Jump to content

Amazon Echo and ISY


madmartian

Recommended Posts

Barry,

 

Thanks to the GREAT documentation you have in your AWS_Config tool package, I was able to get my RPi2 up and running with 2 emulators (all I want for now).  I added a device and it works perfectly.  Just as responsive as the emulator running on my server tower.  

 

However, after doing MOVES of several devices from my old emulator running on my PC, I am finding that Alexa no longer discovers any devices on my RPi2.  I have rebooted several times.  Your configurator sees all the devices, but Alexa only wants to discover devices on my PC, not the RPi2.  I even turned my PC off, but still no joy.

 

Any ideas?  

Link to comment

Try re-registering the Echo to your network by pretending it is a new Echo that you just bought and starting over with it.  I think the Echo may remember things.  You can also call Amazon Echo support. They should be able to assist you in resolving this. Just tell them you have two HUE hubs, they get nervous when you say emulator.

 

The emulators have practically no CPU load when they are running so the RPi is more than adequate. 

Link to comment

Oddly, it turned out that using my RPi wireless was the issue.  If it connected to the right WAP, it was fine.  If not, it did not work well.  So I just went wired for ethernet and all is well.  Barry, your configuration program is wonderful.  Does everything I want.  Only minor change I might suggest is to remove the need to press "OK" every time you move a device between bridges.  I had to move 30 devices, and hitting the extra OK key each time slowed things down.  

 

Great work, Barry.  

Link to comment

The errors on install are either permission or file in use. Ignore is the proper response. The files should already be there but are included in the install package just in case.

 

You need to send me 2 things. The log file that should be in the same folder as the AWS_Config.exe and the .ini file from the same folder.

 

You might want to add the ini option in the systems section as follows:

 

debug = All

 

This will provide a lot more info in the log file.

 

An abort flag of 25 means that the ISY did return "HTTP/1.1 200 OK". With debug = all I ahould be able to see what it got back.

 

As I said I have only run this on native windows machines and suggest that if yiou can run it that way to start till we figure out what is going on.

Barry, first of all thanks so much for your time and efforts. It really contributes a lot to this whole project. I will run it again on PC hardware and will send you those files. Can you please just clarify how should I:

"add the ini option in the systems section as follows: debug = All". Does it mean I should start the program from the term with this option?

It has been a long time since I did something on the windows OS...

Thanks!

Link to comment

In the ini file, a text file, just add the line in the [system] section as an additional line of text. The ini file is read when the configuration program, AWS_Config is started. For some reason the ISY is not returning an http OK response. I want to see what is being sent and exactly what the ISY responds with.

Link to comment

Ferdies: 

 

I am pretty sure I know what your issue is.  The names you are using for devices on the ISY are rather long. I made a design decision on how to handle long names and will review that code. Tell me again exactly what you were doing when then run time error occurred.  Precisely what button did you just click on when the error dialog came up? That will help me pinpoint where in the code the weakness is.

Link to comment

Blueman2,

 

The issue has to do with the selecting of the bridge that is the target for the move. I want to make sure that a bridge is selected and that is why the dialog box shows which requires an OK response/click.  If you don't select a bridge or use the blank option then the move is to the brisge with the most room.  let me think about alternate strategies.

Link to comment

Blueman2,

 

The issue has to do with the selecting of the bridge that is the target for the move. I want to make sure that a bridge is selected and that is why the dialog box shows which requires an OK response/click.  If you don't select a bridge or use the blank option then the move is to the brisge with the most room.  let me think about alternate strategies.

 

Yes, I can see how you got there.  But you seem to have done a very good job with the red lettering 'select emulator' that shows up near the bridge selection area to make it clear this should be filled in.  If the OK dialog is removed, it will still wait for your to select a bridge, right?  And if you do not, no harm since it does what you really want in terms of selecting least used.  

