Simple Changes, Big Differences

[airship]

We've talked around some issues that seem to point out some very, very small changes in the C128 hardware that would have made big differences in how it works. For example:

(1) If the C128 could detect ATN on the serial bus and its peripherals could generate ATN, you'd be able to have peripherals talking directly to each other (for print spooling, for example), CBM computers could easily share peripherals with one another, and you could even implement a simple networking scheme.

(2) If the MMU implemented bit 7 you could add the two left-out RAM banks that  would make the C128 a C256 with full system software support already built-in.

In your opinion, what other small hardware (or system software) changes would have made BIG differences to the C128 or its peripherals?

(Please number your suggestions to make them easier to refer to in our discussions.)

[Guest]

Are you making a wish list for a 128-in-an-FPGA?

[airship]

No, I wasn't really thinking of that. I like the 64DTV for what it is, but I really love the old hardware.

It just struck me that many of the things we've discussed here run into small disconnects with the C128's actual design. I've read comments on everything from unfixed bugs in the ROMs, to design omissions or changes that would be nice, to actual design changes that were widely deemed a mistake (like the change in SID chips).

There are, of course, changes that WERE made that were great, like the move from 16K of VDC RAM to 64K. Kudos, Commodore! So what else could have been fixed, or improved, or made different, that weren't? What small, realistic (not pie-in-the-sky) changes would have made this great little machine even better?

[Guest]

I would say using an 8088 instead of the Z80 would be at the top of my list, then Commodore could have distributed MS-DOS for the 128 (it would have been proprietary to C= due to the hardware).  It probably would have resulted in a RTC being part of the system, which would have been a GREAT thing.

[airship]

NOW you're talkin'! I've had the same thought myself in the past. There was a version of CP/M for the 8088, so you could have still run that if you wanted to. We'll call that idea (3).  

I'll add another, since I already mentioned this in another post:

(4) The ability for the Datasette to play audio cassettes and to record sound from the SID chip.

BTW, one thing I AM thinking is that I may make some of these mods to my own C128, if they're do-able. For example, a new ROM that fixes things like the DS$, things like that. Maybe even hacking a Datasette to do #4. But is #3 out of the realm of possibility? Could you create some kind of 8088 adapter board? Or are they too different?

Edit: #3 Right off the bat I see quite a few differences. The Z80 runs off a single +5v supply, while the 8080 needs +5v, -5v, and +12v. The 8080 needs two clocks, where the Z80 uses one. The good news is I THINK pins 23 & 24 are the same.  

But all of the others would have to be remapped. It's tough to figure out the rest of the non-address, non-data lines. Some may be exactly the same, but Zilog and Intel seem to call them by different names. And we haven't even begun talking about timing, positive/negative logic, etc. And of course almost all the internal operations are different, though they do seem to share at least some commonality in register structure and layout. In short, an adapter would be a tough engineering job, and would require a new ROM for the C128 to boot. (Pun intended.)

[airship]

I give up. I'll just add a bunch myself. It's almost as if you all have actual lives that keep you too busy to post here.

(5) A true RS-232 port. A fast one. (At the very least, see #16.)

(6) Software switchable system and expansion (internal and external) ROMs.

(7) More than two pages of color RAM. Preferably limited only by the amount of RAM you have.

( Built-in support for raster interrupt routines in BASIC. The split-screen modes are nice, but things like swapping sprite definitions or character sets would have been cool, too.

(9) Cursor keys in an inverse T-bar arrangement like on the IBM keyboard, instead of way out there at the top of the keyboard. And a full-function numeric keypad, not a stripped-down one. Add true ASCII support (in addition to PETSCII).

(10) Fast serial saves as well as loads. And why not add a standard fastloader (like Epyx's) for slow drives? System support for reading and writing MS/DOS disks.

(11) BASIC graphics commands on the 80-column screen.

(12) OLD, MOD, and all those other left-out BASIC commands from other BASICs. Matrix math. Double precision math. Make COPY and BACKUP work across units. Useful utilities like MORE. More system variables. Etc. (For the complete list, see my post on the Perfect C128 Expansion ROM.)

(13) The 1572 dual drive.

(14) Cool blinking LED lights for address, data, and status registers. (I love cool blinking LED lights.)

(15) More VIC-II and VDC colors. Like TED, at least. While you're at it, make the 80-column and 40-column colors match.

(16) A couple of expansion slots. (Apple had had them for YEARS by this time!) RE: #2, sockets should have been provided on the motherboard for RAM expansion to four banks.

(17) A parallel printer port. An IEEE port. These could be options, but they should have been available. (See #16.)

(18) Two SID chips. Or at least a socket for a second one.

(19) System level support for mouse, joystick, and lightpen cursor control.

(20) System support for interlace and overscan screens. And automatic FAST mode on retrace.

(21) The ability to genlock/chromakey the 40 and 80 column screens with one another. (Okay, so this one isn't simple - but it WOULD be cool!)

Whew! I'm worn out. Now, time for the rest of you to have some input. Which of these would have been most possible, most inexpensive, and most useful back in the day? Or today, as far as that goes?

[hydrophilic]

#7 would be tuff since the color RAM is a special 4-bit data bus not connected to normal memory.  Connecting it to normal memory would be tricky because the VIC will access the text and color memories AT THE SAME TIME.  Might work if you but color in RAM 1 and confined VIC to RAM 0... on second thought that wouldn't work since the RAM banks still share address and data lines...

#18 Two SIDs for sure.  Although more expensive than simple TTL chips, SID had to be cheaper than the Z80 or VDC.  Very practical also.  Most people have two speakers.  How many have two monitors?

#22 Two extra (i.e., 3-button) joystick.  No extra hardware required for the C64 or C128 either (use POTX and POTY but connect them to +5V instead of GND when pressed).  Why this didn't become standard with the 64 before the 128 was even released is beyond me!  There are so many games that require more than 1 button.  Don't you hate those games where you have to reach to hit the SPACE?
The craziest 'solution' I've seen was by Accolade or Access or Acclaim software with a game called Echelon that had a voice-activated 'gun' called the Lipstick.  Interesting concept that turned loud sounds (think the clapper) into a 'fire' input.  Why they just didn't add a button (or two) to a joystick is beyond me...

[airship]

#22: As far as that goes, a controller with a resistor ladder could have given you a set of buttons that used the POT inputs. One of the problems with the POTs is jitter, which makes them almost unusable for BASIC. But if you have a set of evenly-spaced resistors instead of a pot, that problem goes away. You could have easily added a 10-key pad to each POT line that way. (That gives a range of about 25 for each button in the keypad; very easily done, even with 10% resistors.)