W3C home > Mailing lists > Public > www-ws-arch@w3.org > January 2003

RE: The other concensus problem

From: Assaf Arkin <arkin@intalio.com>
Date: Fri, 10 Jan 2003 13:01:27 -0800
To: "Mark Baker" <distobj@acm.org>
Cc: <www-ws-arch@w3.org>
Message-ID: <IGEJLEPAJBPHKACOOKHNCEFNDAAA.arkin@intalio.com>

PS

If any one is wondering, I am not aware of a lot of high-level readable
documentation on the subject. The unofficial bible on the subject is
Distributed Algorithms by Nancy A Lynch which covers most of these
algorithm, though some interesting optimizations have emerged since the book
was last published. It's not bed-time reading material, though.

arkin

> -----Original Message-----
> From: Mark Baker [mailto:distobj@acm.org]
> Sent: Thursday, January 09, 2003 7:47 PM
> To: Assaf Arkin
> Cc: www-ws-arch@w3.org
> Subject: The other concensus problem
>
>
> On Thu, Jan 09, 2003 at 03:07:46PM -0800, Assaf Arkin wrote:
> > A lot of the comments made with regards to the HTTP GET/PUT
> approach so far
> > seem to mirror these algorithms,
>
> That's no coincidence! 8-)
>
> > though for the most part I would say that
> > the discussion has ignored some of the safety mechanism that is
> required for
> > reliability, such as round identifiers and failure detection.
>
> That's true, we have ignored those in the discussion.  But consider;
>
> http://www.w3.org/1999/04/Editing/
>
> which is a solution to the "lost update" problem, i.e. where subsequent
> rounds modify the actions of prior rounds.  Etags aren't round
> identifiers, but are state identifiers which can be used similarly.
>
> > Just pointing out that this discussion could be elevated if we
> could look at
> > what was already researched/done rather than trying to
> re-invent the wheel.
>
> Excellent idea!  I think I'm doing just that, only not talking about it
> at that detailed a level.
>
> > Personally, I have actually implemented this stuff and I think
> it's superior
> > to traditional coordination protocols in its ability to address failure.
>
> What do you mean by "traditional coordination protocols"?  Most of the
> coordination protocols I know about are quite aware of these issues.
>
> > There's a whole class of use cases where you would want to use these
> > algorithms, and definitely a good learning opportunity for the
> WS community.
>
> Agreed!
>
> MB
> --
> Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
> Web architecture consulting, technical reports, evaluation & analysis
>
Received on Friday, 10 January 2003 16:02:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:13 GMT