The incredible R.V. Guha (who helped create Cyc, RDF and Epinions) has put
together a proposal for Flurry, a Hailstorm replacement.

http://www.alpiri.com/flurry.html

The gist of it is that it allows multiple providers to interoperate and
builds a "network of networks" whereas Hailstorm limits things by making it
all go through Microsoft. A quote:

"Flurry is an interoperable network of user databases. It would almost
instantaneously create something much larger than just Hailstorm can
independently ever hope to become. It would also be a very cool thing."

Here are some comments:

Every piece of information about a user needs an "address" that can tell
us where to find it. Flurry introduces the concept of Data Resource
Locator (DRL) for this purpose. Every piece of data (such as Joe Bob's
wallet or Jane Doe's calendar) gets a DRL.

Why not use a URL instead of inventing a new DRL? It seems like this is a
bit of a waste. Furthermore, I'm not sure why such a location-centric scheme
is used. It doesn't make sense for the use-case I envision. It seems far
more likely that an application wants a piece of data and will attempt to
get it through any means that it can.

I'd tend to imagine a trusted third-party would store all of the important
data and serve it up to sites that were interested. Your vision seems to be
much more distributed, with my data scattered across a thousand providers.

Any way, I'm interested in hearing your thoughts, Guha, and why you went
with this kind of system.

Thanks,

--
[ Aaron Swartz | me@... | http://www.aaronsw.com ]