Talk:2.2.1 INFORMATION DEFINITION

From wiki.dpconline.org
Revision as of 15:59, 20 December 2016 by Hlhours (talk | contribs) (Created page with " "the OAIS must clearly identify the bits and the Representation Information that applies to those bits. This required transparency to the bit level is a distinguishing featur...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
"the OAIS must clearly identify the bits and the Representation Information that applies to those bits. This required transparency to the bit level is a distinguishing feature of digital information preservation, and it runs counter to object-oriented concepts which try to hide these implementation issues. This presents a significant challenge to the preservation of digital information."

I’m not sure if it’s the transparency to the bit level or the fact that it runs counter to object oriented concepts that presents the significant challenge or both. If could be that the implication is that ‘the information object is all important so you shouldn’t generalise in the way you might with other objects in OO’ but unless there’s something really important and specific to say it almost feels redundant to mention OO at all. Is this a case where more information is needed? Or could we lose “and it runs counter to object-oriented concepts which try to hide these implementation issues” without significant impact? While I understand that before you start generating the logic you need to model the data, generalise to a class, imply a type etc it’s not clear to me that this is a problem. Of course your modelling only abstracts the features of the underlying objects necessary to implement the logic and this is usually going to ignore bit-level stuff and concentrate on attributes you care about, but if you define the problem as one that goes down to bit level then your OO modelling would take that into account.

"software [] is used to access the Information Object, and it will incorporate some understanding of the network of Representation Information objects involved. However, this software should not be used as rationale for avoiding identifying and gathering readily understandable Representation Information that defines the Information Object, because it is harder to preserve working software than to preserve information in digital or hardcopy forms."

Does this go too far in dismissing emulation as a solution? The following paragraph then talks about how the reference model focuses on 'preservation of information content' but also talks about various types of emulation.

"Though the OAIS reference model does not focus on these emerging techniques, it should provide an architectural basis for the prototyping and comparison of these techniques." 

Is it fit for this purpose? Is this happening? Should this be updated? Has this ‘emerging technique’ progressed enough since 2002 to be worth a rewrite, or at least to remove the word ‘emerging’?