19 Mar
2007
19 Mar
'07
9:33 a.m.
On Sunday 18 March 2007 23:31, Chuck Guzis wrote:
On 18 Mar 2007 at 22:46, Lyle Bickley wrote:
Not to belabor this, but timing IS the issue. Many early OSes used "timing
loops" - which is bad coding - but easy programtically. OSes also made/make
assumptions of when interrupts will occur - and it what order.
Of course, one could use a clock accurate emulator to deal with this and other
CPU and I/O related timing issues.
> There are many, many cases where software
written for DOS, Windows
3.1x, early
versions of BSD, etc., will NOT run on high speed
systems with more
"modern" OSs (XP, Linux, Solaris, etc.).
Lyle, that's not what I said. I said "a computer from today can run
software written 40 years ago--and generally, much faster." I said
nothing about running under one OS or another.
To say that running older software on newer machines isn't possible
is pretty much equivalent to saying that a new machine isn't Turing-
complete. I can emulate (with whatever software necessary) cycle-for-
cycle operation of an older machine. Maybe not as fast as the
original (depending on the complexity of the emulation)--which is why
I added "generally much faster". I can readily appreciate that I/O devices may well
represent a really
tough nut to crack, but given a free reign on abstraction, it can
probably be done in some fashion (e.g. a disk file instead of a tape
file).
And the reverse is quite true. I can emulate an IA64 architecture on
a 2MHz Z80, given sufficient storage and time. Not fast (or cheap),
but I can do the emulation. To say that it's not possible would
again be equvalent to saying that the Z80 isn't Turing-complete.
It's quite true that one can emulate or model virtually any architecture (I
ran a modeling group at IBM's Poughkeepsie Development Lab. for several
years). Unfortunately, while emulation/simulation of hardware/software is
valuable, it is essentially useless in patent litigation matters.
For example, if one wants to kill a 1983 patent, and there is inadequate
written prior art available, one may be able to use 1983 (or earlier)
physical technology (hardware and software) to do it.
In court, the opposition would "nail" any hardware/software presensted to a
jury that was not contemporaneous with or earlier than the patent being
litigated. Anything not original would "contaminate" an exhibit (because of
the potential for "cheating", etc.). That fact precludes the use of
emulation, simulation, etc. of old hardware/software on modern systems in
patent litigation.
Hey, be glad that some of us use old hardware/software to kill bad patents :-)
Cheers,
Lyle
--
Lyle Bickley
Bickley Consulting West Inc.
Mountain View, CA
http://bickleywest.com
"Black holes are where God is dividing by zero"