I want to share a bit of experience that the New York
          Botanical Garden has gained in its efforts to put
          collections data on line. It's both a cautionary tale, and
          one that might be a useful base upon which others might
          build. I am writing about this as someone who has watched it
          from the position of a fundraiser, not a botanist or
          collections manager.
 
 
          We have been getting our herbarium collections data
          computerized in a two-step process. The first is a
          modification of a program produced at the Harvard University
          Herbaria called HUHpc, which we are adapting to suit our
          collections management and research needs. Our adaptation,
          called (cleverly) NYpc is going on-line as we speak. By the
          way, it is an Advanced Revelation application, and seems to
          be quite functional.
 
          The next step that we are planning is to get this data
          directly onto a client server system on a UNIX platform.
          This is in the planning stages, and we are working with
          several of our sister collections to choose a useful system.
 
          Those of you familiar with natural history collections may
          be aware that the NSF funded a project among several of the
          largest collections to develop automation standards. The
          outcome of this project, the MITRE report, is available from
          Wayt Thomas here at NYBG, the PI for the project.
          ([log in to unmask]).
 
          The next logical step for us, and here's where the story
          becomes fraught, was to work with several other herbaria to
          develop data standards. Well, forget it. Can't be done. Too
          much politics, and investment in existing standards.
 
          We then came up with a compromise, which we called NDBTS
          (National Databases for Botanical Type Specimens). We
          proposed (to the National Science Foundation), a framework
          of common field names and shared authority files which would
          allow all herbarium collections to *export* data from their
          own collections databases in a common ASCII
          character-delimited format. This would work no matter what
          the underlying database was. These ASCII files (which would
          be done first for Types specimens, and later for other
          collections) would be maintained at four institutions which
          would provide Gopher (or WAIS or Mosaic) searching
          facilities.
 
          This proposal ran into all kinds of trouble, and was not
          funded. I still believe that it is an valid concept.
 
          If you are with me this far, and are interested in receiving
          a copy of this plan, please let me know. I think I can post
          it in ASCII format here for individual users, not for the
          whole list (serious bandwidth buster).
 
 
          Eric Siegel
          [log in to unmask]