On Fri, 11 Sep 1998, Simon Spero wrote:
>
> On Fri, 11 Sep 1998, Dean Gaudet wrote:
>
> [some stuff I'll get to later when I'm awake - i'll just do the easy one
> here :)]
>
> > > 7.4) Allow connect_and_write (send data with syn)
> > > 7.5) Allow accept_and_read( read data with syn, delay syn-ack).
> >
> > I disagree... but only because I'm a linuxhead and syscalls are cheap ;)
>
> That's not really the point; the point of connect_and_write is to send
> data in the first packet (i.e. with the syn; avoids 1RTT per request).
> accept_and_read makes that data available to the server and lets you send
> the first packet back with the syn-ack if you delay that.
Erm... wonderful theory, but I'm not so sure it holds up in real life with
real life clients.
|