It's probably exactly the way you suggest, Tony, but I'd say you ay be
overlooking the most obvious way to do this sort of thing, i.e. use volatile
programmable logic. I doubt any patent will be violated if you use an old
schematic as the model for an implementation that couldn't have existed in
the timeframe within which a patent might have been valid. A patent applies
to the implementation of the items that are claimed as original or
innovative. It will be your implementation that is innovative if you do it
this way.
Another point worth considering is the relative risk of being sued. If you
build a single copy of your implementation, even if it is on a standard FPGA
board sold for many purposes, it is a unique effort, yielding a unique
product, no longer really the same board built by the manufacturer, yet it
will be able to execute, and quite precisely if your implementation is good,
the instructions the original one did. However, if there were any money in
this implementation, someone would be building it and you'd have THEM to
worry about, not the original maker. If one's dreaming of making a bundle
supplying an old war-horse that died years back, even at a 50-fold
preformance increase, it's not likely to be worth the filing fees.
If you don't like how it works, you can fiddle with it to your heart's
delight, without any resoldering or cut and paste. Now, there is, actually
a risk you might not be able to preserve pinouts, so what I'd recommend is
to build a board with your FPGA (one is normally better than two or more,
since many pins of I/O are lost if you try to use small devices where a
larger one is indicated) surrounded with dual row pin-fields over which you
can interconnect the signal paths with longer jumpers if need be in
subsequent revisions or other architectures.
Dick
----- Original Message -----
From: Tony Duell <ard_at_p850ug1.demon.co.uk>
To: <classiccmp_at_classiccmp.org>
Sent: Friday, June 09, 2000 4:15 PM
Subject: Re: Wirin' up blinkenlights
> >
> > I haven't had -that- particular fantasy, but another one...
> >
> > ... creating copies of certain "classic" machines. If you were able to
buy,
> > say, an original Nova in a box that for all the detail you wanted was
the
> > same as the original, and inside had the same chips on the same size PC
board,
> > would that be cool? Sure, you're have to use materials you could get
today
> > to do it, but accurately done, it might be interesting.
>
> Yes, I've thought about this too....
>
> Technically, it's not that hard, if you allow 'minor' changes in the
> chips -- using 74HCT chips in place of plain 74xx (and making other
> necessary changes), for example. Or using larger, more modern SRAM chips
> to replace ones that are hard to find and expensive now.
>
> Legally there are at least 2 problems. Firstly, the design is probably
> still covered by some form of copyright (IANAL). Most schematics of these
> old computers have notices on them preventing them from being used to
> make copies of the machine. My guess is that some manufacturers might
> well give the permision, though.
>
> The second problem is that these old machines may well not meet modern
> standard for RFI emissions, etc. Which (at least in the UK) matters even
> if you're only making a 1-off :-(
>
>
> >
> > I'm -not- saying building up clones of certain rare old computers and
trying
> > to pass them off as originals, but rather, trying to make clones that
are
> > undistinguishable from the original.
>
> I would want to make some change that made it impossible to pass it off
> as the original. Perhaps putting the text 'Copy created by <name> <date>'
> in an inside layer of the PCB, so it could be seen if the board were held
> up to the light, but couldn't be removed without destroying the board. I
> have no interest in making fakes, I have every interest in playing with
> machines that I otherwise wouldn't be able to see.
>
> -tony
>
Received on Fri Jun 09 2000 - 21:28:29 BST