--On Wednesday, January 28, 2009 11:40:41 AM -0600 Nicolas Williams <Nicolas.Williams at sun.com> wrote:
> On Wed, Jan 28, 2009 at 09:38:16AM -0800, Glenn Faden wrote: >> Nicolas Williams wrote: >> > One reason to not use a timestamp is to avoid issues with any future >> > fix for the RFE that zones' clocks should be able to have time offsets >> > w.r.t. the global zone. >> >> Since the proposed cron job code only executes in the global zone. the >> zone clock issue isn't relevant. > > What I meant is: would a stat() in the gz of a file in an ngz return > the same timestamps as a stat() in the ngz of the same file? If yes, > then -newer is dangerous. Wanting the answer to be "no" might > complicate that RFE, for all I know. The proposed code executes only in the global zone, on files in the global zone. The clocks of non-global zones are completely irrelevant.
