> Does XML actually "need" a credible linking specification?
            I'm uncomfortable with the word credible in this context. I
suppose that html's simple hyperlinking is non-credible, when does the
status of credibility accrue to a spec. When it's simply too incredible
for common use? [very very sorry for the pun]


>  Did I misunderstand what XML was supposed to be about? Or has the
vision or reality changed irrevocably? 
            I have the suspicion that it has changed, but I am not
capable or indeed credible enough to define that change in language, I
define it in my gut. I suppose that TAG is supposed to in some measure
define that vision or reality, and generic SGML does not seem to be part
of it from the on and off reading I've been doing. I think the vision
now seems to be XML as some component of achieving on one hand the
Semantic Web and on the other the Programmable Web(web services) and
that SGML to most people working on these things does not come into the
discussion. 
 
> If, as seems currently to be the case, the serving of "generic SGML" on
the Web seems to have been largely ignored by the ordinary > user do we
really need a linking specification for XML?
            Perhaps if it were a linking specification that went towards
uniting the bifurcated visions outlined above. I don't know if XLink
does this, I sometimes feel that XLink doesn't do anything I want. I
think there may be a lot of people out there in the "Real World" who
feel the same. 


[