HS4 Feature Poll (Pick 5 features)

[macromark]

So...................

Based on the results of this poll, we set about to implement "Conditional Actions" and work is nearly complete on that. We'll be releasing that for testing very shortly (in approx 2 weeks). Click Here to check our progress

As promised, we're cycling back to this poll to pick the next feature to address. Before doing so, we're going to create another poll which includes the most popular entries on this poll, PLUS any new ones that may come to mind. For example, we didn't include "customizable dashboard" in this first poll. So, that will be added.

Question to the group: Are there any other feature requests that we may have missed? Let me know... and they just might make it to the next poll!

[mrceolla]

I've been in contact with Actions on Google developer support. I explained the utility of sending the entire spoken phrase through to the 3rd party for processing and asked what solutions they currently have that allows for this besides Conversational Actions. After some back and forth, they suggested this:

Fulfillment | Actions on Google Smart Home | Google Developers

This sort of looks like Conversational Actions within the Smart Home SDK. It uses a webhook to send info to a 3rd party fulfillment web service, and allows for a back and forth if more info is needed. The Local Home SDK is listed in the left side menu as well. I very much need to continue to use Jon00's HomeSeer/Echo Skill Helper, which can be hopefully renamed to Jon00's HomeSeer/Echo/Google Helper.

[avpman]

Great! Since the LR doesn't depend on mesh "neighbors" looking forward to more reliability.

If the boards are the updated version, it should be supported.

[randy]

If the boards are the updated version, it should be supported.

[avpman]

Are you saying the chip in the current 700 stick can xmit LR? And that it might only be a code base and possible firmware upgrade to do so? This would be awesome.

Our current 700 series boards should work with it, and I don't believe the firmware needs to be updated, but if so, it can be updated.

We are looking into it. It is doable with our existing code base, but the node ID is now 16 bits vs 8 bits. This is a big code change so it will take some time. I cannot give a timeframe right now but will post something after we scope it out.

[mrceolla]

I'm curious: Outside of local control, what is the use case for sending the spoken text to HomeSeer if using Google Assistant/Speakers/Display?

In short, it allows you to accomplish anything you want, by saying anything you want, and getting the response you want. :-) Ultimate flexibility! You can simply utilize HomeSeer's built in speech processing or create your own commands and responses with Jon00 Homeseer/Echo Skill Helper for Homeseer 3 & HS4 - HomeSeer Message Board.

Google's Smart Home/Device support is far from perfect. They'll never be able to support everything. This type of functionality fills that gap and adds so much more.

[ksum]

I'm curious: Outside of local control, what is the use case for sending the spoken text to HomeSeer if using Google Assistant/Speakers/Display?

[Sgt. Shultz]

OMG! Why would they eliminate such useful functionality? What is their reasoning? What is their angle?

I sure hope HST can bring this ability to Google speakers/displays with the Local Home SDK. Otherwise, Speaker.exe, here I come, again.

Or will Amazon allow the existing skills to function indefinitely? Something tells me no.

As long as we don't try to resubmit the original skill or deactivate it, it's grandfathered in. They won't put in a new one, but they won't take out what's in there now.

The best part of that skill is no discovery is needed..., just put the new device in and allow voice and you're done.

The timers are also useful, as you can set up something to occur for a short time.

Like:

"tell homeseer to turn the light on for 2 hours"
"On monday at ten am unlock the garage side door for 1 hour"
"In thirty minutes, turn off the bedroom air conditioner"

[mrceolla]

OMG! Why would they eliminate such useful functionality? What is their reasoning? What is their angle?

I sure hope HST can bring this ability to Google speakers/displays with the Local Home SDK. Otherwise, Speaker.exe, here I come, again.

Or will Amazon allow the existing skills to function indefinitely? Something tells me no.

[jon00]

I'll switch to Amazon if there is no longer a way to send the spoken text to HomeSeer. Dicks!

Amazon don't allow this either now. It is fortunate that the HS Alexa Homeseer Home Automation skill which was written before the change and why other countries can't get this skill

[mrceolla]

I don't give an F what they're keen on. That functionality is essential. I'll switch to Amazon if there is no longer a way to send the spoken text to HomeSeer. Dicks!

[Sgt. Shultz]

They weren't exactly keen on us parsing the commands instead of them, but in the early days they didn't have a smart home structure yet.

(They also weren't keen on my idea of using a digital radio station process to make unprompted announcements..., go figure.)

[Sgt. Shultz]

Are you sure about that? Conversational actions is different than direct actions. And from what I understand almost nobody was using conversational actions anyways. I believe what we had before was a direct action.

Direct Action is what we have now, where a conversational action is what we had originally.

[mrceolla]

That said, Google has decided to sunset what are called "Conversational Actions" in their development console, which basically means no text passthrough with a standard skill. (It will officially sunset June 2023, but there's no sense building something that'll be obsolete in 6 months)

Are you sure about that? Conversational actions is different than direct actions. And from what I understand almost nobody was using conversational actions anyways. I believe what we had before was a direct action.

[Sgt. Shultz]

I've been waiting to push for this again once HS4 was more mature and you all hopefully had time for something like this.

Please improve Google Home/Assistant integration. I see two potential improvement areas.

1) Remember your first integration with Google Assistant and HomeSeer where you had to say "Hey Google, ask HomeSeer" followed by a command? That should be brought back. The ability to send the spoken text to HomeSeer is powerful. This would allow Google users to be able to make use of Jon00's excellent Echo Skill Helper, or simply let HomeSeer's native speech parsing engine do the work. I believe HST communicated that they needed to ditch the "ask HomeSeer" method in order to make room for the Google Home smart device integration. I was in a long back and forth with Google developer support about this and they told me there is no reason HST couldn't use both of these integrations at the same time. Frustrated by my inability to do with Google what others were able to do with Alexa (direct device control + "ask Homeseer"), I rolled my own solution, but I'd love to see this available for everyone.

2) I understand Google now has a Local Home SDK for local fulfillment. I'm not 100% sure what this is or if HomeSeer can leverage it, but if so this is perfectly inline with HomeSeer's strength's of not needing the clound to function. Additionally, this would allow for much speedier response times. I often experience varying amounts of latency when giving commands to my smart speakers and displays. This Local Home SDK may also allow for something like "ask HomeSeer" functionality so it could kill two birds with one stone.

I hope these can be considered. If not now, I'll revive my old threads in the Google Home sub-forum another day for another attempt.

I haven't been clamoring for anything in that list above, but I will pick my 5 favorite anyway. The two items I wrote about above would be in that 5 if available.

Thank you for reaching out to us to help guide prioritization!

I'm currently looking at the Local HomeSDk to see if we can do something with that.

That said, Google has decided to sunset what are called "Conversational Actions" in their development console, which basically means no text passthrough with a standard skill. (It will officially sunset June 2023, but there's no sense building something that'll be obsolete in 6 months)