Sender: nate AT cartsys DOT com
Message-ID: <36BA72AF.362B4D93@cartsys.com>
Date: Thu, 04 Feb 1999 20:25:19 -0800
From: Nate Eldredge <nate AT cartsys DOT com>
X-Mailer: Mozilla 4.08 [en] (X11; I; Linux 2.0.36 i586)
MIME-Version: 1.0
To: djgpp AT delorie DOT com
Subject: Re: [Semi-OT] Making a transaction.
References: <3 DOT 0 DOT 6 DOT 32 DOT 19990204200656 DOT 00809c20 AT pop DOT netaddress DOT com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp AT delorie DOT com

Paul Derbyshire wrote:
> 
> I've become stumped by a signaling problem.
> 
> Say we have two machines on a network, A and B. We want to send a message
> from A to B, but we want, within a short time after the fact, for one of
> two situations to be the case:
> 
> 1. B received the message successfully and A and B performed some
>    action to do with the message that needs to be atomic, such as an
>    electronic funds tranfer (as an example).
> 2. Neither A nor B performed the action.
> 
> The basic problem then is sending a message and having both sides know for
> sure that either the message succeeds or it fails.

That sounds a lot like the problem TCP addresses for IP (which is an
unreliable protocol; packets may go missing at any time).  I don't know
much about how it's done, but you could read up on networking theory, or
read the RFC's (not that I recommend that).

Btw, maybe you should join some other newsgroups; some of your posts are
getting awfully off-topic, and would probably have a better chance of
getting meaningfully answered where people are experts on their topics.
-- 

Nate Eldredge
nate AT cartsys DOT com