Hi, Am 04.06.2014 09:34, schrieb Martin Pitt: > brltty's systemd unit unconditionally starts the service. This > mis-matches the behaviour of the init.d file which checks > RUN_BRLTTY in /etc/default/brltty. The attached debdiff adjusts the > unit to do the same. It's not pretty, but seems to work well enough. > > In the long run this hackery should stop of course; systemd units are > rather fundamentally incompatible with the idea of enabling/disabling > units with a default file instead of using "update-rc.d > enable/disable" (which works for SysV init, systemd, and upstart). But > as long as we have these options in the default file they should be > respected. > > Alternatively, to fix this bug it also seems fine to me to drop the > RUN_BRLTTY option altogether. I'm against adding such hacks to the .service file. Please don't do that. Please use the migration to native .service files to get rid of it. If you don't want the service to be started by default, simply use dh_systemd_enable --no-enable (or use a Condition*, if that is possible). While unfortunately there is no equivalent (yet) for update-rc.d, you can simply add a short comment to /etc/default/foo, that the RUN_BRLTTY setting has no influence on the systemd service and one should use systemctl enable|disable instead and that RUN_BRLTTY is only for the legacy system. I don't want to see those RUN flags in /etc/default/foo dragged over into systemd service files. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth?
Attachment:
signature.asc
Description: OpenPGP digital signature