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

Re: Semantics (was : Formal Recorded Complaint)

From: Robert Burns <rob@robburns.com>
Date: Mon, 6 Aug 2007 15:36:14 -0500
Message-Id: <209E3517-38B4-4B6A-88C9-ABB261E7D3DE@robburns.com>
Cc: "Philip Taylor (Webmaster)" <P.Taylor@Rhul.Ac.Uk>, Leif Halvard Silli <lhs@malform.no>, James Graham <jg307@cam.ac.uk>, Public-html WG <public-html@w3.org>
To: Andrew Fedoniouk <news@terrainformatica.com>

HI Andrew,

On Aug 6, 2007, at 3:05 PM, Andrew Fedoniouk wrote:

> On Aug 6, 2007, at 11:38 AM, Philip Taylor (Webmaster) wrote:
>
>>> Leif Halvard Silli wrote:
>>>
>>> This argument here doesn't solve «real problems». We are all in  
>>> favour of accessibility. And we []
>>> are also all against semantics for the sake of semantics.
>>
>> No we're not.  Some of us (myself included) believe that
>> semantics for semantics [sake] is fundamental to high-
>> quality markup.
>
> Previous orator meant that "semantics is for human" but not
> "human is for semantics".
>
> Human consumes information from the web page visually -
> directly or with the assistance of Accessibility tools.
> In both cases "semantics of markup" is relevant at some
> pretty small extent to the human/observer. That
> human is not seeing your markup and doesn't want to.

Perhaps you just miswrote here, but just in case this that's not the  
case, there is a very important need for clarification here. The idea  
with HTML is that humans do not "consumes information from the web  
page visually — directly or with the assistance of Accessibility  
tools." That is how screen readers work with RTF TeX and other non- 
semantic constructs. With HTML, the author creates content in a  
device independent, universally presentable way, with a separation of  
concerns (semantics from presentation). This is a very different  
approach: and it cannot be stressed enough. If the distinction isn't  
clear, then continue the dialogue to make sure it becomes clear.

> Semantics is valuable for web developers and scanning robots.
> In their case (semantics == high-quality markup),
> battle for semantics is a fight for the tag system that is
> more compact/expressive so is more understable
> and manageable - more semantical if you wish.

It is also more universally accessible and more device independent by  
design.

> For the consumer of some
> software application it does not matter what
> language/technology was used inside.
> It is the programmer who can assess the
> sematical beauty of C/Java/Ruby/Python/etc.
> code.

It may not be apparent to the user that it matters, but it does  
matter. Having tools that promote best programming practices  can  
make the difference to users in whether the bugs they face or feature  
requests they make will ever get any attention by the application's  
developers. Programming best practices can also  influence whether  
the program is ever accessible to disabled users or whether the  
program is even effective and usable by able-bodied users. These best  
practices all have effects that  matter to the consumer of a software  
application. So much so  that some user's may buy or not buy a  
software application based on it's underlying development practices:

> I would say that CSS is more semantically valuable
> for human than anything else like HTML or any
> other markup language.

Presentation is always necessary to convey semantics (whether that's  
handled through CSS or not). However, the thought of CSS without any  
semantic language to style sounds completely useless to me.

Take care,
Rob
Received on Monday, 6 August 2007 20:36:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:03 GMT