Henry Story, formerly with Sun Microsystems, visited the OII recently and spoke about his work on open social networks. We did a short (5 minute) interview that captures the basic ideas behind his vision and development work. It does appear to me that his vision provides the next logical step for social networking. See: http://webcast.oii.ox.ac.uk/?view=Webcast&ID=20100524_323 for his interview, but you will also be able to listen to his full seminar presentation as soon as it is posted on our OII Webcasting page.
Comments or reactions to his proposals would be most welcome, or contact Henry.
Hi Tim.
You say that you are not convinced that social networking as a fundamental web service is interesting. And yet here we are blogging on the open web, which is an decentralised social space. Consider the user experience of me authenticating myself on this blog.
In order to respond to this post I had to login. Before I could login I had to create an account. I was lucky, my usual account name was not used, or I would have spent some time coining a new one. Then I had to wait for the password to be sent to me – in unencrypted form for many intermediaries to see, who could at a later date come and post misinformation in my voice. This is not such a problem in my case, but what for the head of a corporation: what if someone cleverly wrote something in his name, that implied something that would lead people to sell shares? And what is known about me in the end other than my email address, and what I am saying?
So how could the user experience have been improved? We could have used OpenId. But then I’d have to remember a URL to login. No problem for me, but that is a UI challenge. Furthermore the attributes exchanged in an OpenID are limited and not webby: they don’t form part of the hyperdata web. As a result the fact that they are self asserted is a real problem, leading to the growth of an industry that wishes to vouch for attributes claimed by people.
Now with a Web ID, your browser asks you kindly what identity to choose. The operation is a point and click one ( See the screen cast on the front page of http://webid.myxwiki.org/ ) Now at that point I am at a loss as to see how that is not more user friendly that all the previous solutions that people are currently using. Furthermore because the WebID then ties into hyperdata space, things can be proven: for example my FOAF profile could point to my blog, and my blog back to my Web ID. Then a better blog engine User Interface could show Prof Dutton information about who I am, and tick pieces it was able to find further information about and verify. If I happen to have a friend in common with William, it could highlight that. This would allow the owner of a blog to the put me on a list that no longer need to pass the lengthy comment moderation phase. Perhaps the policy could simply be friend of my friends, or colleagues don’t need moderation. So not only have we simplified the UI experience of the end user but we are now also simplifying that of the blog owner too! And of other people who participate in the thread, as the comment moderation cycle will be much shorter. (One could even imagine a policy where friends and colleagues of william can moderate for him)
As for internet cafes, well they are a priori not user friendly. If you don’t remember all your passwords you are stuck. And if you use the same password for every account, then you are a security liability for all your friends, as it will be used as a spam vector.
I mentioned in the talk and in the FAQ http://esw.w3.org/Foaf ssl/FAQ#How_does_FOAF.2BSSL_work_with_public_terminals_.3F
creating temporary certificates is no problem. Then you could create a temporary cert, your server could log it’s use everywhere, and you would have access to all accounts.
I’m not convinced that the vision of social networking as a fundamental web service (parallel to the Universality of e-mail, and enabled by FOAF SSL) is the logical next step for social networking.
Two reasons:
1) Whilst for some people ‘social networking’ is a function and a service, for many people social networks are best thought of as social spaces. I choose to go to this space to meet these people. I choose to go to that space to meet those people. The choice of space is a simple mechanism for managing privacy in an implicit and lazy, but most of the time effective way.
The cognitive overhead of managing the privacy of each individual relationship when social networks become detached from spaces is high – and potentially problematic for many everyday users of social networks.
2) On a technical level the overview Henry gave of FOAF SSL appeared to require very personalised use of computing devices (i.e. each person having their own log-in, loading certificates in their browsers etc) which appears again burdensome to the non-technical user, who has a shared log-in on a family computer / access in an Internet Cafe etc.
I can see FOAF SSL working as part of the behind the scenes interaction between sites – but wasn’t sure what advantages it had over oAUTH type solutions.
This is a question about implementation though rather than vision. So about vision point (2) here could be reframed as asking whether a vision on which transferring content and connections from one space to another is automated is the right one? Or a vision in which simplifying the way individuals can choose to take their content and connections from one page, and re-create them in another, would be more appropriate….
This second vision is achieved by improving data portability, rather than necessarily focussing on connectivity and interoperability between social network services…
Tim, these are good points, and I hope we can get Henry to respond. The difficulties of implementation are real, but should diminish over time, I say optimistically. However, whether people wish to work in more walled or more open social spaces is a real question. I imagine that many people will want to have some social networks that are not confined to a single space. The most concrete example I often face is when I am invited to join a new social networking site by an old friend. I don’t because I have a limit on the number of sites I want to work with. Ideally, I would have a choice of linking with an individual on another social networking site, if I chose to do so, without needing to navigate to that site. That said, some sites will have a particular culture, say a business networking site versus a more social space, and you should be able to keep them separate. As I understand your concept of data portability, I would still be forced to work in multiple spaces. If analogous to ‘number portability’ in the telecommunications space, then it simply facilitates movement, but does not facilitate connectivity across platforms. Also, it sounds like more work, and less connectivity, for my taste.