Encryption on Mailing lists sensless?

[Garreau, Alexandre]

Le 19/11/2014 à 01h31, Robert J. Hansen a écrit :
>> It’s completely true. However Mark’s right when saying it could help
>> to do it client-side...
>
> No.  Client-side, you get to inspect (fully) only your data, and you
> have to develop a statistical model of spam based on only your data.
> When Gmail filters, it inspects (fully) traffic to *millions* of users,
> and uses that to create a model no individual user can hope to match.

You can do some stats on multiple persons using hashes, meshes,
propagation and this kind of thing. Even better: you can do it F2F, and
ponderate according distance in number of hops. See what try to do
GNUnet. That’s way better than large, politically risky and impersonal
large Google scans.

> Encrypting everything, even Aunt Edna's recipe for potato salad, means a
> significant step backwards in the spam fight.  I love decentralized
> algorithms, but there's something to be said for a God's-eye perspective
> on the problem -- look at decentralized route discovery protocols versus
> Dijkstra's algorithm as an example.

We have to make some sacrifices to get freedom. So yes it can and will
be more complex to stop centralize. But it especially involves an other
thinking model: not a big centralistic individual one, but a
*collective* one, where you think “I have a thousand instance, how
should each of these act so that the whole networks work respecting both
Order and Anarchy?”. It’s a lot more complex, but also a lot more
interesting, and potentially a lot more powerful.

>> But the true solution is this one: use only free software, software 
>> you’re sure you can check the sources.
>
> Maybe one user in ten thousand has the skill to audit a nontrivial
> codebase.  Free software is a good idea, but let's not pretend that
> normal users will realize a real benefit from being able to check their
> source code.

One in ten thousand is enough. And anyway: that was the case too about
written language some centuries ago. How could that not change? For
instance a way greatest amount of Emacs users know several parts of its
code source, and are able to inspect any part at any moment if needed.

And the real benefit is in the *freedom to*, which has only to be
express by the ability to do something, even if « everybody » doesn’t
know how, a sparse minority is enough. That’s the concept of free
software.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 948 bytes
Desc: not available
URL: </pipermail/attachments/20141119/b92d778e/attachment.sig>