Re: Keep-Alive Notes

>What I am left with is a design quandary.  The only solution I can think
>of is to include all three possibilities in the initial drafts and then
>have a bake-off to determine which partitions are left in the standard.

Once servers are allowed to keep state and are allowed to perform
pipelining, the need for supporting roll-backability or ACID properties is
not going to be too far. 

The stateless vs. statey vs. how much state is not a new problem. For a good
exposition on various choices and the problems they bring as it applies to
data transfer, you might want to refer to "Transaction Processing Concepts
and Techniques", by Jim Gray and A. Riuter. 

After reading the HTTP-NG spec. and after observing the debate about
keep-alive, I feel that we are walking down a path that other folk
(particularly TP folk) have walked for a long time. I believe that this will
become more and more of an issue as more interactive applications are
developed using HTTP.


Sankar Virdhagriswaran                                    Phone: (508) 287 4511
Crystaliz Inc.                                            Fax:   (508) 287 4512