Print

Print


*******************************************************
*                                                     *
*    If you subscribe to TEI-L or SGML-L and have     *
*    peviously contacted the SGML Project, you may    *
*    receive duplicate copies of this report.         *
*                                                     *
*    We apologise for any inconvenience.              *
*                                                     *
*******************************************************
 
 
UNIVERSITY OF EXETER COMPUTER UNIT                        SGML/R4
 
THE SGML PROJECT
 
CONFERENCE REPORT:
    ELECTRONIC NETWORKS AND PUBLISHING
    LONDON 4TH OCTOBER 1991                             Issued by
                                                   Michael Popham
                                                  22 October 1991
_________________________________________________________________
 
 
SUBJECT
 
The subject of this one day seminar, as described in the
promotional literature, as follows:
 
"The use of electronic networks for the transfer of scholarly
research information is increasing.  Systems such as BITNET,
JANET and EARN are now widely used, with further developments in
prospect.  In the USA, proposals are before Congress to establish
a National Research and Educational Network and provide a high
speed and wide band network for researchers at an initial cost of
$400 million.  Such networks are intended to support the free
movement of research and other data, text and graphics, and have
the potential of transforming scholarly communication, both
formal and informal, in ways that are hard to foresee.  What will
be the effect on the traditional journal?  Will the customary
academic library cease to collect printed material and serve
largely to provide access to electronic riches?  What will be the
impact on publishers?  What experiments and projects are
underway?  What will be the consequences for libraries?  Who, in
the end, will foot the bill?
 
The day is designed to explore these issues in depth with experts
in this field from Europe and the USA."
 
 
BACKGROUND
 
The seminar was sponsored by The Association of Learned and
Professional Society Publishers and The Institute of Physics'
Publishing Division (IOP Publishing Ltd).  All activities took
place at the Scientific Societies Lecture Theatre, in London.
 
There were well over eighty attendees (including the invited
speakers), drawn mainly from the spheres of publishing,
library/information science, and academic research.  People with
experience of developing electronic networks were noticeably
lacking    but since the aim of the seminar was to examine the
implications of networking, rather than the technical aspects of
how recent goals in network development have been achieved, this
was probably not a bad thing.
 
PROGRAMME
 
The day was short but intensive, with no single presentation
being allowed to exceed forty-five minutes (despite valiant
attempts on the part of some of the speakers!)
 
"The Objectives of the Day" - Anthony Pearce (Chairman IOP
Publishing)
 
Pearce's introduction to the seminar examined the implications of
the ominous-sounding phrase "The Networks are coming!"  With the
advent of increasingly high speed, high capacity networks, Pearce
argued that academic publishers would be forced to reconsider
their production methods.  The first group of publishers that
would be obliged to consider the implications of networking
developments would be those concerned with scientific
publications.
 
Pearce asserted that the current attempts to design and develop
an `electronic journal', would need to provide a product that is
convenient, accessible, and relevant to users in hitherto unknown
ways.  He felt that the coming market would be inherently hostile
to the traditional approaches and concerns of academic and
scientific publishing.  Matters such as  circulation and version
control, copyright, production costings and payments would all
need to be reconsidered and renegotiated.
 
Technological changes will encourage a move away from purely
text- and paper-based communications, towards an environment that
can also include vision and sound.  Pearce stated that it was
with this in mind, that the seminar programme had been designed
to serve as an introduction and inspiration, providing
information about the changes in the near future.
 
"USA - The Key Developments" - Paul Evan Peters (Director of the
Coalition for Networked Information, Washington)
 
Peters offered an account of recent developments in the
technology and use of networks in the USA over the last two
years.  He then went on to talk about the sorts of changes he
expects to see in the near future, and the concept of "Networked
Information"    which he believes will soon become a reality.
 
"Networked Information" is that electronic information which is
accessed or created as a direct result of using a network.  A
jukebox of CD-ROMs contains information which can only be stored
and retrieved electronically, whereas the unique set of
information obtained by performing a search of these and other
stacks of CD-ROMs from a terminal connected over a network, is
called "Networked Information".
 
Peters identified immanent developments in two key areas, these
being a shift away from talking about technologies and towards
their use, and a shift from talking about the future to talking
about how we can actually get there.
 
