[ZPT] Mini proposal: repeat-define and repeat-condition

[Richard Jones]

On Fri, 24 Jan 2003 7:48 pm, Godefroid Chapelle wrote:
> At 12:05 23/01/2003, Richard Jones wrote:
> >ZPT has been used in delivered systems now. If those systems are to ever
> >receive a Zope upgrade (for a critical security or bug fix) then all
> >templates in those delivered systems will have to be checked and possibly
> >repaired too.
>
> How do I manage to live with different versions of Python (Zope, ZPT,
> .........) ?
>
> If it is a problem to me, I avoid to upgrade and I accept not having access
> to new features.
> If I really need those new features, I make the work.

I didn't say features, I specifically said "critical security or bug fix". 
Like it or not, the Zope community is not large enough to support maintenance 
branches for eg. Zope 2.6 for all future "critical security or bug fixes". To 
get those fixes we often have to upgrade to a new major release. That means 
that we have to be mindful in the future to not _arbitrarily_ break backwards 
compatibility like this.

Also, your argument that I'm disallowing adding any features whatsoever to 
Zope in the future is misrepresenting my statements in the extreme. I'm on 
the record as having absolutely _no_ problem with adding functionality like 
the repeat-condition or repeat-define (or whatever it ends up being). Quite 
the opposite, I've come out supporting that proposal.

As far as I'm concerned as a long-time ZPT user, the proposal to change the 
ordering of TAL processing is arbitrary to suit the whim of some users. I've 
never had a problem with the current ordering, and I strongly oppose any 
change now.


    Richard