> Stéphane at least engaged, even if I didn't understand the logic
> ("Fine-tuning of Morph position").

The logic is that, generally speaking, we do not know why a morph is 
where it is. A specialized morph may very well rely on its position to 
have a specific effect (think of a slider controlling sound volume, 
whatever is owner is; just a dumb example). So we do not want to have a 
default behavior for grabbing that does not ensure that the morph stays 
where it is upon grabbing. Else we would be mangling the action of 
"grabbing" with the action of "moving" (actually loosing the morph 
initial position, which is even worse).

I have actual instances of this pattern in muO, but it would be too much 
to describe them here. Play with muO and you'll see them :)


Stef