Squeak default look

Lex Spoon lex at cc.gatech.edu
Thu Apr 18 03:40:21 UTC 2002

Martin Drautzburg <martin.drautzburg at web.de> wrote:
> Doug Way <dway at riskmetrics.com> writes:
> > There was actually someone who posted on the Squeak-Foundation list 
> > awhile ago about offering graphic design help (from his group), so the 
> > possibility is there.  Implementing an arbitrary look created by a 
> > graphic designer may not be easy, though.  (To make this simple you need 
> > something like "skins", but that's overkill in many ways, which I think 
> > you agreed with earlier.)
> I will have to look this up. 
> BTW I did post an article on alt.design.graphics telling them about
> the possibility to do some fundamental work. The response was less
> enthusiastic than I thought it would be. All in all it seems to be not
> exactly easy to attract designers. I may not have hit the right
> "tone" though.

A very important problem before trying to make the UI consistent, is to
figure out *which* UI is being discussed.  IMNSHO, for example,
SystemWindows are messed up by design since we also have halos.  I don't
even think the X and O buttons should be on the SystemWindow title bar,
but many of the discussions are about how to make the X and O look nice.
 We need to resolve questions like that before prettying things up.

But even aside from questions like that, there is a big difference
between etoy-ish interactions and programming interactions.  If I had my
choice, I'd rather the etoy-ish interface was cleaned up, because that's
what makes Squeak the most special.  But, is the etoy interface even
stable??  How long before the MathMorphs ideas of typing-in-air and
drag-and-drop to perform arbitrary messages are incorporated?  These
could have pervasive effects when they are added, and thus override a
lot of any UI that is designed today.

In short, I'm worried that efforts to make the UI consistent and
good-looking, will lock us into a design before we know what the UI is
supposed to be doing.    It's a very worthwhile goal, but let's be
careful not to gold-plate it too early.


More information about the Squeak-dev mailing list