[Grok-dev] Re: A Django-like admin in Grok

Robert Gravina robert at gravina.com
Tue Apr 22 17:57:23 EDT 2008


On 22/04/2008, at 9:49 PM, Martijn Faassen wrote:

[snip]

> So, I'd say, the underlying pieces are in place, but the glue isn't  
> yet. Writing that glue isn't very hard right now for a Grok expert,  
> but we know that newcomers to Grok won't be experts for a while. We  
> don't have enough glue in place yet to tell you: sure, that's easy -  
> right now there are too many details to work out yourself in tying  
> up pieces together, I think.

Yes.. that's basically why I had to use Django instead of Grok. I have  
a project deadline in a few months, and it's hard to convince  
management I can learn Grok, write that functionality *and* get the  
site itself in that time :). And in fact, I'm not sure I can do it  
either.

>
>
> For a while now I've had some thoughts surrounding a project called  
> megrok.bread (bread: browse, read, edit, add, delete, we could also  
> go with megrok.crud) that offers some more high-level components to  
> do this. We could fairly easily supply a set of views that have  
> these functionalities. I could imagine something like supplying your  
> container with the IBread interface and getting a lot of views for  
> free.

That would be great, and pretty much cover most of my "CMS" needs (see  
below).

>
>
> One great bit of news concerning this is that Luciano's project for  
> Flint got accepted as one of the summer of code projects this year.  
> Flint is an attempt to build CMS functionality on top of Grok, and  
> hopefully as a spin-off we'll see bread-like components being  
> developed.
>

That's great news! I saw the Flint project description and thought  
that is exactly what I was looking for :). I'm glad to see it got  
accepted, because it seems to many CMS-related ideas never get off the  
ground because they are trying to do too much, to be another Plone,  
and that is one huge problem to solve in a reasonable amount of time.  
I think that something that allows developers to get CRUD forms and  
listing screens done easily (i.e. like Django) and also provides user  
management (maybe not even group management) - will allow them to get  
something done that satisfies most average content authors/web site  
administrators.

In my experience these people want to a) manage their content and b)  
manage their users, and *that's it*. They don't want to change their  
sites' layout, navigation structure, groups, workflow, or any of that.  
When they do, they are quite happy to ask their web developer to  
handle it.

What the client is most concerned about is that the site keeps working  
(software quality) and what I care about is that development times are  
reasonable and the code is not a jumbled mess (and I believe Grok and  
the Zope CA can really help with this).

I really think this is a tractable problem, and Grok would be a fine  
framework to build it in from what I can see.

I'm a little disappointed about using Django now because the community  
doesn't seem to have the same quality standards as the Zope 3  
community does, and when I talk to them about it they don't seem to  
understand why I feel their API is a mess. I'm also hooked on using  
interfaces and adapters in my programming (got exposure to them in  
Twisted) and having a hard time using them in Django - even if I can  
Django itself doesn't use them anyway.

The problem is though once I settle on a framework it's going to be  
hard (or at least wasteful) to change down the road since so many of  
our upcoming sites will be based on it.. I'm a programmer working for  
a small web design firm, where design is a huge part of the process -  
it can take months. We are then frustrated with the time it's taking  
to customise Plone to the very specific needs of designers and  
clients. Grok seems like it's at the right level - more than a bare- 
bones framework, but less than a full-blown CMS. However, without an  
admin section and user management (and the time and perhaps ability to  
write/test it from scratch) I can't go with Grok... so that leaves me  
with Django as about the only option.

Flint seems to be exactly what I am looking for though... perhaps I  
can start using it in three/six months time :)

Robert


More information about the Grok-dev mailing list