Re: [GNU/consensus] [SocialSwarm-D] Swarm and Consensus

[carlo von lynX]

On Tue, Sep 10, 2013 at 01:49:15PM +0200, Melvin Carvalho wrote:
> I see socialswarm as aiming to make an awesome secure skype replacement, as
> a combination of retroshare, gnunet, secureshare.  This is quite a
> challenge, but if combining forces there may be a chance to gain traction,
> we'll know soon enough.
> 
> I see GNU consensus as more inclusive of technologies such as lorea, elgg,
> gnu social (previously status.net), linked data as well as integration with
> social swarm.  Sort of a challenge to facebook, g+ etc. so a wider scope.

Projects that do not solve the scalability issue will not be able to
implement something apparently simple such as a distributed Twitter.
I mean, yes, they can do it, but when people find out about it and start
using it, it will just crumble apart like Wheetabix. No million user
subscription channels? No Facebook replacement!

As I just posted in "Why the new pubsub & multicast API for GNUnet could
spell revolution" I do not see any of the projects you mentioned taking
on that challenge, so none of them are anywhere close to challenging
the likes of Faceboogle.

> So +1 I think the two efforts are different things.

No, what I think is that at least one functional distribution platform
must be developed. Then everyone can recycle her code to operate on
top of it, as if it were the new openssl or the new pubsubhubbub -
but I don't see how the community can achieve such a huge challenge
twice, especially since many of its promoters still aren't even
perceiving the challenge - understanding why Faceboogle keeps on
winning again and again. Because cloud architectures scale, and at the
same time they are the evil we intend to avoid. There is no such thing
as a safe and globally scalable and trustworthy people's cloud. Or if
there is, it is probably going to be called GNUnet.

See the perspective?

That's why the result from this workshop is basically the same as the
one from the year before: We cannot afford to have parallel developments
going on - we need to work on a common architecture vertically - that is
everyone in charge of his protocol layer.

For example I could envision...

- Briar working on meshnet transports and multiline messaging UIs
- ProjectDanube focusing on an amazing social web interface
- Lorea and Riseup packing tools and making them activist-friendly
- GNU status.net social providing a website generator for life stream channels
- Unhosted developing distributed Javascript apps on top of secushare
- Tox and Mumble teaming up to make obfuscated VoIP conferencing function
- Friendica on an alternative web interface and Faceboogle gateways
- Tor and I2P feeding their long experience into the GNUnet beast
- Melvin & Henry figuring out how linked data can be useful in this architecture
 (probably in the upper payload layers that we are currently so far from having)

Right now it's like all of us are trying to invent SSL, but everyone
differently, and everyone isn't achieving sufficiently good results
to change the world. RetroShare maybe. The way forum posts and chat
lobbies propagate along the network isn't totally different from
multicast - it's just a lot more random.

So back to the vocabulary. From my point of view...

- GNUnet is a routing core, soon with multicast capabilities
- PSYC is an extensible messaging and pubsub signaling engine
- secushare could be an app on top of that
- <yourProject> could be a different app on top of that
- Social Swarm is a meta project making political activism
- consensus the forum where developers fight about details
  or maybe it is the app the way hellekin would do it
- socialnet3 the name of a cluster at the congress and i
  totally prefer the name 'Sovereign Computing' although it
  doesn't hint at the social application dimension at all