oais1.jpgAvailable here on the CCSDS website. ISO approached the CCSDS in 1995 to draw up the standard, which was presented in draft in May 1999. The model was approved in January 2002 and according to its own terms was supposed to expire after five years. There were a few typos corrected in September 2007 but that seems to be all.

The purpose of the model is to define the ISO standard for an open archival information system, that has accepted the responsibility to preserve information, and make it available to a designated community (1.1). “Open” in this context simply means that the model itself was agreed in open discussion: it does not imply that access to the archive is unrestricted. It has nothing to do with the Open Archives Initiative.

CCSDS is a committee of representatives of various space agenices, including NASA, ESA and the UK’s British National Space Centre. Space agencies realised first the need to make data inter-changeable, and available for a long time, so they developed the model rather than archivists. The space bias occasionally emerges, eg “the information may in general be submitted using a wide variety of common and not-so-common forms, such as books, documents, maps, data sets, and moon rocks” (3.2.1.). So in a sense the model really is rocket science. (Ha!)

How prescriptive is it?

In theory the model is applicable to all archives, space-related or anything else, and to all types of archives, whether digital or not. But it’s got a digital emphasis, and if my reading of it is correct then it favous migration over emulation, wherever possible. Take this, in 2.2.1: “it is harder to preserve working software than to preserve information in digital or hardcopy forms.” There’s more on the emulation vs. migration battle in the section on preservation perspectives.

The model only insists that a conforming repository should (a) support the definitions used within the model, and (b) carry out the mandatory responsibilities listed in section 3.1 . Other than those, the model makes no recommendations at all. In particular, it makes no recommendation or endorsement concerning platforms, methodology etc; nor is the 3.1 list of responsibilities a restrictive one – an OAIS can carry out more than this, if it wishes. See the notes on mandatory responsibilities for more about this.

My notes on individual OAIS themes:

The more I read and grapple with the OAIS model, the more impressed I get.