There’s a growing trend for the web to become a (the) platform to be consumed by webapplications themselves. Users can use one login for different sites, sync information between different web applications, import/export information to/from a site. All this is related to identity management and data portability, and on a second level to digital communities/social networks.

Skype proposes that community building applications must have a defined set of areas/scenarios upon which they can interact to facilitate the desired interoperability between them. Here they are (skype journal):

Social Stack’s Six Zones of Interoperability

* ID (Account lifecycles, Login)
* Sync (Profile, Contacts, Objects)
* Permission (Policy, Licensing)
* Find (People Search, Discovery, Gatekeepers)
* Action (Group Actions, Relationship Actions)
* Now (Alerting, Presence)

The idea is that there can be one single sign in (openid), there must be a standard way to sync information between applications - eg. if i export my contacts/friends from facebook to hi5 and then remove that friend in facebook does it get deleted in hi5?, how about the other way round? -, to find people between apps - If i have a friend in facebook and i also have a myspace account, could myspace alert me that my friend is in the network as well? should myspace do it? Maybe the friend wouldn’t like it to because he stores his work colleagues in facebook and his closer friends in myspace. These are questions that must be solved for dataportability to become a reality. Also check Robert Scoble’s post on this topic.

Dataportability also poses the question of the unecessary duplication of content around the web. If a have a blog in wordpress is it really necessary for myspace to store and sync with wordpress my blog posts? Isn’t that just making things difficult, ie. 2 servers now have the same data and must sync this between them, how about if a user changes a post in the wordpress blog and another user changes the same post in myspace, when syncing which version wins? Ouch.. version control management. These are real roadblocks to dataportability.

In some cases it may be simpler to just allow for data access. In the example above, why not just let myspace access the wordpress blog through an rss feed and every change made in wordpress immediately gets reflected in myspace? So instead of dataportability - taking my data from one place (exporting) to another place (importing) -, why not just simple data access?

Or maybe we need both.

Maybe dataportability and data access have different use cases? You want your data to be portable between competing/equal services and you also want to share it with different services. Eg. i want to export my photos from flickr and import them to imageshack and i also want to show my photos in my wordpress blog whether they’re stored in flickr or imageshack.

Ultimately it will be up to the users to decide if they want to take their data with them or just share it. Developers just need to work out a way to make this possible.