> >I'd love to see a manual for the chipset and microcode information.
>
> I think the burning question would be: How does one debug somehting
> like this?
For a commercial company, microcode is tested and debugged on software
simulators before the hardware is even completed. Once the architecture
is defined, software is developed in parallel with the hardware to
simulate the components. Microcode is tested on the simulator, much
easier than using a scope and analyzer. Course, the component
descriptions given to the simulator programmers (these are usually
programmers with a good EE background who understand boards at the chip
level) had better be accurate.
And in parallel to that, an emulator is also done for the target
instruction set. This allows systems programmers to develop the initial
assembler, compiler and operating system for the new CPU. That's how a
completely new CPU design, with no backward compatibility, can appear
all at once. By the time the hardware prototype is ready, there is a
complete set of diagnostics and an OS to boot up. In theory that is ...
Tracey Kidder's book, The Soul Of a New Machine, detailed some of that
process when Data General switched from the 16-bit Nova to the 32-bit
Eclipse.
Jack Peacock
But modern SMP NT boxen can only multiprocess if the task is
especially adjusted for it. Apparently, NT itself will not take
advantage of the second processor, unless the app (e.g. Photoshop)
has the featuer. Is this always the case, or different in other
systems?
>
>
>< Dual processors? Another I've never heard of. My 8086 databook says
>< that up to four can be wired together, but I've never heard of anyone
>< doing it. Did this particular one work well?
>
>You could do more than four, it took a little more work as the simple
>parts to do it were roll your own and it was pretty impressive.
>
>
>later on NCR would have a box with 4 486/50s and it could wail.
>
>
>Allison
>
>
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
Now that I think of it, you're right. I just need the case from the apple
drive to house an old 5.25" full height HD. I needed the TRS-80
drives/boxes for an old TRS-80 model III.
--
-Jason
(roblwill(a)usaor.net)
ICQ#-1730318
----------
> From: Tony Duell <ard(a)p850ug1.demon.co.uk>
> To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
> Subject: Re: Hard drive adapters
> Date: Tuesday, August 25, 1998 1:57 PM
>
> > Apple and trs80 drives are not interchangeable though the case/PS would
> > be. I think I still have one with a drive.
>
> DId Apple drives ever contain a PSU? All the ones I have seen (Apple ][
> single and duodisk, Mac drives) take their power from the main system
PSU.
>
> >
> > Allison
> >
> >
>
> -tony
>
Hi gang. FYI Jameco has power supplies for apple2,2+,2e, and 2GS.
49.95 each, 60 watt units, original equipment style.
If you're thinking of one of these you might ALSO want to think about something
called Bugg power - a guy on the a2 newsgroup (comp.sys.apple2) who takes
pc XT power supplies and adapts them to plug into apples. Since the average
XT PS is 145 watts, the apple no longer has power problems (and you barely
have to run the PS's fan) and it gets the power supply and all that heat
out of the case altogether. (it sits on the floor ala C=64 power brick).
I've never used one of these things, but the reviews from people who have are
glowing. As for heat, I noticed in my GS last night it gets right toasty in
there, and the power supply is almost too hot to touch. I expect to be
PS shopping shortly. :(
--
Jim Strickland
jim(a)calico.litterbox.com
-----------------------------------------------------------------------
Vote Meadocrat! Bill and Opus in 2000 - Who ELSE is there?
-----------------------------------------------------------------------
< Well, mine still works. It just needs a nudge once in a while at
< boot up. Is it going to get worse or can I use it for a while?
It may get worse or better running it all the time is best.
Allison
< Dual processors? Another I've never heard of. My 8086 databook says
< that up to four can be wired together, but I've never heard of anyone
< doing it. Did this particular one work well?
You could do more than four, it took a little more work as the simple
parts to do it were roll your own and it was pretty impressive.
later on NCR would have a box with 4 486/50s and it could wail.
Allison
< Hm, interesting. I had always thought that DEC made all of their own si
< Does anybody know when DEC started rolling their own?
They did and also used harris, AMD and others as silicon foundries as
the then dec facility was capacity limited. DEC did most of their own
design or formed partnerships with others. The earliest was the 6100
series. Later PDP11 and special support chips.
< I think the burning question would be: How does one debug somehting
< like this?
the chipset or the Pcode? Pcode was easy as any other asm lang. The
chipset had enough stuff available externally o debug easily.
Allison
Of course, the amount is irrelevant. Particle accelerator experiments
have not been able to split lusers up into any constituent parts.
In fact, they may be the elemental particles of intelligence. Of
course, if Congress funds my new Me^H^H ultra high-powered particle
accelerator, we'll be able to test this once and for all.
In short, even if your rounded up all of the lusers, you couldn't
destroy them.
>luser-antiluser[1] pairs are produced wherever there's enough
(minimal!)
>energy :-)....
>
>[1] An antiluser is not clueful, right!
>
>-tony
>
>
______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
At 01:18 PM 8/26/98 -0500, you wrote:
>
>< Now *that's* historic! Western Digitals first (and last) CPU effort, th
>< first HLL ever implemented as an instruction set on a chip (afaik, anywa
>< rare as hen's teeth (I'll bet it's scarcity is on the order of the Apple
>< not to mention that it was a *very* early 16-bit system that was actuall
>< available to mere mortals.
>
>BZZT!!! first the PDP-11/03 is the WD chipset, the alpha microsystems
>s100 crate used the same WD chipset. The WD chip set alloed you to create
>your own microcode based cpu. It was the only whole computer WD marketed.
Hm, interesting. I had always thought that DEC made all of their own silicon.
Does anybody know when DEC started rolling their own?
>It didn't run pascal it ran the compiled result P-code which was a stack
>machine.
OIC. Silly Me.
>Scarce, they did make a few. They were expensive though.
>
>< It was a good idea; but as is so common in this business, the old axiom
>< held true:
>< No good deed goes unpunished.
><
>< I think WD lost their shirt on this one . . .
>
>About right. It's was not cheap and hard to expand. However the sales
>of the chipsets to outside producers (DEC and AMS) made them a bundle.
Kinda reminds me of the Moto 88000. The original 88k never gained a
following, and Mot lost their shirt. But much of the 88k's technology
went into the PowerPC, which was a little more successful.
>I'd love to see a manual for the chipset and microcode information.
I think the burning question would be: How does one debug somehting
like this?
Jeff
>
>Allison
>
>Curious. Same goes for people who aren't in a large company and
invent various other things. How did he 'invent' a gun without
owning the necessary tools (at least I doubt a magazine writer would
own this stuff) to process metal? Also, how did Woz make a case for
the Apple ][, etc? Are there companies that take single orders
like this?
There are companies which will build one off prototypes for you. They
are usually listed under something like custom electronics or
fabrication services in the phone book. Building a prototype is not
cheap, the cases, parts, whatever are usually hand machined. Even
plastic cases, since an injection mold would run you over $10K in
startup costs.
Disclaimer: I'm part owner in a company that does this. We have a
substantial investment in machine tools (mills, lathes, welding, plasma
table), a glass shop (neon and cold cathode tubes), plastic machinery
(ovens and router tables), and a paint booth. It's an interesting
business, everything from small computers the size of a deck of cards to
2 ton exhibits for computer companies.
As for Elmer Keith, he didn't start from scratch. He adapted existing
cartridge cases by milling out dies. It's not as complicated as it
sounds, some wildcat (i.e. not production) cartridges can be produced
with nothing more than hand tools.
Jack Peacock
