You ONLY technically need to change it on the first page of the developers site under invocation word. For the examples with the app to use that invocation word, go to Settings>>Personalization and go to the very bottom of that page.
Happy Friday to everyone! Only a couple week until Christmas. I will start off by giving everyone a little Christmas present…I have been working with Brad (@bwyble) to find a few pesky Message Queue issues and get them resolve. Brad has a very interesting use case where his Eufy speakers act not only as a synth device, but also has a mic where it acts as Alexa. He can go into further detail of how he got this to work (maybe start a dedicated thread?), but bottom line is that he can now use NATIVE Alexa Routines and still have them come from the same speaker as his ‘Alexa’. So, basically the output of Ask Alexa (for custom reports, for example) can be initiated by an Alexa Routine and output can come from Ask Alexa…Very nicely done Brad! He also had a suggestion for addition switch state as a restriction within the macros. I did this and added it to the message queue as well. Now you can have simplistic decision making ability within Ask Alexa instead of resorting to something like WebCoRE (which I also love, but has a large learning curve).
As such, I have updated the main app and the message queue extension to fix/add these features. If you integrate your IDE with GitHub, these should already be present. I have also updated the wiki information here: http://thingsthataresmart.wiki/index.php?title=Ask_Alexa#Latest_Versions
Thanks to Brad for his patience while we worked through his set up!
Speaking of updates, did anyone see the new little items within the Alexa routines? While it is not free-form text, you can NOW output little phrases with your Routines. I have been testing this for a while and Amazon JUST turned it on this week. In my household this has replaced my Talking Alarm Clock app that I have been using for almost 3 years! I now wake up with a good morning prompt, along with the weather without having to initiate a command or use my Sonos speakers. Nicely done Amazon!
In other news, I look forward to getting my Amazon Spot in the next couple weeks. For those that don’t know, they are a cross between the Show and the Dot. I have a Show in my bedroom currently, and while it is great as a alarm clock replacement, the screen (even when turned off) still gives off a lot of glow. I am hoping the Spot is a better replacement. It still has video capability and I assume when using Ask Alexa will still have the nice icons I spent time creating for the Show. Using this with the Amazon Routine function, I have the PERFECT alarm clock.
As the last word about Ask Alexa, I STRONGLY advise people now if they intend to use SmartThings as the invocation word to split the word into two in the Amazon developer’s area. I have gotten even BETTER results with it recognizing that than the original “SmartThings” invocation. You can also use the room name or something like house or brain!
Finally, while not directly Alexa related, I updated the DTH for the Aeon MultiSensor 6 that I mentioned last week. All props go to Eric (@erocm1231) for the basis of this code….but I love interface design and I optimized this to give a little bit different user experience. Read about it here: [RELEASE] Aeon Multisensor 6 (Advanced). I also started a wiki page on this, which can be found here: http://thingsthataresmart.wiki/index.php?title=Aeon_Multisensor_6 It is still rough, but I continue to work on it!
Thanks for all hard work Michael. I’ve loved working with Ask Alexa, it’s very well put together.
As for the Eufy Genie, it’s essentially an echo dot with the additional ability to act as a smart speaker. They cost about the same as a dot, and as far as I can tell are essentially the same in terms of function except that the Eufy does not have bluetooth, and it has fewer internal mic’s which could affect its far field ability. I haven’t tested it rigorously but so far I haven’t seen a difference.
In any case, what you get out of the Eufy Genie is exactly what Michael describes, it functions as a dot and a smart speaker, so I hear three different voices out of it depending on what program is speaking:
The alexa voice, sounds the same as ever
The Ask Alexa voice, which I have set as Snarky
And the smart speaker synthesized voice which is Sallie (not sure if this can be changed).
During debugging it’s helpful to know who’s doing the talking.
Anyway, it’s really easy to set up, provided that you’re willing to give the Eufy app your amazon password. You can just hit the mute-mic button to disable the alexa pickup while still permitting the standard voice synthesis.
I have a strange problem which I can’t find a reason for. I have a Alexa Routine called “Bathroom” which came on all by itself. The routine when triggered activates a virtual switch called “Bath” which activates the Ask Alexa Macro. The routine triggered on it’s own. The macro turned on the light, and I heard the acknowledgment message, but nothing was in the message Queue. There is no other smart app connected to this light other than ActionTiles. I don’t know how this is possible as nothing shows in the ST app other than “On command sent”. Any ideas would be helpful. Could it be another ghost in my system?
