Paste and WSGI 2.0 [WAS: Yet another comparison of Python Web Frameworks]

Michele Simionato michele.simionato at
Sun Oct 14 10:46:58 CEST 2007

On Oct 14, 2:52 am, Ian Bicking <i... at> wrote:
> That said, going without a framework (which at least in his article is
> what Michele seems to be comparing Pylons against) isn't always so
> bad.  I started writing an Atompub server in Pylons, but felt like I
> was spending too much time navigating around what the framework setup
> and not enough time just paying attention to what Atompub really is.
> So I ended up writing it (as FlatAtomPub), effectively without a
> framework (though developing it at the same time as WebOb, so I leaned
> on that quite a bit).  The development went quite well, and for a web
> service like Atompub that's probably what I'd recommend (at least to
> experienced developers -- you might find yourself left to drift
> otherwise without a clear idea of where to start).
> But for writing a traditional web application, I'd still use Pylons.
> The choices Pylons have made are with that in mind, and it doesn't at
> all exclude other forms of development in the process.  You could
> actually drop a FlatAtomPub instance right into a Pylons app, for
> instance.  Pylons is a small enough framework that it really is quite
> reasonable to pick and choose and use a very minimal style with it.

I think we do agree entirely, it is just that the application we have
mind is more a collection of web services than a traditional Web
Now, since you are here, there is an unrelated question that I want to
ask you, concerning the future of Paste with respect to WSGI 2.0.
I do realize that at this stage WSGI 2.0, is only a draft, still I
like to know:

1. if you think that WSGI 2.0 is good idea (I expect you will say
2. if you plan to support it in Paste and if yes when (I mean, in a
   in a year, in three years?)
3. if you already have thought of a migration plan and, in that case,
   what your strategy would likely be.

Thanks for sharing,

      Michele Simionato

More information about the Python-list mailing list