A crucial set of data structures that must be translated are the User File Directories (UFD's), which are similar to the directories of today. The UFD's are essential for preserving any file links in a directory. Furthermore, on the incremental backups, one cannot tell what files were on the original file system but were not included on that incremental without decoding this information. Therefore, it is critical to make some sense out of these data structures since we cannot depend on future archivists to decipher this raw information.
The UFD is a much more difficult structure to interpret than the MFD for two reasons. First, the UFD keeps critical information in structures that can be decoded only by interpreting PDP-10 byte pointers. Second, the UFD uses a custom method to track disk block allocation, which must be interpreted to determine file length.
The archivist software translates UFD's into an ASCII text directory listing, and records this and the raw binary data in the TCFS output. The directory listing in Fig. looks nothing like the one an ITS system would produce, but it is meant to include every single bit of information that was originally in the UFD. This directory listing is not particularly valuable to the average user who is searching for files in our TCFS archives: it is useful only to someone who wants to see what files existed on disk at a particular time.