On Tue, Jun 11, 2013 at 05:44:38PM -0400, Richard Brooks wrote:
> This lead me to start thinking about the possibility
> of deploying something like Fidonet as a tool for
> getting around Internet blackouts. Has anyone tried
> something like that?

Usenet has long since demonstrated the ability to route around
amazing amounts of damage and flakiness and to maintain communications
over very slow (including sneakernet) links.

Arguably, that sentence describes the normal operational state of the
network on a typical summer day just like this one, 30 years ago. ;-)

Usenet has some very nice properties for applications like this:

1. There is no centralization.  Thus there is no single target to
shut down or block.

2. Messages are not addressed to individuals.  This frustrates
some traffic analysis.

3. It's transport-agnostic.  Messages can be passed via IP, via UUCP,
by USB stick, CD, DVD, etc.

4. It's highly delay-tolerant.

5. It's content-agnostic.

6. It's highly fault-tolerant.

7. It doesn't require real-time IP connectivity.  In areas where
IP connectivity is scarce, expensive, intermittment, wiretapped
or blocked, this is a big plus.

8. It's standardized.

9. Mature open-source software already exists for it.

10. Peering relationships can be ad-hoc.

Not that it would work for this application as-is: the article
duplication method would need to be replaced because the current
one leaks origin information.  But I think that's a solvable problem.

I submitted a proposal on this very point a few months ago; haven't
heard a thing back, so my guess is that's not going anywhere.  But I
think with a relatively modest investment, the additional code could
be written and a testbed network constructed to figure out if this
really is a viable architecture.  My hunch (of course) is "yes" but
I'd prefer to remain skeptical until there's some experimental
evidence that it'll hold up under the kind of duress we've seen
in various countries during the past few years.

---rsk
--
Too many emails? Unsubscribe, change to digest, or change password by emailing 
moderator at compa...@stanford.edu or changing your settings at 
https://mailman.stanford.edu/mailman/listinfo/liberationtech

Reply via email to