What about using XML Signature ( http://www.w3.org/TR/xmldsig-core )?  This standard can be used not only to prove who is sending a message, but also that the contents message have not been altered by a third party during transmission, which is another common security concern not mentioned in this post.

Simply put, this is done by hashing the contents of a referenced XML block, and then encrypting that hash using a known algorithm/key (private key).  Then on the other side you hash the XML block and decrypt the signature (public key) - if they match then you know the message is unaltered and sent by that party.

The only issue is whether VOS still uses XML messaging at all, or is it now *pure* binary?


On 3/27/06, Peter Amstutz < [EMAIL PROTECTED]> wrote:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Mon, 27 Mar 2006, Hugh Perkins wrote:

> Peter,
>
> Disclaimer: the only thing I'm sure about the problem is that I havent fully
> understood it ;-)
>
> The problem appears to be a stable way of identifying each machine?  Could
> you use an IP address and port as seen by an (internet-based) STUN server?

That would be better than what we it does now, yes.  However it relies on
there being a STUN server that everyone agrees upon to use, which doesn't
really scale.  Also it doesn't really address the other
multiple-protocol/multiple-networks problems I outlined in my mail.  One
of the goals of VOS is for the client/server protocol to be essentially
the same as the server/server interconnection and sychronization protocol.

Actually, a little bit of background might be relevant here.  VOS was
originally conceived as a primarily peer-to-peer system.  The world server
would act merely as a broker which stored the static world geometry and
then linked to the connected clients.  When you joined the space, you
would connect to each client and then exchange updates (position changes,
talk message, etc) directly, without involving the server.

This turned out to be unworkable, for a variety of reasons but starting
with the fact that lots of people are behind firewalls which don't allow
incoming connections.  Thus we have turned towards a more traditional
server-centric model for the end user client.  However, since by
definition world servers are accessable, the peer-to-peer design is still
very useful and relevant -- it allows you to create grids of world
servers, or to knit several servers together to serve a single world!

The expectation is that VOS would form a dense grid of connections between
clients, servers, appliances, hubs etc.  Keep in mind that these are
persistant connections, which is why it is so important to be able to
associate a link with an existing connection: unlike HTTP, they are
stateful connections.

[   Peter Amstutz   ][ [EMAIL PROTECTED] ][ [EMAIL PROTECTED]  ]
[Lead Programmer][Interreality Project][Virtual Reality for the Internet]
[ VOS: Next Generation Internet Communication][ http://interreality.org ]
[ http://interreality.org/~tetron ][ pgpkey:  pgpkeys.mit.edu  18C21DF7 ]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFEJ4uOaeHUyhjCHfcRAjKtAJwJRHtvPiw+OHqy9YsQODEaX3DKogCePEf+
WU//Cm26Q1V8gn0RCumGcvc=
=1ImB
-----END PGP SIGNATURE-----


_______________________________________________
vos-d mailing list
vos-d@interreality.org
http://www.interreality.org/cgi-bin/mailman/listinfo/vos-d

_______________________________________________
vos-d mailing list
vos-d@interreality.org
http://www.interreality.org/cgi-bin/mailman/listinfo/vos-d

Reply via email to