[Coco] What do you make of this non-approved HSCREEN mode?

[Robert Gault]

To round out this thread, we should look at the screen produced by 
HRES=000 or 16-bytes per screen. It behaves normally with CRES set for 2 
or 4 colors. The fun begins with CRES set for 16 colors.
There should be two pixels per byte but you don't see a screen which 
looks like that.

10 HSCREEN2:POKE&HFF99,2
20 PALETTE1,24:POKE&HFFB0,7
30 M=&H60000
40 FOR I=0 TO 3056 STEP 16
50 LPOKE M+I,36:LPOKE M+I+1,36:NEXT
60 FOR I=0TO15:LPOKE M+I,129:NEXT
70 GOTO 70

You should see two vertical bi-colored columns separated by the screen 
background and on the top horizontal line 16 two color dashes.

It seems the screen is 32 bytes wide with 16 bytes accessible for 
drawing. The inaccessible area is filled with the palette #0 color and 
alternates with the accessible areas.

==========================================================

It would be nice if these weird modes were a window on the GIME 
circuitry. I'm not a hardware guy so I can't give an educated guess. To 
me, it looks like some incomplete look-up table is being used. Garbage 
in gives garbage out.