On Thu, Feb 10, 2005 at 06:30:41PM +0100, Marco d'Itri wrote: > On Feb 10, "J. Bruce Fields" <[EMAIL PROTECTED]> wrote: > > > In something like df I think the sensible thing to do is just to > > continue to list all bind mounts, even when it's redundant. > The sensible thing to do is to check major/minor and report file systems > only once.
The default df format doesn't list filesystems by major/minor, it lists them by mountpoint. This makes sense, because most people don't know the major/minor numbers of their filesystems off the top of their head (I don't) but do know the mountpoints. So you have a choice of: 1. As currently, list each mountpoint on a separate line. 2. List each filesystem only once, under one (arbitrarily chosen) path which it is mounted on. 3. List each filesystem only once, with a list of mountpoints. Number 2 seems wrong. Number 3 makes sense to me for a graphical tool, where you can probably make space for a list of names on each filesystem if you lay out the display correctly. But number 3 seems a little cumbersome for something like df. --b. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]