W3C home > Mailing lists > Public > www-tag@w3.org > January 2002

Re: Clark's commentary

From: Tim Bray <tbray@textuality.com>
Date: Mon, 07 Jan 2002 15:43:21 -0800
Message-Id: <5.1.0.14.2.20020107153312.022e68a0@pop.intergate.ca>
To: www-tag@w3.org
At 05:21 PM 07/01/02 -0600, Dan Connolly wrote:

>Meanwhile, I brought up this <e:eacute/> idea up in the
>context of the MathML review of XML schema, and somebody
>objected on the grounds that fulltext search tools
>don't work that way.

I don't buy that.   Yeah, they're going to have to 
put a filter between the raw text and the full-text
indexing engine, but you always need one of those
anyhow.  

Anyhow, the round-trip between the two idioms 
is trivial.  I'd just canonicalize everything into
&#babe; format before I indexed it, and put in the
verbose format on export into an authoring system.

>By they way... I'm really, really sorry I didn't get
>rid of entities back in '91/92 when I first had
>the chance. We could be writing
>        AT&<!>T
>today in stead of
>        AT&amp;T
>if I had done just a little bit more homework.

Bad, bad Dan.  Go to bed with no cookie.  This is a 
horrible sleazy SGML trick that we quite rightly outlawed 
in XML.  Mind you, we had to wrestle CMSMCQ to the ground
first.  There's nothing particularly wrong with 
entity syntax; they'd feel more natural to a lot of 
us if preceded by \ instead of &.  The problem with
entities is what they do, not how they look.

>Hmm... is this the group where we get to re-design
>XML? Whee! 1/2 :-)

It's just irritating that to learn the Really Basic Stuff
That Everybody Uses And You Can't Get By Without, you have
to look in several places.  So James' notion of unifying
the basics and calling it 2.0 is terribly seductive. -Tim
Received on Monday, 7 January 2002 18:43:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:03 GMT