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: David Woolley <forums@david-woolley.me.uk>
Date: Sat, 28 Apr 2007 11:30:00 +0100
Message-ID: <46332228.4000501@david-woolley.me.uk>
To: www-html@w3.org

Tina Holmboe wrote:
> 
>   That's actually the opposite. The concept that "Authors will do
>   idiotic things, so we'll make it easier for them" does not translate
>   into 'user-centric'.
> 
The problem is that if authors *want* to do something, they will either 
find ways of (ab)using existing mechanisms, or the market will provide 
new mechanisms, regardless of the standards.  Unless not implementing 
something puts off most potential users, it can be better to provide 
something that is to some extent controlled.  Think prohibition versus 
licensing of alcohol retail.

Because one, normally, cannot imprison people for silly web authoring 
practices, one has to either make sensible practices more attractive 
than undesirable ones, or support the undesirable ones in a way that 
minimises the resulting harm, without making the authors think they are 
losing something.

If you were writing a standard for something like the C programming 
language, or even writing authoring guidelines for use within a company, 
you can expect people to follow them (although, in the latter case, that 
will depend on whether there is a suitable culture).  If you are writing 
a specification for use on the web, people will only follow it if 
conforms with their *wants*, or there is no alternative that better 
conforms.

(Even with state created laws, people will find the loopholes that allow 
them to do what they want to do, rather than trying to conform to the 
public policy behind the laws.)

If the feature is implemented explicitly, it makes it possible to 
provide a turn off mechanism.  At the moment we are in a sitution where 
many things get implemented with scripting and many sites are unusable 
without it, which means a normal user cannot reasonably disable 
scripting, so cannot turn off any of things done by scripting.  If those 
things were explictly implemented, the browser could allow them to be 
disabled, even though marketing considerations would mean that browsers 
came more or less fully enabled.
Received on Saturday, 28 April 2007 10:30:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:16:09 GMT