 

I think you wrote the software very well, so that it does not need the OK dialog, but I certainly understand why you put it in.  I would prefer it be removed though.  Not a big issue, however.  

 

I was just reading about your background on your web site.  Damn impressive.  And wonderful work on this project.  How far along are you on the "Alexa, ask Sarah to..." functionality.  I did not see any software associated with this yet.  But I am probably just not up to speed with what you are doing there.  

Link to comment

There is a new version of the configuration system on my web site.  I believe it handles all of the issues mentioned in the preceding posts from Blueman2 and Ferdies.

 

Blueman2 - tell me what you think of the Move logic.

 

Ferdies - Let me know if I resolved the error 5 and if not send me the log with debug=all and the ini file.

 

Re Alexa Ask Sarah  . . . I have not progressed much on that as I wanted to get the configuration system to a satisfactory point, and assist those requesting help re putting up the bridge on an RPi. The basic code is written and it handles the one utterance I wrote to test it by responding back with a nice voice command letting me know that the code saw the data from the Echo. I just have to sit down and write all of the Intents and Utterances.  Utterances are spoken phrases and Intents map out what from the spoken phrase should be sent to the endpoint with what label.  It maps what in the spoken phrase are the key words and what they pertain to. E.g. a keyword might be room and the utterance might be "kitchen". This sends the endpoint a JSON object something like {"key":"room", "value":"kitchen"}. A single utterance may have multiple key words as in "Alexa tell Sara to set the Master Air Conditioner to cool to 77 degrees".  "Master Air Conditioner" might be a Device, "cool" might be an action, and 77 a level. which would produce a message containing something like

[{"device":"master air conditioner"}, {"action":"cool"}, {"level":77}].  I just need to spend a day developing the utterances and intents and then a couple of days to write the code to send the information to the final controller.  For HVAC that will be the ISY. 

 

I do have some custom commands in my set of devices on the emulator that are not handled by the ISY; e.g. "Alexa Turn On The SPA". This is much easier to handle as all you need is a program that monitors the IP:port provided in the onURL and offURL in the definition of the custom device. 

 

I am also working with the developer of the bridge to try and get dimming to work.

Link to comment

There is a new version of the configuration system on my web site.  I believe it handles all of the issues mentioned in the preceding posts from Blueman2 and Ferdies.

 

Blueman2 - tell me what you think of the Move logic.

 

 

Just tried it.  I only copied the .exe file; I hope that is OK.   When i tried to move, does bypass the OK dialog, but when I select the new bridge from the drop down, it crashes with the message "Stop Statement Encountered"

Link to comment

Re Alexa Ask Sarah  . . . I have not progressed much on that as I wanted to get the configuration system to a satisfactory point, and assist those requesting help re putting up the bridge on an RPi. The basic code is written and it handles the one utterance I wrote to test it by responding back with a nice voice command letting me know that the code saw the data from the Echo. I just have to sit down and write all of the Intents and Utterances.  Utterances are spoken phrases and Intents map out what from the spoken phrase should be sent to the endpoint with what label.  It maps what in the spoken phrase are the key words and what they pertain to. E.g. a keyword might be room and the utterance might be "kitchen". This sends the endpoint a JSON object something like {"key":"room", "value":"kitchen"}. A single utterance may have multiple key words as in "Alexa tell Sara to set the Master Air Conditioner to cool to 77 degrees".  "Master Air Conditioner" might be a Device, "cool" might be an action, and 77 a level. which would produce a message containing something like

[{"device":"master air conditioner"}, {"action":"cool"}, {"level":77}].  I just need to spend a day developing the utterances and intents and then a couple of days to write the code to send the information to the final controller.  For HVAC that will be the ISY. 

 

I do have some custom commands in my set of devices on the emulator that are not handled by the ISY; e.g. "Alexa Turn On The SPA". This is much easier to handle as all you need is a program that monitors the IP:port provided in the onURL and offURL in the definition of the custom device. 

