Wesley Eddy wrote:
> 
> The draft Adam submitted to the TCPM working group was on only one part
> of Obfuscated TCP that allows data to be carried on the SYN-ACK packet.

I'm sure this was a lively discussion indeed!  Can you summarize the 
major objections to such an approach?

At worst I could see routers dropping these packets, but I can't imagine 
anything really bad would happen (else you could just crash routers by 
sending malformed packets -- something they're obviously designed to 
endure).


> In addition to the questions on design for carrying data on the SYN-ACK,
> there were serious security issues raised with the Obfuscated TCP design
> and its lack of crypto-agility.

Was it tied to a single encryption scheme?  Why not just fall back on 
SSL?  Indeed, I'm not sure why you'd need a payload in the SYN-ACK at 
all: I would think a simple bit saying:

SYN: Hi, let's set up a TCP connection.  Also, I support SSL.
SYN-ACK: Cool, once we're set up, let's switch to SSL.
ACK: Roger that.  Next packet will be SSL; I'm assuming the same of you.

After all, key exchange and all the other SSL junk will take a ton more 
packets than a simple 3-way-handshake.  Perhaps I'll need to read up on 
that spec.

-david

_______________________________________________
p2p-hackers mailing list
[email protected]
http://lists.zooko.com/mailman/listinfo/p2p-hackers

Reply via email to