Re: NCSA implementation of KeepAlive

Jeffrey Mogul writes:
	...
 >    (c) The "stored-state" stuff ought to be carried in a new header
 >    named "Stored-state".
 > 
 > -Jeff


As long as we're making iconoclastic suggestions today, I'd like to
suggest that unless someone can put forward some *quantitative*
results that indicate that there is enough of a savings involved in
this kind of connection state such that these savings would not be
easily swamped by, say, one small GIF file, that introduction of this
mechanism into the spec be delayed until there has been a chance to
experiment with it.


--Shel

Received on Friday, 20 October 1995 17:43:51 UTC