Re: [GNU/consensus] What is a Federated Network?

[Melvin Carvalho]

On 4 April 2013 13:00, hellekin (GNU/consensus) <address@hidden> wrote:

http://networkcultures.org/wpmu/unlikeus/resources/articles/what-is-a-federated-network/

In this article, Stijn Peeters looks at the evolution of the concepts of
centralized, decentralized, federated, and distributed networks. An
interesting read to figure out where we currently stand. From Paul Baran
(Rand Corporation, 1964) to Unlike Us #3 last month in Amsterdam, he
provides an essential review to dissipate confusion about terms we use
often, and often without realizing their overlap, which can make the
conversation less clear than it appears.

A nice article.

In the social graph you have nodes and edges. 

How you name the nodes is the less important thing when you are building a traditional site.  But how you name the nodes becomes critical when you are building a federated system.

If people name nodes differently, which is what has happened to date, there's no real possibility of federation.

Programmers want to program, but I always say that naming is the hardest thing we do.

Some systems that got naming right are:  the internet via the URL, the web via HTTP URLs, git via the git identifier.  What's notable about these are that they are strong, unambigous an universal.  A git identifier on github, is the same on gitlab, is the same on your local machine.  Sounds easy, but the only people that really have done this even reasonably well on the web are facebook, drupal and maybe wordpress. 

FTA:

" According to the announcement of their Federated Social Web Summit in 2010, federation (in the context of social networks) is “letting people on different social networks follow each other” ... So for the W3C, federation can be achieved through standardization of data formats: the precise connection between nodes and server is less relevant."

This is not an accurate statement. 

Firstly he's confusing one particular community group does not speak for the W3C.  Community groups are there to incubate ideas, and if those ideas are liked, they get voted as RECS by the 400 or so member companies.  Nothing in the FSW ever got that far.

The W3C, from what I have seen, first and foremost advises people to name things well and scalably.  Then to use robust protocols for communication.  It's natural to want to invent your own, but the challenge is to persuade everyone to use is.  The advantage of HTTP is that most systems talk it already, so it's developer, client and server friendly. 

 

==
hk