Tom Stepleton is the guy running this project -- I lent him my Lisa for his
work. Here's a wrap-up summary he sent me on the project just before he
returned the system to me.
-- Tony
----------------------
Here's a summary of what I've done. Addressing the unreliability of
the
wiring in the last attempt, I completely rewired the prototyping
board and
switched to a single 84-pin MAX 7000 chip and adapter instead of the
Altera experimentation board used in class. Physically speaking, the
circuit is complete.
I was able to get the circuit to succesfully dupe the Lisa into
believing
that there really was a ProFile attached to the port. Unfortunately,
there
is really no interim step between that and full functionality. I
came a
long way, but I didn't make it. Here's what I managed to do:
Linux side:
On a machine borrowed from Professor Maxwell, I installed a
secondary
Linux kernel - version 2.4.0-test4 - which provides the latest
facilities
for manipulating the parallel port from a userland program (read
more at
http://people.redhat.com/twaugh/parport/html/parportguide.html).
This is
a really neat driver, incidentally, although it is still in
development.
Using this driver, I created MrProfile, a rudimentary ProFile
emulation
program. MrProfile currently responds to interrupts generated by the
adapter, but all calls to read() extract zero bytes of data from the
port. Whether this is due to poor programming, the driver, or a
malprogrammed board is not yet clear.
Adapter side:
Using what I learned in making the program to fake a ProFile's
existence,
I completely reprogrammed the full-featured version of the adapter
circuit. The philosophy behind the device's role has not changed
much from
before, but the code has been altered quite a bit to suit the whims
of
Altera.
What happens:
Activity on the Lisa side causes the adapter to alert the PC with an
interrupt. The PC fails to read any data from the parallel port; the
adapter sits waiting for something to happen, and the Lisa
eventually
times out. There are three reasons for why this is happening, now
that
poor connectivity has been ruled out (via multimeter):
metastability,
flaws in parallel port driver, and poor assumptions about the EPP
parallel
port standard. I don't think it's the first, because the problem is
always
the same. I feel it's about even odds between the second and third
choices.
The future:
I'm not giving up, but the likelihood that I'll be able to put real
work
into it this fall is not all that great. Just in case the hardware
has
been programmed correctly, I'm taking the prototype board home to my
two
Lisas there (hope this is OK, Bruce), where I'll try getting it to
work
with an old Pentium 90. Even if the hardware is slightly faulty, the
low-level programming capability of the parallel port driver
suggests that
I will be able to get around it. However, if I've made an error on
the
Lisa side, I'll be out of luck. I'll see what options I have and
then make
decisions. I'll also turn over all of my code to the folks on the
Lisa and
Lisa Emulator mailing lists at sunder.net and see if they can offer
any
tips.
> ----------
> From: John Lewczyk[SMTP:jlewczyk@his.com]
> Reply To: classiccmp(a)classiccmp.org
> Sent: Thursday, November 09, 2000 11:16 PM
> To: classiccmp(a)classiccmp.org
> Subject: Lisa Profile Drive emulation on a PC
>
> There was someone trying to interface a Lisa to a PC, making the PC act
> like a
> Profile drive. That college class project didn't get completed, but there
> is a
> bunch of info on the project, the Profile and Lisa parallel port at:
> http://www.cs.swarthmore.edu/~tom/cs23/final/
>
> I thought that this was a terrific project, as Profile drives are become
> rarer
> and they will probably all break down before the Lisa's do. Making a PC
> act
> like a Profile is a great solution. Unfortunately, I have never received
> a
> response from the folks at swarthmore college and suspect the project is
> abandoned. :-(
>
> Anyone know what happened to the project, or are there others who have
> done
> this? I worry about my profile drives dying before my Lisa and would love
> to
> have a PC as a backup. If no one (is doing/has done) this, I may take my
> own
> shot at it.
>
> John
> jlewczyk(a)his.com
>
From: Jerome Fine <jhfine(a)idirect.com>
>
>I thought that you said the CE (Circus Engineer) did NOT recognize it??
>
>Did you really mean to say:
>
>"The CE that did NOT recognize it however didn't fare as well. ;)" ???
Actaully I wrote what I ment but I like your version as both were true.
The CE that got bounced, suffered. The next one had a dog on his hands.
Allison
-----Original Message-----
From: Jerome Fine <jhfine(a)idirect.com>
>Jerome Fine replies:
>
>I had to read it 5 times before I realized that the "cab" and the
>"HSC thingie" were the same. Subtle. Very good. I had a
>much better laugh as a result. More at myself, of course.
The CE that did recognize it however didn't fare as well. ;)
Allison
From: Mike Ford <mikeford(a)socal.rr.com>
>>> While the Norden bombsight was smaller, etc., it would be stretching
it
>>>by a
>>> large degree to compare it to a computer.. I'm fairly certain Allison
was
>>> referring to the Mark I...
>>
>>It's not stretching it at all to consider the Norden to be an analog
>>computer. I'm not really sure how it could be considered to be
otherwise.
>
>Its a single purpose device. Computers tend to be flexible. I would call
it
>more of a calculator, or really a mechanical compensator.
Really wrong. First calculator is a fixed program computer (mechanical
or otherwise) And the lack of programability would certainly be
significant
for a large class of machines as to their status.
