No subject

[Gary R Noonan]

     The postings by Jim Beach and Una Smith brought to mind two
comments.
     1. I agree with Jim Beach that data which may be shared with other
people should not be stored in databases built by programs that can't handle
relational data structures and that lack flexibility for third-party
 connections.
And I share Una Smith's concerns  about the problem of sharing data in the
future.  However, we need to consider how we can archive data.  ASCII files
and paper trails by themselves won't solve the problem.
     We need a government funded data repository to store and track data
and provide it to interested parties. My own research data might be of
interest to other biogeographers.  New data are being stored in databases
built with programs I write in FoxPro, and I can easily generate ASCII
output files.  But such files won't guarantee that the data are available to
other researchers.  The paper trails and ASCII files will get lost unless saved
in a central data repository.  When I eventually retire or pass away
(hopefully some years from now!), the data will probably be packed away
somewhere and in time lost.  Even if the data is stored by my institution and
indexed, it will become corrupted if kept on magnetic based disks.  A central
depository could store data on appropriate media (possibly WORM drives)
and ensure it is available to future workers.
     An important function of such a central repository would be to
provide systematists with updates about the stored data.  I subscribe to
several visual news groups and find that the most common question seems
to be where a particular type of data can be found.  Quite often the answers
provided are incorrect, and the data is not located on the servers listed in
the answers.  The central repository might provide information both about
its own stored data and about systematic data stored elsewhere.
     2. A major problem we face is the lack of data standards and data
models.  A few years ago I became concerned about this and mailed
preliminary data standards suggestions to other entomologists.  This led to
my being involved in a workshop funded by the USDA for data standards
in entomology.  A report was issued soon after the data standards workshop
and approved by the Entomology Collections Network.  Entomologists are
using these data standards in developing their own particular applications.
However, we need an interdisciplinary model to handle instances where cross
discipline sharing of data is desirable.  I had been asked to participate in a
workshop on this topic last August but had to decline because of field work.
I'm developing several applications that handle data that people in other
disciplines might want to examine and could use such interdisciplinary
standards now.  Does anyone know when the results of that workshop will
be published?