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

Re: HTML5 script start tag should select appropriate content model according to src

From: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
Date: Mon, 30 Apr 2007 00:37:40 +0100
Message-ID: <46352C44.3030207@googlemail.com>
To: Philip & Le Khanh <Philip-and-LeKhanh@Royal-Tunbridge-Wells.Org>
CC: Henri Sivonen <hsivonen@iki.fi>, tina@greytower.net, David Woolley <forums@david-woolley.me.uk>, www-html@w3.org

Those who think keeping applications out of HTML-next is feasible need 
to realize that their quarrel is not with the WHATWG draft per se, but 
with the WG's charter that offers "A language evolved from HTML4 for 
describing the semantics of documents and applications on the World Wide 
Web" as its main deliverable:


I don't personally think excluding applications from HTML would help 
anymore. But it's crucial for end-users to have an official reference 
documentation for marking up documents in HTML that does not expose them 
to complex functionality intended for application developers.

/Pace/ Philip & Le Khanh, I think HTML should have a semantic emphasis 
precisely in order to ensure authoring and reading HTML is a reasonably 
non-skilled and non-specialist task. Using presentational markup means 
authors and readers need to understand both semantics and typography. 
The testimony of websites including tiny print, poor contrast, and 
scrolling and flashing text suggests even professional web designers 
often fail to grok the later.

But, regardless of technical merits, I think HTML-next is likely to 
include a core set of presentational elements because that idea has 
broad support beyond the key figures of WHATWG:



So I doubt the presence of presentational markup within the WHATWG draft 
would be a good reason to reject the draft wholesale.

The best way to reduce the use of presentational markup would be to 
create tools that do not generate it and other tools that benefit from 
semantic alternatives.

Benjamin Hawkes-Lewis

Philip & Le Khanh wrote:
> Henri Sivonen wrote:
>  >>   My answer to the above is yes. If the tools, which people do 
> appear to
>  >>   root for, become better. I would without hesitation say that an 
> author
>  >>   who takes the trouble of manually writing markup both should and 
> ought
>  >>   have learnt to use the features of the language.
>  >
>  > That seems like a moral argument.
>  From my perspective, a totally rational argument.  How
> many Fortran programmers believe they can ignore the
> syntax, semantics & morphology of the language yet
> still get useful results ?  C programmers ?  Java
> programmers ?  Any-language-you-care-to-mention
> programmers ?  Answer : almost none.  Yet you seem
> hell-bent on believing that HTML authors deserve
> to be treated differently (specially, in fact).
> WHY ?  Just because they are "authors" and not
> "programmers" ?  That argument just doesn't hold
> water.  If a real author chooses to typeset his
> book, then he'd better learn a great deal both
> about typography and about the package he intends
> to use.  If he doesn't, then he can confidently
> expect to produce rubbish.  It may be well-written
> rubbish, but it will be rubbish from a typographic
> perspective.  So why should a web author be treated
> any differently ?  If he /does/ insist on doing
> the markup himself, rather than employ a professional
> to do it, then he can either (to use Tina's words)
> "learn to use the features of the language" or he
> can expect to produce rubbish.  Document markup is
> a skilled and specialist task : it deserves (and
> needs) a language that that allows the person performing
> that task to express himself accurately and clearly.
> Once the markup is complete, another skilled and
> specialist task emerges : converting that markup
> into a beautiful and accessible web page.  They are
> not the same skills, and there is very clear received
> wisdom that they need totally different languages.
> Keep HTML clean and simple for document markup, and
> leave presentation to CSS and its friends.
> Philip Taylor
Received on Sunday, 29 April 2007 23:38:22 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 30 April 2020 16:21:02 UTC