[A86] Re: Async Link Routine

To: <assembly-86@lists.ticalc.org>
Subject: [A86] Re: Async Link Routine
From: "David Phillips" <david@acz.org>
Date: Tue, 12 Feb 2002 05:12:39 -0700
References: <F1142vz0NuQWGhu25aG0001753b@hotmail.com>
Reply-To: assembly-86@lists.ticalc.org


What about doing it the opposite way?  Have the PC side of the graphlink
check before sending.  Does the USB link have the same problem?  You might
also try ditching the traditional protocol and using a timed protocol
similar to RS-232.

> From my experimentation with the graphlink, I have found this sort of
> routine to impossible to write.  From what I can tell, it appears that the
> graphlink "freezes" when it receives incomming RS-232 data, and doesn't
> bother to check for data comming from the calc before relaying the RS-232
> data.  When I try to do this, the "check-before-send" part of the
algorythm
> fails, and the calc tries to send while the graphlink is also trying to
send
> to the calc.  What I end up getting is garbled input.  Event though the
> output timesout, I can copensate for this in software and resend any lost
> outgoing bytes.   But the garbles incomming bytes are lost and are
> unrecoverable.

References:

[A86] Async Link Routine

From: "David West" <bdaddy_mit@hotmail.com>