[NHCOLL-L:2088] RE: Ideas for New Ledger System

[Shirley S Albright]

Databases are great tools for searching and sorting data in a rapid 
manner, but I too like to have a hard copy available for several reasons:
1.  Our collections database is on a server.  Sometimes the server goes 
down - as in the recent case of the SoBig virus.   We were unable to 
access anything on the server for 3 days.  An unwritten law of IT is: 
the server  WILL go down at the most inopportune time.
2.  I think accountability is still an issue with many databases.   Does 
your database, for example, record WHO made changes to the data, when 
those changes were made, and what the changes were?   Ideally, we print 
out a new catalogue sheet whenever something new is added or edited and 
that catalogue sheet is dated and becomes part of the object record.   
We have had issues in the past where well-educated and well-informed 
staff made erroneous edits.
3.  I wish we had the manpower to record absolutely everything about an 
object....but we don't, so having everything on the database is out of 
the question.   We don't have the manpower to digitize field notes, 
images, conservation and preparation reports, publications, x-ray images 
etc. etc. etc.     So....we maintain an object file with all that 
"stuff" along with a catalogue sheet that is fairly detailed.
4.  Some government agencies (we're a part of state government) are 
notoriously slow at keeping pace with upgrading software.   It was 6 
years from our first version until we received money for the second, and 
by that time we had data conversion problems as well as real training 
issues.    Ever try to access your old Multimate documents?

Don't get me wrong....I'm a big fan of computers and databases, but you 
just can't rid yourself of human error.   Unless you've got a 
crackerjack IT team I'd be skeptical of having everything on a server.   
Machines fail and they get old like the rest of us.   Are you willing or 
able to suffer the downtime?   That's one of the real questions to be 
asked.    I'm less concerned about losing data - although a corrupted 
file at the end of a long day of data entry is not unknown - then 
inaccessibility of hours or days.

Shirley Albright
Assistant Curator of Natural History
Database Administrator
New Jersey State Museum