Premise MediaZone Usage or Controlling Whole House Audio

[etc6849]

I am thinking the safest way might be to setup a cloud service with Google or AWS and have it act as a middle man between Premise and IFTTT, adding a long hash to the URL that acts as a password.
 
I am not a web developer and don't work in the tech industry so maybe I am missing something, but at least then the Premise webserver isn't wide open.
 
I'm not sure why the webhook action can't let us edit the header.  It has OPTION as a possible method, but no where to put the header information?  That is how I fixed this issue before when I setup my custom Alexa skill on AWS.
 

Hmm, having the same issues as Ellery now.  Hate to give up the security though since it's wide open then.

 

[samgreco]

Did you every get this to work Ellery?  I'm just not grasping the Webhooks part, I guess.  So if you got it working, please post how.  I learn best by example

 

[etc6849]

I'm not going to run things without any security (it does work then). Someday I will setup a go between website using a cloud service and may use it then.

Hopefully Google comes to their senses and opens up their API instead of having people use this half baked option. It is better than nothing so it is neat Google worked with IFTTT to open things up some, but I'm going to keep using Echo dots for now.

If someone gets IFTTT webhooks talking to Premise using a username and password, hopefully they post it.

samgreco, on 08 Apr 2018 - 18:43, said:
Did you every get this to work Ellery? I'm just not grasping the Webhooks part, I guess. So if you got it working, please post how. I learn best by example

 

[samgreco]

I didn't even know you could upload to their play infrastructure.  Do you have any more info on that?  Maybe we could create some kind of "Auto exporter" similar to what 123 is doing for home remote.
 
As for my google setup, its pretty easy.  I used etc6849's tasker module (http://cocoontech.com/forums/topic/28633-video-premise-tasker-module/?hl=+tasker) This accepts an HTTP command (initally designed to receive from tasker, but can receive from any source) that can control premise objects.
 
He initially designed it to have an android phone using tasker's voice add-on to intercept google assistant requests and send them to premise if it met some criteria.  I used IFTTT.com to create an applet for google assistant.  Mine is configured as a custom google assistant command "Ask the house to ...". Whatever follows that is pushed to my premise setup using the webhooks object.  Using the web address: http://yourip.com:yourport/sys/{B751501F-9DAD-4EAE-8680-5175EBCF8B83}?d??mbNaturalLanguage(<<<" {{TextField}}">>>) where {{TextField}} is whatever follows the command.  Etc6849's tasker module breaks down the natural language and processes the appropriate premise objects.  
 
This way I can say "Hey google, ask the house to turn on the basement lights" and it will find all the light objects in premise in the location basement and set the powerstate property to true.
 
Its pretty slick, although doesn't work for my Nuvo, hence all these questions above.........
 
Questions?

 
Can you share your Webhooks setup so I can see how you did it?  I could not get anything to trigger at all.  Like etc6849, I am concerned about the security side.  But I would love to see how you got it to work.

 

[etc6849]

We can now use Autovoice via Google Assistant across multiple devices to control Premise. I'm not sure when this was added. The commands even work from my smartwatch (Android Wear OS, but had to update it).

I just say "ask autovoice, [some command]" into my Android TV remote, Android Wear OS device, Android phone or Google Home and then Premise responds.

How it works:
The AutoVoice app is linked to an AutoVoice account (that is in turn linked to my Google account), and receives the spoken text from the cloud for processing on my smartphone (so my phone needs to stay on for my TV remote control to control my home, a little confusing I know).

There is a javascript in Tasker (Android app stored on my phone) that receives whatever text was spoken to Autovoice, and sends it to my HA server. Tasker, then reads aloud the response from my HA server on my phone after the command has been processed. In theory, if you wired a spare Android device to a matrix switcher hooked to a whole home audio system, you could get a response in whatever room the command is for (by writing some custom code under the SpeechParser Home object to switch to the smartphone's output to the appropriate room via a matrix switch).

The AutoVoice cloud account (free with app purchase) the AutoVoice app says it also supports Amazon Alexa now, so it seems like an excellent bridge across two of the main voice search services, and is cloud based with nothing to maintain except the Tasker javascript stored on your phone. Plus, you can have HTTPS and use login credentials.

https://joaoapps.com/autovoice/

samgreco, on 21 Apr 2018 - 16:11, said:
Can you share your Webhooks setup so I can see how you did it? I could not get anything to trigger at all. Like etc6849, I am concerned about the security side. But I would love to see how you got it to work.

 

[Shinyshoes]

As requested, here is my webhooks setup.

 

[samgreco]

So here's what I've discovered.  Yes, IFTTT works wonderfully without authentication.  And it does actually pass http://usernameassword@ to my premise server.  But my server/Premise still pops up a box for username and password, it doesn't work.  That's where it's failing.  So the question is: 
 
Does anyone know how to pass Username and Password via a URL that bypasses the popup?
 
Now, I do have Autovoice working as expected.  But I really dislike the "ask Autovoice" invocation.  I find it awkward.  And I thought I had it figured out with the Google Home routines.  But, unfortunately, that doesn't support variables.  Interestingly, Autovoice in Chrome on my PC let's me change the wake word, so a simple "Janet, turn the light on in Sam's Office" works perfectly.  When the mic hears me