On Tue, 2010-12-14 at 08:00 -0500, Elliotte Rusty Harold wrote:
> On Tue, Dec 14, 2010 at 5:54 AM, David Carlisle <> wrote:
>
> > The trouble is that HTML5 parsers will parse a large subset of XML but
> > produce radically different parse trees, leading to silent data corruption.
> >
>
> Really? Is that deliberate? I thought one of the goals of HTML 5 was
> to make sure parsers reported the same parse tree no matter what
> garbage you threw at it.
Not all goals can be met :-) -- e.g. for Microsoft I expect the question
they have to ask when they change parsing behaviour of XML or HTML is,
how many internal corporate systems will break, that they are committed
to support.
However, even if they all do the same thing, it's still not likely to be
what humans expect, because HTML expectations don't carry over well to
generic markup. These rules were designed (as far as I can tell)
primarily for broken RSS feeds, which often contain fragments of HTML
(not even XHTML). A better approach would be to treat mismatched tags
as plain text - over time people would fix their news feed software.
Liam
--
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org www.advogato.org
