Re: validity and other important things

Charles McCathieNevile 08-04-14 13.43:   ­  
> On Fri, 11 Apr 2008 19:34:42 +0200, Dave Singer <> wrote:
>> At 7:57  -0400 11/04/08, David Poehlman wrote:
>>> It's a reasonable approach but I'm not sure it should be designed 
>>> into spec.
>> and as a point of principal here, it seems kinda backward.  On 
>> questions of validity, we would surely want the output of *automated* 
>> HTML production to be 100% valid, yet in some sense allow people to 
>> exercise their judgment.
> Sure. It is reasonable to want everything, but it is not generally 
> likely to happen that we get everything we want. In which case we go 
> about engineering a solution that provides the optimal available outcome.
> In that context, validity is not a holy grail - we first have to have 
> some understanding of what the purpose of validity is, beyond some 
> feel-good factor for developers.

I - ah - feel that the wish to be 100% valid is a result of the 3 x 
"validate progaganda", as Henri called it. If more tools will validate 
because alt is not required anymore, then authors will still perceive it 
as if it is the tool that has improved. And that is also how it will be 

We know what the shops do when the VAT gets reduced. "We've reduced the 
prices", they proclaim.
leif halvard silli

Received on Monday, 14 April 2008 12:04:50 UTC