How about you put some money where your mouth is and e.g. work with ATLAS
to provide them with a better build system. If you don't then you'll have
to agree with me that it would be a great goal but its just is not going to
build you an ATLAS library anytime soon. I've helped a number of upstrea
Le 08/01/2014 14:23, Jean-Pierre Flori a écrit :
> I meant that I'm not sure using python for install scripts will make
> math software more accessible... especially if scripts become more complex.
I get your point, and agree with it.
Snark on #sagemath
--
You received this message because you
On Wednesday, January 8, 2014 2:20:40 PM UTC+1, Snark wrote:
>
> Le 08/01/2014 11:01, Jean-Pierre Flori a �crit :
> > On Wednesday, January 8, 2014 10:52:38 AM UTC+1, Snark wrote:
> >
> > Le 08/01/2014 09:10, Volker Braun a �crit :
> > > But many mathematical packages do need hand
Le 08/01/2014 11:01, Jean-Pierre Flori a écrit :
> On Wednesday, January 8, 2014 10:52:38 AM UTC+1, Snark wrote:
>
> Le 08/01/2014 09:10, Volker Braun a �crit :
> > But many mathematical packages do need hand-holding simply because
> > mathematicians generally don't know how to use e
On Wednesday, January 8, 2014 10:52:38 AM UTC+1, Snark wrote:
>
> Le 08/01/2014 09:10, Volker Braun a �crit :
> > But many mathematical packages do need hand-holding simply because
> > mathematicians generally don't know how to use e.g. autotools. At the
> > end of the day one of the goals o
Le 08/01/2014 09:10, Volker Braun a écrit :
> But many mathematical packages do need hand-holding simply because
> mathematicians generally don't know how to use e.g. autotools. At the
> end of the day one of the goals of Sage is to make mathematical software
> more accessible...
I can't help to n
We also talked about build system changes at SD65 (though this is not what
this thread is about), where we precisely brought up the point that the
build system should be smart enough to not require any build script if the
standard configure/make/make install or setup.py works.
But many mathema
Im not really for this.
Do we actually need horribly compilcated spkg-install scripts?
Most of the time its just patch/configure/make/make install or
patch/build/install.
If we really need something more involved it's a good point that upstream
is kind of broken (I'm actually aware of one python
Le 07/01/2014 10:27, Volker Braun a écrit :
> Slightly off topic, but if you have a particular ticket in mind then
> let's hear
You're right ; I think I confused this thread with another... wasn't
there one about someone complaining he couldn't run some python package
he had installed system-wide
Slightly off topic, but if you have a particular ticket in mind then let's
hear
On Monday, January 6, 2014 11:17:25 PM UTC-10, Snark wrote:
>
> Le 07/01/2014 09:10, Volker Braun a �crit :
> > We were considering the addition of Python as a build-time dependency to
> > be able to use Python
Le 07/01/2014 09:10, Volker Braun a écrit :
> We were considering the addition of Python as a build-time dependency to
> be able to use Python even for scripts that are used at bootstrap time
> here at SD56. And everybody was in favor. We should probably support
> Python 2.5 - 2.7 and 3.3+, which w
+1
On Tue, Jan 7, 2014 at 12:39 AM, Volker Braun wrote:
> To be able to use Python even for scripts that are used at bootstrap time.
> Its a much better language than shell script, better aligned with our core
> strengths, and comes with a decent standard library.
>
>
> On Monday, January 6, 2014
To be able to use Python even for scripts that are used at bootstrap time.
Its a much better language than shell script, better aligned with our core
strengths, and comes with a decent standard library.
On Monday, January 6, 2014 10:31:27 PM UTC-10, Jeroen Demeyer wrote:
>
> What's the motivati
On 2014-01-07 09:10, Volker Braun wrote:
We were considering the addition of Python as a build-time dependency to
be able to use Python even for scripts that are used at bootstrap time
here at SD56. And everybody was in favor. We should probably support
Python 2.5 - 2.7 and 3.3+, which would cove
14 matches
Mail list logo
