[Grok-dev] Re: newbie question: megrok.five status

[Philipp von Weitershausen]

Lennart Regebro wrote:
> On Nov 12, 2007 8:24 PM, Martin Aspeli <optilude at gmx.net> wrote:
>> Because Zope 2 doesn't look exactly like Zope 3? :)
> 
> Well, no, but there are many similarities, and because Grok provides a
> level of abstraction above it would be possible to get much closer to
> the elusive goal of code that run on both.

But what good would that do? Code that's written with Grok is always new 
code. So you might just as well do that new development using the 
regular Grok on Zope 3.

>> The most common use case here is "Grok-with-Plone" or maybe
>> "Grok-with-CMF". I don't see many compelling reasons to use
>> "Grok-with-plain-Zope2", since "Grok-with-Zope3" already fulfils pretty
>> much all of what you could get there anyway.
> 
> This is definitely true. We need Zope2 today pretty much because we
> need Plone. But with Grok-on-Zope2 be very different from
> Grok-on-Plone?

Well, to be honest, I have no idea what Grok-on-Zope2 should look like. 
Functionality-wise, it can't be much different from our regular Grok. So 
Why have it at all then?

> Then again, I'm not sure Grok as a "compatibility layer" is the way
> forward. But I think it may be.

I think Five is the compatibility layer.


I think we should from now on be clear in our terminology. For me, 
"Grok" is *the* web application framework that's distributed in the 
'grok' package. It runs on Zope 3. There's only one Grok. Let's keep it 
that way.

Now, I'm all for reusing Grok's underlying mechanisms and philosophies 
elsewhere (e.g. in Plone), but whatever it is, it should NOT be called 
Grok. We'll just end up confusing ourselves, just like we did with Zope 
2 vs. Zope 3. I bet we would also run into the same marketing problem 
(e.g. if the Plone thing sucked but Grok actually ruled, people would 
condemn Grok because of their experience with the Plone Grok-lookalike.)