[petsc-dev] Do we still need to support python 2.4?

[Satish Balay]

On Wed, 14 Oct 2015, Jeff Hammond wrote:

> On Tuesday, October 13, 2015, Satish Balay <balay at mcs.anl.gov> wrote:
> 
> > On Tue, 13 Oct 2015, Jeff Hammond wrote:
> >
> > > PETSc can build MUMPS. Why not Python? :-)
> >
> > and gcc :)
> >
> >
> I have GCC builds almost completely automated for similar reasons as Python
> 2.4 (eg my CentOS box is stuck with ancient GCC) so I don't actually think
> this is unreasonable. But the build time would be almost as long as ATLAS
> if PETSc did this ;-)

To be more specific - python has lot more dependencies as compared to
MUMPS [or even gcc] and difficult to get them built acorss various
architectures. [ esp when we have different machines with different
system libraries installed]

http://bda.ath.cx/blog/2009/04/08/installing-python-26-in-centos-5-or-rhel5/
For ex: on such a machine - its easier to do the following to get all
the dependent packages installed - then create a bootstrap build tool
that check for - and builds the missing packages from this
list. [perhaps there are equivalent simple routes for other OSes
aswell]

yum install gcc gdbm-devel readline-devel ncurses-devel zlib-devel \
  bzip2-devel sqlite-devel db4-devel openssl-devel tk-devel bluez-libs-devel

And what tool do we we use to build python [and its dependencies]?
Have a separate buildsystem in shell or python2.4 to build python2.6?

Satish