Over the last two to five years, the scale and scope of network
features have changed dramatically    and Peters gave a number of
examples to support this statement.  Top-end bandwidth has shown
a traffic growth rate of 20% per month, whilst low-end and mid-
range services have been improving quickly    offering better
access, availability, reliability and serviceability.  By 1993-94
the US hopes to have in place a network capable of carrying
1Gigabit/sec (=1billion bits/sec), which is equivalent to
transmitting 20,000 pages scanned at 600dpi every second!  (At
present, 50% of the packet traffic on the National Science
Foundation Network consists solely of file transfer and email
data    ie. real information, rather than control messages etc.
which enable the network to function).  Over the last two years,
points of access to the national networks have increased by 500%,
whilst those to international networks have gone up by 1000%.
 
Peters stated that network resources, services and tools would
all develop rapidly    by which he was referring to e-
mail/conferencing/electronic journals, databases and digital
libraries, high volume network addressable print facilities, the
use of A.I. strategies (such as filters and `Knowbots'), and
intelligent databases.  He also expected to see a reduction in
the related complexities and costs, through the evolution of a
more common network infrastructure, that was easier and cheaper
to access than is currently the case.  Similarly, the network
communities of those who simply wanted to move data, and those
who wanted primarily to access remote data, would grow together
to produce the "Networked Information" community.
 
 
In conclusion, Peters identified three cycles of development:
modernization, innovation, and transformation.  He predicted that
the modernization cycle would be characterized by the advance of
technology eg. linked database and information systems, bit-
mapped page images, item level access, `just-in-time' delivery,
delivery to institutions and organizations, on-site printing and
finishing, and new types of site licences.  The innovation cycle
would involve developments of such things as tagged and
searchable text (rather than merely page images), compound
documents, contextual information, direct delivery to the end
user, better capabilities for packaging and re-packaging
information, more self-publishing, and the evolution of item
pricing.  Developments in the transformation cycle, will be of a
different order    concerned with such areas as the increased use
of multimedia and executable objects, access to primary research
materials, knowledge management environments, virtual and
simulated realities, collaborative authorship, and the control of
itterative and derivative works.
 
 
"Publishing Developments, Particularly in the USA" - Myer Kutz
(Myer Kutz Associates)
 
Kutz opened with a comparison of the `old world' of electronic
publishing (circa late 70's - early 80's) and the `new world'
(late 80's - present day).  He gave a number of examples of how
things have changed, with a movement away from limited and
restrictive paper and on-line sources controlled largely by
librarians, towards a more open and powerful environment easily
accessible by end users.
 
Kutz then went on to talk about how developments in electronic
publishing technology would effect the production of research
monographs.  DTP packages will enable more and more authors to
submit camera-ready copy, taking over many of the traditional
responsibilities of the publisher's copy editor.  On-site
printing `on demand', will encourage the movement of texts in an
electronic rather than a physical form; this will have the
advantage that publishers will not have to cost for returns, and
purchasers will to some extent be free to create `user generated'
texts prior to purchase.  However, these developments may simply
lead to publishers' traditional problems being replaced by a
plethora of new (and possibly more difficult) ones.
 
Developments in the processing of electronic journals will have a
number of effects, claimed Kutz.  More `Table of Contents' and
`Document Delivery' services will emerge to service the user
community.  Issues involving copyright legislation will become
more confused and complex, possibly resulting in new laws having
to be put on the statute.  More and more journals will be
available stored on optical disk, but will not necessarily be
available for processing at a digital level.  New technology
should also make it much easier to produce classroom anthologies
of selected journal articles, tailored to meet the needs of
specific academic courses.
 
After touching briefly on a number of related topics, Kutz
finished by discussing the work that has been done to produce
"The Online Journal of Current Clinical Trials".  This is a
document which has been designed from the outset to be available
on-line, and is thus very different from a traditional printed
journal in a number of fundamental ways.  For example, the
journal is no longer `time dependent', and can continuously
receive and publish articles.  Moreover, it is continually
growing    since `old' articles are not removed from the
journal's database, and subscribers are encouraged to submit
comments and rebuttals which can be attached to specific articles
(although it is too early to say whether or not this approach
will actually work successfully).  The entire journal can be
accessed on-line by text, paragraph, page-reference, table etc.,
and formatted for output on paper.  Every subscriber can be
notified of what articles/comments have been added to the journal
since s/he last logged in or s/he can instruct the management
software to only flag new articles on a specific subject.
 
 
"The Projects Under Development in the USA" - Ann Okerson
(Director, Office of Scientific and Academic Publishing
Association of Research Libraries, USA)
 
Okerson began by posing the question of what do we actually mean
by the phrase `electronic journal'?  Is it simply a paper journal
delivered in a different way, or should we regard it as a totally
new kind of document    since it can be automatically
manipulated, searched, and so on, in ways that are impossible
with paper?
 
