database structures

From
Date
Reply-To
RE:  My last posting to Micropaleo and radfolks noted that everyone seems
to be developing their own database.  This is probably more general than
just the micropaleo community, so I am posting this to PaleoNet with the
notice that future discussions should now go to databasenet@nhm.ac.uk.

While most of these developing databases will turn out to be very similar
because we all collect more or less the same data (including images of
specimens, notebooks, maps, or whatever), a tremendous amount of energy is
being duplicated, and large  costs incurred.  There are a number of useable
or slightly modifiable structures and programs out there that will run on
all the common platforms.  Seems a real waste to hear that so many people
are doing the same things.  I count over a half dozen people trying to
develop databases for themselves or their museums.  Lots of duplication!
Surely the time and money are better spent doing science.

The UC Museum of Paleontology database model, which I mentioned ran on
Paradox or Interbase (both by Borland International, Inc), has been
available for two years by ftp from the Museum's site at
ucmp1.berkeley.edu/home/ucmp1/ftp/pub/datamodel.zip  You will need to unzip
this.  If you like this, let us know and you can probably use it.  We have
installed it at the Paleontological Institute in Moscow (Sept. 1993), for
example,  and a few other sites.

This data model will accommodate images and a lot of other things.  It can
be linked to bibliographies and to keys to organisms with little work.  It
is easily modified, added to, and has off-the-shelf support by Borland.
Borland was due out with a final version of Interbase on MicroSoft's
Windows NT and Novell's NetWare last December, and on other operating
systems later this year.  Paradox, which is much like Interbase and with
which data can be exchanged, runs already on Dos and Windows.

All of the data can be loaded onto www or gopher, where it is searchable by
any word or set of words.  See gopher.ucmp1.berkeley.edu under catalogues
to see how it works.

The reason we chose these relational databases was the support level.
Borland has a lot invested and when they upgrade, they make it possible for
users to move up too.  They solve problems for us.  They are a call away
and respond immediately.  And many other things.  We like it because we can
put our efforts into the museum work rather than software support.  We
decided not to go with custom designed museum databases/programs offered by
other institutions because the support level is slow and not always able to
keep up with developments.  Furthermore, when the developer in these
institutions tires of their work , retires or dies, who do we turn to?  Why
should we deal with this stuff, when companies are out there that make a
living at it?  We seek efficiency in dealing with the 10-12 million
specimens we have, ranging from prokaryotes to vertebrates ranging in age
from Precambrian to Recent from every continent and ocean from the depths
of the deepest sea to the tops of the high mountains.  We'll never get the
job done if we spend time developing all these databases.  Flexibility is
the key.  You can always push the delete button or add something if you
need to.  Or just not use it.   In the meantime, I want our staff to get on
with the science, not the bookkeeping except as necessary.


Jere H. Lipps
Professor, Department of Integrative Biology
Director, Museum of Paleontology
University of California at Berkeley
Berkeley, CA 94720
510-642-9006 fax 642-1822
jlipps@ucmp1.berkeley.edu