Re: R: R: Social API: Scope

hello walter.

On 2014-08-01, 9:25 , Goix Laurent Walter wrote:
> [walter] in general if we want (i am speaking in broad terms, maybe beyond the scope of w3c in this field) to have interoperable - federated - social networks in the future we should keep in mind that they are likely to be managing an identity and a list of relationships of ours. This means that they will have to offer interoperable ways of addressing users, and whilst technically not mandatory, access their profile and/or list of relationships (one could always say that this information stays private and/or implementation-/service provider-specific).
> We all know there are already many ways of expressing this information (profile/relationships), and many URI schemes to address users, and it probably would make less sense to define yet others.
> My question is thus wrt w3c social WG/IG: will the social WG provide any "recommendation" on a specific (set of) formats/URI schemes to use, or will the IG provide any "best practice/guidelines" to suggest the use of specific URI schemes (e.g. acct:) and profile/rels format (e.g. foaf, poco?)? or is this fully out of scope and anyone (including other community/sdo) may be free to combine what will be produced here with these other specs?
> I hope I clarified my point.

i think the key point here is to design things loosely coupled. as you 
pointed out, the main point is that we have to be able to link things. 
that can be achieved via link relations, which we can either reuse, or 
we have to maybe register some new ones.

once we can use well-known link relations, we can leave it to runtime 
mechanisms (conneg) to figure out the concrete media type for a 
representation, so that we have decoupled the ability to link to a 
certain concept, and the concrete format that's used to represent that 
concept.

the if the WG believes there's a need to develop a new format, we can do 
so and it will be nicely decoupled from the way it's linked. as we tried 
to point out in our workshop contribution, our main goal is to make sure 
that instead of creating The One True Model, the WG should try to 
develop components that fit into web architecture. these then can be 
developed and evolve independently, and probably (and hopefully) some 
will thrive and some will die. in our point of view, such a loosely 
coupled approach has the best chances of developing something that has 
the ability to evolve, and can be tweaked over time as required, without 
breaking things.

cheers,

dret.

-- 
erik wilde | mailto:dret@berkeley.edu  -  tel:+1-510-2061079 |
            | UC Berkeley  -  School of Information (ISchool) |
            | http://dret.net/netdret http://twitter.com/dret |

Received on Friday, 1 August 2014 16:36:04 UTC