Re: FYI: Lessons from Kai Wu at ArsDigita

> For instance, once the Zope 3 team can migrate zope.org to Zope 3 in a 
> test environment, then we can be convinced that migration is possible.

As I've pointed out before, porting old stuff is really not the issue. 
People tend to not touch working systems. Why should they?

So I'd really focus on making Zope 3 a great product, not on backwards 
compatibility. On of the lessons we can learn from aD is that we should 
not heavily invest in porting stuff.

I don't think the Zope 3X API is stable enough right now to actually be 
able to see all the problems with migration. For me it is important that 
the content/data can be migrated, and that should be no real problem. 
The applications will either be kept in Zope 2 or ported in the process 
of a new feature release.

The main issue is to quickly enable people to write code that will 
basically run on both platforms (new code that is). Time I invested in 
old code is not lost: I can still run it on Zope 2. But time I invest in 
new code for Zope 2 that can not be ported easily will be really lost ...

Maybe we should not call the new beast "Zope" at all ;-) Then there 
would be no problems with numbering ...

Joachim