 

I am also working with the developer of the bridge to try and get dimming to work.

This is very cool.  It would be great to get the HVAC item working as a 'proof of concept'.  In preparation for this, should I go ahead an install stunnel on my RPi?  Let me know if/when you need someone to beta test.  Your current configurator with 2 copies of the 0.4.4 bridge is working perfectly and rock solid on my RPi2.  So I need something else I can break!

Link to comment

All fixed. build 73 on the web site.  Forgot to pull the stop statements that I use for debugging.

 

Copying just the exe file is fine. I generally save the old exe just in case.

 

If you set up stunnel, then you will be further along. I am not sure how Amazon will react as the Skill I am developing is in test mode which allows me to use a self signed certificate.

 

Just remember that you will eventually have to route port 433 to stunnel (port forwarding on your router) and then let stunnel route to where you have the endpoint. 

 

Speaking of routers I just set up my Grandsons house for Ethernet. I installed the wireless version of the router I use, Mikrotik. It is an amazing router that few people know about other than pro installers and custom integrators.  It is in his basement and everywhere in his house he get a minimum of 50 dbm as signal strength. Its not a big house but the router has a 1000 miliwatt radio for wifi. It has no external antenna, although there is a model that allows for it. Mine is a model RB750GL and his is the RB951G - 2HND  IIRC mine was $49 and his was $59 at Amazon 

Link to comment

Thanks, Barry.  1173 indeed works perfectly!!!  

 

As for the Mikrotik, I recall these guys made bare boards that were out of this world.  Some of the best h/w out there.  But I am a long time dd-wrt guy and am not sure I am up for learning a new OS, especially one that is as complex and powerful as the one provided by Mikrotik!

 

Also, I noticed you have some info  below your application's main window (some words on the left, a list of all discovered devices in the middle, and a 'help' text box on right (until you hit okay!).  Just curious what those will be / are used for?

Link to comment

Damm, I forgot to lock the screen height.  The items you see are things I use in the app that are not for the users pervue. The Help box will appear with help info if you right click on a button or other clickable items.  The next update will have the screen size locked.

 

With regard to Mikrotik, They now package one of their boards in a 1x5x5 case. It comes preconfigured to handle the avergae users needs, and there is an app called winbox that is a complete graphic interface for adjusting settings.

Link to comment

In the ini file, a text file, just add the line in the [system] section as an additional line of text. The ini file is read when the configuration program, AWS_Config is started. For some reason the ISY is not returning an http OK response. I want to see what is being sent and exactly what the ISY responds with.

Barry, I have it working. It was my mistake, I did not realize that I have to edit .ini file to enter login/password. Thanks again!

Link to comment

Ferdies: 

 

I am pretty sure I know what your issue is.  The names you are using for devices on the ISY are rather long. I made a design decision on how to handle long names and will review that code. Tell me again exactly what you were doing when then run time error occurred.  Precisely what button did you just click on when the error dialog came up? That will help me pinpoint where in the code the weakness is.

 

Hi Barry,

 

Installed the latest version and "devices" button is working now!

 

 

I might have missed some info though - is the "scene" button working as well? It does not show any list of scenes.

 

Thanks a lot!

Link to comment

Hello everyone,

 

Just a quick update: we just published ISY Skill to Echo and waiting for certification!

 

With kind regards,

Michel

 

Michel,

 

Is there a thread or topic discussing the details of what can be done with this, what is required, potential costs, etc?  For example, will this allow dimming?  Perhaps customer phrases beyond 'turn on' and 'turn off'?

Link to comment

Ferdie,

 

All four buttons work (Devices, Scenes, Programs, Custom) for me.  Put the debug statement in the ini file as described in an earlier post, hit the scene button and send me the log file and the ini file.  Does the Program button work for you? My email address is in my profile.


Michel,

 

As Blueman2 stated, Could you provide a little information about the skill you have developed?

Link to comment

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...