On 12 March 2013 19:26, hellekin (GNU/consensus)
<address@hidden> wrote:
On 03/12/2013 01:51 PM, Melvin Carvalho wrote:
>
> This means understanding the nature of the URL, HTTP and HTML, in that
> order.
>
> The issue is partly that people (with the exception of mark zuckerberg)
> have pre concieved notions about how this works, and do not understand
> that it has the power to do everything you need.
>
*** Are you suggesting we invite him for a talk? <g>
I believe he invested in diaspora!
Seriously, maybe we need to figure out some "How to leverage the Web in
your next Social Web project"? That would probably include comparative
implementations of basic building blocks:
Figuring out the web is certainly key to scaling.
- how to handle identity?
Be lenient. Encourage URLs.
- how to handle messaging?
Get HTTP POST right before looking further.
- how to handle references?
As in REST?
Saying that an URI can be your identity, messages can be passed as URIs,
and references point to RDF graphs does not seems to be sufficient for
most people to grasp the simple beauty of the URI.
A URI *points* to your identity. Messages are exchanged via a protocol using a serialization. References are pointers to data via the "follow your nose" pattern.
Instead that would require working examples of how things work with
XMPP, OStatus, OAuth, etc. and compare to what can be done when
leveraging the powers of URIs instead.
No that only adds confusion and conflation. Understand how the web works and THEN you an understand how other protocols work. Trying to do it all in one go leads to a mess. Notice that Mark Zuckerberg had nice URLs to describe things first, then he got HTTP working well. Only after the system was in good shape did he add XMPP, email etc.
> People have the tendency to think you need something new to be
> successful. The reality and history has proved the opposite. Making
> something new leads to a local minimum that is rarely interoperable.
>
*** +1
> The majority of successful projects come from cloning something existing
> and putting it in a slightly different context.
>
*** Sounds like the theory of evolution.
So, I gather that dissecting Facebook might help finding out what they
do well. I'm still convinced that the means shape the form, and goals
are deeply intertwined with function. Let's contemplate for a minute
that Facebook got "all the right technology for all the wrong reasons".
How can we leverage that technology, or what technology is there to
unfold that social networking platform for freedom?
A good example was StudiVZ. They simply took the facebook pattern and made it in another language. Take the facebook pattern and make it FLOSS / distributed (it doesnt take much more than a tweak). Then add all the bells and whistles.
If you want to OVERTAKE facebook, look at graph and the open graph protocol and complete the missing pieces by adding data freedom
==
hk