From the file format point of view, we note that this technique could be applied to any box-structured file. However, at first glance this would seem to require a schema for each specification (e.g. a schema that detailed all the attributes and contained boxes in ISO/IEC 15444-12). This is both a lot of work, and also risks divergence between the base standard and the XML. We wonder (without any evidence) whether a set of rules could be written that unambiguously defines how to derive such a schema from any specification using box structures. Such a set of rules would cover the conversion of fields, data values, lists and loops, contained boxes, the handling of un-representable binary data, and so on. There is a real tension here between representing the file ‘as it is’, and knowing about e.g. such matters as offsets and data pointers, self-contained and referenced files, and so on.
