When extracting "Any Good" data from a track (dumping or in hexa-editor), there are currently just two possibilities - either absolutely healthy data are found or bad data from the first revolution are returned.
This is not acceptable because:
There may be "less bad" data after the first revolution, i.e. with less errors indicated by the hardware.
Some errors are less severe than others, e.g. CRC error in data is less severe than CRC error in ID.
Some (but not necessarily all) errors may be ignored, e.g. the "Accept" button during dumping. For instance, the "sector not found" error cannot be ignored.
This issue should therefore solve arbitering of "least bad" data, given indication of which errors may be ignored.
When extracting "Any Good" data from a track (dumping or in hexa-editor), there are currently just two possibilities - either absolutely healthy data are found or bad data from the first revolution are returned.
This is not acceptable because:
This issue should therefore solve arbitering of "least bad" data, given indication of which errors may be ignored.