On Fri, Jun 4, 2010 at 3:03 AM, Graham Higgins <[email protected]> wrote:
> I would recommend that all developers who are currently using AuthKit (or
> who plan to do so) should read and inwardly digest the statements on this
> page:
>
> <http://jimmyg.org/work/code/index.html>
>
> in which James lists those of his Python modules that are currently
> maintained and those which he is personally moving away from:
>
> 1. "Currently maintained Python modules:" (the list does not include
> authkit)
>
> 2. "Phasing out of my own projects:
>   * Pylons - I'm using Flows ... instead.
>   * FormBuild 1 and 2 - use FormBuild 3 instead.
>   * AuthKit - I'm using [... other packages ... ] instead."

This is consistent with what James has been telling me over the past
year. I guess we should take AuthKit off the recommendation list and
just say it's an ex-solution. Repoze.who/what does essentially the
same but in a more modular way, and is the only one that has gained
acceptance by several frameworks.

On Fri, Jun 4, 2010 at 4:09 AM, daniel <[email protected]> wrote:
> After a look at <http://jimmyg.org/work/code/index.html> , where
> Pylons appears in the list of phased out projects (by the co-founder,
> although it seems there is still a community maintaining the project)
> and where I can read the following :
> «  Phasing out of my own projects: Pylons - I'm using Flows (which I
> absolutely love) instead. It corrects all the problems with WSGI and
> Pylons which I discovered when writing the Pylons Book and is a
> complete re-write of the whole codebase and core dependencies from
> scratch, with complete conceptual integrity and with a far more
> modular architecture than even Pylons »,

James co-founded Pylons with Ben in 2005 because it met their needs at
the time.  They (or at least Ben) were extending Myghty (a template
engine) with frameworkish features, and that clearly was not a
long-term solution. Paste was new at the time, and looking for a
full-stack framework to go on top of it, and Pylons became it. But
writing a book makes you acutely aware of the pros and cons of a
framework because you have to explain it in full detail (and justify
the bad parts :).  That's why I stared but eventually abandoned a book
on using Pylons with App Engine. App Engine's early flaws just made it
really hard, and by the end I felt like, would I really use App Engine
myself, and if I did, wouldn't I be better off with one of the small
WSGI frameworks that was designed for App Engine?

After the Pylons book was published, James began working on something
that wasn't even a recognizable web framework, much less WSGI
compliant.  I didn't understand it so I can't explain it. If Flows is
it, it may be something too different to just drop in place as a WSGI
framework.

> my questions are :
>
> 1. Is still wise starting new long term projects based on Pylons ?
> 2. How strong and motivated remains the Pylons development community ?
> 3. Which are, if any, the long term python web frameworks to
> eventually consider as alternative to pylons ? (should I re-evaluate
> django , which I don't like technically so much but it seems can count
> on larger stable development community)

Ben and I are using Pylons in production, and I've been spreading it
among the programmers at my organization. We'll be maintaining and
upgrading those applications for the next few years at least.  Phillip
Jenvey has done a lot of work this past year to get Pylons to run on
Jython. Ian Bicking has never used Pylons much himself but remains an
indirect supporter, maintaining half of its dependencies. And the
TurboGears folks tell us what they need, or when something breaks.

I know Ben and myself have been looking beyond Pylons to see what
might be better, but there's nothing concrete at this point, just some
ideas. I originally chose Pylons because it's modular to the core. I
got sick of rewriting applications when they outgrew their framework.
Marco promises to extend that modularization into the Pylons core,
although it sounds like Ben has something non-Marcoish in mind too.
But the zen of Pylons is, "How much can you change before it's not
Pylons anymore?" It's not PasteDeploy or Routes or Beaker or Mako
because those are all optional. To me it's PylonsApp. If you replace
PylonsApp or subclass it so much it's unrecognizable, then it would be
hard to call the thing Pylons. But now we're looking at modularizing
PylonsApp itself. So if I someday wrote my own SOPless dispatcher and
stricter Routes and non-INI front end, would that still be Pylons? It
would interoperate with most of Pylons' components and would adhere to
the spirit of Pylons, so to me it would be, even if had a different
name.

All the WSGI developers have been sprinting together for the past two
or three years. So to us it's one family of frameworks, not
disconnected islands. The islands are Django, Twisted, Plone, and
Zope.  (Zope the entire product, not Repoze or BFG.)


On Fri, Jun 4, 2010 at 5:47 AM, Graham Higgins <[email protected]> wrote:
> One particular characteristic of this community which I personally
> appreciate (but which may act to mislead people at times) is a distinct and
> deliberate absence of evangelism. IMO, that attests to a very level-headed
> engineering perspective and a solid understanding of where Pylons is located
> in the landscape of Python-based web app development frameworks.

It's a personality thing probably. The Pylons developers like to
program more than they like to market. We've also suppressed marketing
so as not to distract from finishing 1.0. Now that that's done, you'll
see development slow down, fewer changes in Pylons, and more emphasis
on marketing.

But I think we've agreed that Django has overmarketed, overpromised,
and gets extremely insistent at times, and we don't want to do that. I
don't know how much of it is due to the core Django developers
(because they're nice to us :), or how much is fanboys and marketroids
on the fringes of the developer base/userbase. But the Pylons tendency
is more to keep our heads down and write code, and let people come to
us. Because that's how we all came to Pylons ourselves.

We know we need somebody with good marketing experience, but we know
we're not it. So we're just waiting for somebody to say, "I'm that
[marketing] person."

-- 
Mike Orr <[email protected]>

-- 
You received this message because you are subscribed to the Google Groups 
"pylons-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/pylons-discuss?hl=en.

Reply via email to