W3C home > Mailing lists > Public > public-html@w3.org > May 2008

Re: UA norms for language content negotiation

From: Robert J Burns <rob@robburns.com>
Date: Wed, 28 May 2008 22:14:53 +0000
Cc: "HTML Issue Tracking WG" <public-html@w3.org>
Message-Id: <8F30C871-9C17-403C-B92E-9139F4C2296B@robburns.com>
To: "Shawn Medero" <shawn@db79.com>

HI Shawn,

Thanks for the constructive feedback on this issue. I welcome any  
feedback you have on these issues.

On May 28, 2008, at 9:58 PM, Shawn Medero wrote:

> On Wed, May 28, 2008 at 1:42 PM, Robert J Burns <rob@robburns.com>  
> wrote:
>
>> Issue:
>> UA norms for content-negotiation (presenting users with choices for
>> multilingual users and whenever appropriate).[1]
>>
>> [1]: <http://esw.w3.org/topic/HTML/ContentNegotiationNorms>
>
> After reading through your issue it is not clear to me where one
> should start debating the merits of it. On surface this issue feels
> like it is outside the domain of the HTML.
>
> The problem is described in terms of author needs (always a good
> starting point) but there's not much substance in terms of how the
> HTML 5 specification would need to be changed to in order to address
> the issue.
>
> Is there an existing error in the specification that would hamper
> user-agents from addressing this issue on their own? Does something
> need to be added to the spec? What are the relevant sections of the
> specification? Also, how do existing user-agents handle this, if at
> all? Are there test cases that would demonstrate the problem? A list
> of websites that currently make use of HTTP Content Negotiation would
> be helpful for others to study.
>
> As you note this feature is facilitated through the use of HTTP
> Content Negotiation - so where can HTML 5 provide additional
> mechanisms that improve matters?

This wiki page is definitely too brief right now. With your feedback I  
should be able to improve on it. My goal here is not to supplant HTTP  
content-negotiation nor to change anything about HTML. Instead I'm  
hoping to add normative language to the draft chapter 4 directed  
specifically at web browsers. Perhaps a subsection of chapter 4 could  
address language content-negotiation or even content-negotiation in  
general (e.g., others have raised issues of the need for content  
negotiation for media resource bitmap resolutions). It is indeed  
possible different UAs could solve this problem on their own, but  
after over a decade of HTTP content-negotiation we haven't seen that  
happen. My hope is that we can provide guidance to implementors to  
help bring about better client support for such content negotiation.

>
>
> (I have similar questions for several other proposals you've sent to
> the WG earlier in the day. I figure you are still updating your
> thoughts on this issue, and several others you've talked about today,
> so apologies in advance if you're already pulling these pieces
> together.)

I am indeed still trying to compile together these issues. Most of  
them reflect my experience and deliberations with others in the first  
6 months of the WG and there's a lot of work in putting these all  
together (over 20 separate issues so far).  My intention is to track  
down the emails in the original discussion of these issues and add  
them to the wiki pages. I also hope to state the problem statement /  
use cases more clearly than currently stated in many cases. The  
solutions themselves are simply my initial attempt to address these  
issues and I would welcome better ideas.

Take care,
Rob
Received on Wednesday, 28 May 2008 22:15:57 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:17 GMT