Re: the possible impact of future changes in webfinger (was Re: Anonymity and multiple identities)

On 9 July 2012 10:41, Michiel de Jong <michiel@unhosted.org> wrote:

> Hi Melvin,
>
> my proposal is for fedsocweb to refer to
> http://tools.ietf.org/html/draft-ietf-appsawg-webfinger-00 right now,
> and then to follow the IETF's updates from there. if at some point
> they put "obsoleted by..." on there, then we follow wherever that link
> points us to.
>
> Are you seriously proposing it's better if we fork away? Please
> consider not only the theoretical but also the practical implications
> of such a suggestion. I see no utility in it.
>
> IMHO, fedsocweb forking away from the webfinger effort would lead to
> total chaos and nothing getting built, and i think you know that, so
> it puzzles me a bit why you keep doing this, it seems so destructive?
>

Seems to be 3 groups for discovery right now.

1. Linked data -- broad term used by facebook open graph and others to
discover data on the Web, I generally class this as the set of projects
that's aligned to the web using HTTP URIs.

2. SWD -- Used by OpenID connect, simple HTTP lookup, easy to implement
stable.  Standards quality.

3. Webfinger -- Used by webmail providers to expose extra information.
Used in OStatus.  Less convincing as a standard, at this point, time will
tell.

4. Other -- there's a number of other competing standards that have been
proposed, and I'm sure will be proposed.

It's really up to each individual project to choose their stack and build
projects around that.  One thing I'm certain about is that (1) is going to
be around and growing.  If you want choose 2-4, as an early adopter, that's
a bet your project needs to take.  Given the latest feedback on webfinger,
I find it less convincing than it was when it had a lot of buzz, a few
years ago.

But I do think webfinger (if it gets approved) is going to be great in a
specific context, and that is to get information out of users in hotmail,
gmail in order to give them a better experience on the web.


>
>
> Cheers,
> Michiel
>