Re: jessie and stretch time problem
On Sun 14 Oct 2018 at 17:21:43 +0300, Reco wrote:
> On Sun, Oct 14, 2018 at 09:22:53AM -0400, Cindy-Sue Causey wrote:
> > On 10/14/18, firstname.lastname@example.org <email@example.com> wrote:
> > > On Sat, Oct 13, 2018 at 07:11:07AM +0000, Long Wind wrote:
> > >> Thank tomas!
> > >>
> > >> i run "cat /etc/adjtime" in jessie and it says LOCALbut in stretch
> > >> /etc/adjtime doesn't seem to exist
> > >
> > > I see. Perhaps this is the root of the problem. On the other
> > > hand, perhaps, systemd is taking care of time in stretch by
> > > default. In that case I can't help, since my knowledge of
> > > systemd is minimal.
> > I was going to bring this up yesterday, just couldn't rope the words
> > in together. With debootstrap, one step the user manually does is
> > this:
> > +++ QUOTE k/t https://www.debian.org/releases/buster/amd64/apds03.en.html +++
> That guide contains curious relics of the past that include, but aren't
> limited to:
> 1) 'lo' interface in /etc/network/interfaces - unneeded even with
> stretch's ifupdown.
Good catch. It could be omitted or tagged as optional, as broadcast
and gateway are in the following stanza.
> 2) 'hostname' record in /etc/hosts - unneeded with libnss-myhostname.
The 127.0.1.1 line? It took us ages to get to this - something which
> 3) '.local' domain at /etc/resolv.conf - a direct violation of RFC 6762.
An example - but a very bad one. Another good catch.
> 4) At controversial (to say the least) recommendation to use 'deb-src'
> of Debian main suite without actually specifying 'deb'.
These are additional examples. Unless I am misunderstanding you, it's
> 5) The /etc/adjtime configuration itself.
> Systemd defaults to UTC - src/timedate/timedated.c in systemd source
> If needed, one can use timedatectl(1) to produce perfectly valid
That's correct but I've never experienced the installer not giving a
user an /etc/adjtime.
At present there is an effort to update documentation for buster. Get
your bug report in. Don't delay!