databases

[Chris Raper]

On Wed, 2 Jul 1997 12:18:49 -0400, Michael Healy
<MJRHEALY at COMPUSERVE.COM> wrote:

>I'm among those using Jim Asher's Levana package, and it works quite well
>for UK butterflies.  It could easily be adapted to other apecies lists.
>However, it does show signs of its age and its life and extension do depend
>rather heavily on Jim's personal survival and interest.  As the originator
>of a one-man statistics package (easier to use, more intelligent,etc than
>the rival ones, but ...) I know what the problems are.  My view is that we
>should be able to agree upon an adaptation of one of the big commercial
>packages such as the one in Microsoft Office.  Then we could rely upon the
>commercial boys for maintenance, and could call upon a vast degree of
>database expertise worldwide.  How do people react to this (including Jim)?
>Michael Healy

Hi Michael,

I agree. I have seen Recorder in action (and not liked it) but haven't
seen Levana. There are other packages like AditSite but Adit Software
don't give out demos and it costs over #100. Instead, I elected to
knock out something in Access myself to store and analyse the data.
It does just what I want (with some limitations) and it is easy to
program / customise.

We all have our likes and dislikes when it comes to using software
packages - some like Windows, others like DOS etc, etc. The key
factor, as far as I see, is the ability to exchange sighting
information between recorders - and to to this you have to have an
agreed _exchange_ format. The recorders should be able to use any
package they want but when they exchange data there should be some
defined minimum requirements - like ASCII CSV text for the file format
and then a list of fields (location, date, species, recorder etc...).

If each package could import data in this agreed format - directly
into the database it would enable the exchange of _data_ rather that
printed lists. Anyone who has had to re-enter someone else's list will
know how long that takes and automatically importing has the added
bonus of no transcription errors.

The tricky bit is agreeing on a minimum standard for the sort of data
we want to exchange. For moths (my pet subject) I would expect the
list to include: Bradley/Fletcher number, genus, species, quantity,
date, location name, os/map ref., recorder name. But for other orders
you might want to include other fields - 'life cycle stage', 'method
of capture', 'field notes' etc...

I have heard it said that the main problem is making species lists and
locations the same on all machines. If one person calls a place
'Hartslock' and the other calls it 'Gatehampton Wood' it would look
like you had two locations. Personally, I don't see this as a problem
because if the OS Ref was given the map plotting (DMAP?) package
should be able to sort it out.

As far as the species names go, if the recorder has any problems with
names changing on one or two species the software should just ignore
those records and print an exception report. You would still be able
to import 99% of the data on the first pass and any others can be
entered manually like before.

Ideas anyone?

Cheers
Chris R.