In message <20050126184701.GA25697@panix.com>Thor Lancelot Simon writes
>On Wed, Jan 26, 2005 at 12:10:32PM -0600, James Chacon wrote:
[...]
>> That wasn't the point. What if I have a stable non-busy system? sync(8)
>> lies currently and doesn't force all data out to the disk...
>
>It always lied, thus "sync, sync, halt".

or  "sync, sync, *sync*, halt"! Heretic! Burn [sic] her!

seriously, in the days of RM05s one could audibly hear the difference
gbetween the second and third sync; on an idle system, the third sync
updates very little (the superblock and maybe a single inode? I forget). 
Subsequent syncs 

If "sync" (or repeated syncs) on an idle system doesn't result in
synching the filesystem (data and metadata) to stable storage, that
strikes me as a bug.