Collecting Interesting Data for Personal Data Stores
http://glennjones.net/ - blog
http://www.alistapart.com/articles/discovering-magic/ - Article on aggregating social media identities
http://qwerly.com/ - Commerical API identiy discovery
Notes from Amber Case
Examples of his work will be online soon, hopefully. http://glennjones.net/
Showed a drag and drop contact system where new data is entered into the system and then the social graph is pulled in as well.
GlenJones: I really hate how things are stored in databases, and when things break I can't fix them very easily.
He showed a self-editing HTML file that could be changed and saved to itself. He's played a lot with storing these systems.
GlenJones: But when I begin to use the cloud I'll write a system that stores a backup in DropBox when I make an edit to the file.
He built a kind of Tinycontacts that's searchable through the address bar with predictive text.
Adactio: This is kind of the dream of the semantic web -- you visit a site and book travel, and another site tells you that your favorite band is playing where you're visiting during this time.
Glenn: I can actually start to build a larger graph of more nodes and more linkages than just Twitter or Facebook alone can build for me. I can then fire this off against my contacts and ask "who have I contacted recently that I haven't followed on Twitter yet?" The idea of getting data that's coming right from the points of your interaction is that it is more valuable and useful and more telling than those sites by themselves.
I'm using Android - and can I use the history on my phone? I could add that, a text message store, and cross-link it to my contacts, and put the lists of text messages together with the contacts in a webpage.
Copying and pasting of semantic objects back and forth from one site to another.
He uses the example of a contact site with modules of content-editable text area where when you click on an object, it selects the text but hides it, and allows you to past the entire object somewhere else.
He then shoes a website where he can type in a Twitter username and press "discover" and it will come up with all of his different profiles and paths. And then he can clear it. Also, it does natural language processing for the person's address.
Identengine -- quite old in code terms that it's not as new as the Draggables he showed.
DiscoveryMagic - Another one out there.
Quirreley - A social graph API, commercial (Glen would go with this one -- and they're quite keen to use open web standards).
Rapportive - Another one
He uses a vertical mouse that reduces his RSI.