Using XLink?
April 13, 2006
Are you currently using XLink, especially in content management applications? If so, I would love to hear from you. Please email me.
Posted by Bill Trippe at April 13, 2006 12:43 PM
I would think it's very unlikely that anyone is using XLink for content management for the several reasons:
1. XLink, with no defined indirection mechanism (other than URL redirection at the server level) is not really usable in an environment where you must manage multiple versions of a hyperdocument as they change over time (the "versioned hyperdocument" problem).
NOTE: See the XIndirect W3C note Innodata submitted to the W3C a couple of years ago for an example of a simple approach to adding indirection to existing XML applications.
2. There is little or no generally-available XLink infrastructure.
3. For most authoring applications (i.e., tech docs) XLink is inappropriate because it imposes too many constraints on how link markup must be structured. In addition, the most popular tech doc formats (DocBook, DITA, etc.) already define their own link syntax and, in some cases, their own addressing syntax (DITA).
XLink was designed for Web-based *delivery* and it is well suited for that but for the reasons stated above is not well suited for authoring or management.
As much as it pains me to say so (as one of the editors of the HyTime standard) it appears that most enterprises simply don't require (or can't apply) the level of link sophistication that XLink provides.
That is, regardless of the quality of the XLink design (which is perfectly good in my estimation) not enough people need it enough to drive development of the necessary infrastructure and practice.
Bill,
I don't think things have changed much in the four years since I wrote "XLink: Who Cares"? (http://www.xml.com/pub/a/2002/03/13/xlink.html) for XML.com. Note Tony Coates' comment underneath it, though; XBRL is about the only large-scale application of XLink that goes beyond trivial usage (XBRL traction being a separate issue.)
Bob
Hi Eliot, Bob:
Thanks for commenting. I can't say I am surprised at the lack of traction, but it's too bad. I see needs for this, especially as people distribute management of content across separate systems.
Bill
Comments for this entry have been closed.
I would think it's very unlikely that anyone is using XLink for content management for the several reasons:
1. XLink, with no defined indirection mechanism (other than URL redirection at the server level) is not really usable in an environment where you must manage multiple versions of a hyperdocument as they change over time (the "versioned hyperdocument" problem).
NOTE: See the XIndirect W3C note Innodata submitted to the W3C a couple of years ago for an example of a simple approach to adding indirection to existing XML applications.
2. There is little or no generally-available XLink infrastructure.
3. For most authoring applications (i.e., tech docs) XLink is inappropriate because it imposes too many constraints on how link markup must be structured. In addition, the most popular tech doc formats (DocBook, DITA, etc.) already define their own link syntax and, in some cases, their own addressing syntax (DITA).
XLink was designed for Web-based *delivery* and it is well suited for that but for the reasons stated above is not well suited for authoring or management.
As much as it pains me to say so (as one of the editors of the HyTime standard) it appears that most enterprises simply don't require (or can't apply) the level of link sophistication that XLink provides.
That is, regardless of the quality of the XLink design (which is perfectly good in my estimation) not enough people need it enough to drive development of the necessary infrastructure and practice.
Posted by Eliot Kimber at April 14, 2006 9:21 AM