[Zope-dev] Defining Zope 3.

Helmut Merz helmutm at cy55.de
Mon Apr 20 06:23:02 EDT 2009


Am Montag, 20. April 2009 09:35 schrieb Martijn Faassen:
> Stephan Richter wrote:
> > ...
> > I never communicated to anyone that I believe that Zope 3 is
> > a successor of Zope 2. Other people pushed that message.
>
> That message has been out there from the start, no matter how
> it arose. One way this conclusion was reached was the obvious
> 3 versus 2. We need to fix that situation.

I was following the discussion from the beginning and was all the 
time thinking about the right time for me to step in, seems this 
now has come, especially as... last night... how to tell the 
story...

Hm, OK, let's try: Last night I had a strange contact with an 
extraterrestrian Zivilization that somehow, I think via a 5D 
transgalactic intertemporal loop (they called it a Z-loop), got 
access to the slides of the 2014 conference of the European Zope 
User Group (EZUG, or GUZE as it is called in France) in Pariz.

And, believe it or not, there was a keynote by Martijn Faassen 
about the last five years of Zope's history! 

I could only manage to get a quick glance at the slides (a lot of 
slides with only little text and without any pictures) before 
the communication line broke down. So I can only give you a 
short summary of what I remember:

Due to the good work of the Zope Toolkit Steering Group and many 
other members of the Zope community the Zope Toolkit flourished 
and was used for the development of a lot of Zope-based 
applications; also more and more of the Zope 2 code base had 
been refactored and ported to use the Zope Toolkit, always 
keeping backward compatibility, of course, and backed by 
informative deprecation messages.

This culminated in the the use of zope.security for Zope 2 and 
even the replacement of the Zope 2 publisher by that from 
Zope 3.

The Zope 3 ZMI was still alive at that time (due to the good work 
of Yusei Tahara and others); then (I think it was about 2012 or 
2013) somebody replaced the Rotterdam skin by something that 
looked exactly like the Plone 5.2 standard theme - whereupon the 
Plone people (with Plone 6) fully switched to Zope 3, thus 
getting rid of the last parts of Plone configuration that needed 
the ugly Zope 2 ZMI, simply by using the anyway superior Zope 3 
ZMI.

In parallel someone - I could not read his full name, I think I 
remember the first letter of his first name as "D", but it may 
also have been an "A" - extracted all Zope2-specific code 
(especially Acquisition but also a lot of other stuff) to a 
namespace package called "zepo" (an acronym for "Zope's 
Eternally Persisting Origins") so that things suddenly changed: 
Now Zope 3 was "the" Zope application server, while Zope 2 was 
still there for those who wanted or needed it, just by using the 
Zope Toolkit together with the zepo Packages (which, by the way, 
also contained ZClasses again). So Zope 2.39 was de facto 
replaced by Zepo 1.0 (which still was branded as "Zope 2"), and 
Zope 3..., hm, see below...

So finally came true what was never intended, thoroughly denied, 
but nevertheless for some unknown reason believed by almost 
everyone: That Zope 3 would be the natural successor of Zope 2.

The only sad aspect of this wonderful story were the 
ongoing "Zope Naming Flame Wars" that started to rage fiercely 
among the members of the Zope community: While some just wanted 
to keep the "Zope 3" brand (it was version 3.6 btw) others 
wanted to call it "Zope 4" or "Zope 5" (you know: 2 + 3 = 5), or 
even came up with new names like "Phoenix" (for obvious reasons) 
or "Ezop" (some strange acronym; I do not remember what it 
meant). 

Some even wanted to call it just "Zope" - a name that was 
considered extremely confusing by others.

So that's my story.

@Martijn: do you still have access to the PSU time machine? It 
would be great if you could verify this somehow. Or maybe you 
can even get clearer and more consistent information...

Helmut


More information about the Zope-Dev mailing list