Premise Amazon Down

[chucklyons]

Somewhat humorous, and perhaps ironic, about Amazon going down last? week.
 
Although I've migrated off of Premise to Alexa for day to day control, I was largely not affected by the Amazon downtime.
 
Premise took over and provided the back-up... ^_^

 

[samgreco]

I was wondering if you were still using Premise.  I think you may be the last man standing.

 

[LarrylLix]

An Amazon that goes down?

Sounds like a male dream.

Sent from my SM-G781W using Tapatalk

 

[chucklyons]

I was wondering if you were still using Premise.  I think you may be the last man standing.

maybe...really haven't done anything with it since I quit working, which is funny because I thot with all of my free time...

 

[samgreco]

I flipped over to Home Assistant and I have a love/hate relationship with it that has been tipping decidedly to hate.  But not sure where else to go.  CQC went open source, but I'm not seeing much action over there.  And it's missing a few things for me.  I may have a go at OpenHab soon(ish).

 

[linuxha]

I flipped over to Home Assistant and I have a love/hate relationship with it that has been tipping decidedly to hate.

 
Oh good I'm not the only one. ;-)

 

[123]

... last man standing.

 
I'm still using Premise as well but together with Home Assistant which now handles most of the automation chores as well as the frontend. The two communicate via MQTT.
 
I use Premise to control two "legacy" devices via native drivers and custom modules that I created a long time ago:
 
Drivers:

• AppliedDigital
• X10

 
CustomDevices:

• Leviton/HAI Omnistat/2 RC-2000 thermostat
• ELK M1
• Environment Canada Weather
• TextToSpeech

Home Assistant doesn't have an integration for HAI thermostats, I prefer my own modules for ELK M1 and EnviroCan, and (for reliability) I prefer local TTS versus using a cloud service (Amazon/Google/Microsoft/IBM). The available (free) speech synthesis applications for Linux have robotic voices so I recently tweaked the TextToSpeech module so I can use it with Home Assistant.
 
Basically, I publish a request (to the TextToSpeech module via MQTT) containing:

• Text to be converted to speech (can include SAPI XML tags)
• Volume
• Speaker group
• Preamble (optional chime played prior to speech)

 
It converts the text to speech and stores it as a local file then publishes a payload (containing all of the received information plus the name of the generated TTS file) as a JSON payload to another topic. That topic is monitored by Home Assistant and it gets the TTS file from the Premise server, forwards it to the specified group of Sonos speakers (pausing any that may already be playing something, then resuming playback afterwards) which play it as the specified volume.

 

[chucklyons]

123, glad to see you're still around!
 
Although I use Alexa, one primary reason I kept Premise, is that putting all of the chips into the cloud seems like a bit risky, as some folks have found during a couple outages. As I have solar, I can stay autonomous during power or cloud outages..
 
I suppose if you have Alexa tied to a life support system, you may want to change your HA approach.

 

[integrate]

Bravo 123!