[gentoo-soc] About Gentoo GSoC 2010 idea g-pypi

[Ling Kun]

Dear Jesus Rivero, Hi all,
     I am a student  from China and  interested in the idea of g-pypi
rewrite. Following is some of my general understand of this idea and some
questions on it. As a fresh student in GSoC , there must be some
misunderstandings about Gentoo's GSoC and  g-pypi. Please feel free to reply
this thread.

     Something of g-pypi:
     It is an excellent tool to create ebuild. The ebuild  will give much
help to whom that want to install package which is listed in Python Package
Index. When there is a ebuild, Python Package Index can cooperate with
portage.
     There is a setuptools which can work will package in Python Package
Index.But it is not convenient for package management, users should have the
package's specific setup.py to manage the package.However, g-pypi is much
like a portage in Python Package Index, and tried to handle these function
using ebuild, without users reverse the package's source.
    G-pypi was started at 2006,the last commit is Aug 22, 2008.and now it
can't work on my laptop(Python 2.6.4, pygtk-2.16.0-r1). That's why Mr. Jesus
Rivero listed it in Gentoo GSoC 2010.

     Is  the following plan  attractive to the Gentoo GSoC2010 organization?

    1,make it work again.Rewriting is prepared too.
    2, More precise dependence manage. g-pypi can only give the DEPEND
package with no version.
    3,Maybe a Python Package Index overlay can be create useing g-pypi.

    Any suggestion or ideas welcome.

    Best wishes to all .

   your,
  Ling Kun