[Coco] The Tri-Annual CoCo 4 Thread

[farna at amc-mag.com]

On Mon, 10 Feb 2014 09:10:21 +1100
Mark McDougall <msmcdoug at iinet.net.au> wrote:

> And you simply will not get any sort of consensus here. Regardless of
> how considered, rational and 'common sense' your proposal is, it's
> simply not going to please everyone. I have absolutely ZERO interest
> in a Raspberry Pi solution, or any other software solution. And that
> really doesn't matter either, because there's just as many people
> that would be happy with it, even though secretly I know my solution
> is "The Best". ;)
>
> Design by committee never works, and it won't here. What is going to
> work is a *number* of people actually building something, and sharing
> the design, and others can choose what they like and adopt it (Gary's
> Coco3FPGA is the perfect example). Hopefully we'll end up with (at
> least) a hardware solution, an embedded software solution, and a
> PC-based emulator solution.
>
> IMHO the kind of consensus that we should be aiming for, are the
> specifications of any 'Coco4' enhancements. Standardise on, say,
> specific enhanced graphics modes, and then support them across all
> manner of implementations. That way there's (hopefully) enough users
> to promote software development utilising those enhancements.
>
> And by all means throw your suggestions around; that's the fun of
> this hobby and you shouldn't be denied that. But please don't get
> surprised or disheartened by it all.

Date: Wed, 12 Feb 2014 10:20:41 -0500
From: Richard E Crislip <rcrislip at neo.rr.com>

THAT is exactly my conclusion! If you build it they will come. Trying
to get a consensus my not be a bad idea, but do it from the archives of
previous discussions, do not let the community know you are doing it
and then spring it on them when it is a finished viable product in what
ever form it is. My thoughts, yours will probably vary ;-).

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

Case in point: The MM/1 was designed "by committee", with the designer
trying to please as many people as possible. He brought a number of people
into the discussion as the machine was being designed. It ended up costing
a lot more than expected and having too many issues/compromises to be very
well received. It also took well over a year just to get the first ones
out.

Frank Hogg listened to the discussions around the MM/1 and CoCo community
in general, then proceeded to design and build the TC-9 "Tomcat" with no
(or little) external help. While it was no more of a success than the
MM/1, it did come out first. It too had too many compromises. In the end
both were pretty good OS-9 machines but had no backwards compatibility
with the CoCo outside of OS-9. The Tomcat was supposed to have a
hardware/software compatibility mode, but as I recall (and I could be
wrong) it was never fully achieved. The Tomcat would run CoCo OS-9
software, but that was all. In that respect it was more compatible than
the MM/1, but not enough. I think the software required some patching to
run due to some of the hardware differences.

We're back to CC3 compatibility at a minimum. I think we can dispense with
CC1/2 compatibility at this point, as there are only a few useful programs
that won't run on the CC3. Enhancements are fine as long as there is
transparent CC3 compatibility as well. Supporting enhanced modes will
limit those programs to the new machines of course, but that should be
expected. Once a standard is set and a couple programs using the
enhancements appear (even if just games), then interest in the enhanced
version will grow.

Right now I'd like to see a two-pronged approach to best cover the current
(and future) market -- a hardware solution like the DE-1/FPGA along with
an emulated solution. Develop the FPGA version with compatibility and
enhancements, then alter one of the emulators to match the hardware. That
will solve the dilemma of cost and help both camps. Those who want the
look and feel of real hardware have it, and those who would be satisfied
with, prefer, or need (usually due to lack of space and/or funds) a
software solution will have it. Since new software will run equally well
on both the user base is increased dramatically, though probably still not
to a commercially viable level. Still, the larger the potential base the
better.