[Coco] Verilog CoCo 3?

[Joel Ewy]

David Gacke wrote:
> Hi Guys,
>
>  
>
> Whatever happened to the guy with the Verilog CoCo3 implementation?  I found
> a link in the archives, but the website that it pointed to is gone it seems.
>
>   
That was Gary Becker, I believe.  I haven't been to the web site for a
while, so I didn't know it was completely gone.  I had noticed with
great disappointment, that the downloads had been removed and he had
left a cryptic note about "rethinking" his FPGA projects, whatever that
means.  But it didn't disappear before I had downloaded the source and
the documentation PDF.
>  
>
> Also, was that code he had posted GPL'd or similar?  If so, could someone
> please send it to me, or post a link to it.  I'd like to do some poking
> around in my spare time.
>
>   
I am fairly certain that he mentioned the GPL by name on his web site. 
Unfortunately, I saw no "COPYING" file, and no mention of any license
whatsoever in the source or the documentation PDF, so I would say that
means that it has not, technically, been released under the terms of the
GPL, or any other recognized Open Source license.

I hope it's not the case, but it looks like we may have another
marvelous project, 99% complete, abandoned by its original author,
hiding in plain sight, just out of reach of a potential community of
collaborators, behind a glass wall of copyright law.

But before we get too fired up...  My recollection was that in Gary's
posts to this list, he seemed to express an interest in making his FPGA
CoCo a community project.  It seemed clear to me that, at least at one
time, he was very open to collaboration, and wanted to share his work. 
He did post the source code after all.  I also got the feeling that his
attitude soured as he saw the tone of some of the discussion on this
list, but I would have to look back at old posts.  I can understand
being put off by some of the sniping that goes on.  Perhaps we could all
take a civility pill, and wash it down with a dose of skin thickener.

It would certainly be worth having somebody very politely and tactfully
contact Gary and see what his present intentions are and what he would
have us do with any copies of his source code that we may already have
downloaded.  Needless to say, it is up to us individually to choose
whether we comply with his wishes on this point or not.

On the positive side, there are already other CoCo-related chunks of
Verilog/VHDL out there.  The CPU core Gary used with his design was
written by John Kent, and is quite definitely licensed under the GPL. 
It has been criticized as amateurish, and isn't cycle-accurate (neither
is the 6309 running in native mode), but can apparently be clocked to at
least 12 MHz, and most importantly of all, it exists, it works, and it
is actually available under an Open Source license.  I'd rather have a
sucky, amateurish, functional, and available CPU core that can't be
yanked away on a whim than nothing at all.  If it's bad, make it
better.  If nothing else, it can serve as a stand-in processor until
somebody comes up with a better one.  Good commercial software is a fine
thing, but I'll take bad Open Source over proprietary vaporware any day.

In addition to John Kent's 6809, I believe that Mark McDougall has done
quite a bit of work on an FPGA CoCo 1/2, and James Daggett has been
working on a GIME replacement.  So even if Gary Becker's code becomes
lost for the purposes of collaboration (which we don't definitively know
yet), there is other work that could be brought together, if the
respective authors are willing to share.

I think that part of the problem the CoCo community has in finishing up
projects is that they all tend to stay one-man-shows.  I'll admit that
in my weaker moments I still have ludicrous fantasies about starting
something big with a project that begins in the basement in my spare
time.  (Spare time?  That's the biggest joke of it all!)  The mythology
of Gates, Wozniak, and Jobs looms large in the imaginations of us all,
so it's hard to give up the notion that our hobby could become our
livelihood, or at least could pay for itself.  It's difficult to give up
control of an idea and let a bunch of other people get in there and
monkey around with it.  But the more we focus on how much time we are
spending working on our babies, and how little others appreciate our
efforts, the more bitter, angry, and depressed we become.  This is no
fundamentalist free software screed, and it is not an attempt to tell
anybody else what to do with their projects, nor is it a criticism in
any way of those who put in monumental efforts and ask for payment for
their work.  I have great respect for Boisy Pitre, Mark Marlette, Roger
Taylor, and anybody else who makes and sells a product for the CoCo.

