[Freedombox-discuss] "What's a Distributed Social Network?" -- the comic

[Matt Willsher]

On 25 February 2011 10:37, stephen white <steve at adam.com.au> wrote:
> I commented on the comic via the buddycloud developer area, and Jonas suggested that I repost here. I
apologise for my first post being a criticism:

> That comic makes it a lot harder than it needs to be. If I was explaining this to my grandma, I would say "everyone's been using email for decades. It works and it's proven... Now Google and Facebook want your information so they want you to use their website, and now we have all of these privacy problems you keep reading about in the news. Buddycloud does the same as Facebook but like email, so there are no privacy problems because it's being done the way email worked for decades"

I think the comic is an excellent piece of work. It is lucid and
explains the concepts well without being preachy.

'Everyone' hasn't been using email for decades and the the email model
has problems which centralised systems can solve: spam, data retention
and search, validation that messages come from the same source. With
more personal information they could even go as far as to address the
authentication entirely.

> That version of the explanation points out the established solid history, then says it's continuing that history. The problem with the comic is that it argues the merits of both cases, making people need to consider one or the other as though they're on an equal basis. They're not. The centralised Facebook approach is the design that doesn't have the history, hasn't proven itself, and has the privacy issues. This is a substantial advantage that shouldn't be wasted by "fair and balanced" arguments.

You mean it allows the reader to draw their own conclusions by
presenting an unbiased view? Heaven forbid!

The centralised Facebook history is a similar model to the
server+terminal. That's very well proven. The idea of distributed
nodes is not at all well proven and has only really existed in any
volume since the mid 1990s. How do you think people accessed email
before desktop PCs? Via a dumb terminal to a central server. Okay,
this was own by an organisation but the data was not under a users
control and arguably they had less privacy protection than they do
now.

The degree to which an individual values their privacy varies from one
person to the next. Let's not preach at any one. It just creates
friction.

 I don't like the current models and would like more control over my
personal information but that's my personal view. A mesh model is an
entirely new bread of model which is proving rather difficult to
implement on any sort of scale.

Let us not view history with rose tinted glasses.