You can see my feeds at
I'd be grateful for any comments or criticisms.
If anyone is working on consumers for the people Data API I'd be really interested in some joint effort.
There's no authentication, but I'm taking the same approach as we did for FOAF. If it's available as public HTML, I'm showing it in the XML. At some time in the future I hope to extend this with oAuth or AuthSub so that the owner of the data can get more data out.
One of the (many) things that is missing here is a GUID for people. The People Data API and the examples seem to show something like
<id>http://www.ecademy.com/feeds/people/1<id>. As long as the trailing ID number is unique
for the domain this ends up being a globally unique GUID.
What this doesn't do is to match http://www.ecademy.com/feeds/people/1 with
http://www.orkut.com/Profile.aspx?uid=11056323781708313949 who is actually the
What the foaf people did was to use sha1(mailto:emailAddress@domain.com) and named it mbox_sha1sum. It's assumed that any one email address (used for identification) and hence it's hash, maps to only one person. And the Hash obfuscates the email address just enough to get over email privacy issues. And further that someone's primary email address will turn up somewhere on their profile on every social network they belong to.
The other big problem I'm grappling with is that the Google spec and definition of a person is really trivial and doesn't go nearly far enough. And we really need people to extend the spec in standardised ways that developers working on Apps that consume this data can make sense of it.
[ << An answer to Gabe Wachob: Google's Open Social: The missing pieces? ] [ Refining the OpenSocial Definition and more speculation >> ]
[ 09-Nov-07 11:26am ] [ G ] [ # ] [ Google , opensocial , ecademy ]