My opinion on the problem of identifying ID semantics

I think that a lont-term solution is to develop an 
independent language for specifying integrity constraints.
That language should be developed on the basis of 
recent research such as Wenfei Fan's works [1] and [2].

Such a long-term solution takes a lot of time, however.  
We need a short-term solution.  It should not cause any 
problems to long-term solutions.

I believe that xml:id is the best short-term solution.  
It is ad-hoc but does the job at minimum cost.  I 
think that the inline declaration solutions sketched 
in the Draft TAG finding document is not powerful 
as a long-term solution and takes too much cost 
as a short-term solution.



[1] http://www.bell-labs.com/user/wenfei/publication.html
[2] http://www.research.avayalabs.com/user/wadler/planx/planx-eproceed/papers/E00-1591816349.ps


Cheers,

Makoto

Received on Tuesday, 13 May 2003 00:18:23 UTC