[Freedombox-discuss] In-the-cloud infrastructure and business involvement

Charles N Wyble charles at knownelement.com
Thu Mar 17 17:52:37 UTC 2011


On 3/17/2011 11:34 AM, Yannick wrote:
>
> Let's clarify one thing: end to end principle is broken by ISP,

It's broken by access network providers. The "backbone"/transit networks 
are wide open.
Well there are peering agreements and so forth in place that we aren't 
privy to, but for the
most part things flow fairly well.

>   it is
> not broken by design.
Correct. NAT was created to extend Ipv4 by many years.

>   In some case they do not give you a real internet
> address, i.e. a world wide IP address (even better a fixed one ; I do
> have one with my ISP),

Right. Several ISPs now hand out an RFC1918 IP address on the WAN 
interface.

> in some other case they filter content, e.g.
> forbidding some protocols.

Inbound or outbound? I have seen a few cases of that.

>   It is because they do act as an administrator
> of the network by taking measure against what you can do with the
> network, against your freedom, not as a service provider.

Right.

> Let's be even more clear: an IP address does not cost anything; you just
> ask to the regulation authority, and if found valid for technical reason
> they give it to you for free.

This isn't true. All RIR have small fees per year for 
assignments/allocations. Also assignments have a
cost that most be borne by operators around the world. Please see 
http://bill.herrin.us/network/bgpcost.html for
a great explanation of this. Routing isn't free. However we could 
certainly spread the cost around to far more end points.

>   The issue with current implementation
> (i.e. IPv4) is there is not enough address for all because the field
> defining the address is too short (something easily fixed in IPv6, the
> next iteration of the base internet protocol).

Yes this is true. Though v6 has far more problems that will have us 
moving to new protocols (probably using same address system) before
we run out of addresses.

> Even more: filtering protocols for security reasons or legal issues?
> They fail: there is much spam and they do not prevent copyrighted files
> sharing.

So true.
>>
> It is not that moving bits is evil, it is that a business model based
> almost exclusively on it wont pay much, and will pay even less in the
> future.

So true.

>   One of the main idea behind the freedombox is mesh wifi, which
> mean we do have yet the technology to let the people moving bits by
> itself at a cost far much cheaper than the 30 euros I do pay each month
> to my ISP in france.

If it can be properly managed and scaled, yes we can build a very large 
scale parallel network that will evolve the internet to where it needs 
to be.

> I do not see the reason for IPv6 tunneling using a centralized service,
> when the freedombox itself can do it. As I told you I do have an IPv4
> fixed address, I can do this tunneling too. I do not see anything
> preventing me to use IPv6 in my local network and be connected to the
> actual internet based on IPv4.

Um.... you need native v6 connectivity for this to work. If you have 
that then great. Tunnel away!

However if you have native v6 connectivity, wouldn't those who are 
network close to you also have native connectivity and not need to tunnel
through you to get to the ipv6 net? For those that need to connect to 
the v4 net, wouldn't they just use the default route that already exists?
Certain economies of scale favor going through a built up infrastructure.





More information about the Freedombox-discuss mailing list