nasty side effect of a failed simon_two_descent search

[williamstein]

This could seriously confuse a person:

deep:sage-4.7.rc1 wstein$ ./sage
----------------------------------------------------------------------
| Sage Version 4.7.rc1, Release Date: 2011-04-29                     |
| Type notebook() for the GUI, and license() for information.        |
----------------------------------------------------------------------
sage: E = EllipticCurve([1, 1, 0, -23611790086, 1396491910863060])
sage: E.simon_two_descent()
(1, 2, [(88716 : -44358 : 1)])
sage: E.gens()
[]
sage: E.rank()
0
sage: E = EllipticCurve([1, 1, 0, -23611790086, 1396491910863060])
sage: E.gens()
[(4311692542083/48594841 : -13035144436525227/338754636611 : 1)]

The point that simon_two_descent finds is a torsion point. But for some reason E then thinks it has rank 0! Even though simon_two_descent's first output was 1, and in fact E has rank 1. This is pretty scary.

Related: #11005 (Update Simon's GP scripts and convert to use via gp2c)

Component: elliptic curves

Keywords: rd2

Author: William Stein

Reviewer: Jennifer Balakrishnan

Merged: sage-5.0.beta11

Issue created by migration from https://trac.sagemath.org/ticket/11372

[williamstein]

comment:2

This may be a dup, because Cremona says: "Yet another one of many related bugs being looked at and/or already fixed, see #10832 and #11005 and #11130 and ...."

[jbalakrishnan]

Stopgaps: #12695

[williamstein]

comment:4

Attachment: trac_11372.patch.gz

[williamstein]

Reviewer: Jennifer Balakrishnan

[jbalakrishnan]

Changed stopgaps from #12695 to none

[jbalakrishnan]

Changed keywords from none to rd2

[jbalakrishnan]

comment:6

Looks great, and all tests pass. 

Could someone else go to the stopgap ticket (#12695) and mark that it's no longer needed?

[jdemeyer]

Author: William Stein

[jdemeyer]

Merged: sage-5.0.beta11