Wanted: 5 people to look at an embedded book idea.

Hello all,

I'm working on the idea/presentation for my next book. It relates to
retrogaming from an embedded developer's standpoint (not an emulation
book, but more along the lines of "this is how you could develop a
modern version of the 2D tile-based graphics ASIC in XYZ old arcade
machine", and "this is how you can simulate a YM2159 synthesizer in
VHDL"). I feel many of these circuits have modern applications beyond
gaming. I will include /some/ historical material in the book, but it's
mostly concentrated on applications.

The book will be based around the Xilinx ML403 EDK [I haven't yet
decided fersure which CPU core I will be using but in order to exploit
synergies with other work I'm doing it is likely to be PowerPC]. Very
strong C and digital design skills are presumed. This is assuredly not
a novice-level book; it is intermediate to advanced material.

I'm looking for five people (only) to review my outline and make
suggestions, particularly of additional pieces of arcade, console or
home computer hardware I can implement in VHDL or C. There is a small
gift certificate in it for you if you participate (your choice of
amazon.com or cabelas.com).

During the official submission process, I also have to provide at least
two reviewer contacts to my publisher. I will be selecting those two
reviewers from the pool of five people gathered in my first-round
review. My publisher typically pays a cash honorarium of ~$80 or twice
that amount in books.

If you're interested, please contact me via email.

Muchos gracias.

Just a few initial observations. First, what exactly are you trying to
do with your book? Is this a lesson in VHDL/FPGA/Digital via
retrogaming design examples? Or how historical hardware paved the way
for today's technology? Or..?

Second, if you're intent is to accurately model vintage gaming
hardware, why on Earth are you trying to bastardize them into a PowerPC
or MicroBlaze? There's plenty of folks who have done free models of the
older processors (6502, Z80, 8080, etc). It's not unlike trying to
drive a nail with a belt sander.

Next, who is your intended audience? Anything Virtex-4 is quite
expensive for anyone other than corporate, and cost-prohibitive for the
casual programmer or student. Spartan-3E's are 1/50th the price and are
more than enough ponies for emulating 80's LSI and MSI hardware. Ever
try a belt sander with balsa?

Companies such as www.digilentinc.com have decent demo boards for a
song with plenty of firepower to do what you need.

Lastly, be very careful of asynchronous logic when modeling some of the
older game boards, especially ones with no CPU. Atari's PONG comes to
mind. :-)  FPGA's do not lend themselves readily to the asynchronous
Rube Goldberg designs of yore because they (and I can only speak for
XIlinx) cannot guarantee a minimum propagation delay anywhere. So,
where the emulated design relies on race conditions you kinda have to
fudge it and play tricks to get it to match a good synchronous design
practice without losing the intent of the circuit.

Hope that helps at least a bit.

Cheers!

- Craig

larwe wrote:

> I'm working on the idea/presentation for my next book. It relates to
> retrogaming from an embedded developer's standpoint (not an emulation
> book, but more along the lines of "this is how you could develop a
> modern version of the 2D tile-based graphics ASIC in XYZ old arcade
> machine"...

Please don't restrict yourself to such 'modern' hardware; _real_ arcade
games were TTL MSI :-)

Regards,

Michael Grigoni
Cybertheque Museum

On 2006-04-28, msg <msg@_cybertheque.org_> wrote:

>> I'm working on the idea/presentation for my next book. It relates to
>> retrogaming from an embedded developer's standpoint (not an emulation
>> book, but more along the lines of "this is how you could develop a
>> modern version of the 2D tile-based graphics ASIC in XYZ old arcade
>> machine"...
>
> Please don't restrict yourself to such 'modern' hardware; _real_ arcade
> games were TTL MSI :-)

I think I've still got the first book I bought on video game
design.  A counter for the ball's X position, a counter for the
ball's Y position, some logic to detect collisions and toggle
the counter direction flip-flop, a couple A/D converters for
paddle positions, and away you go...

-- 
Grant Edwards                   grante             Yow!  Boys, you have ALL
                                  at               been selected to LEAVE th'
                               visi.com            PLANET in 15 minutes!!

> Just a few initial observations. First, what exactly are you trying to
> do with your book? Is this a lesson in VHDL/FPGA/Digital via
> retrogaming design examples? Or how historical hardware paved the way
> for today's technology? Or..?

None of the above exactly, but most like the first option.

