PaceBatch and pipelining

[Mark Baker <distobj@xxxxxxx>]

Kudos to the authors for clearly doing their homework on this one.  I
do, however, have a significant criticism to make.

The use of HTTP pipelining appears to be rejected for the following
reason;

"The fact that POST operations cannot be considered idempotent limits
the usefulness of HTTP 1.1 pipelining for APP, as it means that member
creation requests cannot be pipelined."

That's not true.  RFC 2616 sec 8.1.2.2 only says that non-idempotent
operations "SHOULD NOT" be pipelined, not "MUST NOT".  Moreover, the
problems with batching non-idempotent requests isn't at all specific to
HTTP pipelining, it's just inherrent to any batch mechanism, including
PaceBatch.  That is, if you POST a BatchRequest with two app:Post
requests embedded, you're going to have exactly the same issues as
with two pipelined HTTP POST requests.

So I'm -1 on PaceBatch, but I'd support APP calling out the use of HTTP
pipelining for batching purposes (with a caveat about going against the
SHOULD NOT recommendation from 2616, of course).

Mark.

On Sat, Jun 25, 2005 at 03:38:57PM -0500, kmarvin wrote:
> 
> I've written up two new PACEs for consideration, based upon the the
> atom-protocol thread on support for batching changes and optimized
> categorization [1].
> 
> You can find them at the following location:
> 
>   http://www.intertwingly.net/wiki/pie/PaceBatch
>   http://www.intertwingly.net/wiki/pie/PaceCategorize
> 
> Thanks to all who have provided input, feedback, and help to get the
> proposals this far...
> 
> -- Kyle
> 
> [1] - http://www.imc.org/atom-protocol/mail-archive/msg00932.html
> 

-- 
Mark Baker.  Ottawa, Ontario, CANADA.          http://www.markbaker.ca
Coactus; Web-inspired integration strategies   http://www.coactus.com