But people keep on stressing that this is a hobby, and that 15-25 year
old publications for an obsolete computer will never again have enough
monetary value to be worth suing over, et cetera.  Previous discussions
of FPGA CoCos have become mired in speculation about the difficulties of
turning such a thing into a commercial product.  Fine.  Forget it. 
There will likely be no Next Generation CoCo that you can buy, already
assembled, in a box, from a single vendor.  There are undoubtedly people
willing to buy peripheral devices, software, accessories, an so on, but
there probably isn't a market for a mass-produced, conventional product
on the scale of a whole CoCo-compatible computer.

But why on earth do we have to limit our thinking to such grandiose
outcomes?  Just because it'll never become a shrink-wrapped product,
produced by the boat loads in China, filling the shelves of your local
Wal-Ly World, played with for five hours, and then dumped in the
landfill, doesn't mean it's not a project worth doing.  In fact, in my
mind, the rough and open nature of such a project makes it much more
interesting  than some closed, polished box.  One of the best features
of the CoCo was that it was a diamond in the rough.  All those little
ads in the back of Rainbow Magazine were because the CoCo had so much
potential in the form of "room for improvement".  I know my capabilities
are humble.  But I just might be able to contribute something of worth,
however minimal, to a real, honest-to-goodness, collaborative Next-Gen
CoCo effort.  The rougher and more open it is, the more chance there is
for me to contribute -- and that's what would motivate me to take part.

Keep your expectations low and you'll be less likely to be
disappointed.  Gary Becker built his FPGA CoCo on a ready-made FPGA
development board.  It was somewhere in the $100-$200 range.  We don't
need an assembled, tested, branded CoCo in a box.  We need a freely
distributable, collaboration friendly, HDL CoCo core that individual
users can, if they can supply the necessary hardware and expertise,
build into their own custom FPGA CoCo-compatible system.  Once that is
available, there will at least be the groundwork in place for a
community of developers to grow around it.  There will be plenty of
opportunity to develop and market little kits that make it easier for
ordinary users to set one up for themselves, or add new functionality. 
Keep the design modular and people can even sell commercial plug-in
replacements for the HDL components.

You can put hours and hours of solitary effort into a thing, get mad,
quit, and have wasted a chunk of your life.  End of story.  Or you can
put hours and hours of effort in, slap a GPL (or Open Source license of
your choice) on it, put it out there, and open yourself up to the
possibility, however slight, that somebody else might just give
something back to you in the form of collaboration.  Sure, you're taking
a chance that nobody cares, or has the time or expertise to do any more
with your baby than what you have done yourself.  In that instance,
you're no worse off than if you had just picked up your marbles and gone
home.  Better yet, you can make a real effort to get others to
contribute and collaborate.  Sure, it may in the end amount to nothing. 
But the chances of a project going somewhere are far better if you don't
insist on going it alone, and if others share a vested interest in
seeing their work realized as well. 

Let us take it for granted that, for the most part, the market for
CoCo-related items on the scale of entire computer systems is tiny and
sporadic at best.  Let us simply assume that any such development must
either be the colossal effort of one entirely dedicated individual, who
is likely to be profoundly disappointed at the lack of a massive
groundswell of eager and wealthy buyers; or else an entirely
collaborative labor of love by the CoCo community to produce a successor
to the Color Computer line that is not tied to any particular hardware
implementation, or to the fortunes of any company or individual.  The
rewards will likely not be monetary, though if such a project comes to
fruition, I do see opportunities for enterprising developers.  Instead,
what one would get for participating in such a project would be the
enjoyment of collaboration, the expression of creativity, and the thrill
of the hack.

Obviously if Gary's project had been licensed under the GPL, and if
Rainbow Magazine was re-released with a Creative Commons license, we
wouldn't have all this hand-wringing, sniping, flaming, accusation,
threatening, and whining.  We could simply all get down to the business
of playing, working, and talking about the Color Computer and its
relatives, which is what I think we all really want to do on this list
anyway.  Those things are what they are, and we may be able to do little
or nothing about them.  But if we wanted to, we could take what we do
have and begin new projects that are collaborative and open from the outset.

FWIW, IMHO, the views expressed here, bla, bla, bla...
JCE
>  
>
> Thanks a lot,
>
>  
>
> Dave Gacke
>
>  
>
>
> --
> Coco mailing list
> Coco at maltedmedia.com
> http://five.pairlist.net/mailman/listinfo/coco
>
>