In recent years, US libraries have found that the cost of
purchasing journals has been greater that the cost of purchasing
any other type of information source.  The proliferation of paper
journals has caused great storage problems, as has the fact that
any delays in their processing can soon make them out of date.
In comparison, therefore, electronic journals look very
attractive to libraries    since they appear to offer cheap,
accessible, easy means to store and retrieve large volumes of
information.
 
Okerson identified four general types of electronic journal.
Firstly, sophisticated high-tech/high-quality commercial
publication for profit.  Secondly, grass root journals and
bulletin boards created/accessed by network users and
researchers.  Thirdly, high quality on-line formatted documents
produced by `not-for-profit' organizations.  Lastly, subject-
oriented citation databases.
 
The growth of networked electronic publications may cause a
number of difficulties    quite apart from the technical problems
which will need to be overcome.  For example, it is inefficient
(and confusing) to store multiple copies of the same article in a
number of different commercial, private and research databases
not only from the point of view of wasting storage space, but
also because it might lead to general searches of the network
returning numerous copies of the same text!
 
Okerson then went on to discuss two sample uses of electronic
journals.  She cited the case of a commercial organization in the
US which claims it can make available an on-line version of a
complete electronic journal within two hours of receipt of a
request    less, for a well-known journal.  She then talked about
moderated, `not-for-profit' bulletin boards and discussion lists
(such as those available on LISTSERV), and their usefulness in
providing academics with the means to publish, access and discuss
current information.  Whilst the latter have the advantage of
being straightforward to use, they only offer facilities to
circulate crude linear documents, unaccompanied by any
sophisticated manipulation or searching tools.  However, even
bulletin boards/discussion lists such as these are able to
attract large numbers of enthusiastic readers and contributors
who offer a valuable service of peer review and support.
Moderators of these boards/lists often express the desire to
produce them in a more sophisticated form (offering search tools,
hypertext interfaces, the ability to easily extract and print
formatted output, etc)    but this would be expensive to set up
in terms of time and resources. Yet once in place, the basic
requirements for making good use of such boards/lists would not
be unduly prohibitive; for example, the minimum kit could be a
286 PC with 2M main memory, Windows 3.0, and a graphics display
monitor (although other users would still be able to read the
text using any plain ASCII text editor).
 
Okerson said that at present, it looks as though the on-line
bulletin of the American Mathematical Society will be the first
highly sophisticated electronic journal available.  All the text
is marked up with a set of SGML tags, and tools are provided to
down-load articles and format them using TeX.
 
Okerson concluded by asserting that although paper is still the
predominate repository for storing information, it will seem
increasingly uneconomic as the relative costs of electronic
production and storage decline.  Soon, the best source of up-to-
date information will be the on-line electronic journal, and
libraries will need to plan how they are going to adapt to this
situation.
 
 
"JANET, SuperJANET and Europe" - James Hutton (Joint Network
Team, Rutherford Appleton Laboratory)
 
Hutton gave a non-technical account of the development of the
JANET and SuperJANET networks, as well as an insight into how
this might tie in with work being carried out in Europe and the
US.  I shall not reproduce the section of his presentation
concerning JANET, as I assume that this will be familiar to most
readers.
 
Before beginning to discuss SuperJANET, Hutton made the
observation that as things stand at present, the international
links between the various national networks will need to be
upgraded.  This is because the gateways will not be capable of
handling the greater volume of high speed network traffic that
will be generated by the introduction of faster and faster
national networks.  If the situation remains unchanged, then
there will certainly be a `bottle-neck' effect as network traffic
tries to move from one national network to another    and this
will often appear to the user as if the whole network is slowing
down (and s/he will perceive no benefits to have been gained from
the upgrading of his/her national network).
 
SuperJANET is intended to be a high performance, optical fibre
Wide Area Network (WAN), supporting the work of UK research and
education.  The new network should be capable of supporting such
facilities as multi-media electronic mail, advanced visualisation
techniques, electronic libraries and multi-media information
services, and also distributed systems.
 
Hutton set the work on SuperJANET in the context of other
European research in the field of networking. He mentioned
briefly the national efforts being undertaken in Germany, Spain,
and Switzerland, the subject-specific network activities for
astrophysics and particle physics, and the international networks
being developed by EUnet and EARN.
 
