[Grok-dev] Re: grok zmi
faassen at startifact.com
Mon Jan 7 20:40:10 EST 2008
Rene Bourgoin wrote:
> But if Grok does not find a way to add these powerful zope 3
> components easy for people like me then it will probably remain a
> framework for experienced Zope developers only.
Agreed. This has been, and still is, and will be, the main challenge of
the Grok project.
Grok has already made quite a lot of powerful Zope 3 components more
easy to use, from forms to indexes. Unfortunately all this is severely
underdocumented. And it's just scratching the surface.
What we have been doing, and should continue to do is this:
* try to do something with Grok.
* there's nothing obvious yet in Grok, so we dive into Zope 3.
* after possibly a long hard struggle, we figure out how to use the Zope
* we see a pattern in the usage, and see a way to use defaults to make
it more simple.
* we implement these.
* we *document* how to use this stuff (we need to get a lot better at this)
A UI project (such as a ZMI) could, eventually, help with discovery:
instead of having to read through web sites or, worse, source code,
you'll see the components available in a drop-down list. Creating such a
ZMI is a very significant challenge, however. It's not obvious how a
filesystem component translates into a UI. My mail was intended to start
up the thought process on this.
That doesn't take away at all from the pattern I described above. Quite
aside from the ZMI we should continue to try to make things easier.
More information about the Grok-dev