Honest Jim.
Similar issues arrise in our consumption of subscription databases, how do we know that we will have access to the same journals this time next year? Will we still have access to journals older than 5 years? Do we really have to have what the publisher wants us to have? (my eye’s have been opened by Adrienne Muir, lecturer at LU)
Perhaps we need some sort of Electronic Service Level Agreement, that sets out how long information will be available for, who it is available to, what is available, and how accurate we can expect it to be – some sort of margin of error.
If these are made available as an xml schema, services could display the information to users or developers when requested. If the ESLA contained controlled vocabulary, watches could be placed from within software to alert developers and users when something changes.
Consider a scenario where a service is relying on data that needs to cover a particular timespan to be useful. Maybe a Financial site which consumes data from another site to analyse trends over the last x years. if that data suddenly becomes unavialable, the consuming site is going to have a problem, and possibly not be aware of it until lots of users have had a bad experience and complained.
If the consuming site could periodically check an ESLA to see that nothing had changed, then the developers could be alerted and take appropriate action.
I knew it… minutes after posting this I googled and found this:
Heiko Ludwig, Alexander Keller, Asit Dan, Richard King. “A Service Level Agreement Language for Dynamic Electronic Services,†wecwis, p. 25, Fourth IEEE International Workshop on Advanced Issues of E-Commerce and Web-Based Information Systems (WECWIS’02), 2002.
Technorati Tags: web 2.0, truth, data publishing, Library 2.0, perpetual beta, electonic service level agreements