ell_finite_field.py "long" doctest fails

[sagetrac-GeorgSWeber]

In the file "ell_finite_field.py" change the line 1013 from

        sage: for p in prime_range(10000):           #long time (~20s)

to

        sage: for p in prime_range(32768, 42768):           #long time (~20s)

to achieve the same intended amount of testing for the elliptic cirves code as such.
(But do not run into an --- as of this writing --- outstanding bug related to 16-Bit signed integers on Mac OS X 10.4.)

Component: doctest coverage

Issue created by migration from https://trac.sagemath.org/ticket/4179

[sagetrac-mabshoff]

comment:2

Georg,

are you planning to post an actual hg patch?

Cheers,

Michael

[JohnCremona]

comment:3

I'm not so keen on making this change, since this test was put in originally to show that a previous bug was fixed. It is natural to use a sequence of primes starting at 2, but not so natural to use a sequence like prime_range(32768, 42768).

Given that we are tracking the root problem anyway, can we not live with this one doctest failure which only occurs on one type of machine with the long option anyway?

[JohnCremona]

comment:4

PS this ticket also duplicates #3760.

[sagetrac-mabshoff]

comment:5

Replying to @JohnCremona:

PS this ticket also duplicates #3760.

Ok, I agree with John here and am closing this as a duplicate of #3760. I did comment on that ticket and mentioned this ticket, so the info should not get lost.

Cheers,

Michael