Difference between Morphs and bad programming

Ned Konz ned at bike-nomad.com
Tue Mar 27 23:23:43 UTC 2001


On Tuesday 27 March 2001 12:28, Ross Boylan wrote:

> It seems to me the argument for separating the model from the
> view/controller is a good one.  And it seems to me the "cool" thing about
> morphs is that they don't separate these things.

The appropriateness of any design pattern (in this case, MVC (Observer)) has 
to be determined in context. One of the useful things that the so-called 
"Design Patterns" movement did was to introduce a convention for discussing 
the appropriateness of such patterns.

These are engineering decisions, not moral judgements. There is no absolute 
moral superiority of model/view separation. Similarly, there is no moral 
superiority of "object-orientedness".

> All the buzz at least suggests that morphs represent some kind of
> mind-shift from the usual way of doing things.  My mind seems not to have
> shifted!  Could anyone help me out on this one?

Morphs are indeed different, but not necessarily with respect to the 
model/view separation. There is, in fact, a MorphicModel hierarchy that 
allows for MVC designs using Morphic GUI elements. The browsers, etc. all are 
designed this way (look at the Pluggable*Morph classes).

> Do you agree that morphs include both the GUI and the model? 

They can, if you want them to.

> Or are morphs
> intended to be a more capable GUI layer that will still operate in tandem
> with underlying, non-graphical model classes?  If not, why is combining
> model and GUI a good thing?
>
> Regardless of the first point, what is or are the core innovations of
> morphs?

-- 
Ned Konz
currently: Stanwood, WA
email:     ned at bike-nomad.com
homepage:  http://bike-nomad.com





More information about the Squeak-dev mailing list