DigId Again…
Eric does a good job anticipating my replies to his replies about digitalID.
The people I trust who know about such things (including Eric) agree that the Liberty Spec will give users fine-grained control over their identity information. Many of the supporters of it like it precisely because it gives customers a shield against the commercial interests that are without conscience or soul. Having said that, however:
Adopting the Liberty Spec gives users a technical means to protect their ID but simultaneously and in exactly the same way creates the infrastructure by which ID information can be demanded and delivered. So, the defense is via technology but the demand is via economic forces. Guess which will win?
Categories: Uncategorized dw
The liberty spec unfortunately does not adress a customer demand, but rather a merchant demand, where I federate my “AIRLINE” id with my “CAR-RENTAL” id, but not my id anywhere. Its about ‘delivering consumers’, even though the nyms used at the federating sites maybe known only to the user. Its good in the sense that it prevents large scale database building (except by tracking ip’s).
On the other hand, as a customer, I still have to explicitly federate identities at two sites. I still have to enter my credit card info at multiple places. I still have a different shopping cart at different sites. So it changes diddly squat for me. Yawn.
I am not unreasonable and willing to trade some privacy for convenience. But its not clear to me how Liberty helps me here. What I need is user-centric identity, not site centric.
You might object, saying that I want to centralize everything and thats a far worse alternative. I’d agree with you. Federation has its place, but for it to be useful to me, it must be between the spaces my life revolve around, and it must be under my control. I must be able to federate my home, and office, and shopping identities in the direction I want..for eg federate home and shopping to office, at the office identity providers server, witht he same pseudonymous guarantees that Liberty provides. Or I must be able to maintain it separate, like my ‘porn sites’ identity.
Paradoxically, what I am saying is that removing
the identity management to identity managers paid explicitly and retained by the user (or his/her company) may actually preserve the user’s privacy more than the present site-site model, if there is economic incentive directly from the users, rather than from the delivery of users. And therein lies the necessity of compelling apps which make digID relevant to the user, for example, not having to type my info on comments anywhere on blogs, internet wide shopping carts and product lists, ad hoc formation of groups, convenient payment like paypal, etc.
Its all a non-starter until we brainstorm what these apps are, and why me, as a common user would care about digID in general.
couple of things:
1. Liberty’s spec can be twisted to do exactly what you’re talking about.
2. a non-starter? you mean like email? the internet? the typewriter? there was NO “user demand” for these things before they appeared.
No, I mean like pointcast. RSS delievery of news achieves much the same result, but on users terms rather than content provider’s terms. Pointcast was a non-starter. RSS is famously successful. The latter puts the user at the center of the universe, the former the channel. That, IMHO, is precisely the difference between liberty conceptualized the way it is in the tech document and the twisted liberty you are talking about.
The app for RSS was my need to get my info fast and across ‘channels’. There wasnt necessarily a business case for it. Email and the internet solved problems of expensive, non-persistable communication. What does digID do for me, the user, is the question we ought to be asking..
A start might be to reconceptualize the scenario’s in liberty’s tech document along the lines i described in the last post…
I’ll beg to differ with your example Rahul. There were far more copies of pointcast installed on the internet than there are RSS aggregators today. But I believe I understand what you are saying.
Just as the John Perry Barlow Paradox says that all things tend towards the public domain. It seems that in the long haul any software service that requires centralization will fail if it is designed to delivered to maximize business value instead of customer value. This means that all services on the internet will tend towards a Peer-to-Peer architecture over time.