[Gsoc] [GSoC 2008] Zope 3 on Jython improvements

[Georgy Berdyshev]

Dear Zope community and Zope developers, I would like to receive some
feedback to the undertaking of Zope 3 improvements for Jython as a
Google Summer of Code project.

There are many benefits for running Zope on Jython, as the origin of
Jython is the Java programming language. In the past more and more
companies have been started to deploy solutions based on Python,
mainly using Zope. Many companies have Zope and also Java projects
deployed at the same time, but there was always a good connector
between those technologies missing. This lack should be fixed by
introducing Zope on Jython.

The Java platform offers a huge amount of libraries, which could be
used from Jython together with Zope and are not available for CPython.
Java's major position as industry standard leads to the fact, that
where Python is going to be used, Java applications have already been
deployed for years. An other fact is that Java VMs are available for a
lot of platforms and therefore there is a good portability from the
start.

For those environments, where Java products have already been
deployed, it will be much easier to promote Zope running on Jython, as
it will not need a big restructering or change of the major root
technology being used, Java.


In the last days Jython has been really active by introducing
improvements for Python 2.5 compability and also the port of several
frameworks to Jython.
It would be a great thing to start to work on porting Zope 3 to Jython.


This Summer of Code project has a lot of potential for Zope by
starting to sail to new land and solve problems within Zope, that
could help to introduce Zope to other Python implementations later on.

The new all-rounder character will spot more light on Zope and
introduce the possibility to interact with major technlogies and be a
worthier competitor on the market.

Zope consists of many components and there will be a need to make the
decision, which components will be important for the first steps of
the port to Jython.
I would really like to receive feedback on this from Zope developers
and the Zope community.


Several Zope components have C code included for speedups, which will
not run by default on Jython and will have to be rewritten in Java, to
make those speedups available for Zope on Jython.
At the moment there are some Zope C libraries with equivalent parts
written in Python and could be used as workaround.

After being able to run important parts of Zope, like zope.interface,
zope.component and maybe grok, there is the possibility to port the
parts written in C to Java and make some of the speedups available on
Jython


The Jython community is willing to assign a main mentor to this Summer
of Code project. But there is still need for a mentor from the Zope
community, willing to support Zope related issues, having a great
knowledge of Zope internals and its components.


Thanks in advance!

Kind regards, Georgy



-- 
Georgy Berdyshev

GPG key: 830F68C5
Fingerprint: 0379 ED5A BEE5 65A8 7BD5 31E7 F5B4 1EC7 830F 68C5