Data portability

It's some months since I began to meditate about the concept of data portability. You might have read already my concerns about proprietary file formats, but the data portability stuff I am referring to is slightly different; it's not about an specific proprietary program that you use for writing the greatest bestseller ever, but rather about external websites where you participate and contribute with your fair amount of data. And the problem (the need?) is to be able to import and export that data freely, since it belongs to us.

I am not sure what ticked me off. Maybe it was that my otherwise creative mind was beginning to run out of ideas for inventing passwords for each new website I had to sign up to, or maybe it was the captchas plague.

And OpenID came to the rescue. It's not perfect yet, and I still haven't managed to have a good play with it, but I like the idea. We don't confide our passwords to potentially catastrophic websites but we are still able to log in those sites and all that. Looks nice.

A little bit later I got hit by the facebook craze. I must confess I logged in just for poking friends, but it soon got tiring. It demands constant attention, like a tamagotchi, but in a perverted way, trying to steal all your possible data for lucrative data mining thereafter. That was easy to verify; I had a conversation via facebook's messaging system on a very specific topic and all of a sudden I noticed extremely targeted ads showing on the site. It was so accurate that it looked as if they had a mass of slaves personally reading every single message and assigning the best ads to them.

So I wondered: if we weren't to use facebook, where could we go? Not that I need to go anywhere; I am fortunate enough to share my name with only another well-known person (that I am aware of) and if someone looked for me, would find me quite easily.

But what if I built a new application which allowed you to perform all your social network duties, say, something in the lines of UberSocializr, and I didn't want to enter all the data from scratch, but rather import that from facebook? Would that be possible? And what about other sites like bebo, friendster, blablah?

Even more: why are social utilities useful? Which needs do they fulfill? They are useful for non techy users on first instance. They just register and have everything they might need: pictures, messages, videos, games, forums and friends. Oh and party events!

There are already several applications which can do the same, although not in the same seamless, integrated manner. An experienced user can build his homepage which acts as their profile page, with his blog, pictures, etc. But that's leaving the social side out. I thought that the FOAF concept could work for that, but I didn't buy it because you may like to have private friends and not to declare them publicly in your website, and I think that's not possible as of today.

More or less at that time, the DataPortability project popped out. I was happy to find that they presented some of the same solutions I had thought of. I can also say I strongly agree with what one of their members says:

If your service relies on capturing users, then I'm sorry, you suck
(Ian Forrester)

I like the whole data portability concept and I would love to see it materialise in a practical and effective manner soon. It's also good to see that more and more services are not only following but also empowering the trend, for example, allowing the use of OpenID while also acting as service providers for it. That will lower the entry barrier for lots of people which cannot or do not want to mess with technical details to enjoy this new way of doing things, and hence they won't get discriminated.

It is also going to need some evangelism to spread the word about these tools, although they itself could get easier and less cryptic. And I have the feeling that this won't work very well if the people in our near circles do not use these technologies, so I'm curious: are you actively using any of the solutions proposed by DataPortability yet?

Oh and in a slightly off topic way... it can be very fun to play with this data. 2D/3D visualization of data is just the obvious beginning, but if we aggregate stuff from several sources I think we can come up with some interesting results. Hpricot is calling...