Re: DRP and content identifiers.

Hi Yaron,

Thanks for the pointer. I was aware of the earlier work done 
by Jeff Mogul. It is referenced from the DRP proposal because
it validates a lot of the observations that we have made during
our work on differential updates here at Marimba. I wasn't aware
of the HTTP implementation proposal which is described in Jeff
Mogul's paper. 

Jeff's proposal is similar to the solution proposed by DRP. Although
there are a lots of similarities, there are also some differences. 
The DRP proposal suggests some new HTTP headers, like the
"Delta-base" header, we suggest the use of "Differential-ID". This,
combined with the "Content-ID" header gives us a way to make delta 
caching work on existing HTTP/1.1 servers (the details are described 
in the DRP proposal). Unfortunately, it appears that defining
new headers is not acceptable to the HTTP working group, so we are 
moving away from that.

One important difference is that we would like to introduce 
the notion of globally unique entity tags. That will allow
intermediate proxies identify resources in the cache by content
rather than just by name and version. We have found that there
are significant benifits when resources from different locations
can be substituted when entity tags match. This is often the
case in software distribution where many applications may contain 
the same libraries. Although Jeff's proposal doesn't address this
specifically, it seems that it would be compatible with globally 
unique entity tags.

By the way, besides notational differences it seems like the
two proposals achieve similar goals. I would be happy to use
Jeff's proposal, if it is accepted. However, I hope we can get
globally unique entity tags incorporated into it.

Have fun,

	Arthur van Hoff