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

Re: Version information

From: Maciej Stachowiak <mjs@apple.com>
Date: Fri, 6 Apr 2007 12:43:20 -0700
Message-Id: <1835EC15-410D-4081-8922-8AEDC0282FF8@apple.com>
Cc: Anne van Kesteren <annevk@opera.com>, "L. David Baron" <dbaron@dbaron.org>, "public-html@w3.org" <public-html@w3.org>
To: Chris Wilson <Chris.Wilson@microsoft.com>

Hi Chris,

On Apr 6, 2007, at 11:17 AM, Chris Wilson wrote:

> Anne van Kesteren [mailto:annevk@opera.com] wrote:
>> Another problem is that user agents incrementally evolve and not  
>> implement
>> a complete specification perfectly in one go. They'll have bugs in  
>> their
>> initial implementation of features, et cetera.
> Look, we absolutely cannot change our behavior in rendering HTML  
> 4.01 as we do today.  It is just a fantastically bad idea, for  
> document authors and users.  As I said:
>> I'm happy to make HTML >4.01 have behavior based on "compromising
>> details where existing implementations disagree," even when that  
>> means
>> IE needs to change behavior - but not in a situation where that would
>> break compatibility with a current web page.
> If the document author is opting into those changes, I have no  
> problem with it.  We cannot have a continuously evolving HTML spec,  
> that no one (document author or UA implementer) can rely on not  
> changing.  That is a recipe for continuous compatibility problems.

I think you are conflating the issue of changes in the standard with  
bug-compatibility for past UA bugs. The intent is that HTML6 would  
not make any changes that break web content relative to HTML5, and  
HTML5 tries not to break significant amounts of actual known content  
relative to HTML4.01.

The only way a UA can promise to really never change anything for old  
content is to never fix bugs in older versions of a specification  
after shipping the first implementation. In practice a browser  
following this policy would never be conformant to anything, since  
it's impossible to ship a perfect implementation on the first try.

Internet Explorer is in a special position with HTML4.01. It has been  
a market leader for a long time, and for many years did not fix  
standards conformance bugs. In the past, Microsoft also actively  
encouraged building MSIE-specific sites. The combination of these  
factors means that a lot of web content depends on MSIE conformance  
bugs and quirks (I should know, I've spent plenty of time reverse- 
engineering them). Even cross-browser content depends on IE quirks  
via browser switches like CSS hacks.

One would hope that Microsoft will not repeat these mistakes with  
HTML5 and future versions, but instead would set out on a policy of  
continuously improving standards compliance except where breakage  
would be too extreme, as the other major browsers do.

Received on Friday, 6 April 2007 19:44:00 UTC

This archive was generated by hypermail 2.4.0 : Saturday, 9 October 2021 18:44:08 UTC