RE: Reliable messaging

+1, it can be easily proven. Have done it several years ago.
But, it doesn't mean there can't be useful but imperfect schemes that solve
this problem in some special cases...

Cheers,

-Suresh
Sterling Commerce   



-----Original Message-----
From: Francis McCabe [mailto:fgm@fla.fujitsu.com]
Sent: Thursday, August 29, 2002 4:14 PM
To: Cutler, Roger (RogerCutler)
Cc: 'Christopher B Ferris'; Mark Baker; Burdett, David;
www-ws-arch@w3.org; www-ws-arch-request@w3.org
Subject: Re: Reliable messaging



Yes, the `distributed systems corollary' of the halting theorem might 
be phrased as: "you cannot distinguish a lazy node from a dead node"

Frank

On Thursday, August 29, 2002, at 01:29  PM, Cutler, Roger (RogerCutler) 
wrote:

> I know of mechanisms that, if successful, will assure the sender that 
> the message HAS been received.  I do not know of any mechanism that 
> will allow the sender to know that the message has NOT been received.  
> The ebXML spec most certainly does not.  So I believe that the word 
> "whether" below is inappropriate.
>  
>
> -----Original Message-----
> From: Christopher B Ferris [mailto:chrisfer@us.ibm.com]
> Sent: Thursday, August 29, 2002 3:25 PM
> To: Mark Baker
> Cc: Burdett, David; www-ws-arch@w3.org; www-ws-arch-request@w3.org
> Subject: Re: Reliable messaging
>
>
> #1 in my definition reads:
>
> the ability of a sender to be able to determine whether a given
> message has been received by its intended receiver ...
>
> It doesn't speak of a mechanism, but there are many means of achieving 
> this.
>
> Cheers,
>
> Christopher Ferris
> Architect, Emerging e-business Industry Architecture
> email: chrisfer@us.ibm.com
> phone: +1 508 234 3624
>
> www-ws-arch-request@w3.org wrote on 08/29/2002 04:01:41 PM:
>
> >
> > On Thu, Aug 29, 2002 at 11:48:41AM -0700, Burdett, David wrote:
> > > I like your definitions,  however, they do not address what I 
> think is the
> > > certainty that although you can be sure a message was received, 
> you can
> > > never be absolutely sure that it was not.
> >
> > How can you be sure that a message was received?  Because there's 
> always
> > a chance that the response to a message doesn't make it, and leaves 
> the
> > two parties out of synch (i.e. two army problem).
> >
> > MB
> > --
> > Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
> > Ottawa, Ontario, CANADA.               distobj@acm.org
> > http://www.markbaker.ca        http://www.idokorro.com
> >

Received on Monday, 2 September 2002 23:53:17 UTC