Originally posted by per View Post
When a file in a CP/M directory fills up it's last directory entry completely (typical when the filesize is an integer multiple of 16K), should a final "empty" entry be added for it?

From what I can read, when a directory entry is full, another might be expected. Observing a couple of disk images, I see that files of 16k or 32k might have a final entry without any allocations, but it is not always so. Is the empty entry a more proper/safe way, or is it just the waste of an entry?
I've seen it happen--but saying 16 or 32K is the maximum size of an extent is quite misleading; depending on the disk parameter definition, they can be considerably larger.

Even more curious is the case of "sparse files"--that is, random files with a "hole" in them. You can have a file which, logically, is larger than the physical disk. AFAIK, MS-DOS has no analogous mechanism.