Amazon EchoControl LastVoiceCommand doesn't update anmore

I can confirm lastVoiceCommand works fine with OH4.3.

this means you didn’t get any “too many requests”-errors in your log?

I increased the delays to ridiculous 36000 sec but still get the error. So it seems it’s independent of those settings. If 4.3 (I assume it’s the smarthomej-version) works fine then something is wrong in the 5.0 version.

I don’t, or should I check the verbose logs for those? I use the smarthomej version and OH 4.3.5, delay is practically nonexistent.

Regarding the echo models, I had bad experience with the newer “half-ball” models so switched back to the flat ones. The half-balls didn’t seem to react or hear me as well, or somehow they just didn’t work as well, so I now use them in rooms that are not that much used.

do you trigger the refresh channel or do you have even no delay while not using the refresh channel at all?

That’s funny. I went the other way and got rid of the old ones due to the worse sound quality.

But I experience that the other trigger words (except ”Alexa”) are not always detected too.

just to chime in on “not working”! :wink:
I’m running OH5.0.1 on docker with the core “Amazon Echo Control Binding”.

I upgraded to OH5.0 since it was released and it did not work, also now on 5.0.1.
I do get both: 429 and 503 errors from Alexa API, mostly 429: too many requests.

sometimes it comes through, but most of the cases it doesn’t. I do not use some kind of “trigger”, but an item linked to the corresponding lastVoiceCommand-channel. There’s no visible change since the upgrade von 4.3 to 5.

I trigger the refresh channel at the end of Alexa routines like this.

thanks, that’s the same way that I’m doing it. Overall it’s working pretty fine now.

@binderth : I don’t think you will get reliable lastVoiceCommands without using the refresh channel.

didn’t need it for years. It just appeared with the whole API change thingy - and I’m still convinced it’s some kind of excess quota problem, as the multiple 429-errors indicate.

For sure. Unfortunately Amazon dropped to push the lastVoiceCommand immediately after the command via API. It seems the binding requests updates at least every 10sec, independent of the settings. Maybe there’s a relation to the 429-error.