Re: Federation protocols

Hi there,

Dnia piątek, 31 maja 2013 o 17:32:16 Simon Tennant napisał(a):
> On 31 May 2013 17:09, Darrell Prince` <prince.darrell@gmail.com> wrote:
> > I also hear a lot of pushback on Facebook. Zuckerbergs recent involvement
> > in some hare brained conservative PAC doesn't help. People want to get to
> > something better. but there needs to be something out there noticeably
> > better, 3 steps ahead, and provide more advantages than, better friend
> > requesting.
> 
> Betamax was technically better than VHS.
> 
> I know I'm the downer here - but I'm trying to be realistic: people will
> not go through the friction of switching for a small feature change. A new
> product has to be an order of magnitude better to have users switch.
> 
> Here's how I see the future of federated social networks playing out:
> 
> I start with the following assumptions:
> 
>    1. unique features drive user adoption (Instagram)
>    2. that federation and open are just features that some developers see
>    as useful / we don't sell a solution by pounding these down someone's
> throat 3. the majority of developers care about quickly building a
> solution that works and gives them existing libraries and tools / any
> solution will need to make their life easier than using the Twitter of
> Facebook API and SDKs.
>    4. developers do care about building on APIs and protocols that don't
>    get removed from underneath them (Twitter API example from Michał).
> 
> With that in mind here's how I see us getting open and distributed social
> networks adopted:
> 
>    1. we build great easy to use tools, libraries and provide fantastic
>    documentation using existing established standards (eg Activity streams,
>    XMPP)

And that is PRECISELY the point I am arguing for all this time. We do not HAVE 
such an established standard!

>    2. define the additional protocols for each of the functions of social
>    networking (follower management, post management, following management,
>    media sharing, inbox to client synchronisation) and work them through
> XSF and W3C committees.

Yes, yes indeed!

>    3. build great reference implementations (we're not build a clone of
>    facebook) of specific features - eg meda sharing between domains.

Exactly. MediaGoblin[1] could add social integration. Drupal could add social 
integration. As could many others.

[1] http://www.mediagoblin.org/

>    4. blog extensively about building on these tools.
>    5. Other developers that are less concerned about federation and
>    openness start building on these tools and creating their own apps.

Here's the problem. They won't as long as there are no users there. We need to 
get users on the network. That's a chicken and egg that I think we can tackle 
once we have the protocol:
http://rys.io/en/88 [yes, already posted here, sorry]

>    6. now we start to see an ecosystem around the apps and users are using
>    their one identity to sign into existing social federated apps.
>    7. at this poing perhaps someone comes along and writes an app that
>    looks like a facebook wall - and it has more of a chance of
>    working because users can reuse their existing social login from all the
>    single use apps already written.
> 
> At no point have we started out to create a facebook-like wall and copy
> what facebook does. Instead we provide a great framework for developers who
> have a problem building on the existing social APIs.

I am not advocating creating a "Facebook killer" directly. I am advocating 
that we need to create a single, workable protocol that would give the 
developers a chance of building interoperable, federated solutions so that 
users get an alternative to walled gardens.

-- 
Pozdrawiam
Michał "rysiek" Woźniak

Fundacja Wolnego i Otwartego Oprogramowania