On 2013-02-05, Keshav Kini wrote:
> John Cremona writes:
>> 32-bit machines are getting rarer, but still exist -- e.g. my 2008
>> laptop! Is there a 32-bit testing platform we can access at sagemath,
>> skynet or similar?
>
> There is arando, the 32-bit Ubuntu 12.04 buildbot running at Nanyang
>
On Feb 5, 12:34 am, Robert Bradshaw
wrote:
> By default, such morphisms would be kept alive by their codomains,
> which are typically more complicated structures that often have direct
> references back to the domains as well (e.g. R[x] -> R). Embeddings
> (e.g. K -> CC) would be kept alive by th
On Feb 4, 10:57 pm, Simon King wrote:
> There is a base class of maps, and the first
> thing I'd try is to use the weakref there!
Some notes for this I found upon initial inspection:
- coercion discovery is littered with references to mor.domain() and
mor._domain, so these maps are really assume
John Cremona writes:
> 32-bit machines are getting rarer, but still exist -- e.g. my 2008
> laptop! Is there a 32-bit testing platform we can access at sagemath,
> skynet or similar?
There is arando, the 32-bit Ubuntu 12.04 buildbot running at Nanyang
Technological University and administered by
On Tuesday, February 5, 2013 11:14:32 PM UTC+1, Simon King wrote:
>
> Hi,
>
> On 2013-02-05, Jean-Pierre Flori > wrote:
> > It seems that somehow the Integer's created before the hooked functions
> are
> > in place do not respect the Python debug memory API, or the marker that
> > this API p
Kwankyu Lee writes:
> I solved the problem by moving out the ".sage" folder. Thanks for all
> your attention!
I was wondering what could be causing your problem but didn't comment
because I didn't have a solution. Could you elaborate on how you fixed
it? What was your current directory, where was
Hi,
On 2013-02-05, Jean-Pierre Flori wrote:
> It seems that somehow the Integer's created before the hooked functions are
> in place do not respect the Python debug memory API, or the marker that
> this API puts around mallocated memory get overwritten.
> As a consequence the deallocation funct
On Tuesday, February 5, 2013 8:59:13 PM UTC+1, Volker Braun wrote:
>
> This looks like http://trac.sagemath.org/13868 which was merged in
> sage-5.7.beta3
Nope, it was 5.6.beta3.
It seems that somehow the Integer's created before the hooked functions are
in place do not respect the Python deb
This looks like http://trac.sagemath.org/13868 which was merged in
sage-5.7.beta3
On Tuesday, February 5, 2013 8:05:16 AM UTC, Simon King wrote:
>
> Cython backtrace (newest frame = last)
> --
> #0 0x00400794 in main()
> #1 0x7f9a771218aa in
On Friday, February 1, 2013 1:03:55 PM UTC, jori.ma...@uta.fi wrote:
> Should Sage aim to efficient code whenever something is added, or should
> we just put in something that works and lock "user interface", i.e.
> command or function name, order of arguments etc?
>
I don't think we have to c
On Tuesday, February 5, 2013 9:26:40 AM UTC-5, John Cremona wrote:
>
> Thanks. To start with can you check that
>
> EllipticCurve([0, -1, 0,
> -360796668980,83414685883377400]).minimal_quadratic_twist()
>
> fails on these? Thanks,
>
> John
>
> PS ticket is #14060
>
>
>
On my ~2004 eMac I g
Thanks. To start with can you check that
EllipticCurve([0, -1, 0,
-360796668980,83414685883377400]).minimal_quadratic_twist()
fails on these? Thanks,
John
PS ticket is #14060
On 5 February 2013 14:19, Travis Scrimshaw wrote:
> For the record, my host machine is a 64-bit (Windows Vista), but
For the record, my host machine is a 64-bit (Windows Vista), but the
virtual machine (through VirtualBox) I do my development is a 32-bit. I
don't know of a way to have the VM have a 64-bit architecture...
Best,
Travis
On Tuesday, February 5, 2013 6:49:55 AM UTC-5, John Cremona wrote:
>
> 32-b
On Tuesday, February 5, 2013 12:49:55 PM UTC+1, John Cremona wrote:
>
> 32-bit machines are getting rarer, but still exist -- e.g. my 2008
> laptop! Is there a 32-bit testing platform we can access at sagemath,
> skynet or similar? I just found out that
>
There should be on skynet according
32-bit machines are getting rarer, but still exist -- e.g. my 2008
laptop! Is there a 32-bit testing platform we can access at sagemath,
skynet or similar? I just found out that
sage: EllipticCurve([0, -1, 0, -360796668980,
83414685883377400]).minimal_quadratic_twist()
fails on 32-bit since the
On 2/3/13 11:39 AM, Thierry wrote:
Hi,
On Fri, Feb 01, 2013 at 03:38:22PM -0600, Jason Grout wrote:
On 1/7/13 11:09 AM, sage-googlesu...@lma.metelu.net wrote:
here is a link to the self-replicating live USB we discussed about in the Bobo
2012 sage days thread.
http://sagedebianlive.metelu.net
Hi,
On 2013-02-05, Simon King wrote:
> I suspect we're running into this problem here because of some circular
> import issue that prevents it from being completely cleaned up the first
> time around.
> """
> (that's why #2435 introduces a Py_INCREF(ONE) to make the problem
> disappear)
>
> Perha
Hi Jean-Pierre,
On 2013-02-05, Jean-Pierre Flori wrote:
>> When I build sage-5.7.beta2 with SAGE_DEBUG then I get a crash, with the
>> following Cython backtrace, upon quitting Sage:
>>
> Did you also used the Python spkg from
> http://trac.sagemath.org/sage_trac/ticket/13867 ?
Yes.
Regards,
On Mon, Feb 4, 2013 at 2:24 PM, Nils Bruin wrote:
> On Feb 4, 1:53 pm, Simon King wrote:
>
>> But is a weak reference to the domain of a map really enough?
>
> I don't think so. Consider
>
> K=QQ(z)
>
> P=place(K,z-1)
>
> KP. = Completion(K,P,precision = 10)
>
> i.e., KP is the Laurent series rin
On Tuesday, February 5, 2013 9:05:16 AM UTC+1, Simon King wrote:
>
> Hi!
>
> When I build sage-5.7.beta2 with SAGE_DEBUG then I get a crash, with the
> following Cython backtrace, upon quitting Sage:
>
Did you also used the Python spkg from
http://trac.sagemath.org/sage_trac/ticket/13867 ?
>
Hi,
On 2013-02-05, Simon King wrote:
> 6048
> 6049# Free the object. This assumes that Py_TPFLAGS_HAVE_GC is not
> 6050# set. If it was set another free function would need to be
> 6051# called.
> 6052
>> 6053PyObject_FREE(o)
> 6054
> 6055
On Mon, Feb 4, 2013 at 11:30 PM, Nils Bruin wrote:
> On Feb 4, 10:57 pm, Robert Bradshaw
> wrote:
>> Currently the cdef _call_ assumes the argument is in the domain
>> (sometimes giving incorrect results or segfaulting if it's not, for
>> performance reasons), whereas __call__ coerces its argumen
This is a regression due to the ECL upgrade at #13324:
Execute
time print(len(maxima._commands(verbose=False)))
Before: about 2.5s
After: about 20s
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving
Hi!
When I build sage-5.7.beta2 with SAGE_DEBUG then I get a crash, with the
following Cython backtrace, upon quitting Sage:
Cython backtrace (newest frame = last)
--
#0 0x00400794 in main()
#1 0x7f9a771218aa in Py_Main()
#2 0x7f9a77105b2e in
24 matches
Mail list logo
