Re: Generic link handling

On Monday, December 2, 2002, 5:03:14 PM, Ann wrote:




>> -----Original Message-----
>> From: www-tag-request@w3.org [mailto:www-tag-request@w3.org]On Behalf Of
>> Chris Lilley
>> >>> Part of the issue is that some folks do not reliably distinuish
>> >> between an attribute of type anyURI, which there can be lots of per
>> >> element, and a link (which can have only one such attribute that forms
>> >> the link URI, if there is to be any metadata also associated with
>> >> that link  using standard and readily recognisable attribute names).
>>
>> AN> Actually I think it's more of a problem of some people assuming
>> AN> that a link is restricted in such a manner (to one attribute
>> AN> forming a link URI). That's NOT been the case until XLink decided
>> AN> to make it so, despite significant outcry.
>>
>> Could you give some examples of a language which does not have such a
>> restriction?

AN> Sure, any language not based on XLink.

Well, no. I asked that question because I was not aware of any
language that has just a single link attribute with no other
attributes along with it.

For example, HTML would not be such an example. The link element and
the a element and the img element are links, sure, but they are
elements and they all have other attributes on those elements that
relate to the link, besides the actual attribute containing the URI.

AN> The problem with the argument asserted above, is that it presumes
AN> a commonly defined semantic for a "link".

Yes. It presumes that a link is a URI plus some associated metadata.
As does, for example, HTML along with every other markup language I
can think of.

AN> XLink has provided *a* definition of a link, more precisely it
AN> defines "an XLink link", and an XLink linking element. [1]

It does, though you don't have to use that element if you have another
element you would rather use.  Besides, I didn't mention Xlink.

AN> It should be noted that that this passage goes on to state:

AN> ..."The term "link" as used in this specification refers only to
AN> an XLink link, though nothing prevents non-XLink constructs from
AN> serving as links.)..."

AN> Given that there's not a larger, global definition of a link, we
AN> cannot say that XLink's interpretation, which has been
AN> controversial, is the definition to be applied to all languages.
AN> That would be like asserting "You don't know how to play chess.
AN> You used to know, be we have changed it so that the queen can only
AN> move forward and backward now."

I don't find that particularly helpful. Amusing, but not especially
accurate.

Instead of trudging through the same well worn ruts to the same dead
ends, how about looking at what a generic XML linking system would
look like?  And as I said, a link would consist of a URI plus
associated metadata.

I don't see anything contentious there; I don't see anything radical
or difficult or dangerous, and I don't see any mention of XLink or
chess in there either.

From that assertion, and moving from vertical-market specific linking
systems to a more generic one, my assertion that

"a link (which can have only one such attribute that forms the link
URI, if there is to be any metadata also associated with that link
using standard and readily recognisable attribute names)"

still seems to stand, unless you can provide a mechanism that would
resolve the ambiguity?


-- 
 Chris                            mailto:chris@w3.org

Received on Monday, 2 December 2002 11:39:57 UTC