With all this discussion about the creation of archives and indexing,
I figured it was time to add my 2 cents.  (It's actually more like 50
cents before inflation, so if you're not interested in this subject,
please skip to the next posting now!)

I think that creating a permanent, well-indexed record of the publica-
tions of various mechanical music publications is a great idea.  Who
could really argue with that.  However, I think that spending time on
the index is much more important than worrying about how to turn the
articles into PDFs (the obvious format of choice for such a project).
Yes, having PDFs or just the unformatted full text of the documents
would allow incredible searching to take place, but it sounds like a
lofty and expensive goal.  I'm not saying don't aim for that goal, but
it's beyond our reach today.

I'd vote for a simple database containing Article Title, Publication
Date, Publication Name, Page Citation and Author Name(s).  Once you've
found the article you want, then you can figure out who to contact to
obtain a copy of the publication or article.  Certain people would be
granted update permission, restricted by publication name and date
range.  Updates would be made through the web.  Obviously, building on
any data that others have already compiled would be a good start.  Yes,
it won't be perfect, but at least by making it a cooperative effort, it
will probably get done.

I'd suggest adding the ability to assign keywords as a second phase,
and I would let *anyone* (from within a closed community, probably)
contribute keywords.  The Internet Movie Database (http://www.imdb.com)
or CuisineNet (http://www.cuisinet.com) for two good examples of
databases where individuals contribute ratings of movies and
restaurants, respectively.  That's how I would treat the keyword issue.

My single greatest concern is ongoing maintenance of the data storage.
If you care about the durability of the end result of such a project,
read the article that appeared in yesterday's NY Times (April 7, 1998)
about digital archives (the second such article they've printed in
about six weeks).  They first quote Shakespeare's 18th sonnet ("Shall I
compare thee to a summer day") which ends

   "So long as men can breathe or eyes can see
    So long lives this and this gives life to thee."

The article continues by quoting from a paper by Jeff Rothenberg (of
the Rand Corporation) who says that were Shakespeare writing today, he
might be forced to end the sonnet,

   "So long as the magnetic flux on this disk
       has not been disturbed,
    and so long as humans retain the appropriate
       size and speed disk drives,
    and so long as they have hardware controllers and software
       device drivers capable of reading bits from this disk,
    and so long as they have access to the software
      that encoded the file structure and character codes
      employed in the bit stream of this document,
    and so long as they can still find or recreate the
      computing environment necessary to run that software
    *and* so long as they can still breathe or see
    So long lives this and this gives life to thee."

The problem is explained quite clearly in the article.  Computer and
storage technology are advancing at rates that make data inaccessible
in short time.  Storage technology formats change rapidly (Anyone still
have any 8-inch diskettes around?) and the life of electronic storage
media is not nearly as great as that of paper.

The point is not that we should abandon computers, but creating and
maintaining electronic archives requires a committed process that
insures future usability.  It's not like a book: publish a database and
put it on the [electronic book-] shelf and it will become unreadable in
a short period of time as technology marches on.

End of rant.

Bob Fitterman
bobf@ilx.com