[mod_python] Re: Future of mod_python.

2009/3/30 Graham Dumpleton <graham.dumpleton at gmail.com>:
> I will not be writing a book on mod_python. IMHO mod_python is dying
> and the quicker people stop using it and shift to WSGI based Python
> web applications the better.

Let me add a few thoughts to this.

Yes mod_python provides a simple option in the form of publisher and
being in one package means a low barrier to getting something going,
unlike other options where various dependencies need to be installed.
Even if you were to write a publisher equivalent on top of something
like webob or Werkzeug, you are still going to face the problem of
having to install the dependencies it requires as well as a means to
host it. This is one of the things that many people dislike about
other frameworks or toolkits, although Django has a reasonable balance
and tries to be self contained as possible. So I can understand where
people are coming from in preferring mod_python in this respect.

If you stay with mod_python though, there are huge uncertainties as to
whether it can be easily or sensible be ported to Python 3.0 without
making porting applications to it from Python 2.X quite a lot of work
and possibly not worth the effort. So, use mod_python and you may have
to accept that you would never be able to use Python 3.0. This is one
aspect of why it might be said that mod_python is dying.

The bigger problem though is that there are no current active
developers working on mod_python. Me answering questions does not
constitute an active developer. Already the latest tarballs for
mod_python will not compile on the most up to date Apache and APR
libraries, nor build/run properly on MacOS X Leopard with Apple's
version of Apache. The changes are in the subversion repository, but
don't even have enough active developers to satisfy Apache rules
regarding voting on releases.

So, mod_python may still through its ownership of the name that ties
Python to Apache appear to be a logical choice for most people and so
may for some things still be seeing increasing use, the developer
community side of things has basically gone.

As I have said before, I may well be in part to blame for no interest
on the development side. That is, because I answer questions on the
user list still, people see things as all going fine and think that
there are people looking after the development side when there isn't.
Thus I raise my conjecture that maybe what might be best for
mod_python is for me to unsubscribe from this list. That way it will
give others the chance to step up and start helping and grow
sufficient knowledge themselves that they could take over development
of mod_python if they were interested enough. In unsubscribing from
user list, I would still stay on the developer list, but only so that
at some point a transition could occur to anyone who steps up and says
they are interested in taking over.

Graham