[Coco] Why do we need a CoCo 4? (Long irrelevant rant)

[jdaggett at gate.net]

On 29 Dec 2008 at 18:05, BookWorm wrote:

> A new GIME chip might be practical. So would a web browser. If TCP/IP
> is to much, skip it for now, and just do HTML. With CoCoNet, we can
> get online through a $#%& pee sea, so why not take it a step further
> and display a page? Or use DOS Lynx, like a shell account in the old
> days?

A new GIME chip is farther along that you might think. In fact about 60% 
coded. Several of the hurdles that I have struggled with over the past 6 
months have now beed resolved and design is moving slowly forward. I 
finally came to the realization that SDRAM will be needed over SRAM and 
DRAM. for size constraints mostly. I have finally decided on a desktop screen 
resolution of 800x600 at 256 colors. What I am coding will probably be better 
suited for OS9/NitrOS9 than DECB. I have concepts in mind and need to get 
time to finish codeing and synthesis to get to a point where I can be more 
comfortable of further discusions. Let's say that I would love to move the 
video section of the GIME chip into a small GPU. That would take burden off 
of DECB as well as OS9. 

An usefull web browser would require better video than what a Coco has. 
Also more memory either physical or virtual. That work is in progress. Many 
of the niceties would also possibly mean the backwards compatability with 
DECB is in peril. 

TCP/IP is best done via a peripheral processor. Even an 6809 in an FPGA at 
25MHz is hard pressed to do al lot of graphics and computation along with 
TCP/IP. There are several micros out there that a TCP/IP stack can be 
written for and a Coco can just use a parallel interface of some sort to 
transfer data. 

When one works on this as a hobby and at rather low priority, then progress 
is slow. I hope to free more of my time in the next 30 days to finish to a point 
of saying I have something to really offer. 

james