MoMA
May 13, 2015  |  Media Conservation
Open-Sourcing MoMA’s Digital Vault

Binder3

In my previous post we introduced MoMA’s digital art vault, and two of its components: Archivematica (aka “the packager”) and Arkivum (aka “the warehouse”). These systems meet two essential challenges: how to ensure that, far into the future, people will be able to understand the bits and bytes that constitute a digital item in our collection, and how to ensure that we can keep a bit-for-bit authentic copy of this digital item indefinitely. These two systems work together to facilitate digital preservation designed for the long haul; it’s a bit like carving something into stone in a universal language and storing it in an underground vault. While this is great for preservation, it renders active management, access and big-picture analysis quite difficult and cumbersome. We talked to experts from the private sector, libraries, archives, museums, and more to see what others were doing to address this issue. Our findings uncovered plenty of systems designed to address this issue, but the digital preservation requirements of media conservators working with museum collections far outstripped current state-of-the-art options. In other words, there was a gap between what we needed and what was available. So we rolled up our sleeves and did something about it!

Today I am very pleased to introduce Binder, a Web application designed for overseeing and managing the active preservation of digital collections, developed by MoMA and Artefactual Systems. Today we have released Binder as free, open-source software for anyone to use, adopt, modify, and redistribute.

As we’ve discussed, Archivematica (“the packager”) produces all kinds of invaluable information about our digital collections: what file formats are present, the granular characteristics of each file, information that allows us to verify the authenticity, and more. This information is packaged with the collections in the warehouse, in a format that requires no special tools or techniques to understand—it can be read easily by the naked eye, and is also very easy for computers to understand because it is written in a structured, standardized format. Yet this format makes it difficult to run quick and effective analyses across the entire collection. What is good for preservation is not always a boon to access or management. Therefore, before packages are sent to “the warehouse,” Binder sifts through them, indexes their contents, and stores what it finds in a database that is built to be very good at queries across large sets of data. Binder allows us to see the bigger picture in our collection.

Binder also helps us understand the smaller picture. As explored by Peter Oleksik, a single video artwork can exist in many different formats—Digital Betacam, U-matic, DVD, and multiple digital video file formats. In conservation, a huge part of what we do is to manage change and variability, so conservators need the ability to understand the past lives of artworks. Binder helps us do this.

Media components of artworks always possess some kind of dependency—for instance, a device or piece of software that is required in order to render the material properly. This can vary from a software-based artwork requiring a certain operating system to a digital video file that only plays properly in a particular video player (which is in turn dependent on a particular operating system). Binder helps us draw, describe, and understand these sorts of relationships, and to preserve the dependency itself.

For the nerds in the audience, you will be excited to know that Binder provides a REST API, which means that not only does Binder provide unparalleled control and knowledge over archival packages of digital materials, it also opens doors to new applications that can be built on top. We have been using Binder for about five months now, and already we understand our digital collections in new and incredibly exciting ways. But we are more excited to now be giving it away as free, open-source software. You can find Binder on GitHub, and documentation of its features, technologies, and API on readthedocs. Feel free to send any questions to Binder’s newly created Google Group. Now get downloading!