Creating a simple digital archive

March 23, 2020 4:18 AM

I'm putting together a simple digital archive for my workplace. It will store pdfs of newspaper articles, blueprints, a few books, as well as image files or various sorts of photographs, etc. I'm not a librarian or archivist so I'm seeking advice on the types of information to have associated with these files.

Currently I have:
File name
Date
Title
Author
Source (ie which newspaper)
Keywords
Format (ie photo or newspaper)
Physical Location (which file drawer or box has the original blueprint/photo etc)
Image resolution
Collection (we have a couple of collections of photos given to us by individuals)
Notes

What am I missing? What am I doing wrong that is going to bite me in the butt later? What questions should I be asking that I'm totally oblivious to?

I'm using Airtable to do this if that is relevant.

posted by sciencegeek to Grab Bag (1 answer total)

As long as you retain the ability to export all the data and all its associated metadata in some simple and/or widely supported and/or non-proprietary format, like PDF for documents and JPEG and PNG for images and well-structured flat text files for metadata, then the fact that a subscription-based proprietary online database service is always going to do its level best to lock you in before going belly-up won't bite you too badly.

It would also be good to have some automated method for adding extra kinds of metadata in bulk, should the need arise. This would be more reliable with some kind of guaranteed-unique ID that you control assigned to each data file to tie it to its associated metadata. Burying the ID inside the data file using whatever tagging scheme its file format supports is better than relying on making it part of a filename, because people will rename things. Doing both is better still.

And if this is an archive, you want local copies. More than two, for preference. The iron law of digital information is that unless you can always put your actual meat hooks on two different offline storage media, each of which contains a bit-for-bit identical copy of every piece of digital information you care about, it doesn't really exist.

Yes, cloud providers will pinky swear to back your stuff up carefully and look after it, and you should absolutely let them do that and expect that they will do it properly. But if the shit ever hits the fan, your locally maintained copies are going to be a much more useful fallback than whatever financial compensation your lawyers can extract from the remains of your cloud provider.
posted by flabdablet at 5:18 AM on March 23, 2020