W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2008

Re: qvalue *

From: Henrik Nordstrom <henrik@henriknordstrom.net>
Date: Tue, 05 Aug 2008 01:37:48 +0200
To: ryah dahl <ry@tinyclouds.org>
Cc: Julian Reschke <julian.reschke@gmx.de>, Frank Ellermann <hmdmhdfmhdjmzdtjmzdtzktdkztdjz@gmail.com>, ietf-http-wg@w3.org
Message-Id: <1217893068.12221.103.camel@henriknordstrom.net>

On mån, 2008-08-04 at 22:52 +0200, ryah dahl wrote:

> In the case where this might be meaningful, Accept-Language, UA do not
> give the ability of the user to specify something more detailed than
> an ordered list of perfered languages. Accept, Accept-Charset - who
> cares? Is an ordered list ever not enough?

No, as the differen variants on the server do not need to be of equal

This is best explained in the description of Accept. Read that part of
the spec again, then continue thinking on the topic of preference..

   The example

       Accept: audio/*; q=0.2, audio/basic

   SHOULD be interpreted as "I prefer audio/basic, but send me any audio
   type if it is the best available after an 80% mark-down in quality."

Which also means that even if audio/basic IS avilable it does not need
to be the prefferred if the quality of that representation on the server
is far worse than something else (more than 80% mark-down in quality)..

charset does not have the same quality level. But language may, even if
most user-agents currently does not give the user any easy way to
indicate his level of understanding of the language. But many
multi-lingual users read several languages fluently and really want's
the best representation in any of those languages with no strict

If you know the language reasonably well then the original language
something was written in is often the preferred variant even if
translations of varying quality is available.  For example in technical
literature I have a strong prefernce for english over my native swedish
language as english is often the original language and of better quality
than the swedish translations..

So while we are not quite there yet in actual deployments the framework
is there to be used and does it's thin without complicating life for
those who don't care. Don't toss it out.

Received on Monday, 4 August 2008 23:38:35 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:46 UTC