No one wants to hear this, but this DOES happen occasionally. I have had it happen a few times, including to a connected garage door. Hence, I have a switch lockout for my garage door that is a faildsafe.
Bottom line, it DOES happen, but shouldn’t happen all of the time. If it does, then it probably means SOMETHING is tripping the switch…even if it is someone in the mobile apps scrolling and accidentally tripping something (That happens a lot too).
This has happened before with other devices. If it happens a huge amount I would be concerned, but the fact it wasn’t in the Queue was what confused me. I am sure I have a “GHOST” living in my Hub.
I’m sure this is explained somewhere but I can’t find a definite answer. I have Ask Alexa and Big Talker up and running. Love Ask Alexa for the detailed status. However, trying to get notifications (door open/close mainly) through echo without asking. The confusion is can this happen with Ask Alexa and Big Talker alone or do I need to set up a DNLA speaker? http://thingsthataresmart.wiki/index.php?title=How_to_Set_Up_a_Door_Open_Chime/Siren #9 implies a speaker outside of echo is not needed. However, other posts contradict.
So, can I get this to work or do I need to get a separate speaker (samsung R1)?
Thanks in advance!!!
Short answer, if you only have Alexa, is no…you can not wake the speaker up with a skill…you must initiate the conversation. That will occur sometime after the first of the year, but until then, you must have some sort of Sonos or ST connected speaker to hear uninitiated prompts.
Thanks for the quick response! After playing with it that was my conclusion (just some confusion on some of the boards). First time actually playing with Lambda. Love it.
I’m trying to get pistons to rattle off a series of verbal responses using ask alexa and am having some trouble. The pistons and the messages queues work as expected, but when two voice outputs fire in sequence, even with a 10 second delay in between, the second one is usually lost (sometimes I hear a syllable or two). Any idea? Here’s the piston I’m using. Each of these virtual switches triggers a voice report. Those voice reports work individually, which I can test by tripping the switches using Alexa. Moreover, when using those pistons, both those voice reports add messages to the queue even though I only hear the first one on the speaker.
Without knowing the whole sequence of events it is hard to determine what is happening. If this is with your speakers, it probably has to do with two different streams hitting the TTS API close to the same time without one completely finishing. And why run two voice reports? Why not just combine them?
Right, it’s pretty simple though. It triggers one voice report, then triggers the second one 10 seconds later. I agree that they’re probably conflicting but I was surprised that they’d conflict so far apart in time. Even with 20 seconds it wasn’t very reliable. I was wondering if there’s some lingering process after a voice report that is making it hard to trigger again, but it’s probably hard to know what’s going on inside the speaker’s DNLA code.
Re: why not combine two voice reports,
I was running into a different problem there, in which the combined utterance was too long and got clipped. This piston is an attempt to avoid that problem.
The control of the speakers is definitely in the DTH, not Ask Alexa. There might be some ‘recovery time’ or something else that the speaker is going through when it switches from one voice stream to another. I did look and I could probably find a way to output an Extension Group out to the message queue if you think that would help. It would basically allow you to combine voice reports into one report that could then be sent to the message queue. What I am afraid of is that there is a limit to how much the speaker can speak before it clips the end of a stream. Have you tried doing a VERY long weather report and see if you run into issues. If not, I will see if I can get the output of an extension group to output to a message queue.
Nope, haven’t tried that yet. I’ll run some more tests. Thanks!
No problem…adding the output of an extension group is on my roadmap. I can simply speed up the development of that if you find that a REALLY long TTS will work through your speakers. I would test the long weather report through the normal Alexa output as well…there is a limit to what it can speak as well and there is no use bandaiding one problem for another.
I haven’t done the weather report yet, but if I put the delay between the two webcore switch commands to 20 seconds, they both fire every time. That’s reassuring, at least. It’s not completely obvious how to ensure that it executes these things in sequence, rather than in parallel.
How are you putting a delay in? Unfortunately, from a developer side there is no ‘delay(20)’ command.
I did play with adding an ‘output to message queue’ in the extension group macro, so that might also be a solution long term.
I’m adding the delay through webcore. There’s a command to wait for some specified length of time. Here’s an example. Note the “async do” and the “async with” Those are important to make sure all of that stuff happens in sequence.
Ah…within WebCoRE…that makes sense.
Thanks. Even without the more advanced webcore integration, it’s really nice to be able to bounce back and forth between Ask Alexa and Webcore.