History: A long time ago, I was interested in the Sega System 16
hardware for various reasons (Thierry Lescot, author of the Sys16
emulator, has at least one of my old Golden Axe boards :) I got all
fired up and began writing an "emulator for a nonexistent piece of
arcade hardware". This was called System N, and the intent was to use
it as a generic 2D gaming engine. It was written in C with PPC assembly
(for MacOS 9). It implemented two 8-way joystick interfaces with three
buttons apiece, three 16bpp 32x32 tile layers for parallax effects, and
a text layer on top of that.

More recently I've become interested in lashing some of the same sorts
of capabilities to various micro systems. It occurred to me that it
would be extremely useful to have a VHDL model of, say, the ZX
Spectrum's video hardware. It would also be useful to me to have a
64x64 3-layer palettized tile graphics system with an 8x8 text overlay
and at least 16 128x128 sprite channels. But not specifically for
anything to do with retrogaming.

Using the old game hardware as historical info puts the ideas in
context.

> Second, if you're intent is to accurately model vintage gaming
> hardware

No, it's not the intent. I did say this was not an emulation book. The
intent is to develop various graphics and audio output hardware that
does the same sorts of things as various arcade, console and
home-computer equipment.

This sort of material has applications in non-gaming sorts of arenas.
For instance, simulations (of the industrial-equipment-training
variety, not flight sims), digital instrumentation, and test equipment.

> or MicroBlaze? There's plenty of folks who have done free models of the
> older processors (6502, Z80, 8080, etc). It's not unlike trying to

Again, the aim is not to emulate anything. I'm well aware of the free
cores - but that's not what I need.

> Next, who is your intended audience? Anything Virtex-4 is quite
> expensive for anyone other than corporate, and cost-prohibitive for the

The ML403 EDK is < $1,000. While not cheap, it's not prohibitive
either. I did say this wasn't an introductory-level book... :)

It really boiled down to this board or an equivalent monster from
Altera. I already have the Xilinx board as part of some other writing
I'm doing, and Xilinx is -probably- going to be an avenue for marketing
the book. Done deal.

> more than enough ponies for emulating 80's LSI and MSI hardware. Ever
> try a belt sander with balsa?

Ever try making a 1:1, fully operational model of a C-130 out of balsa?
:)))

msg wrote:

> > modern version of the 2D tile-based graphics ASIC in XYZ old arcade
>
> Please don't restrict yourself to such 'modern' hardware; _real_ arcade

The oldest part that's currently in the TOC is the AY-3-8500, is that
ancient enough for you? ;)

In comp.arch.embedded Dave <dave@comteck.com> wrote:
> That EDK is what, $900-$1000 USD?  I don't picture many individuals
> of intermediate to advanced skill levels shelling out that much just to
> recreate old games.  And I don't see people who want to recreate old games
> shelling out that much.  Good luck!
> 

The only place I've had access to them was in Uni, and they get it at 
academic prices too.

-- 
Wing.

On Thu, 27 Apr 2006 19:03:48 -0700, larwe wrote:

> The book will be based around the Xilinx ML403 EDK [I haven't yet
> decided fersure which CPU core I will be using but in order to exploit
> synergies with other work I'm doing it is likely to be PowerPC]. Very
> strong C and digital design skills are presumed. This is assuredly not
> a novice-level book; it is intermediate to advanced material.

That EDK is what, $900-$1000 USD?  I don't picture many individuals
of intermediate to advanced skill levels shelling out that much just to
recreate old games.  And I don't see people who want to recreate old games
shelling out that much.  Good luck!


    ~Dave~

"larwe" <zwsdotcom@gmail.com> escribi&#4294967295; en el mensaje 
news:1146195916.743345.180090@i40g2000cwc.googlegroups.com...
>
> The ML403 EDK is < $1,000. While not cheap, it's not prohibitive
> either. I did say this wasn't an introductory-level book... :)
>

Interesting subject.

Please, make sure the whole thing can be programmed with the Webpack, or 
some other free package. I don't think hobbyists, and students are willing 
to pay an anual subscription for a full ISE in addition to the development 
board.


Best regards

Josep Dur&#4294967295;n


* It is  Much_a_s  Gracias   unless you want to refer to Terminator or David 
Beckam or ...   ;-)

Dave wrote:
> On Thu, 27 Apr 2006 19:03:48 -0700, larwe wrote:

> > The book will be based around the Xilinx ML403 EDK [I haven't yet
>
> That EDK is what, $900-$1000 USD?  I don't picture many individuals
> of intermediate to advanced skill levels shelling out that much just to

It's just under $1k retail. Assume that you already have suitable
hardware, though. The idea is to lift segments out of the book for your
own evil purposes. I'm using the ML403 because I like it and I got it
"free".