Re: Suggested revised text for HTML/XML report intro

On 16/08/2011 11:32, Anne van Kesteren wrote:
>
> Your principle is wrong. HTML is not repaired; processing just does
> not stop.

I think that this point isn't really well enough explained (or at least
not well enough sunk-in to the xml community)

The view that html markup has errors which may be repaired (which was
really the sgml/html4 view) doesn't really fit with the html(5) parsing
model which is simply designed to parse anything and produce a document
tree.

Even the current report draft says

> Where HTML goes to great lengths to defined how an agent must
> recover  from markup errors,

Which is really referring back to the earlier model.

I don't agree actually with John that XML5 is impractical. XML error
reporting works great when it is reporting errors to the person who can
do something about them. (Like compiler errors, having the process stop
until things are fixed is what you want). Browsers are not in that
position; so personally I'd be quite happy for browsers to use an "xml5"
style parser for anything servered as application/xhtml+xml (or in fact
anything served as application/xml or text/xml), so long as the parsing
that was specified met the constraint that well formed xml parsed to the
same tree using "xml5" oas using xml (1.x). If the xml5 parse produces a
possibly useful document instead of a yellow screen of death on non well
formed input, that is a good thing in a browser context.

David


-- 
google plus: https:/profiles.google.com/d.p.carlisle

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs. 
________________________________________________________________________

Received on Tuesday, 16 August 2011 11:32:00 UTC