Re: RFC : SOME IDEAS FOR THE APPLE II FPGA'ers

["mdj" <mdj.mdj@gmail.com>]

Michael J. Mahon wrote:

> But there are lots of other RGB adapters and RAM cards.  I've never
> used an RGB adapter, so it's pretty easy for me to live without one. ;-)

There are, you're right, and they're all harder to find than the
Applied Engineering gear, as they were produced in lower quantities. I
have a Checkmate MultiRAM lying around that I don't use, because the
flicker that occurs when the banks are switched drives me nutty.

> That almost certainly will not prolong their life.  Go ahead and
> use them.  Almost all IC failures after infant mortality (3 months)
> and before old age (>30 years ;-) are very low incidence random
> failures--unless they are severely mistreated.  And most of the
> IC's in Apples and their peripherals are standard parts (but not all).
>
> It's better to die booted than sitting in a desk drawer.  ;-)

Agreed. Although there are other forms of failure: lightning, wierd
power supply failure modes that blow something/everything up, etc. And
there's still the scarcity issue. Eventually, operational heat will
kill components though; I'm convinced every failure I've had is due to
temperature variation eventually either damaging the silicon substrate,
or the interconnects from the packaging to the silicon, so from this
perspective they may 'last longer' in a sealed static bag.

> Exactly.  My CFFA cards have enabled me to unplug my SCSI cards
> without any feeling of loss (not that I don't want to keep them ;-).

Yeah... I never got around to purchasing one of those... I thought I'd
wait it out until the FPGA era replaces many more things. The CFFA is
very cool, but uses a lot of hardware to accomplish a very simple job
in todays terms.

> Bummer--ProSEL utilities are among my favorites, too.  They work
> fine on a CFFA card.  ;-)

Yep. The issue with the Workstation card is that it uses the jump hook
in the ProDOS MLI to attach arbitrary devices. An AppleShare volume,
while it shows up as volume, is not a ProDOS block device, as the
filesystem operations are actually performed by the server, not ProDOS.
This, plus the fact that an AppleShare volume can exceed the normal
ProDOS device size limits, means that block level operations on the
device aren't possible. (The actual filesystem on the other end would
typically be HFS, although in my case it's emulated HFS via AppleDouble
conventions on a UNIX filesystem).

It's a fair tradeoff though, to have Asimov accessible as a ProDOS
filesystem fulltime :-)

> Eventually, of course, they will be right--but they haven't been
> right yet!  Frankly, if we were going to be inundated with mass
> chip failures in five years, there would be a lot of them going
> bad *today*.  Since that apparently hasn't happened, I feel it's
> safe to conclude that doomsday is still way in the future.

See above, but I think for the most part, you are right. It won't stop
me wishing to circumvent potential failure, and guaranteeing
reproduceability for myself, and anyone else who may want it.

> And by the time that you've reverse-engineered exactly what an ASIC
> does, you have all the information required to simply replace it
> with a modern programmable part (perhaps on a socket adapter board).

True, but it I think it's both more efficient and fun to just reverse
engineer the logical behavior of a peripheral and replace it
holas-bolas with a modern component of equivalent functionality.

> > So at a minimum, what I'd *like* is a replacement motherboard for my
> > IIe that includes accelerator functionality, RamWorks style expansion
> > memory, and ability to drive more modern displays. It doesn't have to
> > clock at stupidly fast speeds. It should include more than 1mb of
> > auxiliary RAM (oh how I've longed for one of those 2mb expander
> > boards).
>
> Hmmm, what exactly do you *do* with all that RAM?  I've found that
> I'd hardly miss any RAM beyond what's needed to hold a 3.5" disk
> nibble image (for 1-pass copies).

It's more a case of if you're going to do it, do as much as you can
without increasing complexity. And since it's (at least in my mind)
easiest to implement the functionality with a generic 24-bit address
bus to the FPGA I figure there may as well be 16mb of RAM on it. It
opens the device up to a plethora of potential other fun uses without
increased complexity or significantly increased cost (16mb FP mode
SIMMS are a dime a dozen still)

In terms of actually using the memory on the Apple II, I basically
don't. Although I have been known to use both a slinky card AND the
RamWorks, as the slinky makes a better ramdisk. It means I can keep a
copy of my Apple Pascal environment on my hard disk, and simply image
it onto the slinky card and reboot to that slot. Afterward, I can just
image it back.

The same goes, actually, for DOS3.3 based applications, although I've
never bothered to set this process up. I did consider for a time
imaging most of my favorite DOS3.3 based applications onto my hard disk
and setting up ProSEL menus to 'boot' them via the slinky card, but if
I went that far, I'd be too addicted to the slinky to ever remove it
from the system, and it gets mighty hot in there as it is :-)

> Yes, the display issue seems to keep cropping up...yet there appears
> to be a huge pool of high-quality RGB-capable NTSC-scan monitors out
> there from all the TV studios converting to digital...  A little
> stockpiling (distributed or centralized) could easily satisfy the
> entire Apple II community's needs for the next 40 years or so.  ;-)

True, but to be honest I'm getting sick of keeping them around. They're
big, heavy, and blow up a lot. It'd be far easier to just avoid ever
having to keep them around, IMHO

> I think you'll find that when the system is the FPGA, then integrating
> a lot of uncoordinated subsystem designs will be both difficult and
> ugly.
>
> And coordinating subsystem designs requires a *serious* design
> methodology that *all* designers buy into--not easy to do.

If *I* was building it, I'd adopt the OpenCores project's Wishbone Bus

http://www.opencores.org/projects.cgi/web/wishbone/wishbone

These guys have already amassed a phenomenal amount if IP, all there
for the using.

> I understand why you are not excited by the C-one itself.  And I think
> you can see why I'm worried about the necessary excitement--and
> committment--in the Apple community being hard to come by, too.

Sure. Although even if the excitement doesn't follow up, and all that
gets created is the means to easily recreate Apple II systems + some
nice to have peripherals I can live with the fact that there's
boundless untapped potential there.

Of course, things might never get that far, hence trying my level best
to push things that way using only my hands and a keyboard ;-)

> Please don't apologise--some "pragmatic dreaming" is very much in order!
> That's why I keep quiet until it seems like it's all just spinning out
> of control, rather than converging on something achievable.

No problems... I was just being facetious ;-)

> Also, I must admit, I've been hanging around here long enough that it
> just seems like every year we get spring rains and Apple fancies...
> After the seventh or eighth time, with no real results, it gets a
> little less enjoyable--particularly as a lurker.  ;-)

Indeed. although also every year the means to make such a thing a
reality gets that much closer, so the effort doesn't go totally to
waste.

Matt