We’ve had enough of digital monopolies and surveillance capitalism. We want an alternative world that works for everyone, just like the original intention of the web and net.

We seek a world of open platforms and protocols with real choices of applications and services for people. We care about privacy, transparency and autonomy. Our tools and organisations should fundamentally be accountable and resilient.


Nicholas H.Tollervey [LibreList] Re: [redecentralize] Thoughts on decentralization: "I want to believe." 2014-08-02 15:59:09 (4 years 11 mons 17 days 08:22:00 ago)
Hash: SHA1

On 02/08/14 00:07, Adam Ierymenko wrote:
> I just started a personal blog, and my first post includes some 
> thoughts I've wanted to get down for a while:
> http://adamierymenko.com/decentralization-i-want-to-believe/

Bravo Adam,

You have succinctly put into words many of the fuzzy "hand wavy"
thoughts I've been reaching for. Reading your blog post has allowed them
to become more concrete.

Regarding peer-(super)peer-peer: when I read this turn of phrase I had
an "aha!" moment. My work/thinking about DHT (via the drogulus project)
has led me to wonder about the nature of hierarchy - when someone or
some node in a network is more important than another. I skirt around it
in my recent Europython talk.

Given the way the Kademlia DHT algorithm I'm using works I expect
several things to happen in this context:

* Individual nodes prefer peers that are reliable (for example, they're
always on the network and reply in a timely fashion). The reliable peers
are the ones that end up in the local node's routing table (that keeps
track of who is out there on the network).

* Nodes share information from their routing tables with each other to
discover who else is on the network and keep themselves up-to-date (it's
part of the process of a lookup in the DHT).

* I would expect (super)peers to *emerge* from such interactions (note
my comments in the Europython talk on hierarchy based upon evidence
rather than architecture).

* If a (super)peer fails or doesn't "perform", the algorithm works
around it - i.e. I expect (super)peers to both emerge via evidence and
for the network to ignore them if or when they fall over or die. This
addresses the "how do we get rid of you?" question from the blackboard
slide in my talk.

Also, I like your use of the word "feudal". I've been exchanging emails
with an old school friend who (surprisingly to me at least) is
interested in P2P. Here's a quote from a recent (private) exchange via
email about the Europython talk:

"Consider my slide about hierarchy and power: it works when the people
with authority derive their power through evidence. Unfortunately,
technology can be used to manipulate power in a way that is analogous
to the way aristocratic power works: "Why are you my King?", "Because
my father was your King!" It's the result of an imposed system
(Feudalism or a certain technical architecture) rather than merit or
consensus of opinion based upon tangible evidence (the king has
authority via accident of birth, the website has authority because of
the client/server model; contrast that to a doctor who has authority
because they have years of training and demonstrate a certain skill -
making ill people feel better)."

Finally, you end with "in the meantime, please do your own hacking and
ask your own questions". For this very reason I'm sitting in my shed on
my 17th wedding anniversary hacking on the drogulus (I have a young
family, a "real" job and organise UK Python community stuff - so I get
very little time to work on it; this needs to change). I'd be interested
to know how you'd see this sort of decentralised / peer to peer work
being funded. The best plan I can come up with is to save money and then
take some months off (likely around March next year).

Once again, congratulations on such an interesting and thought provoking
blog post..!

All the best,

Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/