[Coco] Hidden 256-color mode

[Nickolas Marentes]

James wrote:
I thought enabling the 256 color mode was difficult because they had run 
out of physical pins on the chip rather than to hide the mode from 
Tandy. Not releasing the info was because of Tandy for sure.

Nick replied:
My contact at Tandy said the chip designer had run out of pins. This is 
why the mode was a little weird to enter. It was more of a kludge mode, 
exactly as you said, a function that the engineer said, "Hey! I can add 
256 colors by just adding this!". Or maybe, at Tandy's request at 
removing the mode he simply "hid" the mode rather than a major reworking 
of the design. I also heard that originally, the GIME was to be an RGB 
output only device and at the insistance of Tandy, they were forced to 
add composite output at a later stage. This is why the composite mode 
has a very different pallette system to the RGB mode. Maybe there was so 
little space left of the chip that there was little left to fully 
implement a proper palette system that matched the RGB output. Maybe 
sections were sacrificed from the 256 mode in order to implement the 
composite part. Interestingly, PAL based CoCo3's do not use the 
composite output of the GIME. Instead, there is a daughter board hanging 
below the mainboard that takes the GIME's RGB output and creates PAL 
composite video. The palette table used for composite PAL video is the 
same as the RGB palette...a much better system of only having the one 
palette to program for.



James wrote:
However if each byte had the color info built in then the hardware just 
bypasses some of the existing hardware when in the 256 color mode and 
passes on the info dirrectly.

Nick replied:
Once again, read my web page. It describes an observation that John 
Kowalski made of the GIME chip block diagram within the Tandy CoCo3 
service manual. John, in analyzing the block diagram, in particular 
where the video data is passed to the 6 bit color map ram, noticed an 
extra register before the color ram component that appears as a bypass. 
It appears to pass the video data past the 6 bit color palette ram and 
into another register just after it. John couldn't work out what 
register this was. It was after John's observation and his suspicions of 
a 256 color mode that we happen to make contact with a person at Tandy 
who was actually involved with the design of the CoCo3 (a contact whome 
we have lost since). Without mentioning to this person about John's 
suspicions, he came out (on an IRC chat at the time) with a comment, 
"Have you found the hidden 256 color mode"? Johns and my jaw dropped 
because it it was the first piece of possible proof that the mode may 
actually exist. Next was the official Tandy R&D document (also on my web 
site) from Mark Hawkins that showed the mode was certainly in the design 
stage.



Roger wrote:
Nick, I would like to post all or some of your articles on this subject 
on CoCo3.com and see what happens. We're getting over 15,000 hits a 
month and I'm sure we have some old Tandy officials or designers who 
stumble through. It's worth a try to stir something up on this and get 
some more information that might lead to the answer.

Nick replied:
No need. Just link to my 256 color page off my web site. Everything we 
know so far is there, described clearly.

Link to it via:   
http://members.optusnet.com.au/nickma/ProjectArchive/256mode.html



At this stage, we need to either...

1) Find someone who was actually part of the CoCo3 development team who 
actually knows about the mode.

2) Have the CoCo3 prototype analyzed more closely to see if more clues 
could be found.


Nickolas Marentes