[pulseaudio-discuss] [PATCH] Added set-(sink|source)-latency-offset commands to pactl and pacmd.

[Tanu Kaskinen]

On Tue, 2016-01-12 at 17:05 +1100, Chris Billington wrote:
> Hi all,
> 
> Just to bump this thread, what are the odds of getting this patch included
> in pulseaudio? It would be useful to many of the network audio streaming
> services that use pulseaudio (specifically my one :p).

The odds are good, since this is a welcome addition. I don't know when
the patch will be merged, however (in any case it's too late for 8.0).
I don't promise a quick review myself due to lack of time. Peter
already commented, though, and he is able to approve patches. Peter, do
you plan to review this newest version of the patch?

I do have some comments from just glancing at the patch, though:

It would have been nice to separate the patch to the part that modifies
the protocol and the part(s) that add the new feature to pactl and
pacmd.

The commit message should contain the justification for the change
(which you provided in a separate mail).

The references to 8.0 need to be changed to 9.0.

It would be good to have better separation between the sink/source
latency offset and the port latency offset. The patch doesn't seem to
handle it particularly well if a sink has a latency offset set and the
same sink also has a port that has a latency offset set. It would make
sense to add those two together when querying the sink latency.

It would be good to save the offset on disk. The place to do that would
be module-device-restore. If you try to implement this, however, it
might turn out to be a bit tricky, because module-device-restore deals
with both per-sink and per-port entries depending on whether the sink
has ports. The sink latency offset should not be saved separately for
each port, but it's not obvious how to achieve that when module-device-
restore is in the "per-port mode".

-- 
Tanu