W3C home > Mailing lists > Public > public-html@w3.org > September 2012

Re: Issue 30 (Was: RE: Getting HTML5 to Recommendation in 2014)

From: Roy T. Fielding <fielding@gbiv.com>
Date: Thu, 20 Sep 2012 15:51:52 -0700
Cc: Maciej Stachowiak <mjs@apple.com>, Adrian Roselli <Roselli@algonquinstudios.com>, "public-html@w3.org" <public-html@w3.org>
Message-Id: <8D22A6CC-D89B-49CA-98A2-5F71F2F1106D@gbiv.com>
To: Sam Ruby <rubys@intertwingly.net>
On Sep 20, 2012, at 2:54 PM, Sam Ruby wrote:
> On 09/20/2012 05:41 PM, Roy T. Fielding wrote:
>> Whether or not the top 10,000 web sites home pages is a target market
>> for longdesc is not relevant to the definition of HTML.  The Web is
>> not that shallow, and HTML is expected to handle everyone's needs.
>> I don't know why it is even being considered a rational objection.
> 
> It is relevant given the details of the proposal being offered:
> 
> http://www.d.umn.edu/~lcarlson/research/ld-rendering2.html
> 
> (Changes are marked with <ZZZ>...</ZZZ>)

I am not seeing the relevance.  Those are examples of how a UA might render
a longdesc, if such an option were chosen by the user.  There doesn't seem to
me to be any implication that it is one of the top 10,000 web sites home
pages (or any home page, for that matter -- it looks like a scientific
paper is being used as the example), nor does it seem to prevent the same
rendering being applicable to a home page if that does occur.

>> Maybe someone should just ask TimBL is he thinks HTML should be limited
>> to the needs of the top 10,000 home pages?  It would save time.
> 
> That's not the right question.
> 
> If longdesc is not intended to be used on home pages, then I would have expected the proposal for longdesc to indicate that it would be a conformance error for it to be used in such places.

Why would you have expected that?  In the entire history of HTML, I am
not aware of any case that would call for such an expectation.  Rarely used
attributes can be used within any page that an author might want to use them;
doing so is not a conformance error.  That does not imply there is anything
wrong when 10,000 home pages don't happen to use them.

>  And such a requirement would likely be more general than just simply home pages.
> 
> As the people who are advocating the longdesc proposal that has been presented support both the rendering section that I point to above and strenuously object to conformance errors being produced by longdesc, I can only conclude that others may disagree with your assessment on whether longdesc is intended to meet everybody's needs.

It seems to me that your logic is based on an entirely false premise,
both in terms of a mysterious need to be applicable to home pages and
an even more mysterious need to have mark-up conformance be based on the
type of page being served.

Regardless, I would strenuously object to any conformance requirement,
on any element or attribute, that is based on what kind of page is
intended.  How would any implementation conform to such a thing?
I can deal with requirements on transmission, such as limiting what
is allowed to be sent in HTML email messages, but I'd never look to
the language definition to limit what I can or cannot have on my
home page.

....Roy
Received on Thursday, 20 September 2012 22:52:15 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:39:34 UTC