Those of us who work with digital assets know that one day we’ll face format obsolescence. The formats we have in our care will no longer be rendered by the applications that created them or by readily obtainable alternatives. This applies to all formats not just JPEG2000. As a relatively new and untried format planning for the long term management of JPEG2000 will require some work.
The key challenge with migration as a strategy is not deciding how to do migration but how to identify and maintain the significant properties of the format being migrated. The danger is that some property of the format may be lost during the process. The biggest fear with images being that quality will deteriorate over time. This loss of quality, whilst insignificant in the initial migration, may have a detrimental cumulative and irreversible effect over time.
So, do we have a plan for the future migration of obsolete JPEG2000 files? No, we do not. We are still trying to develop the specifications for the types of JPEG2000 that we want to use. Beyond the pale or not we have accepted that our images will be lossy. What we are trying to do is create JPEG2000 images that are consistent, have a minimal range of compression ratios and have a few variations in technical specifications as we can provide for. As a start this will make long term management simpler, but we are aware that we still have a way to go.
Our promotion of JPEG2000 as a format will hopefully make it more widely accepted and therefore the format will attract more research into possible migration options. We’re pleased to see that already individuals and organisations have been thinking about future migration of JPEG2000. The development of tools such as Planets in recent years has been a great step forward in supporting decision making around the long term management of formats.
Obsolescence is not something totally beyond our control. We are free to decide when obsolescence actually occurs, when it becomes a problem we need to deal with, and, with proper long term management strategies how we plan to migrate from obsolete formats to current ones. The choice of JPEG2000 as a master format supports this broader approach to data management.
The long term management of JPEG2000 as a format is part of our overall strategy for the creation of a digital library. Ease of use, the ability to automate processes and the flexibility of JPEG2000 have all been factors in our decision to use the format.
We’re clear that the choice we have made in the specification of our JPEG2000 images is a pragmatic one. Its also clear that the decision to use JPEG2000 in a lossy format has consequences. However, we have a format that we can afford to store and one that offers flexibility in the way that we can deliver material to end users. For us this balance is important, probably more important than any single decision about one aspect of a formats long term management.