We have machines (electronic, electrical, mechanical, water, air,
whatever)
that can:
Perform repeated steps (cams, timers)
Perfom compensating controls (servos, thermostats)
Perform a calcuation based on several variables
Subgroup fixed program (calculators)
Subgroup programable
by wires, jumpers, cams, gears, relays, roms
(calculators, embedded computers)
Variable program ("true computers", programable
calculators)
See the linkage, as soon as the system takes on multiple variables and
procudes
a results based on them we are doing calculation... Computing is just a
different
name.
Oh, the example I posted was written with Norden Bombsight, Mark-1A and
most any other tool used to calculate ballistic trajectories.
Allison
Does anyone happen to have a list of what the various layered product PAKs
in the OpenVMS Hobbyist program go to? I'm specifically wanting to know
what the SQL-DEV PAK is for, but am interested in the rest if anyone has
such a list.
Zane
On Nov 8, 18:07, Richard Erlacher wrote:
> If I remember correctly, the format difference between PAL and NTSC lies
not
> only in the difference in the vertical refresh rate but also in that the
> NTSC format is based on 525 horizontal scans per frame, while PAL uses
625.
> I'm not at all sure about whether PAL interlaces in the same way as NTSC
> does it, and, in fact, I'm not sure that they do that at all.
One is a consequence of the other. The most obvious difference between PAL
and NTSC is the use of different vertical scan rates (50Hz vs 60Hz) but
they use the same horizontal scan rate (15.something kHz), so if you do the
arithmetic, you'll see why they use different numbers of scan lines (625 vs
525). They both use interleave; so the number of full frames is 25 per
second or 30 per second. They also differ in the way they encode colour
information. The basic scheme is similar but the subcarrier frequencies
differ, and in the case of PAL, there is a "colour burst" of pure
subcarrier at the start of every line. This allows the circuitry to
synchronise better. Also, the colour signal is 180deg different in phase
on every second line (hence the name, Phased Alternate Line) which makes it
a little more stable.
In the context of broadcast TV, they also use diffent audio subcarriers (in
fact, there are several PAL variants, used in different countries. They're
called PAL/G, PAL/K, PAL/I, etc).
The net result is that you can often use a PAL display instead of an NTSC
one or vice-versa, but you would expect to lose the colour and sound. You
might possibly have to tweak the vertical hold to get a steady picture.
Since the sync pulses are encoded in the same way, any monitor or TV
should lock on to the horizontal signal perfectly, but the vertical might
be beyond the range of the sync circuit (it's usually OK in practice,
though).
I bet Tony knows more about this than I do.
--
Pete Peter Turnbull
Dept. of Computer Science
University of York
> > An older elector mechanical device did this for a different
> problem.
> >
> > Given variables: speed over surface,
> > altitude,
> > gravity,
> > an objects drag
> >
> > Problem: At what time do you release an object to hit a
> > predetermined spot.
> >
> > Hint: it was top secret in the early 1940s.
>
> And I would like to see an accurate simulation of one. I've seen
> the real thing, but being fastened to a metal post set into a concrete
> floor, it wasn't in operable condition. Not to mention that none of the
> dials were labeled, undoubtedly because the functions of each were also
> top secret at the time it was manufactured.
I think I've seen a photo of this, with Vannevar Bush standing over it,
looking at it.
-dq
>Ever since the late 60's, very large simulations of power
>grids were being performed on analog computers for the purposes of
>contingency analysis and dynamic stability assessment. These
>special-purpose analog computers filled entired rooms and sometimes
>had thousands of dynamic states (one integrator per dynamic state).
Exactly!
An older elector mechanical device did this for a different problem.
Given variables: speed over surface,
altitude,
gravity,
an objects drag
Problem: At what time do you release an object to hit a
predetermined spot.
Hint: it was top secret in the early 1940s.
Allison
> On Wed, 08 Nov 2000 15:07:20 -0600 Bill Richman
> <bill_r(a)inetnebr.com> wrote:
> > Our mainframe guys recently got rid of some big hard drive chassis,
> > and in the process of cleaning up they were going to thow away (!) a
> > couple of little Memorex/Termiflex one-handed terminals that I've
> > lusted after ever since I saw the service engineer using them. I
> > think they're a 1 or 2-line LED dot-matrix display, and
> they have some
> > form of "chording" keyboard on the front.
>
> Sounds like the old Microwriter, or even the original chord
> keyboard that Doug Engelbart used with the mouse.
>
> I'd love to see a picture of one of these, to get some idea
> of how the chording works. I have a couple of Microwriters:
There are two photos of Doug's chord keyset in issue #4 of
MacWorld. If you can't find a copy, let me know, I'll see if
I can remeber to bring it in to the orifice and scan it for
you.
-dq
From: Bill Pechter <pechter(a)pechter.dyndns.org>
>The only problem with CSSE is they made it so easy to work on the DEC
>stuff Field Service management began hiring the clueless to work cheap
>and they lost a lot of the clued folks to people who changed the board
>with the RED led lit. One day they worked on VT220's the next they
>were promoted to your VAX cluster!
Management... They took the work and tried to understand it. Field
Circus
management was a bunch of ex-bag carriers that felt they were the best
and the rest were dumb. The end result is that first you couldn't find
people that knew what a scope was and you still had to fix it due to
committed response time. The latter often led to get a body there,
NOW, I dont care who.
Allison
