Re: Packet size s on CCID3

[Eddie Kohler <kohler@xxxxxxxxxxx>]

> I believe you are asking for (3) 's = MSS' to be officially deprecated.  
> I see no need for that.  In general 's = MSS' is conservative (yes?) so 
> there is no need to deprecate it.

In particular, the simulations that show 's = MSS' as extremely 
non-conservative in Widmer et al. 2004 involve drop probabilities that are 
per-packet, not per-byte.  I.e. larger and smaller packets are dropped with 
the same probability.  I don't find those simulations particularly realistic 
representations of today's internet.  They would argue that TCPs could send 
smaller packets to get higher throughput, when in fact the opposite effect is 
generally observed (as far as I know).

Perhaps you would like to help work on an ECN-nonce-verifiable option that 
would implement Virtual Packets.

Eddie