Its always good to hear Marc Canter speak especially when it's from the front of the room as opposed to bellowing heckles from the floor, and in moderating this panel he was in his element.
A panel consisting of Jonathan Hare from resilient, Chad Dickenson from Yahoo! and Jeff Barr from Amazon discussed with great aplomb the notions of the importance of open, interoperable infrastructures to the ongoing development of the web.
Open was defined as:
- Freely licensed - to anyone
- Freely hostable - by anyone
- Open APIs - to provide extensibility and interoperability
The ideas of what constitutes 'infrastructure' ranged from discussion around the 'undifferentiated muck' that Amazon is seeking to provide though its EC3 to the notions of user data components such as 'reviews' which in many cases are still locked into the systems of the provider.
Yahoo! gave examples of companies based upon its APIs such as Qoop and made the point that now companies are building their systems using yahoo open API's before comming to Yahoo to talk about deals.
There was a lot of discussion over the notion of identify which gave Marc a chance to to outline his people aggregator product but also to challenge the panel on their ideas of where shared identify and related data should or could live.
There was also firm agreement that data ownership should lie with the user/creater and that sucessful systems must allow both imprprt and export features.
The recurrent theme of the session was that openess is its own reward and that companies and systems that fail to open will fall by the wayside.
A panel consisting of Jonathan Hare from resilient, Chad Dickenson from Yahoo! and Jeff Barr from Amazon discussed with great aplomb the notions of the importance of open, interoperable infrastructures to the ongoing development of the web.
Open was defined as:
- Freely licensed - to anyone
- Freely hostable - by anyone
- Open APIs - to provide extensibility and interoperability
The ideas of what constitutes 'infrastructure' ranged from discussion around the 'undifferentiated muck' that Amazon is seeking to provide though its EC3 to the notions of user data components such as 'reviews' which in many cases are still locked into the systems of the provider.
Yahoo! gave examples of companies based upon its APIs such as Qoop and made the point that now companies are building their systems using yahoo open API's before comming to Yahoo to talk about deals.
There was a lot of discussion over the notion of identify which gave Marc a chance to to outline his people aggregator product but also to challenge the panel on their ideas of where shared identify and related data should or could live.
There was also firm agreement that data ownership should lie with the user/creater and that sucessful systems must allow both imprprt and export features.
The recurrent theme of the session was that openess is its own reward and that companies and systems that fail to open will fall by the wayside.
Comments