file size not available for deferred datasets

I don't know what should happen then, but I suppose a lot of rules make the assumption file size is available, which will be broken with deferred datasets. I suppose you'd want to document that somewhere ?

Start by locating the rules that assume file size is available and the handling of deferred datasets. Determine which assumptions need to be documented and whether any behavior must be clarified first. Done means the deferred-dataset file-size limitations and affected rules are documented, with any required behavior decisions recorded.

Written by the indexing model from the issue text.