[Tsvwg] Re: [e2e] What's the benefit of out-of-orderprocessi

[Hilarie Orman]

One could regard the pointer as a "covert channel" and interpret it
as data.  

Hilarie

>>> "David P. Reed" <dpreed at reed.com> 09/18/01 08:45AM >>>
The urgent pointer delivers no out of band data, except for the pointer itself.
What it does do is tell the receiver where in the upcoming stream it needs 
to "read" to in order to get urgent data.  The TCP recipient still needs to 
acknowledge all bytes up to that point - thus the stream remains 
ordered.  In the design of TCP (I designed the urgent pointer mechanism in 
1977 or so, as part of eliminating the entanglement of TELNET functionality 
with TCP), out-of-band data was deliberately not supported, since the same 
function could be achieved more simply and generally, by opening one or 
more parallel TCP connections for the out-of-band data.


At 06:36 PM 9/17/2001 -0700, Jacob Heitz wrote:
>The urgent pointer does not point at a range, only at a single
>byte. It can only deliver a single byte of out-of-band data.
>
>Sam Liang wrote:
> >
> >   Also, doesn't TCP's urgent pointer offer a means to deliver out-of-band
> > data?

- David
--------------------------------------------
WWW Page: http://www.reed.com/dpr.html