[xml-h] link typing

Bob DuCharme bobdc at snee.com
Tue Jan 21 18:24:45 GMT 2003

At 1/21/03 11:48 AM, Simon St.Laurent wrote:
>jcowan at reutershealth.com (John Cowan) writes:
> >The true XLink representation of "link types" in the sense you mean is
> >xlink:role attributes.
>XLink piles an awful lot of stuff into attriibutes, and roles are pretty

John has a good point, that what I want to see (user definable link 
categorization to ease large-scale link use and maintenance) is there in 
XLink, it's just called "roles" instead of "types." My search of the XLink 
spec was a bit too perfunctory.

Simon, I thought you'd object more to the fact that their values have to be 
URIs. I personally don't have a problem with this; if I'm going to assign 
all of my citation link elements a type/role value of "affirm", "reverse", 
or "concur" and needed to do it with URIs I'd document them at 
http://whatever/citationRoles and give them values of 
http://whatever/citationRoles#reverse, etc.

XHTML 2 has a nice concept of link types, although I don't think anyone 
uses it: the rel and rev attributes, which (according to 
http://www.w3.org/TR/xhtml2/abstraction.html#dt_LinkTypes) can have values 
of Alternate, Stylesheet, Start, Next, Prev,
Contents, Index, Glossary, Copyright, Chapter, Section, Subsection, 
Appendix, Help, and Bookmark. Just think of the possibilities if lots of 
web page links had values for these attributes:   it would make the current 
versions of web crawling look absolutely blindfolded. ('a' elements in 
XHTML 1 had these attributes, but they were defined as being NMTOKENS, with 
no values that I know of suggested for them.)

Bob DuCharme          www.snee.com/bob           <bob@
snee.com>  "The elements be kind to thee, and make thy
spirits all of comfort!" Anthony and Cleopatra, III ii
(NOTE: bobdc e-mail address used only for mailing lists;
please send private e-mail to bob@)

More information about the xml-hypertext mailing list