[sc34wg3] Re: [topicmaps-comment] RE: paradigmatic PSIs

[Murray Altheim]

Mason, James David (MXM) wrote:

[...]
 > I believe that anything in the standard is there by definition and not by
 > being addressable. It is therefore not necessary for it to *be anywhere*. It
 > simply *is*. An implementer just writes it directly into a program (assuming
 > the program is doing something that is needs that part of the standard).
[...]

I quite agree. Rather than write a long response to two long and thoughtful
messages, I'll try to keep it short.

What I see being discussed here is the same issue that has plagued the
W3C since its inception, namely the issues surrounding use of names and
addresses (i.e., locations). TimBL threw his steak on the ground on
this topic long ago, claiming that names and addresses are the same thing
(see his "Myth of Names and Addresses" treatise [1]). Many of you have
heard this rant from me before, so I'll not elaborate.

Despite the best efforts of some folks, because the Web has never developed
a reasonable and accepted name resolution mechanism, there continues to be
confusion about addressability, and I think we're seeing that here. Whether
a location is resolvable is important if the resource *needs* to be available
in order for essential functionality to occur. When something like a PSI is
used as a *name* there is little need to resolve it, unless that "infinite
regress" is actually required, and in most cases (IMO) this is distinctly
not the case. Most applications are local to a specific domain, and having
one's laptop send out a webbot to be sure it has merged all its subjects is
the last thing we'd want as a general rule.

[...and I might add that I think a generally accepted, reasonable name
resolution mechanism (like PURLs) is an extremely important development
and likely represents a necessary condition for the next stage of the
Web.]

I seem to remember us discussing this issue during that year of XTM's
birth, and agreeing that this would be an application-level decision,
i.e., that a topic map-based application would allow a user to set the
level of regress, and in some applications the default might be zero
(as in Jim's example, where local use or security issues predominate).

So I hate to fob this one off as an "application decision" I think this
is really the most appropriate way to approach it. Leave it to the users
to set in their applications, not write it into the standard.

Murray

[1] http://www.w3.org/DesignIssues/NameMyth.html
......................................................................
Murray Altheim                         <mailto:m.altheim @ open.ac.uk>
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK

      In the evening
      The rice leaves in the garden
      Rustle in the autumn wind
      That blows through my reed hut.  -- Minamoto no Tsunenobu