I've just taken a look at the copy of the database that NYPL filled out
and sent back, and I have some general comments. Rather than sending
these just to NYPL, I'm posting them to the list. Hopefully this will be
helpful for everyone.
Stanton field notes: You have text checked for every page. This
indicates that you will have a seperate transcription file for each page.
I think what you will want to do, in almost every case, is to have a
transcription for the object and not at the page level.
Mammoth prints: The structure for this object is collection -> photo. If
these are loose photos, I believe that each photo should constitute an
item in the database. If the collection is an album, then the object is
an album, with subobjects which are pages, containing photos. Does this
make sense?
Stanton field notes: Here again, your structure is collection -> folder ->
item. This heirarchy is handled by the finding aid, so you don't need to
replicate this in the database. Instead, you'd want to make each item
that is going to be digitized a object in the database. You will probably
want to store information about where in the collection this items lives
(folder, box, series?) elsewhere in the database. What are people's
thoughts about how this should be done?
Each item has a transcription and an image, are each of the
foldered items one page each? If not, you are going to want to rethink
this.
Hudson River Portfolio: alternating image, text, image, text... I
couldn't tell if someone was just fooling around, or if the document
really has this structure!
All for now. Looks great, and I'm glad you took the time to play with the
imaging portion of the db.
We should have the final copy of the database for you this week (today or
tomorrow, I hope). Documentation is being written, and this should be
ready for you in January.
Merrilee