W3C home > Mailing lists > Public > w3c-sgml-wg@w3.org > June 1997

RE: !BEHAVIOR

From: Jon Bosak <Jon.Bosak@Eng.Sun.COM>
Date: Fri, 13 Jun 1997 10:04:01 -0700
Message-Id: <199706131704.KAA09500@boethius.eng.sun.com>
To: w3c-sgml-wg@w3.org
[Andrew Layman:]

| I agree with Mr. Rivers-Moore that one of the principal
| accomplishments of SGML was to separate processing from
| content. Therefore I oppose suggestions that would mix these
| two. Formatting is a kind of processing. All processing belongs
| separate from the document's content, for example, in a DSSSL style
| sheet. ACTUATE and SHOW should not be in XML-LINK (or, if there,
| clearly separated into a processing layer equivalent to a style
| sheet).

This was discussed in great detail some time ago.  The upshot of the
discussion was basically as follows:

1. Everyone agrees that the separation of processing from content is a
Good Thing.

2. Most people think (again, after a great deal of discussion) that
most link behavior should be specified externally, the obvious and
most popular candidate in publishing applications being in the
stylesheet.  This is one of the key reasons that xml-style is an
integral (though optional) part of the XML effort.

3. Point 2 notwithstanding, most people (after a great deal of
discussion!) concluded that you can't have links that interoperate on
the Web without the ability to associate some very general kinds of
intentions with the link.  After a great deal of further discussion
(!) we finally agreed on two axes of specification, SHOW and ACTUATE,
and some very general settings for those axes.  There is general
agreement that, as a practical matter, we cannot do with less than
this and still have links that work for us.

Having put the time in on this that we already have, we are not going
to re-open this discussion for XML 1.0 unless someone comes up with a
completely new reason for taking another look at it.  All of the
reasons advanced lately are very old ones.

Jon
Received on Friday, 13 June 1997 13:04:46 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:04:41 EDT