Re: "Co-Processor" Apple ?

[mjmahon@aol.com (Michael J. Mahon)]

Robert C. replied:

>"Michael J. Mahon" <mjmahon@aol.com> wrote in message
>20040817220555.19157.00003638@mb-m06.aol.com">news:20040817220555.19157.00003638@mb-m06.aol.com...
>> Robert C. replied:
>>
>> >"Michael J. Mahon" <mjmahon@aol.com> wrote in message
>> >
>> >> All these issues caused me to prefer using an annunciator output
>> >> and a pushbutton input for NadaNet--it is both simpler and more
>> >> capable.
>> >
>> >The more I read about NadaNet, the more I think it is "the next best
>thing"
>> >to true Ethernet.
>>
>> I'm glad you like what you've read.  ;-)
>>
>> The practical bandwidth is about 8 kilobytes/sec, which is a lot
>> lower than the peak bandwidth of Ethernet, but the latency is
>> very low and Apple II's don't need screaming bandwidth.
>>
>> NadaNet delivers data at about the same rate as a 5.25" disk,
>> which is quite practical for many purposes.
>>
>> And you can't beat the price...  ;-)
>>
>> I've just written a Applesoft program "prun" which will do a
>> "parallel run" of any Applesoft program on all networked
>> machines running the service loop.  The slaves complete
>> their computation and return to the service loop by doing
>> any input operation.
>>
>> I've been pretty busy reorganizing my lab area for the last
>> few weeks, and expect that to go on for another few weeks,
>> but my next NadaNet project is assembling the code to
>> run at the top of RAM (at $B800 for AppleCrate slaves),
>> so that HIMEM: can easily protect it, and all of low memory
>> will be contiguous for program(s).
>>
>> I'll give some thought to self-relocation in the process...
>
>I have downloaded and read the documentation and code and am in the process
>of entering the code one byte at at time. I know the files are available for
>download too, but I figured: what better way to learn the code than to enter
>it manually.

Entering the code in the monitor is certainly an excercise in
patience!  At least you can "List" the code to verify that it is
correct.  (Errors in hand entry are a major issue.)

> I will be using my 2 Apples as a "Master" / "Slave" setup, so I
>was wondering, is there any parts of the code that should be changed? The
>Applesoft program you wrote can be easily changed so that it does not use
>"machine 2" as a message center, and the Slave polls the Master upon
>completion of its job.

Yes, it can be changed, but as written it requires at least three
machines:  the master, the message server, and at least one
"worker" slave.  The master initializes the system and puts
"jobs" in a message queue, "workers" receive jobs from the
message queue and place results in another message queue,
and the master receives the results and prints them.

As I note in the documentation, the primary function of the message
server is to permit queued asynchronous communication between
machines.  If you don't use a message server, then any two
communicating machines must "rendezvous", usually by one
of them polling in its service loop.

If you are only connecting two machines, then you may want to try
some different things.  The "parallel run" program would allow you
to initiate an arbitrary Applesoft program on the slave machine from
the master.  I'll look at putting it up on the web in the next few days.
(Although you can probably write your own using "PARALLEL.SIM"
as a template.)

You will not need the NadaBoot code, since you will not be
running a "diskless" slave.  That also helps with debugging,
since you will be able to debug your slave machine(s) locally.
(Some network debug capability is getting higher on my want
list.  ;-)

(As I've been cleaning and reorganizing my lab, I discovered
that it has a _floor_!!  Amazing!  ;-)

-michael

Check out parallel computing for 8-bit Apples on my
Home page:  http://members.aol.com/MJMahon/