glennjones.net

Microformats and portable social network

Over the last few months there has been an ongoing thread of conversation about centrally managing your identity for social networking sites using microformats. In the last couple of days both Derek Fetherstone and Jeremy Keith have blogged about their frustration with having to re-enter the contact details and relationship links and the subject has re-surfaced.

What many people seem to want is a portable social networks, we are all fed up with social networking as endless one-way streets for data. As the designer of one these “fad-sites” I have spent a great deal of time thinking about this issue.

I really agree with Jeremy’s summary about the direction of some of the comments

“A lot of people are talking about the need for some kind of centralised service (ala Gravatar) for storing a social network. But surely the last thing we need is yet another walled garden or roach motel?”

As the debate is so lively at the moment, I thought it would be good to outline the concept I am working on for backnetwork.

A publish and subscribe portable social network
What we need is not only a way to prefill social networking information, but also to dynamically and automatically transfer data between systems. At the same time the solution needs to be simple to understand, use and build.

How would a system like this work – The power of rel=”me”
A social networking site would ask for a URL. It could be a page on your blog or hosted for you by a third party. The important thing is that you can control the content.

To this page you add a XFN list of profile pages on other social networking sites. My list looks something like this:

• d.Construct - Backnetwork
• BarCamp Ireland - Backnetwork
• Flash on the Beach - Backnetwork
• del.icio.us
• flickr
• linkedin

The important thing is that we mark the links correctly with the XNF rel=”me”. This will allow any spider to understand that these external pages may have data about you. e.g. http://www.glennjones.net/about/

<li><a href="http://dconstruct06.madgex.com/people/person.aspx?person=glennjones" 
rel="me">d.Construct Backnetwork</a></li>

The spider will then look for contact information formatted as hcard and any XFN relationship links. It will crawl all the other sites and aggregate the results. It will also parse the URL you gave it.
The result would be a number of versions of your contact information from which you could choose and an aggregated list of XFN relationships from all the profiles which use microformats.

RSS like auto discovery
There has been some discussion on the microformats list about using an auto discovery meta tag like the RSS one

<link rel="alternate" title="RSS" href=" http://www.glennjones.net/ rss/" 
type="application/rss+xml" />

We can achive the same goal with microformats. At the bottom of all the pages on my blog you will find the following code.

<address><a rel="me" href="http://www.glennjones.net/about">Glenn 
Jones</a></address>

The spider I am building for backnetwork allows me to just add www.glennjones.net, it then follows the rel="me" link to my about page and on to the other profile pages.

Handling identity through URL or email
Traditionally the web application uses email addresses as a unique identifier for individuals. This convention works well in a closed system where the data is private, unfortunately the public nature of microformats would make this a spammer’s heaven. As such URL is a much better identifier.

Having a URL start point which you control is important as it means the spider will only collect data from sources you specify. It also provides a definitive source, so hcard contact information held at the start point is weighted above data found on secondary profiles.

Issues
There are many issues such as:

• how much contact information people are willing to publish openly on the internet
• the consistent formatting of XFN links
• dealing with duplication and defining primary source
• designing an interface that will help bypass an complexity.

But by far is getting the likes of Linkedin to open up your data into a format that would be useful to you.

backnetwork
I will be adding the functionity to backnetwork. I have already half-built a proof of concept, but need a lot more time to get the idea into the production code.

Navigation

• Home
• Archive
• Links
• About

Copyright ©

Glenn Jones

Creative commons license: Some rights reserved