Hutton went on to argue that Europe should be forward-looking,
and make sure that in networking terms she is closely involved
with both the eastern European states, and the US.  However,
efforts at intercontinental co-ordination are hampered in two
main respects: a lack of resources (especially people), and the
difficulty in assigning responsibility (ie. who should speak for
"Europe", the US, or Japan?)
 
From a practical perspective,  Hutton felt that it is vitally
important that any developments in the networking field should be
based on internationally agreed standards    so as not to favour
a single vendor or a specific country.  Unfortunately, such
standards can take a long time to develop and implement, and have
difficulty in keeping pace with the continual developments in the
networking world.  Thus the current, and somewhat unsatisfactory,
course of action has been to propose interim standards    which
prove expensive to support (as developers move from one interim
standard to the next), and lead to problems of scale during
actual implementation.
 
Despite these reservations, Hutton concluded by suggesting that
the most important uses of SuperJANET are likely to be uses that
we have not yet identified (ie. it represents an enabling
technology).
 
 
"The Consequences for Traditional Methods of Publishing and
Communication in the UK" - Bill Tuck (University College London).
 
It was pleasantly refreshing to hear from someone who did not see
networking and electronic publishing as the apparent cure-all for
our paper-based ills.  Tuck gave an entertaining presentation in
which although he acknowledged the potential of networking and
electronic publishing, he also highlighted the limitations and
practical difficulties involved.
 
Tuck began by giving an example for his own particular area of
interest, the study of early-medieval illuminated manuscripts.
Since these documents are now too valuable/delicate for the
majority of researchers to even touch, scholars are forced to use
very high quality photographs of the original documents. However,
it is quite difficult and expensive to make photographs such as
these widely available.  Thus in an ideal electronic world, Tuck
would like to have these high-quality images stored
electronically and available on-line (ideally with additional
information which makes it possible to codify their content and
store them in a searchable database).  Unfortunately, current
everyday technology is not quite up to the level of
sophistication required to this; for example, it is not yet
possible to email a high-quality colour image in a standard way.
 
Despite his personal disappointment with the current state of
electronic publishing/networking technology, Tuck cited a few
examples of work currently being undertaken in this field.  He
then went on to suggest that the areas of electronic publishing
most likely to develop in the near future, would be those
involving novel variations on existing technologies.  He gave as
examples the creation of large-scale CD-ROM storage via the use
of networked jukeboxes, the dramatic growth of phone/fax
publishing in the US, and the ability of data broadcasting
services (such as CEEFAX and ORACLE) to uncover/create potential
new markets.
 
Tuck envisaged three basic formats of electronic publishing
Wide Area Information Services (WAIS), fax, and Document Image
Processing (DIP).  He also envisaged that each format would have
its own particular market of users.  Thus WAIS will be the
province of academics, heavy computer/internet users, and `info
junkies'; fax will belong to the business/financial community and
others with `time-critical' needs; DIP will be mainly used by
those involved in formal industrial or academic research, users
of local information networks, and so on.  Tuck suggested that
although the three formats of electronic publishing will probably
move closer together (in terms of the sorts of
information/services they can provide and support), the different
user communities will probably remain separate    and so
electronic publishers will need to match their format/medium with
their intended market.
 
Tuck then went on to identify what he considered to be some of
the problems with electronic publishing.  He felt that the
capabilities of technology are often overrated (whereas the costs
of achieving a desired result are often underestimated!).
Moreover, the take-up of existing technology for
producing/reading/accessing electronic publications is still
surprisingly low    which Tuck felt was not unconnected with most
people's perception that the productivity gains from using I.T.
are still elusive.  Lastly, he argued that current electronic
source documents are a mess    they usually look dreadful, are
far too large, and consist of low-quality content.
 
Tuck concluded by suggesting that the `advent' of electronic
publishing will have no perceptible consequences for traditional
publishers.  Rather, it will parallel the introduction of
television    which instead of leading to fewer books, radio
programmes etc, simply meant that there was now something else to
impinge on everyone's available time.
 
 
"Challenges and Opportunities for Publishers" - Discussion
involving all the speakers and the audience.
 
Each of the day's speakers were invited to list key points which
they felt were relevant to the topic under discussion.  Their
remarks have been summarized below.
 
