It is Mac OSX yosemite 10.10.2.
But it is a special computer being monitored by the institute.
I am not sure what kind of software they use. But all the activities in the
computer will be recorded.

On Tue, Mar 31, 2015 at 5:54 PM, Pádraig Brady <P@draigbrady.com> wrote:

> On 31/03/15 21:29, Dake Zhang wrote:
>
> > FAIL: tests/misc/ls-time
> > ++ ls -ut a b c
> > + set b c a
> > + test 'b c a' = 'c b a'
> > + fail=1
> > + test 1 = 1
> > + ls -l --full-time --time=access a b c
> > -rw-r--r-- 1 dakez BCM\Domain Users 0 1998-01-14 11:00:00.000000000
> +0000 a
> > -rw-r--r-- 1 dakez BCM\Domain Users 0 2015-03-31 20:18:25.000000000
> +0000 b
> > -rw-r--r-- 2 dakez BCM\Domain Users 0 2015-03-31 20:18:25.000000000
> +0000 c
>
> Strange. The access time for b and c should have been set to 1998 like
> it was for "a".  Now we only verify that a's access time was set,
> so we could be more robust and test all 3.  Does you know if there
> might be any scanners etc. on your system accessing new files?
> What file system is this?
>
> > FAIL: tests/cp/link-symlink
> > + touch file
> > + ln -s file link
> > + touch -m -h -d 2011-01-01 link
> > + case $(stat --format=%y link) in
> > ++ stat --format=%y link
> > + cp -al link link.cp
> > + case $(stat --format=%y link.cp) in
> > ++ stat --format=%y link.cp
> > + fail=1
>
> The above is again a timestamp issue.
> Now there are extra gating checks in tests/cp/link-deref.sh
> which might be appropriate to add to this test,
> though perhaps again there is something else interfering with timestamps?
>
> What operating system/version is this exactly?
>
> thanks,
> Pádraig.
>