I’ve run into what might euphemistically be called an ‘undocumented breaking change’ in 2023.05 although I consider it a bug.
If you use tts.service_say
then the media player(s) used for the announcement do not change state to reflect that they are now playing
. Neither do any of the attributes update to reflect the real world situation.
The most likely noticeable effect of this for average users will be ‘announcement clash’ where one announcement is playing when another is requested, so cutting the first one short.
For me this is a fairly big issue as I need to know when announcements have finished playing and with this change there is no way to do so.
I’ve had a brief informal conversation with the developer involved with implementing the announce
parameter and they are aware of the situation and they have suggested that they might be able to look into it but that there no promises
I’d like to raise an issue/bug for this so as to raise it’s visibility/profile but I don’t want to,
- waste my time (I have heard how issues/bugs can get shut down without any real explanation)
- ruffle the devs feathers because just for the avoidance of doubt, I think the new
announce
feature is a good thing.
However…
I think it is fairly a fundamental principle that device states (and attributes) should reflect the real world
What does the panel think?
Is it a bug worthy of escalation or am I getting upset over nothing?
PS. I use Sonos. I don’t know if it affects all media players.