[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: HXTML5

• From: John Cowan <cowan@mercury.ccil.org>
• To: Stephen Green <stephengreenubl@gmail.com>
• Date: Wed, 23 Feb 2011 14:40:59 -0500

Stephen Green scripsit:

> Elements do not have to have closing tags (rules determine where
> missing tags would have been).

Such rules are *always* going to make the language more complicated, but
no less rigid, than XML is.  In XML the rule is simple: if the tags aren't
matched and balanced, it's not XML.  The HTML rules co-evolved over many
years, depending on the HTML produced by broken tools and naive users,
the rules given by the SGML schema and the basic recovery strategies of
SGML, and the behavior of browsers.  Trying to make up rules to fix XML
today is pretty much trying to invent a kludge in advance of usage.

Show me ill-formed XML plus what the user meant it to mean, and I
can painfully figure out some rules for repairing generic XML, maybe.
Without that, I can't even start, which is why TagSoup has a schema that
defines HTML to it.

> Empty elements do not have to have a backslash e.g. <h> above, in the
> spirit of HTML. They are presumed to be empty as determined by parsing
> rules, as with HTML.

The rule in HTML is that there is a fixed list of element names with
this behavior.  That isn't flexible enough for XML.

> Attribute values do not have be surrounded by quotes.

Except when they do, and that's a more complicated rule than
"they always do", the XML rule.

-- 
John Cowan                                cowan@ccil.org
At times of peril or dubitation,          http://www.ccil.org/~cowan
Perform swift circular ambulation,
With loud and high-pitched ululation.

• References:
  • HXTML5
    • From: Stephen Green <stephengreenubl@gmail.com>