Re: [ARSCLIST] Database template - Artist Instrumentation Project

["Steven C. Barr" <stevenc@xxxxxxxxxxxxxx>]

----- Original Message -----
From: <Mwcpc6@xxxxxxx>
> In a message dated 4/16/2005 10:16:46 PM Eastern Standard Time,
> seubert@xxxxxxxxxxxxxxxx writes:
> FileMaker can import and export just
> about any data format including csv, xml, dbf, xls and can talk to SQL
> databases through ODBC. That being said, any relational database
> solution (including ones built in FileMaker, Access, dBase etc.) cannot
> be easily migrated to another database platform
> ****************
> For an institution with a professional information technology staff and
the
> grants to fund them, this shouldn't be a problem. However for an
individual or
> small volunteer group, it can lead to frustration.
>
> In the last twenty years the museum group I work with has been through
three
> generations of computer platforms and database attempts.  Each time the
> "computer expert" or curator position has changed hands, an attempt has
been made to
> place the museum records in a new, favorite, database.  Even though the
raw
> data has been converted to .csv form, use was never made of this data,
mostly
> due to political issues over format and definitions.
>
> The result is that after twenty years of computer technology, the museum
> still doesn't have a collection database.
>
> One solution would be to stick strictly to Microsoft products, in the hope
> that their domination of the market will assure continuity. The other
solution
> (mine) is to avoid proprietary software altogether.  I can access and edit
my
> files on any computer made.
>
> Both of the above meet strong resistance. Most feel that one must use some
> sort of "official" approved software. But many will insist that it must be
> anything but Microsoft.  I don't see that changing soon.
>
> Perhaps helpful would be for an organization like ARSC to publish an
official
> list of those items of information that are of most use in identifying and
> finding sound recordings, in a simple outline form, ranked in order of
importanc
> e or usefulness, that people can use to structure their own database
> application. This might lead to enough standardization to make file
conversions
> practical.
>
> Meanwhile most collectors find their records by remembering which milk
crate
> they are in.
Actually, I use a separate .dbf file to keep track of which milk crate
they are in...and (more or less) where that milk crate is stored!

Anyway, xBASE files are about as close to a universal "LCD" as exists.
I don't know how long this will continue, but so far just about every
database application can both import and export .dbf files...and they
can, as well, be opened, read and printed legibly in text editors and
word processors, as an added benefit. Just ignore the header!

Now...when I started this whole thing...and this point to some extent
proves that point...what ARSC should be doing is NOT to identify one
particular application as standard...rather, they (or someone) needs
to establishy standards for a minimum set of core fields used to
create data records for phonorecords (that is, field formats and
minimum sizes). This way, if I catalog my records using Lotus Approach,
and you use FileMaker, we can select an interchange format (probably
xBASE) and know that TITLE (or whatever you call it) is a 50-byte text
field (if you use 100 bytes the last 50 will be lost in the transfer).
For example, the Abrams files all use 50-byte title fields, making
it possible for any cataloguing program with 50+ byte text fields
to import Abrams data records and see the titles. If we have that
degree of standarization, you can use Hooflungdung v. 7.3 if you
wish...and whatever it can export that I can import, I know that
I can view the first 50 bytes of your title fields!

You may have additional fields that identify your copy of the
disc as being stored in shelf 7, cabinet 3, closet 4 of the
Wilbert Hoohah Memorial Library attic...I don't need that data.
However, if I'm trying to establish who has copies of "Who
Threw the Overalls In Mrs. Murphy's Chowder" I need to access
your title field. If your title field happens to be a binary
field containing a scan of the label, that tells me nothing!
That's why we need this minimal degree of standardization...

Steven C. Barr