[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [DotGNU]Regarding your message "SOAPtalk" roundtable discussi on in

From: James Mc Parlane
Subject: RE: [DotGNU]Regarding your message "SOAPtalk" roundtable discussi on in Zurich
Date: Fri, 10 May 2002 12:05:19 +1000

> >- business models for web services?

Apart from an extension of the current mature? model of the electronic
distribution of raw knowledge for a fee. (Rueters,Bloomberg via Tibco etal.)
A formal schema for web services makes several new models more practical,
simply by making it easier for people to access and inter-operate.

1) A return to centralised specialised processing. eg. Send me a renderman
file via XMLRPC or SOAP and I will send you back a MJPEG frame, which comes
down to the interplay between bandwidth charges vs. the cost of computing
locally. "The granularity problem".

I expect that entire RPC API's for rent will slowly pop up all over the
world. The speed of emergence of this trend will be dependent on the ratio
of the cost of computing locally vs. the cost of computing remotely. So we
will probably initially see the rise of centralised processing centers of
cheaply clustered computers and FPGA based solutions, and eventually see
that technology become more distributed.

I expect that one day someone will invent the "Killer RPC" that everyone has
to use at least once per day and its completely closed. I think this is what
MS wants to do with Passport. So if I was able to invent a RPC that would
for instance factor any integer instantly.. It would be in my financial
interest to sit on the algorithm and charge access to it and not risk
releasing any software based on it into the wild where it could be reverse

I expect at some point someone will set up a sweat-shop where every RPC call
is answered by a real person and .. oh that right.. the pr0_n industry has
done that already... :)

2) The next model is the data-warehouse model. This is where instead of
taking advantage of inter-operability, you take advantage of reachability.
eg. I can get to my data from anywhere in the world. This is the concept of
the global desktop. I believe this was the target of MS's Hailstorm. 

The major disadvantage this suffers from is the requirement of trust on the
part of the end user in the service provider.

Just my thoughts.. anyone else got any opinions..


James Mc Parlane LPAL The MetaWrap Project
irc: Oxd00fd00f

reply via email to

[Prev in Thread] Current Thread [Next in Thread]