Subject: Re: large inode numbers
To: None <tech-kern@NetBSD.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-kern
Date: 12/17/2003 01:37:43
>>> I thought that too, but: 0-sized files to not have a first block
>>> number..
>> Oh, feh.  Ok, back to the drawing board.
> Huh?  Can't one reserve inode "0" for all empty files and offset all
> other inodes from "1", say?  What am I missing in my ignorance?

Do all zero-size files have the same permissions, owner, or other
attributes that Unices associate with inodes?

Even though I think this was talking about FAT, where there really
aren't any inodes (or at least, to the extent that there are, they're
synonymous with directory entries), I think software may get upset at
statting two different files, with different permissions - or even
different types! - and getting the same inumber.  (Unlike the
inumber-collision problems, though, I have no specific examples to
cite.  Since inumber collisions were considered acceptable enough to
commit unionfs in its present form, presumably this risk can be
considered ignorable since it's surely even less.)

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B