xml-dev - XLink WD question (steps=n)

["John E. Simpson" <simpson@polaris.net>]

I'd asked this question once before on XML-L, but got no replies. (Posted
it as it happens during the week of the SGML/XML thing in Paris, so maybe
it was missed. Anyhow....)

The XLink WD (19980303), section 5, discusses extended link groups and in
particular how to control the depth of recursion with the steps="n" attribute.
As far as I can tell, though, it doesn’t explain how different steps
attributes in successively-visited extended link groups should affect the
*overall* search for extended link groups. 

For instance, suppose there’s a hierarchy of documents with extended link
groups that looks, schematically, something like this (using the "-->" to
indicate an extended link group):
	A --> B
	     B --> C, D, E
	          C --> G

If the extended link group in A has steps="1," and the one in B has
steps="2," and the one in C has no steps attribute at all, where does the
search stop (assuming it to have been initiated from A)? 

(1) With B? (That would be one step, which would mean that the "master
extended link group" -- A's in this case -- takes precedence over any
succeeding ones dug up in the search.) 

(2) With G? (That would seem to imply that B's steps="2" attribute had
overridden A’s steps="1.") 

(3) Or somewhere else -- perhaps even (assuming that the absence of a steps
attribute for C implies what it seems to imply) going on indefinitely?

Has anyone working on an XLink engine grappled with this yet? Any other
voices that would like to be heard? :)

Thanks.
John E. Simpson			| It's no disgrace t'be poor,
simpson@polaris.net		| but it might as well be.
					| 		-- "Kin" Hubbard

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo@ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo@ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa@ic.ac.uk)