Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!zaphod.mps.ohio-state.edu!swrinde!ucsd!hub.ucsb.edu!gnome
From: gnome@lemon.ucsb.edu (Andy Perrine;6512 Segovia #313;961-7466;968-0403)
Newsgroups: comp.os.msdos.programmer
Subject: Interrupt Driven Serial IO...
Message-ID: <8871@hub.ucsb.edu>
Date: 6 Feb 91 04:42:51 GMT
Sender: news@hub.ucsb.edu
Reply-To: gnome@cornu.ucsb.edu
Distribution: comp
Organization: UC Santa Barbara Computer Science
Lines: 30

I've recently been playing with doing interrupt driven serial IO. I've
consulted several books, and actually after getting all sorts of
bizarre bugs, finally have something that resembles a plain run of the
mill terminal program. The odd thing is this: Somehow keystrokes seem
to be interfering with the receiving of data.  Is the keyboard
interrupt somehow interfering with my serial port interrupt? 

Here is an example:
Call local BBS. Ask for long menu of options. Displays perfectly.
Ask for menu again. This time hold down a key while menu is
displaying. This time, I'll notice that two or three characters have
been dropped from the menu.

Another example:
This one is much more annoying.  Call local BBS. Enter a message for
the sysop, typing quickly (not TOO fast, say 40wpm).  Look back at
message, and hmmm... one or two characters seemed to be missing,
almost as if I'd made a typo.  But upon having the BBS list the
message back to me, the characters are there. They just were dropped
by my program when they were echoed back. 

So is this bug related to keyboard interrupts? Do I need to do
anything special with keyboard interrupts to prevent this from
happening? Or has noone ever heard of this sort of thing, and I must
have a bug somewhere else? Whatcha think? -Andy

gnome@cs.ucsb.edu
6600andy@ucsbuxa.bitnet
6600andy@ucsbuxa.ucsb.edu
---