Paul Evan Peters:
        Key issues are beginning to emerge
        The necessary technology is becoming available
        Publishers must take risks and try to move forward;
        trying to preserve their current
        situation will result in many of them being left behind
 
Myer Kutz:
        Publishers must be aware of current developments
        whilst also keeping up
        traditional good practices (ie. listening to the
        needs/feedback of the market-place)
        Many publishers are experimenting with electronic
        publishing by collaborating with
        universities and libraries
        Keep looking for new ways to produce, package and market
        publications.
Ann Okerson:
        Cited "The Publisher's Weekly"    (1) book production
        fell by 17% last year (2) the
        coming of the hand-held electronic book (3) MacMillan can
        now go from manuscript
        right through to final production without having to print
        a paper version of a text.
        Publishers should subscribe to email/distribution lists
        They should follow work in specialist journals
        Publishers should think of libraries as partners
        Whether network developer or publisher, always pay
        attention to recognized standards
        Networking is as much an art as a science
 
James Hutton:
        There is no guarantee that technology will be able to
        meet all the requirements or
        solve all the problems that we have envisaged
        Publishers must get onto the networks and make serious
        attempts at electronic publishing
        Technologically speaking, the whole area of networking
        and electronic publishing is still immature.
 
Bill Tuck:
        The electronic medium offers the potential to do/find
        something new (in ways that
        were not possible when using paper-based texts.
        New types of publication will be created on the basis of
        utilizing the available
        technology    ie. journals with articles which can
        contain large numbers of high
        quality images, sets of statistical data etc.
        To be successful, electronic publishing will need to
        provide journals that are more
        usable than the current attempts/prototypes (a technical
        issue concerned with
        questions of document delivery and manipulation)
        Publishers have yet to appreciate the potential for using
        networks as an advertising
        medium (cf. word of mouth)
 
The speakers' remarks were followed by an open question and
answer session,  which began with someone asking what are the
journal user's reasons for wanting to change over to using an
electronic publication?  The consensus amongst the speakers was
that readers in many spheres would benefit from having easier
access to larger and more up-to-date sources of information
but especially those who were involved in large scale
collaborative scientific research (who would find it much easier
to publish, dissmeninate and access information than is currently
the case).
 
Discussion then moved on to consider the suggestion that the
models of publishing for the future will not come from publishers
or readers, but rather from developments in technology and how
readily they are taken up.  Various speakers argued that once
readers have access to news networks etc., they will not be
content with plain ASCII text for very long; similarly,
publishers will want to be seen to be producing high quality,
sophisticated documents, accompanied by useful manipulation tools
etc.  One of the speakers argued that although it appears that
new technologies are undermining the traditional conventions and
institutions familiar to both readers and publishers, it may
instead by the case that they are simply providing new
conventions and institutions (eg. self-publishing on a
distribution list subject to peer review, as opposed to
submitting an article to a recognized journal).  Moreover,
participants on a networked distribution/discussion list, and
subscribers to an on-line journal or news system, seem to feel
that they belong to a `community' of readers very early on
and actively seek to control/influence the way `their'
list/journal etc. develops.
 
The highly problematic question of the economics of electronic
publishing was also raised.  For example, how easy will it be for
publishers to recoup their costs should they decide to publish
electronically?  Do users really want the `added value' inherent
in electronic publications, and are they prepared to pay for it?
All the speakers found it difficult to answer a question with so
many implications    particularly with regard to practical
matters such as copyright and costings, and the difficulties of
charging/pricing electronic publications.
 
Another questioner suggested that it was the reliability of
networking technology which, in his opinion, would be the biggest
single influence on the success or failure of electronic
publishing.  The response to this was that the previous pattern
of networks increasing in capacity whilst falling in cost is
expected to continue in the future.  However, even if all the
technical difficulties could be overcome, it is still not clear
how users should be charged for their use of the network (ie. by
a scheme of fixed pricing, or on a cost-per-unit basis).  The
panel of speakers argued in favour of a fixed price scheme,
primarily because using a cost-per-unit approach necessitates
that a great deal of time and effort has to be spent simply
monitoring the network and calculating charges to the users, and
the expense of all this effort will need to be recouped in the
prices charged.
 
SUMMARY
 
The seminar offered a valuable overview of the current situation
and immanent developments in the field of electronic networks and
publishing.  Whilst it might have been slightly unrewarding for
anyone hoping to gain some insight in to the functioning of the
separate worlds of networking development, publishing, or
libraries, it was very interesting to learn how the three overlap
and interact.
 
The field of networked electronic publishing is developing so
rapidly that it must be a very difficult time for any publisher,
librarian, or network expert that has the task of trying to
anticipate which way things will progress in the next few years.
However, from the standpoint of the work of the SGML Project, it
was good to realize that many developments will almost certainly
require that the systems involved can support and process texts
marked up with SGML (or something very similar ).