Bil Herd comments and writes about the old days

[RobertB]

(On Facebook, CBM engineer Bil Herd has been reminiscing about the old days at Commodore.  I've taken some of his postings and combined them here.  Check out the link at the end for an article he wrote!  I hope Bil doesn't mind my putting his messages here.)

Bil Herd
Dec. 10 --

Just exchanged some emails with an old friend, Yash Terakura, who was extremely intertwined with early Commodore machines and proponet of the Toi/Ultramax which led into the Vic 20. Been trying to hunt that sucker down for 12 years.

Dec. 14 --

In the Olde Commodore Engineer department, I found out more about Yash Terakura, stuff I didn't know while working there at the time. In addtion to Color Pet he did the Vic 20 and C64 designs with help from MOS guys. Yup. I got this from Chuck Peddle, designer of the 6502 and PET (and arguable home computers in general) just an hour ago.

The problem with these guys is you have to beat out of them their accomplishments, they weren't out to "change the world" they were just getting shit done and in the process making Jack happy for a couple of days. (Nothing like being driven to new levels of accomplishments from the top)

In addition it was from Yash that I learned how to talk that "Osakan Street Trash" in Japanese.


Yeah there is always where one story leaves off and the next starts when one regime leaves and a new one comes in which explains why I didn't know a lot of things that came before me when I worked there. (People like Yash, Freddy and Bob Russell where multi-generational. )


Yash pretty much confirmed when he said he did the Vic and C64, I was checking with Chuck to see what his POV was. These things are team efforts as well of course, a Lot of people would have been involved in a lot of ways, I just never got whose desk the piece of paper came off of for these two designs. I heard Russell take the blame for the overheating Power Supply in the early Vics, typical self-effacing attitude form the early guys who freely talk about what went wrong as well as what went right.


Oddly enough there would be enough stuff for a small book, but it would be 15 years too late to be relevant to anythings. The link below is something I wrote for Compuserve back before spell checkers were available to the folks at home, I just grepped for this one, I forget who posted the original.


http://home.datacomm.ch/fmeyer/c64/c128_story.html

[airship]

I'd read the linked story, but this is the first I've 'herd' of Terakura.

I love these old CBM stories. Keep them coming!

[RobertB]

...this is the first I've 'herd' of Terakura.

In regards to VIC-20 development, Yashi Terakura is spoken of in five paragraphs in the book, "On the Edge: The Spectacular Rise and Fall of Commodore", p. 240.

Now on vacation,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
The Other Group of Amigoids
http://www.calweb.com/~rabel1/
Southern California Commodore & Amiga Network
http://www.sccaners.org

[carlsson]

As I haven't read "On the Edge", it was a new name to me too.

Now I want to know who was responsible for the VIC-20 memory map. :-D As being a non-engineering programmer, it seems to me it could have been laid out differently and made room for up to 39.5K of continuous RAM without a lot of pointer changes and confusion.

[RobertB]

Now I want to know who was responsible for the VIC-20 memory map. :-D As being a non-engineering programmer, it seems to me it could have been laid out differently and made room for up to 39.5K of continuous RAM without a lot of pointer changes and confusion.

A clue from p. 173 of "On the Edge" --

Memory was the most important consideration in order to meet Jack's target of a three-hundred dollar computer.  MOS Technology had a surplus of one-kilobyte RAM chips, which engineers no longer used in PET computers.  Jack made sure his designers used these one-kilobyte RAM chips in the new computer [i.e., the VIC-20] to clear out any inventory.

And another clue from p. 175 --

"There should be no question in anyone's mind that the primary architect behind the VIC-20 was Bill Seiler," says Peddle.  "We didn't have to go back and invent all kinds of new things except for the I/O stuff."

"Seiler had most of the hand in what was really the design," says Russell.  "Seiler was the guy who made sure there was production hardware."

Bob Yannes affirms that the final design of the VIC-20 was nothing like his prototype.  "If you look at the VIC-20, it is a PET, except instead of the 6845 video controller, it has the VIC chip in it," says Yannes.  "It has the same I/O structure and the same processor, and the kernal code that was running in there was the PET operating system kernel."

Truly,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
The Other Group of Amigoids
http://www.calweb.com/~rabel1/
Southern California Commodore & Amiga Network
http://www.sccaners.org

[carlsson]

This is how the (VIC-20) memory map is laid out:

$0000-$03FF : System RAM (1K)
$0400-$0FFF : Unconnected (3K expansion)
$1000-$1FFF : Basic RAM (3.5K) + screen matrix (0.5K)
$2000-$7FFF : Unconnected (3*8K expansion)
$8000-$8FFF : Char ROM (4K)
$9000-$93FF : I/O Block 0 : VIC, VIA (1K)
$9400-$97FF : I/O Block 1 : Colour memory (1K nybbles)
$9800-$9BFF : I/O Block 2 : Unconnected (1K)
$9C00-$9FFF : I/O Block 3 : Unconnected (1K)
$A000-$BFFF : Cartridge RAM/ROM (8K)
$C000-$DFFF : Basic ROM (8K)
$E000-$FFFF : Kernal ROM (8K)

I assume something like this would've been doable:

$0000-$03FF : System RAM (1K)
$0400-$07FF : I/O Block 0 : VIC, VIA (1K)
$0800-$0BFF : I/O Block 1 : Colour memory (1K nybbles)
$0C00-$0FFF : I/O Block 2 : Unconnected (1K)
$1000-$1FFF : Cartridge RAM/ROM (4K)
$2000-$9FFF : Unconnected (4*8K expansion)
$A000-$AFFF : Basic RAM (3.5K) + Screen matrix (0.5K)
$B000-$BFFF : Char ROM (4K)
$C000-$DFFF : Basic ROM (8K)
$E000-$FFFF : Kernal ROM (8K)

If that was possible, memory expansions would have been added backwards so the first +8K would start at $8000, +16K start at $6000, +24K at $4000, +32K at $2000 and still 4K RAM/ROM. It would have maintained the screen at a given position, custom characters could always be put in the same place.. No reason to release both 3K and 8K expansions unless it was entirely for cost reasons. The extra I/O blocks are so rarely used that most probably one could be sacrificed.

Err.. sorry for getting so much into detail, I just felt the need to clarify.

Of course one drawback with this would have been games written for +8K could not be loaded with +16K unless a special routine was used to make sure it loads to its intended position. 

[Andrew Sutton]

I remember having ram expansion for the VIC that used a "backplane" where the ram had to be plugged in a certain way for software to load properly. Being pretty young at the time, I didn't fully understand why... now I do, sort of!