> 
> Weigh this against the alternative:  those with a dynamic IP address
> must run a transient node, which does not offer plausible deniability
> for the operator.
How so?  Running a transient node is just as secure as running a real
node, if not more so since you dont store any data.

> Fingerprint addressing would not be incompatible with the existing
> addressing scheme.  The paranoid could use fingerprint addresses at the
> expense of performance; others could stick with the current addressing
> scheme.
No.  The current scheme wouldn't know how to handle the new
addresses.  Having the option to strip the fingerprints would allow a
'creeping' security hole in that parts of the network would become
lax.  Fingerprint addressing would probably not be a performance problem
either.


> 
> Seems to me that a fingerprint should suffice.  Can you elaborate?  What
> am I missing?
Every node in Freenet is untrusted.  The only trust issues come in when
you have been speaking to a node for a long time and suddenly it appears
that its a different person (doesnt authenticate).  To do that, you have
to have some concept as to what you expect the fingerprint to be,
otherwise you just have a whole shitload of meaningless public keys.

You must have an Address->Key link so that you can say "I expect Address
to have this key", at which point you can verify something.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
URL: 
<https://emu.freenetproject.org/pipermail/devl/attachments/20000816/2ddffd46/attachment.pgp>

Reply via email to