Re: [i] XPointer and addressing in OHP-Nav

From: Peter Nuernberg (
Date: Tue 21 Jul 1998 - 09:00:30 CDT

Hi all-

   I think we are mistaking two issues:

  1. The OHSWG should adopt WWW standards. That is, we (the OHSWG) can use the W3C stuff to base our OHS stuff around. This is demonstably false, since (for example) XPointer doesn't define non-XML inter-document locations.
  2. The OHSWG implementations should be W3C standards compliant. Particular wrappers that wrap, say, WWW browsers should know about the neatest, coolest WWW stuff. This is demostrably doable, since many of us have already done it (for most of us, that was probably a previous year's neatest coolest stuff :-)

   Nobody said here that the WWW is evil or bad or that we should ignore it or anything even remotely like this, although attributions of such remarks seem to abound lately. So, since I'm the receiving end of these attributions, let me clarify what I said:

   We can't use XPointer as our LocSpecs. I think we should be careful about using XPointer specs even as a _type_ of LocSpec, since it is by definition tied to one data type, one system.

   OK, note this says nothing about supporting XPointer in those places where appropriate (which seems to me to be the wrapper of an integrated WWW browser). It seems like a grand idea to study XPointer in-depth and see how we can map back and forth between our LocSpecs and XPointer specs. This is how I understood Sigi's first post.

   If we do use XPointer stuff as a LocSpec type, then I'd like to know why we _don't_ use RTF- or PS-specific (very widely used data formats) location schemes as types of LocSpecs. The whole intention of these is to promote interoperability between different systems, different wrappers, different applications. The LocSpecs we use are a good step in this direction. I'm not sure XPointer helps us in this regard.

   Also, on the "we can't survive without the WWW" stuff that's been flying lately. Again, nobody has said this. By the same token, one can argue we can't survive with MS Word, or Emacs, or any other very popular app. This is of course the entire point of open hypermedia - you don't want to have to choose your app - you want to be able to integrate (potentially) any application.

   So, I think we can stop the discussion on whether a particular app should or should not be integrated, since there is no discussion point - we all think all apps that people want to use should be integrated. The relevant points here seem to be:

   (I think these are orthogoinal, although clearly "yes" to the first probably makes the answer to the second more important).


Peter J. Nuernberg
Department of Computer Science, Aarhus University, Denmark

This archive was generated by hypermail 2.1.5 : Tue 13 Aug 2002 - 07:21:02 CDT