Looking for good examples of morph programming

Ned Konz ned at bike-nomad.com
Tue May 20 15:48:34 UTC 2003


On Tuesday 20 May 2003 12:47 am, Lex Spoon wrote:
> > > Why do it this way?  I tend to lean the opposite direction. 
> > > Using mouseDown: methods makes the code easier to browse.
>
> [...]
>
> > reassign the event behavior
>
> [ ... ]
>
> > it handles the waitForClicksOrDrag: stuff
>
> [ ... ]
>
> > eToys
>
> Okay.  I would have thought this kind of thing was handled in
> handleMouseDown: , but you're right.  I don't know why
> handleMouseDown: *doesn't* do these things.

Because (I think) it's relatively high overhead to make a 
MouseClickState object for every mouse click.

> There are 80 implementors of mouseDown: in my image.  Are none of
> them supposed to be amenable to, say, having EToys override the
> mouseDown activity?  That doesn't seem right, but it's the way it
> is.

Most of them are either widgets with very specialized needs (and that 
are never "out there" alone in a World to be scripted), or they call 
"super mouseDown:" somewhere.

Generally, the ones that should respond to #on:send:to: are the ones 
that might be useful by themselves (as opposed to, say, a

There *are* some that don't play well (my Connectors come to mind here 
-- cough, cough --) with eToys.

-- 
Ned Konz
http://bike-nomad.com
GPG key ID: BEEA7EFE



More information about the Squeak-dev mailing list