Re: Re REL/REV links

Murray Maloney (murray@sco.COM)
Fri, 19 May 95 19:56:44 EDT

David - Morris writes:
> On Fri, 19 May 1995, Terry Allen wrote:
>
> > Your example suggests, though doesn't assert, that the sequential model
> > would be the ordinary reading sequence. Here's another example:
> > the book has within it several paths (selections from
> > the contents), which I want to indicate by BEGIN-->NEXT-->END markup.
> > Now I know I could use PATH/NODE for this purpose (in which case
> > perhaps we don't need both PATH/NODE and BEGIN/END). But if I
> > did mark up paths with BEGIN>END, then the BEGIN wouldn't necessarily
> > be the TOP of the entire document (the book).

Yes, there is an implication that there is a sequential
path through a document that utilizes BEGIN/NEXT/PREVIOUS/END
<LINK> values. I certainly would not assert that it would be
the "ordinary" or even the usual reading sequence, only
that it might be if the author/publisher chose to specify
a sequence.

You can't really have multiple BEGIN or END <LINK>s
pointing to different URLs. How the heck would a
user agent deal with that? I'll have to put in a
note to that effect when I republish my proposal.

Agreed that BEGIN would not necessarily be the TOP.
But when the top and beginning of a documemnt are the
same, I am suggesting that you would only need to specify
one or the other.
>
> This is exactly the approach taken by Hiltz & Turoff in their book
> "The Network Nation" (1st edition '78, revised '93). They encourage
> readers with different interests (social vs. computer scientist
> for example) to read subsets of the book AND their preface provides
> the suggested reading order for each of four pathes. I have
> refered many people to this book over the years and all have
> appreciated this effort on the part of the authors. Only one
> example so I suppose one shouldn't generalize but it can work.

Very good! Yes. This is what I imagined PATH/NODE being used for.

After suggesting it publically, I have taken my own advice
and begin re-reading "Literary Machines". Ted Nelson makes
some very good points about hypertext which many people
seem to have forgotten or overlooked -- at least some of
the people that I talk to. A key one is that hypertext
does not insist on non-sequential reading of information,
it merely allows it. Another is that hypertext publications
of his imagined future might have multiple paths through them,
created by the author/originator and also by readers as
they discover their own prefered routes through a docuverse.

Reading Nelson's book has also led me to re-discover that he
had aleady enumerated a set of link types -- incomplete by
his own admission. So, y'all can expect that aside from
some clarifications and deletions in my proposal, there
will be some new link types which I had overlooked.

Murray