[Coco] CC-Five (was Re: Pseudo CoCo4???) (LONG)

[Joel Ewy]

Mark McDougall wrote:
> ...
> You forgot one thing that's going to be the make-or-break of any such
> project - namely Software.
>
> Without software you can't make use of a new MMU or expanded memory.
> Without software you can't access the new peripherals; storage devices,
> sound chips, mice.
> Without software you can't make use of the new graphics modes. There
> will be *NO* application software that can use these new modes. New
> versions of BASIC(/09) will have to be written. New paint programs. New
> text mode drivers for enhanced text modes.
>
> Writing all this software is going to be a majorly time-consuming task.
>   
I think there would need to be a sufficient level of backwards
compatibility that people would have software to run on the new platform
while developing the new stuff.  This also means that much of what now
exists in software could just be enhanced rather than being entirely
rewritten from scratch.  This is also why I'm suggesting modest,
attainable improvements over the CoCo 3.  If a next-gen CoCo has better
built-in sound, then build in an Orch-90cc circuit, so that one could
run software that already uses that hardware.  Here's where emulators
and real hardware could work synergistically.  If the emulator is free,
or very inexpensive, then lots of people can quickly get their hands on
the new system and write new software, so that there will be something
to run on the real hardware.
> And I can foresee other problems:
>
> * Who decides what's "in" and what "out" of a community-based spec?
> Design-by-committee is historically plagued by failures.
>   
Yeah.  I think that the design principles need to be clear at the
outset, and some kind of process would have to be put in place to ensure
a specification process that doesn't spin out of control.  Maybe a small
group could make proposals and present them to this mail list (plus put
them on a web site) for a period of public comment and voting.
> * Hardware development is expensive. 
Yet the System09 was developed by a single individual.  And since it is
GPL, the rest of us could benefit from what he has already put in place.
> And technology is moving so rapidly
> it very quickly becomes obsolete. Just look at the C-one - only about
> 1/2 the initial production run has been sold with little prospect of
> selling more; who wants to pay hundreds of dollars for an FPGA-dev board
> with little more power than modern CPLDs, when you can get much more
> powerful devices on a $99 board these days?
>   
This is also a plus.  This is why the PC is so ubiquitous.  It keeps
getting more powerful for the same amount of money.  We shouldn't spend
time on designing a custom FPGA board.  Just design a CoCo "personality"
unit to plug into off-the-shelf development kits, or educator boards, or
whatever they're called.  And the software emulator can always be there
for those who can't (yet) afford the hardware.
> It's a noble vision - don't get me wrong - but I question the
> practicality of 'selling it to the masses' - at least in a
> shrink-wrapped form. HDL source packages that can be adapted to various
> platforms may indeed prove the solution, but there's always the question
> of software support IMHO.
>
>   
I agree that this probably isn't going to be shrink-wrapped for the
masses, though an emulated enhanced CoCo could be.  That's why any
future CoCo development should exist as a specification, not a product.
> Regards,
> Mark
>   
Thanks for your thoughtful replies.

JCE