[p2p-hackers] Google cheating on TCP slow start

Ian G iang at iang.org
Sat Nov 27 15:44:50 PST 2010

On 27/11/10 9:59 AM, David Barrett wrote:
> Interesting article:
> http://blog.benstrong.com/2010/11/google-and-microsoft-cheat-on-slow.html
> I know a lot of people on this list are interested in this topic.  But
> I'm curious: if all sites were to start adopting *ahem* "alternative"
> congestion strategies like this, would would the real-world
> ramifications be?  Indeed, it seems reasonable to assume that before
> long it'll be a standard Apache option to do what Google does.
> Is this the end of the gentleman's internet?  Should ISPs detect and
> block/throttle this behavior -- essentially punishing (or overriding)
> this type of behavior to re-establish normalcy?

One could say that, as an emerging force eating away at the web.

The problem is a layering one.  TCP/IP was a fabulously successful 
product but is now showing the end-of-life blues.

As p2p apps have discovered, the proper software engineering is to 
bypass the TCP/IP + TLS + HTTP bloat completely and develop a custom UDP 
security protocol.  This is relatively easy to do, and I would argue 
this results in better performance (at all levels including human coding 
costs) than trying to improve TCP/IP/S.

At some stage, those with browser and server access [0] trying to 
improve the HTTPS experience are going to realise that what is needed is 
a complete bypass.  They'll hack up a secure datagram mode in the 
server, and get Chrome or IE to flip across to it by means of some trick 
in DNS or HTTP headers.

Hopefully they'll do it correctly, and eliminate all the pointless 
negotiation about security policy in TLS.  It's all on, all the time. 
There is only one mode, and it is secure :P  One True Cipher Suite!

 From what I've seen when I've done this, resiliance goes up an order of 
magnitude, and speed for ordinary stuff doubles [1].


[0] yeah, same two players.
[1] it wouldn't solve the congestion issue, but nor should it.
p2p-hackers mailing list
p2p-hackers at lists.zooko.com

----- End forwarded message -----
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE

More information about the cypherpunks-legacy mailing list