Description:
For background, please read the main paper on this topic.
There also is a shorter but newer discussion.
This effort is on technical problems related to preservation.
IBM would give us software for the UVA assembler and interpreter,
as well as documentation, for Windows. One possible plan is:
- get familiar with the methodology
- identify a sound file format for input
- determine how to get at the information that it contains (either by writing a C or Java program or using some available tool)
- the data has two parts:
- part 1 contains some descriptive data: track, duration, etc
- part 2 contains the sound data itself (assume it is in a simple wave form).
- choose the data structure of the stream that will be archived to save both types of data sequentially in a single data stream.
- at the same time, choose the logical format (XML-like) that we want the future user to see.
- write a UVC program that decodes the data and returns it in its logical form.
- Similarly,
- choose the data structure of the stream that will be
archived to save the metadata (the description of the logical view).
- the logical form (XML-like) for the metadata is fixed by our architecture.
- write a UVC program that decodes the metadata and returns it in its logical form.
- Develop an example of a Restore program (would normally be written and executed in the future).
- it invokes the UVC interpreter and executes the programs to get the data and metadata;
- it would display the metadata; it would also display the part 1 data and recreate a sound file (of any format) that can be played;
- it should sound like the original one!