W3C home > Mailing lists > Public > www-html@w3.org > January 2003

Re: script tag - archive attribute

From: Roland Bluethgen <calocybe@web.de>
Date: Fri, 24 Jan 2003 02:13:17 +0100
Message-ID: <3E30932D.5020304@web.de>
To: Boris Zbarsky <bzbarsky@MIT.EDU>
CC: Stephen Battey <Stephen.Battey@vega.co.uk>, www-html@w3.org

Boris Zbarsky wrote:
> Surely you mean Transfer-Encoding, not Content-Encoding...
> Content-Encoding means the content is actually encoded, should be
> cached encoded, and should be saved encoded if the user decides to
> save it.  Transfer-Encoding means the content is just encoded for the
> duration of the transfer and is decoded immediately upor
> reception....

As to my knowledge, Content-Encoding is regularly used to transmit
content compressed, for example by mod_gzip for Apache.
Transfer-Encoding, if used, is set to "chunked" most of the time, which 
has a certain advantage for transferring dynamically generated content. 
As I understand RFC 2616, both could be used if compressed transmission 
is desired, though the RFC is not very clear about this: OTOH it says 
"the content-coding is a characteristic of the entity identified by the 
Request-URI" (which supports your claims regarding caching and storing), 
on the other hand "a non-transparent proxy MAY modify the 
content-coding", which is somehow contradictory. And it says, the 
entity-body "is [typically] only decoded before rendering or analogous 
usage". IMHO it's arguable that storing a downloaded page on disk is 
kind of "analogous usage".
Received on Thursday, 23 January 2003 20:12:24 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:20:48 UTC