Stable Framework (was Re: [Zope-CMF] Zope component model and CMF)

Chris Withers chrisw@nipltd.com
Fri, 06 Jul 2001 14:25:39 +0100


(negative hat response ;-)

seb bacon wrote:

> Non-incremental changes can be so
> destabilising to 3rd party development efforts,

I'll say! I've got two major products I want to base on the CMF and by the
sounds of it, they'll both need to be completely re-written when this happens
:-(

Don't get me wrong, I really like the NR stuff and the stuff in Shane's
document, but it makes the CMF a hard choice to currently justify as a framework
to use, purely for stability reasons. Having come in this afternoon and read
that, I feel a bit like I've had the rug pulled out from under me :-S (but in as
good a way as possible, IYKWIM :-)

> For example, as I understand it part of the NewReligion is to find a
> solution to the ZOBD/filesystem dichotomy.  Would a move to
> NewReligionCMF be predicated on Zope moving to things like this
> *first*?

...and, what the overall timescales we should be expecting for these things?

Finally, will there be a 'pre-NR' branch of the CMF maintained so that if
products are implemented now, they won't _have_ to move until time can be found
to re-implement them?

How are DC managing this process WRT to installed (and customized) CMF sites
which customers are using?
Do you (DC) have any customized or alternate tool implementations? If so, how
hard to you see it being to migrate them?

cheers,

Chris