This report is also available as an Acrobat file.
Contents
Authoring and Design for the WWW
PRINCIPLES AND THEIR IMPLICATIONS
The future: a better hypertext system?
Web browsers have introduced some valuable hypertext
features to a mass of users, notably the indication of links to documents which
have already been opened, but there is still plenty of room for improvement.
Nielsen (see Web Resources) has proposed a list of
facilities which browser ought to provide, which includes
- Better searching facilities
- Overview diagrams of hyper-linked document structures
- Tabletops' allowing some hypertext nodes and the ones connected to them to be seen simultaneously (at least some browsers now allow additional windows to be opened)
- Pictorial history' cache (in addition to the names of documents as now)
- Relevance ratings on anchors
- Guided tours
- Filtering of anchors
Of all these, in HE it is perhaps the last three which are of
greatest importance.
Relevance rating of anchors
Once unusual, it is now common for retrievals from
databases to be marked with relevance ratings. There is a need for such a system
for the link anchors of the Web. This is one special case of a general need for
more types of link than the ubiquitous go to of the current Web.
Guided tours
Laurillard (1993) has expressed the view that hypertext is
not educational technology, on the grounds that it does not encapsulate a
learning strategy. Laurillard's view is perhaps too dismissive given the many
different kinds of hypertext which can be created, but nevertheless there is often
a need to bind a hypertext more closely to a learning strategy. An obvious way
of doing this is to devise virtual tours. Some of the characteristics of such a
facility would need to include:
- tours could be created with minimal expertise
- tours could include any Web material
- staff or students would be equally free to construct tours
- students could leave and rejoin tours at any point
- cross-overs between one tour and another should be visible
A postgraduate student project exploring some of these
ideas is written up in Bulmer 1995. One way to make the idea work might be
through the filtering of anchors.
Filtering of anchors
A fundamental characteristic of HTML technology is that
the link anchors are embedded in the document. This has the advantage of
technical simplicity, since it is robust in the face of editing changes. But the
disadvantage is that each file has only one set of links, and those links cannot be
added or removed without editing the source file
We said (Uses of hypertext p12)) that in 1945, Vannevar
Bush recognised that the trails which users would build would have their own
intellectual (and commercial) value, which was built on the documents, but was
not part of those documents. We can easily see how part of a page might form a
part of one user-author's trail while another part of that page might be included
in someone else's. Or two critics might gloss the same text, offering different
links to commentary and supporting materials. It is clear that to be able to
invoke and suppress particular sets of links would be useful for any Web user,
but especially for educational users. The current technical structure of HTML
makes this impossible.
Other hypertext systems have explored a technically
different approach, where link information is kept separately from content
information, in principle allowing any set of links to be summoned up or
dismissed at will. Microcosm is a system developed at Southampton University
which in addition to these advantages, is also enabled by its alternative
architecture to link documents of any kind, even if the software applications
which deliver them are not hypertext-capable.
Contents
Graphics Multimedia
Virtual Environments Visualisation
Contents
