[SML] Whether to support Attribute or not?

[Sean McGrath]

[Steve Newcomb]
>
>The value of the nameless attribute (the "tag name" or "generic
>identifier") is only minimally constrained; it can be can be almost
>any token, and its value can be used for any application-defined
>purpose whatsoever.
>
"Minimially constrained" in Unicode-land is a lot more resource
intensive to police than "minimally constrined" in 7 bit ASCII
land. (But that is beside the point.)

>Therefore, if anyone thinks SML can be made to work entirely without
>attributes, they should think again.

"Therefore"? A non-sequiter surely?

>Attributes are of the essence of
>markup, and you can't get rid of them entirely.  Since we can't get
>rid of them entirely, where should SML draw the line, and why?

Mu. I reject the premise of the question. You have not established
a case for retention of attributes yet you are proceeding to
discuss questions that presuppose you have established a case.

>Personally, I'm not easily persuaded that parsing an "=" sign is so
>very difficult that it deserves to be discarded in the name of
>simplicity.
>

Parsing the "=" sign has nothing to do with it. The problem is
that attribute value literals are second class citizens. As
containers for data they are constrained to carrier bag of
character data and entity references.

>If you really want simplicity, why not make *all* the attribute names
>explicit, instead of exempting only the generic identifier ("GI").

Turn it around for a moment. Think of it this way:
	1) elements are elements
	2) attribute values are kinda-sort elements except that they
	are constrained in various ways.
Simplification:
	1) Make all attributes elements

Expressive power lost = 0

>   <z id="z1">

becomes
	<z><id>z1</id>...

>I mean, for example:
>
>   <sml:GI="z">      instead of     <z>
>
>Then it's simpler to process:
>
>   <sml:GI="z" id="z1">
>
>than it is to process:
>
>   <z id="z1">
>
> ...because everything that's space-delimited in a start tag is parsed
>in exactly the same way.  Furthermore, we no longer care about the
>placement, within the start tag, of the generic identifier, because
>
>   <id="z1" sml:GI="z">Hello, simple world!</sgml:GI="z">
>
>can be a perfectly valid start tag for a z element.  That's simpler,
>right?
>

<Sean uri="http://www.digitome.com/sean.html">
Developers Day co-Chair WWW9, April 2000, Amsterdam
<uri>http://www.www9.org</uri>
</Sean>



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)