[Zope] RIP Zope 2!

Lennart Regebro regebro at gmail.com
Fri Apr 10 10:04:30 EDT 2009

On Fri, Apr 10, 2009 at 08:55, Dieter Maurer <dieter at handshake.de> wrote:
> With my current experience, the Zope 3 way to handle skins is considerably more
> work then the previous CMF skinning. True, you have a separate namespace
> for views and thus avoid name clashes in edge cases -- but is this
> really worth the extra effort?

Yes, without any doubt whatsoever.

And when it comes to effort, you have a point. Zope 3 in itself is too
fragmented, too low level and too XML-y. Grok solves that. Zope 3 was
also too big and monolithic. The eggification process solved that (and
made Zope3 pointless as an application server, and it became a
toolkit/framework). And some central parts of Zope 3, in particular
the publisher, are too complex. Repoze and Repoze.bfg solved that.

That means that for most cases, except when you need Zope 2
compatibility, The Thing That Once Was Known as Zope 3 are now finally
ready. Obviously there is not much point in porting projects, but if
you start a new project, the extra work of learning Grok or repoze.BFG
could very well be worth the effort.

I love Zope 2 as well, although I forget sometimes, since I never work
with Zope 2, I work with Plone. Which I don't love (but Plone 4 looks
like I will love it again). But with the Zope Toolkit I can do
everything I want to do with Zope 2, with less code and less magic
handwaving, and less (un)expected problems.

We who know Zope 2 can develop in it easily and without problems. But
it WAS a pain to get to that point. Zope 3 had a completely different
set of pains. IMO, Grok has a much lesser pain level.

Yes, Zope 3 did kill off a lot of interest in Zope 2, and was a
contributing factor to the fact that Zope 2 doesn't attract new
developers.  But it wasn't the only one. It was already losing
mindshare because it was too painful to use, and Python people didn't
like it. People went from Zope to Python, not the other way around.

With a time machine, much could have been done differently. But it's
too late now. Time has ran away from Zope 2, Zope 3 never took off.
It's time to take the experiences and the vast codebase, and move
forward. And I guess Zope Toolkit, Grok and BFG is that way forward.

Lennart Regebro: Python, Zope, Plone, Grok
+33 661 58 14 64

More information about the Zope mailing list