Humanist Discussion Group, Vol. 14, No. 466.
Centre for Computing in the Humanities, King's College London
<http://www.princeton.edu/~mccarty/humanist/>
<http://www.kcl.ac.uk/humanities/cch/humanist/>
[1] From: Wendell Piez <wapiez@mulberrytech.com> (81)
Subject: Re: 14.0449 XML and the Web
[2] From: Willard McCarty <willard.mccarty@kcl.ac.uk> (40)
Subject: hypertext and open hypertext
--[1]------------------------------------------------------------------
Date: Fri, 03 Nov 2000 07:57:15 +0000
From: Wendell Piez <wapiez@mulberrytech.com>
Subject: Re: 14.0449 XML and the Web
Willard,
You asked whether XML would be able to expand the accessible
functionalities of the web to include the kinds of high-powered hypertext
that are the subject of hypertext research, some of which have existed
off-line for years.
I essentially agree with the assessment of Fotis Jannidis. I'd add
[non-technical readers stop here!] --
* Until we have real XML/XSLT/scripting functionality in the client
(preferably with direct support for XLink/XPointer), we are locked into
(X)HTML/Javascript/Java. (What Fotis said.)
* As Fotis indicates, this means that the real data model is held on the
server, which from the point of view of requirements common in the academic
Humanities community, is a poor architecture for a number of reasons. We
need XML-based processing on the client -- not just for linking but for all
kinds of processing. Ironically, the best platform for this currently is in
Internet Explorer/MSXML3. But we need something platform-independent to
assure longevity of our projects and resources (both the data and its
functionalities).
* W3C XPointer/XPath semantics (specifications still in draft), which
should be able to express most or all of the kinds of complex linking
structures the hypertext community has developed, can in theory be bound to
Javascript/Jscript/ECMAscript and run in today's browsers. But it may be
clunky, difficult to implement, and brittle. On the other hand, this
should be enough to serve as a *demonstration* of what is possible to
achieve cleanly.
* Much depends on the ways the browser vendors approach this problem.
Microsoft, we are seeing, likes XML, but is not above wanting to own it; we
can assume the same will be true of complex linking. Netscape is a very odd
case at present; we could be lucky and see this stuff emerge in
Mozilla/Netscape, but it's hard to know when or how. There are also
third-party approaches, pure XML browsers and others (such as Adobe Acrobat).
* Much also depends on how specifications that are currently in late stages
of revision, but not yet completed, materialize. The relevant ones are
XLink and XPointer themselves, and (I'd add) XSL formatting objects. Oh,
and XSLT 1.1, whose Requirements document addresses some scripting issues
related to all this.
Bottom line is, yes XML can and should help deliver advanced hypertext
capabilities, but it's hard to know whether it will be over the medium or
only long term. At the moment, it is all very much "in play" -- like the
U.S. elections this year.
Fotis is also correct that a framework -- a specification of a range of
linking semantics we need, along with the markup/modeling infrastructures
to support it, ideally with at least one reference implementation -- would
be a huge step forward. XLink/XPointer/XSL on their own might be enough to
start the bottom-up work even without XML on the client; developing a
framework would approach the problem top-down (sometimes a very good idea).
Such a specification would also be a good proactive step to head off any
proprietary developments that emerge.
Christian Wittern suggests we consider ISO Topic Maps in this context.
Whatever we do, it is fair to say that it should be able to take advantage
of a Topic Maps framework. But we also need to be able to describe
functional requirements for the kinds of links we want (links with
fallbacks, links with pop-up notes etc. etc.), and how their components map
to an information set such as TEI with or without Topic Maps.
XLink/XPointer/XPath will give us a start on this.
This is a tough one partly because hypertext linking doesn't just mean
linking; it gets us directly into issues about *how* a text is searched,
made accessible, and represented -- the kinds of thing that a purist
Generic Encoding is designed to avoid....
Regards,
Wendell
At 07:08 AM 10/30/00 +0000, Fotis wrote:
>It seems to that for now and for some time to come XML won't
>change the visible side of the net, because most xml users use xml
>on the server but serve html files to the clients. They may switch to
>serving xhtml, the xml conform version of html, but this won't change
>the rather sad state of affairs concerning interoperability of open
>accessible scholarly edition on the net. As long as one cannot
>access the xml structure of an edition from the outside, but has only
>the data chunks, which fit into a browser window, the whole power of
>XPointer, XLink and XPath can't be used.
....
======================================================================
Wendell Piez mailto:wapiez@mulberrytech.com
Mulberry Technologies, Inc. http://www.mulberrytech.com
17 West Jefferson Street Direct Phone: 301/315-9635
Suite 207 Phone: 301/315-9631
Rockville, MD 20850 Fax: 301/315-8285
----------------------------------------------------------------------
Mulberry Technologies: A Consultancy Specializing in SGML and XML
======================================================================
--[2]------------------------------------------------------------------
Date: Fri, 03 Nov 2000 07:57:32 +0000
From: Willard McCarty <willard.mccarty@kcl.ac.uk>
Subject: hypertext and open hypertext
Thanks to a note from David Durand I've discovered that access to the
Hypertext conference proceedings is possible at a much lower rate than the
ACM Digital Library main gate seems to enforce -- via the SIGWEB group, for
which see <http://www.acm.org/sigweb>. There seem to be some problems with
the ACM Web site at the moment, however; these are under investigation.
Back to my rather broad-brush and probably naive questioning. Let me ask
this: are the goals of the Open Hypertext movement realistic for the WWW?
As I understand it, the goal of the movement, principally instantiated in
the Open Hypermedia Systems Working Group <http://www.csdl.tamu.edu/ohs/>,
is to make accessible the layer of hypermedia software in which links and
nodes are defined and their actions specified. This seems to this scholarly
outsider to be exactly what one needs. I don't see how hypermedia can be
*fully adequately* deployed in the service of scholarly resource
construction without the direct involvement of active scholars in the
disciplines of application. One reason for this is that at least for some
time to come we will be inventing new ways of realising our scholarly forms
-- not by thinking them up abstractly, as so often seems to happen at the
CS end of tool-building, but as Jerome McGann says discovering what we do
not know by making it. Which, I'd guess, means that we've got to get our
hands on the definitions of links and nodes.
I'm sure there are loads of problems here, however. One surely is, as John
Bradley pointed out to me in conversation the other day, the problem of
communicating what one is doing. If I build a commentary, say, and in the
process of doing this invent a bunch of link-types and node-actions, the
scholarly overhead in learning how to use my new gizmo will be
discouraging. Yet I don't see how we can arrive at a good working set of
fully adequate types and actions without a lot of experimenting -- by
working scholars. (Ah, it seems I'm in a vicus of recirculation, back to
the question of primitives....)
I have just run across the paper by Gary Hill et al., "Applying Open
Hypertext Principles to the WWW", at
<http://www.bib.ecs.soton.ac.uk/records/1304>; perhaps this will tell me
whether to hope.
Comments?
Yours,
WM
-----
Dr Willard McCarty / Senior Lecturer /
Centre for Computing in the Humanities / King's College London /
Strand / London WC2R 2LS / U.K. /
+44 (0)20 7848-2784 / ilex.cc.kcl.ac.uk/wlm/
This archive was generated by hypermail 2b30 : 11/03/00 EST