2009/6/28 William Stein <wst...@gmail.com>:
>
> 2009/6/28 John Cremona <john.crem...@gmail.com>:
>>
>> Interesting. My guess is that withe the new automated merging system,
>> it becomes easier to merge the wrong patches form a ticket by mistake.
>> We need a more failsafe system (e.g. the ability of people who have
>> uploaded patches to delete them -- can we do that?), and also it would
>> be preferable if the "merged" tag said exactly which patches had been
>> merged.
>
> Whether you chose from a menu (with the script) or by manually
> downloading and applying requires an equivalent decision based on
> identical information. In any case, I was in the same room as Robert
> Miller when he was merging patches, and I don't think he even used the
> automated script, since he considered the script too new or something.
>
> I've reopened http://trac.sagemath.org/sage_trac/ticket/5080, marked
> it a blocker for sage-4.1, pasted David Loeffler's comment from above
> in, and cc'd Tom and Robert.
>
> Regarding deleting patches from tickets, some users can and some
> can't. I'm sure it's possible to configure trac so that *anybody* can
> delete any patch from trac. Should we do that? Since we do have
> regular snapshots/backups of the whole trac server, I vote for yes,
> since if somebody deletes everything, we could recover.
That seems a bit extreme! I was only suggesting that the person who
uploads a patch should be allowed to delete it, but I can see that
that might be impossible (or not without a change to trac itself).
John
>
> -- William
>
>>
>> In the meantim I hope someone knows how to unmerge!
>>
>> John
>>
>> 2009/6/28 davidloeffler <dave.loeff...@gmail.com>:
>>>
>>> On Jun 27, 11:54 pm, davidloeffler <dave.loeff...@gmail.com> wrote:
>>>> On SuSE, 32-bit, sage -testall -long passes except for errors in the
>>>> same three files Jaap reported above (and a harmless timeout in
>>>> elliptic curves).
>>>
>>> I spoke too soon. Something rather harmful has in fact happened: the
>>> wrong patches have been merged for track #5080. My first attempt at
>>> fixing this problem caused a catastrophic slowdown in elliptic curve
>>> Sha routines, so I started again from scratch and did a new patch that
>>> worked differently. It seems that the old patch has been merged, with
>>> the result that
>>>
>>> sage: EllipticCurve("858k1").sha().an_padic(7)
>>>
>>> has been slowed down by *several orders of magnitude*. That was why I
>>> was seeing timeouts in that file.
>>>
>>> To reiterate: the patch "trac_5080.patch" on that ticket is evil, bad
>>> and wrong, should not have been merged, and must be removed from Sage
>>> ASAP.
>>>
>>> David
>>> >
>>>
>>
>> >
>>
>
>
>
> --
> William Stein
> Associate Professor of Mathematics
> University of Washington
> http://wstein.org
>
> >
>
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
