The Standards Project Announcement

[Virginia Tech]

I guess, from the increasing activity on this list now that the academic
semester is over, we're all finding time to turn to our special interests!
It is great to see things happening on this list.

Actually, Jim Hart and I have been working all year on the project which we
agreed to take on last fall at Roanoke.  As participants will recall, the
Standards Meeting at the Roanoke Conference decided to adopt -- as a
starting model for a WWW-based Archive of Watermarks and Papers -- the
Bates College Archive of Watermarks and Papers in Greek Manuscripts.  The
objective of the project was to expand it by adapting the IPH Standard to
WWW use and internationalizing the resulting standard.

The results of our year's work should be on line on a new web site sometime
within the next two weeks.  The new web site for this project is called the
Watermark Initiative. We will post the URL here as soon as it is ready.  It
is intended as a catalyst for discussion about what a WWW-based Archive
system ought to be and do. Then, hopefully, discussions on this listserve
will identify ways to improve it and make it better serve our needs.

********

The model we have constructed envisages an international system of
web-based archives, all (from the user's perspective) seamlessly
interlinked, and all using the same core web page design, so that a search
>from anywhere in the world would search all participating archives.

It tries to accommodate the needs of people in many diverse fields,
including, for example, forensics specialists, curators and librarians,
auction houses, as well as academicians in historical and literary fields
or in the arts.

It conceives of the web-based archive as a research tool AND as a mode of
publication of watermark images and paper descriptions, which was an
important feature of the Greek Watermark Archive.

It proposes a way for dealing with "ownership" of data by the institutions
that will mount databases and for crediting those who publish data on
watermarks or paper anywhere in the system.  Institutions will frame the
core html pages in their own presentation (Institutional headers & footers,
copyright information, contact persons, etc.) and graphical presentation
(logos, backgrounds, other images).

It continues the emphasis of the Greek Watermark Archive on encouraging
researchers to publish the data they collect in the course of their work by
keeping the process simple and intuitive (IPH codes are not needed any
more) and using standardized language (pretty much as the Greek Watermark
Archive does).

It allows for freedom of choice with respect to both computer platform and
underlying database software.

It continues the emphasis of the Greek Watermark Archive on the distinction
between raw data and analysis in a database, and provides ways for
accommodating analysis.  (The IPH standard does not distinguishin its
database between "raw data" and scholarly or professional judgments based
on analysis of raw data, and fails to provide for recording the bases for
such judgments or analyses.)

It raises questions about how best to handle the updates and revisions,
and differing assessments of data, and proposes a way for including updates
and revisions and differing analyses in the databases and in the search
process.

Finally, it adopts the IPH standard, but reorganizes it on the basis of
accepted database (Entity-Relationship) design principles and uses fuller,
more intuitive field names. (Disk space is no longer the factor that it was
in 1992 when the IPH Standard was issued.)

Can you believe that as recently as 1992, when the IPH Standard was
published, the editors prefaced it by saying:

. . . it is assumed that the digital/optical storing and reproduction of a
watermark on a computer screen will not find widespread use in the
foreseeable future.

WHO could have predicted the revolution that has happened in only 5 years!

*********

The project we undertook to adapt the IPH Standard to a model for a
WWW-based archive system was MUCH more demanding than we had anticipated.
It was really a kind of translation project, because the media are so
different, and the potential uses (objectives) so different, from a simple,
self-contained data base.  And the IPH Standard was not very fulsome in its
statements about how it was supposed to work, what its data fields were
supposed to contain, or what its categories were. The printed IPH Standard
does not constitute a real, systematic Database Dictionary.

To be assured that we would adress diverse kinds of use of such an archive,
we consulted individually with some of those who were present at the
Roanoke conference. We also requested permission from the IPH to post the
Standard on the announced web site for study purposes and reference, but so
far (after two attempts and several months) have not received any reply
>from them.

We decided, once we were well into it, that we would have to approach this
project in two Phases:

Phase one would be to get a working model based on HTML (as the Greek
Watermark Archive was), but using instead the IPH Standard as our basis.

Phase two (for the future) would be to shift to a real relational database
operating on the Web via a database interface.

To accomplish Phase One meant that we had to work out an interpretation of
the IPH standard with the principles of a relational database in mind.  The
more we worked with the IPH Standard, the more we appreciated the
thoroughness of its creators. Nevertheless, it was surprising how many
problems and inconsistencies we found when pressed to apply it
systematically in a different medium.  We concentrated on developing a set
of model output pages or data presentation pages for two reasons. First we
felt they would serve the purpose of defining how a WWW-based archive
should be organized, and how it should function. Secondly, the model pages
would provide a tangible and readily understandable  basis for discussion
among interested participants in this project on the Watermarks Discussion
Group.  Creation of a set of sample pages was, in turn, a catalyst for
identifying and facing all the issues I have alluded to above.

Some of you may have discovered our developmental server, where we have
been working on this project, and may have been able to see how much the
looks of the model WWW-based database has changed from week to week, and
lately from day to day. We are now in the process of a final review of our
model pages, testing them against the standards of Entity-Relationship
database design.

But to whet your appetite, in case you haven't found it --

It contains introductory and/or "theoretical" pages on the following subjects:

      Origin and Objectives of the Project
      The Watermark Initiative Discussion Group
         (we wanted to get people who found the site involved
         in discussion on this Discussion Group by setting up
         a link to it.)
      The Relational Database with diagrammatic overview
      List of IPH Database Fields and their
         locations in our model WWW-based archive
      Commentary, Interpretation and Critique of the IPH Standard

Then the heart of the project -- the series of model pages indicating what
the presentation of data might look like.

The following list of model pages is not yet the final form of our
proposal/model, but it gives you an idea.

      Paper
      Watermark
      Facsimile
      People (scribes, artists, printers, illuminators, binders,
papermakers, etc.)
      Institutions (papermills, publishers, printing houses, binderies, etc.)
      Mould (i.e. data on moulds which are still in existence)
      Reference (i.e., editions, articles, literary references)


If you are familiar with the IPH Standard, you will see this involves some
different ways of organizing data than what that standard envisaged. Our
pages revise and clarify the section divisions of the IPH Standard
according to basic principles of Entity-Relationship database design. In
addition, some of our pages represent some of the subsidiary "special
databases" which the IPH Standard suggests but left for future development
and to torment people like us.


*********

As I said, this is to get you all started thinking about these issues. Jim
& I hope we'll have it ready for you sometime in the next 2 weeks.

Bob Allison

Robert W. Allison
Dept. of Philosophy & Religion and
Chair, Classical & Medieval Studies
Bates College,
Lewiston, Maine, USA 04240

E-MAIL:         rallison@bates.edu
TEL:            (207) 786-6307
FAX:            (207) 786-6123