On Monday, 14 September 2015 13:35:15 UTC-7, Nathann Cohen wrote:
>
> > I don't quite see how this is a practical strategy in general. There
> might
> > be algorithms and procedures that can be implemented in a relatively
> > stand-alone fashion, but I expect that most non-trivial mathematical
On Monday, September 14, 2015, Volker Braun wrote:
> On Monday, September 14, 2015 at 7:56:58 PM UTC+2, Nils Bruin wrote:
>>
>> I don't quite see how this is a practical strategy in general.
>>
>
> There should of course be a central 'sage-core'. E.g. category setup,
> integers, ... But many peri
> I don't quite see how this is a practical strategy in general. There might
> be algorithms and procedures that can be implemented in a relatively
> stand-alone fashion, but I expect that most non-trivial mathematical
> functionality will work a lot better if it's built to make use of the
> infras
On Monday, September 14, 2015 at 7:56:58 PM UTC+2, Nils Bruin wrote:
>
> I don't quite see how this is a practical strategy in general.
>
There should of course be a central 'sage-core'. E.g. category setup,
integers, ... But many peripheral modules could easily be split off, e.g.
various interf
On 2015-09-14 20:20, William Stein wrote:
It is possible to create a body of code (e.g., the
Sage library) using standard packaging managers, which is properly
tested, has dependencies, and is very useful.
I think you are describing exactly how the Sage library works *today*.
--
You received t
On Mon, Sep 14, 2015 at 10:56 AM, Nils Bruin wrote:
> On Sunday, September 13, 2015 at 11:31:57 AM UTC-7, William wrote:
>>
>> We should break sage up into dozens of smaller libraries that are hosted
>> on pypi. I think pip is good enough these to support
>> install/uninstall/versions/dependenci
On Sunday, September 13, 2015 at 11:31:57 AM UTC-7, William wrote:
>
> We should break sage up into dozens of smaller libraries that are hosted
> on pypi. I think pip is good enough these to support
> install/uninstall/versions/dependencies/c++ code, etc. Then the code
> you're talking about
On Monday, September 14, 2015, Jeroen Demeyer
wrote:
> On 2015-09-14 16:44, William Stein wrote:
>
>> I guess I will have to lead the way on dragging sage into using the
>> standard Python package manager properly.
>>
> What do you mean with "using the standard Python package manager properly"?
On 2015-09-14 16:44, William Stein wrote:
I guess I will have to lead the way on dragging sage into using the
standard Python package manager properly.
What do you mean with "using the standard Python package manager properly"?
The Sage library is already using the standard Python tools. When
On Monday, September 14, 2015, Francesco Biscani
wrote:
> On 13 September 2015 at 20:31, William Stein > wrote:
>
>> I think pip is good enough these to support
>> install/uninstall/versions/dependencies/c++ code, etc.
>>
>
> I think the situation is still dire for c/c++ extensions. Compiled wh
On 13 September 2015 at 20:31, William Stein wrote:
> I think pip is good enough these to support
> install/uninstall/versions/dependencies/c++ code, etc.
>
I think the situation is still dire for c/c++ extensions. Compiled wheels
seem to be viable for platforms like OSX and Windows, but for so
On 2015-09-13 20:22, Nathann Cohen wrote:
What do you think of that ? Just wondering.
I have no objection in principle. We already have some .c/.cpp files in
the Sage library and it seems to work fine. I don't think it is worth
the effort though...
--
You received this message because you a
On Sunday, September 13, 2015, Nathann Cohen
wrote:
> Hello everybody,
>
> From time to time, I get these emails asking whether this or that
> feature of our graph theory library if available independently from
> Sage.
>
> I often answer 'no' or 'it isn't had to strip the python file of its
> Sag
13 matches
Mail list logo
