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

Re: relevance of diverse HTML authoring practices [was: Versioning re-visited ...]

From: Philip Taylor (Webmaster) <P.Taylor@Rhul.Ac.Uk>
Date: Mon, 25 Jun 2007 17:41:48 +0100
Message-ID: <467FF04C.501@Rhul.Ac.Uk>
To: Maciej Stachowiak <mjs@apple.com>
CC: Philip & Le Khanh <Philip-and-LeKhanh@Royal-Tunbridge-Wells.Org>, Dan Connolly <connolly@w3.org>, Roger Johansson <roger@456bereastreet.com>, HTML Working Group <public-html@w3.org>



Maciej Stachowiak wrote:

> 
> If I understand your proposal correctly, it is to have a specification 
> for what browsers should implement, and then a separate real HTML5 that 
> is completely uninformed by the browser specification. 

Correct.  The specification for the putative language HTML 5
should, IMHO, be based not on what either current or future
browsers do / are expected to do, but rather on what the
current received wisdom tells us is both desirable and necessary
in a hypertext markup language intended for use in the early part
of the 21st century.

In practice, I would expect such a language to be very
firmly based on HTML 4.01 Strict, which represents the received
wisdom that obtained during the final part of the 20th century,
augmented by information that was either unavailable at that
time or which  the current perspective has significantly changed
since then.  Any proposal to add an element to those in HTML 4.01,
or to delete an element therefrom, or to modify the syntax or semantics
of an existing element, would need to be fully justified in terms of
"desirability" and "need" rather than by merely citing
"widespread current practice" [1]

> It seems to me, then, that this second specification [the
 > spefification for HTML 5] would be an excercise in pointlessness,
 > since by design browsers would not support it, and therefore content
 > would have no reason to follow it.

"Since /by design/ (my italics) browsers would not support it"
is a demonstrably false premiss.  Nowhere did I (or do I) suggest
that HTML 5 should be designed in such a way that browsers would
not support it : rather I suggested/suggest that browsers be
required to support it, rather than it supporting them.

To do the converse (as Maciej appears to propose) would be
like designing a computer programming language based on
the ?mis?behaviour of current compilers rather than by
standing back and researching what is /required/ of the
language.

>> The way I think of it, the living, existing World Wide Web is the dog 
>> and written specs about how it should work are the tail. I imagine you 
>> may be thinking of it the other way around.

"The living, existing World Wide Web" consist largely of tag soup,
occasionally interspersed with little gems of well-written,
full conformant, fully accessible, HTML or XHTML.  To write
any specification based on how current browsers parse and
render tag soup is, IMHO, an exercise in futility.

Philip Taylor
--------
[1] I am sure no-one on this list would advocate that
"accommodation" should be spelled "accomodation" in the
future simply because Google statistics shew that
"accommodation" is mis-spelled in this way in about
16% of all relevant web pages.
Received on Monday, 25 June 2007 16:46:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:01 GMT