Re: make check

[marco atzeri]

On 7/31/2012 6:53 AM, Robinson, Melvin D wrote:
> Hello,
>
> Isn't this the same test?  It fails on one, but not the other
>
>   scripts/sparse/spstats.m ............................... PASS    1/1
>    scripts/sparse/spy.m ................................... PASS    1/1
>    scripts/sparse/svds.m .................................. PASS    5/6
>     FAIL 1
>    scripts/sparse/treelayout.m ............................ PASS    3/3
>    scripts/specfun/bessel.m ............................... PASS    1/1
>    scripts/specfun/beta.m ................................. PASS    6/6
>    scripts/specfun/betaln.m ............................... PASS    3/3
>    scripts/specfun/factor.m ............................... PASS    2/2
>    scripts/specfun/factorial.m ............................ PASS    5/5
>    scripts/specfun/isprime.m .............................. PASS    5/5
>    scripts/specfun/lcm.m .................................. PASS    3/3
>    scripts/specfun/legendre.m .............................^C
> make: *** [check] Interrupt

> octave:1> test svds
> PASSES 6 out of 6 tests
> octave:2>
>
>
> I must be missing something, right?

the test is base on a random starting point, so it is
not fully repeatable.
The chance of success depends on specific system / library

we had a discussion some time ago:

http://thread.gmane.org/gmane.comp.gnu.octave.maintainers/23255/focus=23304

Regards
Marco