W3C home > Mailing lists > Public > public-html@w3.org > April 2007

Re: Support Existing Content (was: Proposed Design Principles review)

From: David Hyatt <hyatt@apple.com>
Date: Mon, 30 Apr 2007 16:25:22 -0700
Message-Id: <B0350D46-2FB7-4AA8-8909-85A756FD8F65@apple.com>
Cc: Maciej Stachowiak <mjs@apple.com>, "Philip Taylor (Webmaster)" <P.Taylor@Rhul.Ac.Uk>, HTML WG <public-html@w3.org>
To: tina@greytower.net

On Apr 30, 2007, at 3:59 PM, Tina Holmboe wrote:

> On 30 Apr, David Hyatt wrote:
>> I would like the HTML standard to document the reality of the Web,
>> not some idealized imaginary version of the Web.  What good is a
>   Then you confirm that you would like the HTML standard to document
>   poor quality code, bad practice, and lacking accessibility?
>   THAT is the reality of the Web today. Personally I'd like to see the
>   worst mistakes taken out.

Maybe we fundamentally disagree on this point, but to me an HTML5  
document should be a superset of an HTML4 document.  Any HTML4  
document should render correctly as an HTML5 document.

Yes, these features should be documented.  They can be labeled as  
poor practice or as deprecated, but if they are widely used on the  
Web, taking them out results in a spec that can't handle real-world  

>   Or are you seriously telling me that IF we create a HTML 5 which
>   contain only structural elements, no presentational tags, is  
> backwards
>   compatible (except for, yes, presentational tags. It's 2007), and  
> add
>   richer semantics, then the browsers will suddenly and overnight stop
>   implementing HTML 4.01 complete with the current level of error
>   correction?

Yes you could make a compatibility break, but then you would have to  
have two completely different parsing and rendering modes.  This  
would also cause extreme author confusion as many things authors did  
in HTML4 would just stop working when they opted in to your strict  
version of HTML5.  This is akin to the HTML -> XML switch, which many  
people just aren't willing to make because it is too difficult to get  
all their content working again.

I don't see any reason why version 5 of the language should be  
incompatible with previous versions of the language or with the  
existing Web.

>>>   I'm afraid that if we /do/ make HTML 5 what the browser vendors  
>>> are
>>>   willing, able, and eager to implement then we'll not get anything
>>>   /other/ than what they want.
>> What browser vendors want is usually what Web site authors are
>> telling us that they want.
>   Great. When do you guys implement proper kerning? We've screamed for
>   that since 1996. It has nothing to do with the markup, but it has  
> alot
>   to do with readability.

We would like to do proper kerning in WebKit, so that is a bad  
example.  Browser vendors do want to have proper kerning, but at this  
time implementing kerning incurs too great a performance hit, and so  
we have to keep it disabled.  If such a feature were added to CSS, we  
might consider supporting it (e.g., having an opt-in kerning mode).

>   It is far more complicated than how you make it sound. On occation
>   browser vendors add things that may not *be* a good idea, regardless
>   of who asks for it.

Usually such features fail in the marketplace.  There are plenty of  
examples of this.  That is why HTML5 is taking the features that are  
widely used and incorporating them into the language.

Received on Monday, 30 April 2007 23:26:35 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 29 October 2015 10:15:19 UTC