From: Spiro Trikaliotis (ml-cbmhackers_at_trikaliotis.net)
Date: 2004-10-21 08:59:05
Hello,
* On Wed, Oct 20, 2004 at 10:44:56PM -0500 Jim Brain wrote:
> Spiro Trikaliotis wrote:
> >Anyway, as JD seems to be quite synchronous [...]
> I think your terms are backwards (JiffyDOS is very timing dependent,
> like RS232, which is called asynchronous operation, but I know what
> you are saying.
I tend to disagree. ;-)
The sending of a single byte in JD is synchronous, as it highly depends
upon sender and receiver being synchronous, knowing exactly what should
happen at which instance of time. This is called time-triggered, or
synchronous.
Anyway, JD is asynchronous in when each byte is sent. You cannot know
beforehand when it will be sent. There is some "mark" that a transfer
starts, and it starts afterwards. This is asynchronous, or
event-triggered.
The same holds for RS232 (asynchronous operation): Every byte is sent
asynchronously, thus, you cannot know beforehand when a byte will be
sent. Anyway, as soon as it starts, the bits are just sent
synchronously. You need a synchronous system to catch every sent bit,
that is, the timing must be correct. That's the reason why you have to
setup the bps rate on the receiver site beforehand. (RS232 is something
special is the receiver can determine with bps rate with the help of the
start bit, if it knows how much start bits there are. But the details
remain.)
> on the trace I have:
>
> time 0 data goes high to signal new byte. (not sure which side does
> this. I think receiving side)
> time 7 first 2 bits
How much time is there in between? 7 us? I will never be able to react
this fast on NT if I do not know beforehand when the data byte will be
started on the transmitter. Well, I could, but this would mean this I
must block the complete NT system for a non-trivial amount of time.
(Something like wait-for-listener comes to mind).
Or is there some "block handshake" before a first byte, which is not
that timing critical?
Regards,
Spiro.
--
Spiro R. Trikaliotis
http://www.trikaliotis.net/
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.