[Zope3-dev] Using parent references rather than context wrappers to represent containment

[Phillip J. Eby]

At 08:30 AM 8/11/03 -0400, Jim Fulton wrote:

>I find this unacceptable for component lookup,
>as context awareness should not be a prerequisite for using
>the component architecture.

I thought this too, in PEAK, and originally had a global context, plus an 
optional "local" context that was intended for per-thread context.  Then I 
found that the cost of having to explain to people why things didn't work 
as they expected, was a higher overhead than what I was "saving" them by 
trying to DWIM a context.

My two cents: trying to push this issue into a DWIM, just steepens the 
learning curve, because you still bump into it *later*.  I've been finding 
it preferable to just treat the issue up front as a requirement, because 
the cost to the user is lower than when they have already started trying to 
figure out why something didn't work the way they expected.  When starting 
to learn something, the user is open to learning.  But, when they thought 
they knew how it worked, and they are surprised, they begin to distrust the 
framework and their knowledge of it.  After all, what *else* might they 
think they understand, but really don't?

I'm not arguing against the current proposal, just commenting on something 
that appears to be a direction (DWIM) that has historically led to the 
infamous "Zope Zen" learning curve.  I think your current proposals don't 
have this issue as a major concern, though, since per-request default 
global context seems unlikely to bite anyone who hasn't already reached 
Zope 3 guru status.  Unless, that is, it leads to people ignoring context 
issues (because they don't have to think about it), and thus getting bitten 
sooner.  But I don't feel strongly one way or the other about that, which 
is why I'm +1 on the proposal as a reasonable compromise between being 
completely explicit and completely implicit.