From: <bgardner@ATHENA.MIT.EDU>
Date: Wed, 20 Sep 89 23:37:56 -0400
Message-Id: <8909210337.AA04923@PINK.MIT.EDU>
To: geer@ATHENA.MIT.EDU
Subject: Re: Re: Athena Look&Feel
Cc: staff@ATHENA.MIT.EDU

   Oh, ok, I'll put in my two cents, too.
(As he puts of his Perceptual Pyschologist hat)...
Using the diffuse lighting model adopted by Motif, only 3 stages
of the day could be destinctly perceived; which isn't enough to
make the Athena Look&Feel something worth suing over. If we
added specular highlights to the lighting model, we could up
the number of stages to 5 (only Apple's lawyers could win this suit).
Unfortunately, many of the computer displayable stages
(over the 3 of the diffuse model) are not noticably perceivable, due to the
averaging effect of the eye over constant global shading changes.
[see Dr. Geer's paper on the "Effects of the Retina on Shading Suability", '89]
The problem here is we are trying to show subtle lighting effects on 
boring rectangular window borders.
   What I'd suggest is that we make the labels on the title bars look
like they are engraved. (Instead of the boring 2D titles Motif now has on
its neat 3D titlebars.) The titles should give us enough 3D contours
for more informative shading. Then we could visually locate Dan's "Kerberos
Sun" as it tracks across the sky.
   We could accomplish this easily, by simply triple drawing the title text;
once in black, then once in white (diagonally offset by several pixels),
and then once in grey (diagonally offset to overlap between the black and
white copies)... (Get ready to blurr your eyes.) ...
So, an "A" (with the sun at 2 o'clock) would look roughly like:

 ::::::::::::::::
 :::::    :::::::
 ::::,,,,, ::::::
 ::*,, **,, :::::
 ::*,, :*,, :::::
 ::*,,,,,,, :::::
 ::*,, :*,, :::::
 ::*,, :*,, :::::
 ::*,,::*,,::::::
 ::**:::**:::::::
 ::::::::::::::::

   (Looks better when drawn...trust me.  :-)
   Then, we can simulate the lightsource moving by redrawing the 3 copies of 
title text at slightly different offsets. This would make it appear as though
the highlights and shadows on the engaved letters are moving around as
a result of a moving lightsource.
   For example, the 3 drawn copies of the title; white (w), grey (g), and
black (b) text could be placed at the following linear offsets:


                w       w      w      w     w      w       w
                   g      g     g     g    g     g      g  
        w  g  b       b     b    b    b   b    b     b        b  g  w


  This gives us 9 distinct light source directions, even if we only use an 
offset distance of three pixels.
  
    Absurdly yours,
         Brian