vt2-6 still not available with base system

Simon McVittie smcv at debian.org
Sun Nov 9 20:43:36 GMT 2014


On 09/11/14 18:56, Michael Biebl wrote:
> Running for i in `seq 2 6` ; do systemctl enable
> getty at tty$i.service; done on a system without dbus, would
> statically enable a getty on vt2-6. So we could simply document
> this fact (in README.Debian, release notes etc).

I'm very tempted to say systemd in Debian should just make these 5
symlinks to preserve traditional Debian behaviour, at least for
jessie. It doesn't really even need to run systemctl, it can just use
ln -s or whatever. People who are so desperately short of memory that
5 agetty instances are undesired can always disable them.

If you want it to be clear that it's a Debian-specific thing that's
pulling them in, maybe it would be better to add a new
debian-gettys.target or debian-gettys.service that Wants them, and make
it WantedBy an appropriate target?

We can reconsider in a few months for jessie+1, at which point these
will hopefully be clearer:

* how many people switch to systemd and how many stick to sysvinit
* whether kdbus is likely to be merged by the time the kernel
  for jessie+1 freezes, making dbus-daemon entirely
  unnecessary on systemd systems
* whether we can get rid of init from Essential so chroots don't need a
  /sbin/init (yes I know chroots don't necessarily have
  Priority:important)

> The alternative would be, to bump dbus to priority [important], so
> it's installed by default (on new installations). I think Simon
> wasn't to keen on raising the prio of dbus further. I've CCed him
> for his input.

I'm not delighted by the idea of dbus-daemon escalating from standard
to important. I install it on my own headless machines, but I know a
lot of sysadmins are more change-averse than me.

Also, installations with sysvinit (or no init at all) really shouldn't
need it, and we don't have a way to make it "conditionally important"
depending on init system (other than the Recommends that systemd
already has).

> A third idea would be to have a generator, which creates the 
> getty at tty$i.service symlinks in
> /run/systemd/system/getty.target.wants Implementing that in shell,
> would only take a couple of lines.

My instinct is that that just sounds like an unnecessarily complicated
way to add 5 symlinks or a Wants that could equally well be there
statically.

A generator could be conditional on the absence of dbus.service, or
some such, but then again so could a debian-gettys.service with
ConditionPathExists=!/lib/systemd/system/dbus.service.

    S





More information about the Pkg-systemd-maintainers mailing list