OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] XLink olden days

[ Lists Home | Date Index | Thread Index ]

> Uche Ogbuji writes:
> > Thanks for the history.  It's a fun read.  I was hoping that it would
> > shed some light on the technical problems the XHTML folks encountered
> > in trying to use XLink.  IOW, they don't really help explain current
> > arguments to me. And why in particular do you think namespaces are a
> > problem in XLink?  The only point I've heard from the XHTML folks so
> > far wrt XMLNS are that the XLink namespace is extra to type.  Surely
> > this isn't what you mean?
> I think the notion is that XLink was originally a toolkit for describing
> linking semantics, which used an architectural forms [based|like]
> transformation to connect those semantics to other vocabularies.  There
> was a vocabulary, but you didn't have to use that vocabulary explicitly
> thanks to the remapping.
> In later drafts, post-namespaces, XLink became just a vocabulary.  To
> use XLink, you must use attributes in the XLink namespace.  While in
> some ways this just a shift from the abstract to the concrete, it's a
> pretty large imposition on vocabularies that already have linking
> semantics (like HTML - not just A but images, forms, objects, etc.).

Why is using attributes with namespaces a bigger imposition than a requirement 
for implicit transform in order to avoid attribute clashing?  I guess it comes 
down to your recent post about not being sure about declarative versus 
procedure.  The namespaces approach is as purely declarative as it gets.  The 
olden-days-XLink approach is declarative, but with a strong hint of procedure. 
 I personally almost always prefer declarative to procedural, and this is no 

Nevertheless, I can understand your preferring the older approach.  I think a 
lot of that comes down to general preference.  What I don't understand is 
where the matter is that is grave enough for the XHTML WG to entirely ditch 

> For yet a different approach to squaring this circle, see:
> http://www.w3.org/TR/xlink-naming/
> I'm not sure I particularly love any of these mechanisms, but I can
> certainly see why the XHTML folks are angry.

Ann hinted that attempts were made to discuss/resolve issues with the XLink 
issue, and that these were rebuffed.  If this is the case, then I could 
understand the WG members' being angry.  This would just tend to indicate that 
the W3C is tending towards uselessness.  But I can't understand "anger" at 
these supposed technical issues.  The XSLT and XPath WGs faced down all the 
namespace issues, real and imagined, and still managed to emerge with 
brilliant specs which have stood the test of time.  Without thinking too hard 
about it, I'd say the only significant wheel re-invention they did was 
xsl:include, which I encourage people not to use (if you really need it, use 
entities).  I am personally disappointed (not angry) that the XHTML WG seems 
to have found the oddest sorts of excuses not to rise to the occasion in the 
same way.

Listen, folks.  XML *needs* a general and credible linking specification.  
XLink is not perfect, but with the emergence of things such as XHTML 2.0, 
there is an opportunity to improve it in the light of practical needs of 
another WG.  The two technical matters Micah brought up that make sense to me: 
the restrictions on the values of actuate and other such attributes, and the 
weasly wording about the content of complex linking elements, should be 
matters for relatively non-controversial fixes.  If the XLink and XHTML WGs 
were not able to sort this out in conference, shame on them.  This is exactly 
the reason that W3C omerta infuriates me: whatever blackguard was standing in 
the path of resolution is safe from the deserved criticism of his or her peers.

BTW, XML Topic Maps was able to use XLink quite well.  It seems their demands would be more exacting than those of a hypertext presentation format.  I have long expressed support/desire for an RDF serialization that uses XLink, and in the little experimentation I've undertaken, I also don't see the insurmountable problems XLink poses.

Uche Ogbuji                                    Fourthought, Inc.
http://uche.ogbuji.net    http://4Suite.org    http://fourthought.com
Track chair, XML/Web Services One Boston: http://www.xmlconference.com/
Basic XML and RDF techniques for knowledge management, Part 7 - http://www-106.ibm.com/developerworks/xml/library/x-think12.html
Keeping pace with James Clark - http://www-106.ibm.com/developerworks/xml/library/x-jclark.html
Python and XML development using 4Suite, Part 3: 4RDF - http://www-105.ibm.com/developerworks/education.nsf/xml-onlinecourse-bytitle/8A1EA5A2CF4621C386256BBB006F4CEC


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS