X-Authentication-Warning:	delorie.com: mailnull set sender to opendos-bounces using -f
Message-Id:	<5.0.2.1.0.20020221093735.00a60470@mail.amtelecom.net>
X-Sender:	sboyd AT mail DOT amtelecom DOT net
X-Mailer:	QUALCOMM Windows Eudora Version 5.0.2
Date:	Thu, 21 Feb 2002 09:40:00 -0500
To:	opendos AT delorie DOT com
From:	"Stephen W. Boyd" <sboyd AT amtelecom DOT net>
Subject:	Re: Serial I/O
In-Reply-To:	<00a101c1b73f$5ab02340$c03dfea9@atlantis>
References:	<5 DOT 0 DOT 2 DOT 1 DOT 0 DOT 20020215155258 DOT 00a5d200 AT mail DOT amtelecom DOT net>
Mime-Version:	1.0
X-MIME-Autoconverted:	from quoted-printable to 8bit by delorie.com id g1LEaZ814075
Reply-To:	opendos AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	opendos AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

At 17:27 2002/02/16 +0100, you wrote:
>On 2002-02-15, Stephen W. Boyd wrote:
>
> > I am using a very old version of DR/DOS that is embedded in the
> > ROM of a hand held device.
>
>May I ask you which version you have exactly? I´m just curious...
>Is this an issue of PalmDOS?

It says 3.4 when it boots.

> > Everything works fine unless I lose DSR during a write operation.
> > In that case I am getting the infamous 'Abort/Retry/Fail' message
> > rather than just getting a bad status code returned.  Does anyone
> > know of anyway to suppress the error message?
>
>The proper way would be to install your own Critical Error Handler
>(INT 24h) in your application, of course.
>
>But here are a few other things to try:
>
>Well, "old" is relative, so if you would be using the OpenDOS 7.01+
>(1997-03-21) or newer, you could try COMMAND.COM /N to suppress
>the shell´s Critical Error handler. The default handler in the
>kernel will always assume Fail without displaying a message.
>OpenDOS 7.02 BETA (1997-11-13) and anything newer also supports
>the more standard option /F.
>
>As Alian Mouette suggested, MODE COM#:b,p,d,s,P is also worth
>a try. The ,P (not RETRY=P) will retry instead of timeout.
>
>Finally, if your hand held would be connected with the server
>through a modem, you could use AT commands to fiddle around
>with the configuration of the DSR wire there.

Unfortunately none of this worked.  I finally got around the problem by 
checking the state of the output queue and not writing until there was 
space available in the queue and then only writing enough to fill the 
queue.  Before each write I check to see if DSR is still asserted.  This 
seems to get around the problem.

Thanks to all for the help.

>Hope it helps,
>
>  Matthias
>
>--
><mailto:Matthias DOT Paul AT post DOT rwth-aachen DOT de>; <mailto:mpaul AT drdos DOT org>
>http://www.uni-bonn.de/~uzs180/mpdokeng.html; http://mpaul.drdos.org

Stephen W. Boyd
e-mail: sboyd AT amtelecom DOT net

  ___   O
(_^_)   o
   o--- .
I'd rather be under the sea!

- Raw text -