Re: New issue: error recovery practices (Re: Proposed TAG Finding: Internet Media Type registration, consistency of use)

On Fri, 2002-05-31 at 18:56, Keith Moore wrote:
> not useless - they do serve as a specification that implementors 
> at least consider attempting to adhere to - but this prevents neither 
> bugs nor proprietary extensions nor failure to implement new features
> in a timely fashion.

Sure, but we've got ALL of those problems right now, and in spades. 
Perhaps uselessness is in the eye of the beholder?

> otoh, a requirement in the specifications to change functionality in 
> a way which causes more pain to users (e.g. forbidding browser 
> interpretation of improperly-labelled content) is highly likely 
> to be ignored.  

I have to disagree.  Those kinds of "pain to users" rules are in fact
standard practice for XML work pretty much across the board.  (RSS is
about the only place I can think of where HTML-style looseness is even
occasionally permitted.)  They make it very clear where things are going
wrong in a given communications process.  When recipient software howls
(and I don't mean sending a polite message that has no effect on the
transaction), senders listen.

Maybe it's too late for text/html, but perhaps the transition to
application/xhtml+xml is an opportunity to make clear that solid Web
architecture can only be built on trustworthy foundations.
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!

Received on Friday, 31 May 2002 19:37:59 UTC