[Freedombox-discuss] Network effect
On 3/21/2011 3:41 PM, Paul Gardner-Stephen wrote:
> A further comment on leveraging the network effect is to<broken
> record mode>collaborate with various related projects to create a much
> larger combined community earlier than any of us can do alone</broken
> record mode>.
Yep. This is very true.
> My suggestion is to pick something that people have to pay for today,
> that we can offer for free tomorrow.
> To me the obvious choice is telephony: VoIP is already fairly well
> proven, and people hate phone bills.
This is true. It's only recently that we have the necessary
systems/software/low cost per port that allows us
to compete with the traditional telcos in an end to end fashion
(operation and business support systems, network (global SIP platform),
devices (android on open handsets) .
It won't be free though. However it can be delivered at a substantially
I've talked with a few people about building a global scale IP telephony
platform that can compete with the existing SS7 system. We are at the
infancy of technology that can give
us a competitive solution. It's on the drawing board, and I'm convinced
that it can be done in the next 5 years.
We now have google experimenting with delivering a global SIP platform.
Others are working on it as well.
I feel that it's necessary to bootstrap the system, instead of it being
done by google etc.
> This means you have to be able to use existing phone numbers, and just
> have it work.
Hmmmm. Yes this is true. Existing phone numbers, completely new routing
> (This introduces some authentication issues, which I can talk about
> separately, and also some tragedy of the commons issues, which I can
> also talk about, i.e., how to create incentives for people to keep
> their device on and contributing to the mesh).
Well they want to be able to receive phone calls. This would mean their
phone is on and meshed.
I like it.
> The number management has to be completely distributed as well, and
> there has to be support for using the internet to limit the local mesh
> size to deal with scalability.
Yeah. No more NANPA. Though I am curious as to the limits of self
we need some sort of distributed authority? Hmmm. I guess I should be
discussions on the serval project/vt list? :)