A Foole's Thoughts on the windowBoundingRect

[Rob Cozens]

Hi Everyone,

Not too long ago I challenged the conventional wisdom that focused on 
an image's imageData rather than its text property as the essence of 
the image.

Today, I'm foolesh enough to challenge referencing the 
windowBoundingRect when maximizing a stack to full screen.

1.  The windowBoundingRect can be set from the Message Box or any 
handler to be any value you choose.

1a.  Rev Dictionary: "By default, the windowBoundingRect is set to the 
screenRect. On Mac OS systems, the default windowBoundingRect is 
adjusted to leave room for the menu bar. On Windows systems, the 
default windowBoundingRect is adjusted to leave room for the task bar." 
  But that's just the default.

2.  The windowBoundingRect changes depending on whether you are running 
a stack or a standalone, because the Rev IDE resets the 
windowBoundingRect to some value other than the default.  Create a 
stack with the openStack handler "put the windowBoundingRect into field 
1".  Run it in the IDE and then as a standalone.  In the first instance 
the windowBoundingRect is (on my TPC in portrait screen orientation)  
89,114,768,994; in the second it is 0,0,768,865 with the TPC Input 
Panel visible and 0,0,768,994 with it hidden.


 From Rev Dictionary v2.7.1, re screenRect: "Adding the working 
adjective to either form returns the virtual co-ordinates of each 
screen's working-area.  The working-area of a screen is defined to be 
the area not covered by OS furniture (such as the task bar [and TPC 
Input Panel] on Windows, and the Dock and Menubar on Mac OSX)."

If one replaces "windowBoundingRect" with "working screenRect" in the 
test above, the result is consistent whether running from a stack or a 
standalone.

As an aside, it appears the Rev IDE does not reference the 
windowBoundingRect when it maximizes a stack, because we see above that 
the wBR changes depending on visibility of the TPC Input Panel while 
the maximize logic never grows the stack to full screen if the TPC IP 
is hidden when the M/M button is clicked.

I believe using the working screenRect rather than the 
windowBoundingRect is preferable in these cases.

Rob Cozens
CCW, Serendipity Software Company

"And I, which was two fooles, do so grow three;
Who are a little wise, the best fooles bee."

from "The Triple Foole" by John Donne (1572-1631)