- From: Jeffrey Mogul <mogul@pa.dec.com>
- Date: Fri, 14 Nov 97 14:43:27 PST
- To: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Larry Masinter:
This is off-topic on this issue, but there was some question as to
whether a delta coding was best though of as a 'range' or an
'encoding'.
So, can you delta-encode a range? I mean, does the same argument
hold?
This isn't exactly off-topic, since the reason that I realized that
there is an ambiguity with respect to Range + compression is that
a group of us are trying to figure out how to write a Delta-encoding
spec, and the same kinds of issue comes up.
For those of you who don't know what "Delta-encoding" means, see
Jeffrey C. Mogul, Fred Douglis, Anja Feldmann, and Balachander
Krishnamurthy. Potential benefits of delta encoding and data
compression for HTTP. In Proc. SIGCOMM '97 Conference, pages
181-194. ACM SIGCOMM, Cannes, France, September, 1997
or check out the expanded (& somewhat corrected) version at
http://www.research.digital.com/wrl/techreports/abstracts/97.4.html
Anyway, delta-encoding is most certainly NOT best thought of as a
"range"; the algorithms that people actually use are really not
describable in this way.
I think it could be described as "content-coding" (this is in fact
what we proposed in the paper). However, it is definitely of interest
to be able to apply Range retrievals together with delta-encoding,
and we are beginning to wrestle with how to actually specify that.
But for the purposes of HTTP/1.1, please don't think too hard about
delta-encoding; it will only confuse the issue. I.e., I don't think
it is necessary to solve the delta-encoding problems in order to
figure out how Range interacts with compression. Delta-encoding is
not exactly analogous to compression, and it would tremendously confuse
things to pretend that there is an exactly analogy.
-Jeff
Received on Friday, 14 November 1997 14:47:15 UTC