode failures

[Jean de Largentaye]

On 8/9/06, Jason Perkins <starkos at gmail.com> wrote:
> On 8/9/06, Bram Stolk <bram at sara.nl> wrote:
> > Anyways.... to make more functions pass the tests, all you need to
> > do is raise the tolerance.
> >
> > We may want to make this change in svn, so that we stop
> > people from worrying about their builds.
>
> Does anyone actually use test_ode? Perhaps it should be removed from the build.
>
> Jason

It seems to me that test_ode is similar to the ODE internals (the
physics part, not the collision, I mean) in that it's pretty much a
black box to most people on the list.
It might be considered useless in that no-one uses it's output, but
that's because no-one is currently fiddling around the physics
internals.
I think that it's best to keep the test around, at least so that when
someone comes along who'll want to work on the physics internals,
they'll have a test harness, or at worse it'll serve as a hint that a
test of the physics part is necessary. Of course, since the current
tests fail on some points and are ignored, it kind of reduces that
last point :/

Jean