[Grok-dev] Re: First Experience with Grok

[Martijn Faassen]

Hey,

Shane Hathaway wrote:
[snip]
> Still, I feel like Zope 3 is a block of cheese that's moldy on the edges
> but otherwise good.  We need to trim away the parts of Zope 3 that
> didn't work out well (such as the Rotterdam skin) so that people who
> don't spend 8 hours per day on Zope 3 don't have to dig so much to find
> the good stuff.  I'm hoping that Grok makes it easier to figure out what
> belongs in Zope 3 and what doesn't.

Yes, this is indeed our goal. Grok should be a Zope 3 that just works 
for you out of the box, and contains whatever is useful and strips away 
stuff that doesn't work (like the ZMI). Of course opinions on what is 
useful differ, so to avoid having to worry about convincing other Zope 3 
users, we'll just call the result "Grok". :)

I sometimes call Grok an "integrated megaframework". TurboGears made 
headway by saying they're a megaframework, containing best of breed 
components that could evolve independently. Then they told the Django 
people that they were inventing everything themselves. The Django people 
parry by saying their framework is integrated, so you don't have to deal 
with a disparate mix of components but can deal with a unified 
experience with good documentation, etc.

Grok's a megaframework: it selects best of breed components, mostly from 
the vast pool of what's there for Zope 3, but also outside (hopefully 
more so once we get WSGI going). It also explicitly aims to offer an 
integrated experience, and since most of our components come from the 
Zope 3 pool, we are quite far along already.

Regards,

Martijn