Re: Apple II Interrupt card

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

In article <3E7D4683.4060904@noneofyourbusiness.com>, Exegete
<millers@noneofyourbusiness.com> writes:

>Michael J. Mahon wrote:
>> In article <wUFea.1008$oY6.85397@news20.bellglobal.com>, Yves McDonald
>> <""yves.mcdonald\"@NO SPAM sympatico.ca> writes:
>> 
>> 
>>>>What was the no slot clock made of, and can that be used to create an int?
>>>>
>>>
>>>The not-slot clock was an hybrid component consisting of clock 
>>>hardware/socket/battery you would install between a ROM chip and its 
>>>socket. Its interface was on a single bit (serial) and it would be 
>>>transparent when reading EPROM.  A program (like ProDOS) could set and 
>>>read time.  The battery let the chip tick while the computer is turned 
>>>off.  It didn't have interrupt output or even a counter output since ROM 
>>>chips are not connected to IRQ* line normally :)
>> 
>> 
>> Which is fine, since no interrupt is required.
>> 
>> Adam said that the 50/60Hz interrupt was used only to synthesize
>> a "seconds" time of day clock, which is polled only in the idle loop.
>> 
>> CONTIKI DOES NOT NEED INTERRUPTS!!!
>
>I wouldn't have mentioned it Michael, except some have posted that it 
>did. I certainly don't know one way or another.

Some ASSUMED that it did, and Adam corrected that assumption.
Then, observing the continued divergence, I pointed out that interrupts
were NOT required at least twice.

>IF however, interrupts are needed, or are a more efficent way to get 
>Contiki to work (again, I'm clueless on this) this little card seems to 
>be the cheapest way to go.

I'm not convinced, unless you find a bushel basket full of them that
you're willing to sell for $1 plus 37 cents postage.  ;-)  I think that
by the time a PC board is made, parts acquired, stuffed, and tested,
that only a fool could charge less than $10 for one of those little
cards--more than the typical mouse card (and there are _lots_
of those out there--just not all at once).

Further (as I also posted) this job could be done with a couple of
wires and a couple of milled-pin 16-pin DIP sockets, together with
a 2-input open-collector gate (and that even provides for selective
disabling of the VBL interrupt).  I suspect that someone has already
designed this hack, which could be constructed for $0.50-$2 (depending
on whether you have to buy the sockets new--or use them at all).

BUT...this is all irrelevant to getting Contiki up on Apple II's.

>I'm almost sorry I tried to help....

My "blast" wasn't directed specifically at you, Roy, but at the runaway
divergence of this thread from solving the problem at hand--running
Contiki on a (minimally assisted) Apple II.

Since the "clock tick" constant was undefined for the Apple, Contiki
would not compile.  Then it surfaced that this constant was used in
other implementations to implement an interrupt-driven real time clock.
Adam stated that the clock() function was called in Contiki only in
the idle loop, suggesting that it is polled to determine when enough
time has elapsed to cause one or more timeouts.

Meanwhile, a headlong rush toward finding a mechanism to generate
interrupts suitable for implementing a Commodore-style real-time
clock on the Apple II was started.  Interesting, certainly--fun, maybe,
but apparently unnecessary to solving THE PROBLEM AT HAND.

This kind of "herd" behavior is interesting from a psychological point
of view, but it is frustrating in the sense that effort is diverted from two
more critical areas: 1) determining with precision EXACTLY what
requirement(s) Contiki has for the clock() function, and 2) finding the
simplest, most universal way to meet those needs for the widest
set of Apple II's.

"Focus, people, focus!"--if you want to solve the problem.

-michael

Check out amazing quality 8-bit Apple sound on my
Home page:  http://members.aol.com/MJMahon/