[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Strange issue with brltty



Shérab <Sebastien.Hinderer@ens-lyon.org> writes:

> Responding to myself to provide a bit of additional information:
>
> Shérab (2015/08/20 14:19 +0200):
>> Dear all,
>> 
>> I'm observing something with brltty that I do not understand.
>> The brltty that is started at boot is configured one way, but if I do
>> /etc/init.d/brltty stop && sleep 2 && /etc/init.d/brltty start
>> 
>> Then I get an instance of brltty that is configured differently.
>> 
>> As if the two processes were not using the same preferences file.
>
> Indeed. The brltty that is run at boot time uses /etc/brltty.prefs,
> whereas the brltty started manually uses /var/lib/brltty/brltty.pres. So
> what remains to be understood is why it is so and any idea on this point
> will be warmly appreciated.

Do you have START_IN_INITRAMFS enabled?  If so, the problem is likely
due to the fact that /var/lib/brltty/brltty.prefs is not copied to the
initramfs during creation.  This is arguably a bug.  However, BRLTTY
used to restart during boot if START_IN_INITRAMFS is enabled, to avoid
other problems around BrlAPI sockets going lost.  I guess we need to
re-evaluate the whole START_IN_INITRAMFS situation now that systemd is
being used.  The systemd services file does *not* do the restart.

It appears as if Debian is planning to switch away from initramfs-tools
to dracut.  If this happens, we will have to rewrite the
START_IN_INITRAMFS support anyway.

-- 
CYa,
  ⡍⠁⠗⠊⠕

Attachment: signature.asc
Description: PGP signature


Reply to: