[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: "clean" to be replaced with "flawless" or "faultless" ?

Hello Justin,

thank you for reading further sources. But I doubt whether the point
is worth further correspondence. Though, is a see a silver lining as
Theodore Ts'o, the expert for the filesystem tools fsck, e2fsck, etc.

- is the author of manpages fsck(8) e2fsck(8) (released August 2014) 

- answered as well to the bug report you found
  https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=48368#9, dated
  from October 1999.

- explained there that there is no "dirty bit" in ext2 filesystem (but
  a valid bit) that simply reflects whether regular umount has been

- suggests, with his remarks on option -c in tune2fs(8), that the
  journaling filesystem (afaik ext3 and ext4) have no "dirty bit"

- does not use the word "dirty" in fsck(8)

- concedes that "dirty" is "vernacular"

- refers with respect to the ext2 file system to a "valid bit" in the
  superblock. This is in accordance to the explanations in the
  mentioned book of Michael Kofler.

Thus there is a certain probability that Theodore would accept
alternatives for clean / dirty, e.g.

attributes for the state of a file systems with respect to usage
+                          -
clear                      not clear
ready for mounting         not ready for mounting
                           needs being checked

attributes for the state of a file systems with integrity
+                          -
flawless                   has flaws 
faultless                  faulty
consistent                 not consistent

Best regards

Justin B Rye schrieb am 10. August 2016 um 01:01

> Markus.Hiereth@freenet.de wrote:

> > If, as you explain, "clean" just means "dirty bit not set" and
> > actually does not reflect the result of a file system check, the
> > adjective "clean" should not be used.
> I have to admit I'm not completely confident about what fsck does mean
> by "clean".  Usually when fsck finishes without an error, that means
> that it has rendered the filesystem "flawless", but does it intend us
> to understand "clean" as covering all of that or just the relatively
> trivial final step?

> A further complication is that a filesystem which is "clean" in the
> sense of having been unmounted in an orderly fashion may nonetheless
> have errors - that's why filesystems get periodic checks.
> >> Perhaps some equivalent of "unmarked" would work?

> > I would not use "unmarked" (en) or "unmarkiert" (de) as it raises the
> > question what kind of mark it refers to.

> > In a Linux-Book for Beginners, the author Michael Kofler speeks of a
> > "valid bit" in the superblock of an ext2 file system. The description
> > there would favor usage of the adjective "clear" in a message that
> > states that there is no need to check the data on a file system.
> Googling, I find an ext2-era comment in the BTS:
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=48368#9
> which leaves me even more uncertain.
> Also, tune2fs(8) on the -c option claims that "If you are using
> journaling on your filesystem, your filesystem will never be marked
> dirty".  So why does fsck.ext4 still declare it "clean"?

Reply to: