Eric A. Bolden
eabolden at FACSTAFF.WISC.EDU
Mon Sep 15 11:43:04 CDT 1997
I would like to make some additional comments on databases:
There is currently a great difference between a database used to manage
local information for a small group, and a database used for a large site.
* Often times one database is a small flat file. Indeed sometimes this even
means a simple word processing document. These have the advantage of being
relatively easily understood and managed. When necessary this type of
database may even be printed out on paper and function nearly as well in
that hard-copy state.
* The other type of database, the large interconnected relational model; is
very elegantly done technically. There is usually an 'expert' who designed
and managed the system. The database is sometimes used by local scientists
and staff, and it is sometimes bypassed in favor of a more familiar tool
like that listed above.
I) A modification to these database options has started to appear. The new
scenario looks a little like the old mainframe models many have tried so
hard to get avoid.
The information is held on a single machine.
Interaction with the database is tightly controlled by a central group.
There are predefined screens used to view the information.
II) There are, however; some key important differences between this and the
previous mainframe model.
The computer hardware is relatively inexpensive. Servers for
between five thousand US and fifteen thousand US, versus the 50 to 100
thousand plus maintainance fees found previously.
The reports may be modified in a short interval of days, not the
months or years needed in the past.
The information is being kept in a format that is portable to other
systems from the start. The migration nightmares encountered several times
in the past have been remembered and dealt with.
The user of the system is considered important and is asked for
imput in the design and funtioning of the database. This is an important
III) Why are these changes possible? The incredible speed of new computer
hardware, the increased size of hard drives, and the more educated and
targeted demands of users should be credited.
CONCLUSION: What will this mean in the future?
Databases are expensive and the people who manage them are even
The Internet is getting faster. Intranets are getting more reliable.
The internet is one large computer.
The next change to database management will occur across institutions. One
or two very well designed mainframe-like internet based systems could
support hundreds of institutions in a very consistent, extensible, and
reliable way with a minimal personnel cost.
Eric A. Bolden
University of Wisconsin - Madison
600 Highland Ave.
Madison, WI 53792-4108
eabolden at facstaff.wisc.edu
More information about the Taxacom