Talk: Representation Information

Revision as of 18:02, 15 December 2015 by Euanc (talk | contribs) (→‎Should software be part of the digital object?)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Should software be part of the digital object?

Software, necessary for enabling information within a digital object to be interacted with by a user, is often considered Representation Information. Without such software the object cannot exist as information to be interacted with. Furthermore, any change in the software may/will change the information that is made available for interaction by the user. Therefore the OAIS model should enable and require such software to be considered part of the digital object itself. In cases where the software is truly just providing additional meaning then it should still be considered Representation Information

Euanc (talk) 14:25, 26 October 2015 (UTC)

I'm very much in favour of recognising the importance of software in digital preservation (and for preserving software to enable data preservation for that matter), but I'm not sure I understand the motivation for this proposal. What benefit does this change bring, and why is it a problem currently?

In OAIS there is some flexibility as to what RI is captured, so that RI capture is based upon need (the whole designated community thing, ensuring RI is enables a particular audience to use some data in the OAIS). So leaving software in the RI camp would ensure that this flexibility is there. Otherwise aren't we heading to a situation where every object in OAIS requires software to be provided? Why not the OS as well? Would this not be overly prescriptive, and potentially a waste of precious resources?

PRWheatley (talk) 12:23, 18 November 2015 (UTC)

After reading Dinah's great blog post I realized what I think might be best would be if the OAIS recongised that representation information is made up of data objects and that the only difference between those data objects and others in an AIP is that they may be shared with other AIPs but otherwise need to be preserved just as much as the other data objects do. The benefit of this is that the OAIS would recognise that information objects cannot exist without the representation information (at least when that RI is software) and that therefore the representation information should be considered to be of the same importance as the other data objects in any AIP. Right now it implies that the representation information is less important than the other data objects in an AIP.

And I do think that every object in an OAIS should require software to be provided with it. Without the software you can't interact with the information object so you haven't preserved it. I think the interesting consideration should be whether the software has to be within the AIP or just referenced by the AIP. I can see representation information being used to reference other AIPs that include the software. I suspect others believe this also. For example, Ex Libris Rosetta basically assumes this requirement to have the software available as it assumes that an object is at risk only if there is no software available to interact with it. In summary, I think software should be considered a first class object in the OAIS just like the current "data objects" are. It is as important to an information object as the current "data objects" are so why do we relegate it to an effective after-thought? --Euanc (talk) 18:02, 15 December 2015 (UTC)