Python 3 vs Python 2.7 dilemma

[Eric Snow]

On Mon, May 16, 2011 at 10:48 PM, Navkirat Singh <n4vpython at gmail.com>wrote:

> Hi Guys,
>
> I have been trying to fight this issue for sometime now. I know that a
> large part of the python 3rd party software base has not been ported to
> python 3 yet. I am trying to build a web-based enterprise solution for my
> client. Most of reputed frameworks like Django and Turbo gears are yet in
> the 2.x stage. I know that these frameworks are extremely good. But I wanted
> to build my base with python 3 as that is what is going to prevail in the
> future.
>
> I have built my own little architecture using python3. Here is what I have
> accomplished till now:
>
> a) A multiprocessing webserver built directly using low level sockets for
> maximum control, conforming to RFC 2616 (not completely right now).
> b) A HTTP message parser  for parsing HTTP/1.1 requests and generating
> response messages
> c) A session control base using python multiprocessing dictionary manager
> d) A partially build MVC model, without a templating engine at the moment.
> I am planning to put Jinja 3 there.
>
> I have spent months of free time doing this. I have learnt a lot, but well
> I am not sure if the path I am on is the right one.
>
> My question to everyone is whether I should go ahead with this approach, or
> should I just use 2.x technology? I am not sure if I will be able to port
> all the code to python3 later.
>
> I will really appreciate any input.
>
> Thanks and regards,
> Navkirat
>
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
I would say go Python 3, unless you have a dependency that prevents it.
 There was a lot of talk at pycon this year about how the rate of Python 3
transition is drastically increasing.   Many of the large projects (as well
as pypy, ironpython, and jython) talked about their transition plans for
Python 3 in the next year or two.

The expectation is that this year will see a switch across many projects.
 It is a snowball effect, as more dependencies transition the remaining ones
have less reason to stay on Python 2.  The anticipation was to see everyone
on Python 3 by 5 years after its release.  It was released  just over 2.5
years ago.

Here are some references that you might find helpful:

http://wiki.python.org/moin/Python2orPython3
http://py3ksupport.appspot.com/
http://dev.pocoo.org/~gbrandl/py3pkgs.png

Cheers,

-eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20110517/d94f8876/attachment-0001.html>