Web Evolution: Implementors Needed !

My response to:
Web 3.0 Might Be Really Stupid

Given that the first implementations of the activitystrea.ms standard went live about 2 months ago. I think its premature to call the lack of analytics around this new offering stupid. The analytics are evolving and many of those will be feeding into the recommendations engine. There is a lot of opportunity in this field.

As described in this recent article exploring how friends influence purchases
>> "Moderately connected" users exhibit "keeping up with the Joneses" behavior. On average, this social influence translates into a 5 percent increase in revenues.

Additionally, we should not shy away from investigating hyper targetting and ad placement models. After all every item in your activity stream can be considered a personal ad that you want your followers/friends to read. How to make it more memorable ? How to get the user to interact ?

Incidentally, I was just mentioning to Chris Messina last week that the Activities team at MySpace has already gotten initial approval to provide a user's public activitystrea.ms feed in order to encourage adoption. And no, not because no one wants to consume MySpace's feed.

We have several large partners and don't forget the fact that MySpace has more than 70 million total unique users in the US (as of the March 2009 comScore data). However, I do agree a public feed would provide more portability so we are reviewing this proposal.

Bottom line is we are doing everything possible to promote a smarter web and we need everyone's collaboration. Whether its producing smarter feeds or consuming this feeds and creating another consumable a bi-product. Stop pondering on how far others will get and jump on the implementors side.

Collaboration of Opensocial apps via the Stream

Last week at IIW #8 Scott Seely, Martin Atkins and myself worked to add the hooks to the Opensocial APIs (0.9) in order to facilitate developers adding the full fidelity version of their activities into the stream.

The results of our collaborative brainstorming is here:
http://wiki.activitystrea.ms/OpenSocial-Activity-Publishing-Integration

The main concept behind this was to allow developers to reference an optional tranformation bundle which maps the fields in the activity raised to the proper objects in activitystrea.ms like for example the author, verb, object type and most importantly all the properties of such object: unique id, name, annotations, url, thumbnail, etc

The main pros of this approach is that it allows for the activitystrea.ms standard to evolve orthogonally.

All this led me to thinking:
Could we in fact provide such flexibility to our app developers where their products (the activities) could cross reference one another ? After all the ids are gobally unique. WHy could we not have the activities from one app reference the object of another domain ?

Imagine this scenario:

John Mayer Vanderbilt show on May 22nd 2009 photo stream

• Deb loves the new look on John >> from IFan Just Now
• Monica is wondering why John takes so long ! >> from Ticketberry 2 minutes ago
• John is getting ready to go on stage >> from IRockstar 4 minutes ago

Similar effects have been achieved with machine tags however there is no need for that given that the activities stream provides global unique identifiers. Surely one can produce a consumer which can leverage this and properly group based on the object id ?