[RELEASE] Echo Speaks V4

Docker hub would be the best place to post one of these.

https://hub.docker.com/

As an example, here is one of the best HomeBridge images:

https://hub.docker.com/r/oznu/homebridge

1 Like

I was thinking of iterating on it as a group, to come up with the best solution.

2 Likes

It's in the docker folder this script file start.sh. It does a replacment of the IP address in one of the configuration files of the node.js code. If you take a complete zip of Tony's repository:

Unzip it on a pi that has docker-compose, then go into the docker folder you will file the shell script. The docker-compose up -d will build a new container image with all the node.js stuff and start it. You should then be able to use the ip address of your pi and the right address will be sent to the Echo Speaks speaks for it's configuration.


#!/bin/bash

# I added the line below and give it the Pi's IP Address

IPADDRESS=192.168.20.251

if [[ (-z "${IPADDRESS}") ]]; then
    echo "Please set IPADDRESS environment variables"
    echo "example: IPADDRESS=192.168.1.100 ./start.sh"
    exit 1
fi

if [ $(grep -c "IPADDRESS_TO_CHANGE" env.echo-speaks-server.dist) -ne 0 ]; then
   sed "s/IPADDRESS_TO_CHANGE/$IPADDRESS/g" env.echo-speaks-server.dist > env.echo-speaks-server
fi

echo "Starting docker-compose"
docker-compose up -d

1 Like

Thanks,
I was just looking at the master branch. I wondered why the IP was not just passed in through the env like the rest of the variables.
I ended up doing a patch on master for my image.

I'll update my change to match the new node change.

-D

Excellent, I will see if I can get it going on linux

I swear you guys are very impatient!
Just relax and use the current server... we have almost 3 months!
I will have the new version ready before then

30 Likes

Not impatient, just prefer to run things locally. Getting the notice from Heroku just got me off my butt to try again. Fully understand the "local server" that we are debating on setting up and with some success is just transient.

1 Like

To each their own I suppose!

The new server will be a complete rewrite along with how it interacts with the ES Hubitat app

16 Likes

Thanks for all you work @tonesto7 . Really appreciate all your contributions! We just love your apps so much (anxiety felt when there is even a hint that we may have to go without them, LOL).

1 Like

I noticed that the callback url provided by ES window is the cloud endpoint, would it work if we provide the local IP endpoint of the hub?

lol. I am waiting and in no rush :slight_smile:

…but still reading the thread to keep up to date.

1 Like

@tonesto7 I agree with you. Thus, my previous comments in this post. I did cut over to a dorker setup. However, I do see some functionality issues that I wasn't seeing before. I'm happy to test and I agree we have three months to sort this out. I'm probably going to flip back to the original server. And retest to make sure that all the functions work again. Thank you again for all your help and effort with this solution. I know many people appreciate your efforts all across this platform.

@moh my previous response regarding the authentication was response to the general thread not to a particular person. There are many people having authentication issues and I wanted to present that as a alternative possible concern/solution that they might be running into.

I'll wait! But I have been looking to set up an rPi with docker anyway. You just gave me a reason.

1 Like

I wonder what API or functionality is needed in Hubitat so that the echo-speaks-server could be written in groovy. Some sort of general purpose key-value store for saving the cookie/token, perhaps.

1 Like

Honestly, I'm hoping it never goes there. My hub is already pitching a fit about too many events, so I find it much better to offload as much responsibility as I can.

1 Like

I keep a Mac running to host Homebridge and will likely just use this to run the new local server. I’ll probably wait for Tonesto7’s server update, but I may get impatient because right now my Heroku server is crapping out every 10 days, forcing me to manually re-deploy it. Did you use the information here to configure yours?

It's not crunch time yet but I do need to sort out the best way to keep this running.

Based on how we do things at work we only have two sprints left and one of them is for dark deploy and defect identification and test while in dark deploy. I miss the days when it would be ready when it's ready. :grinning:

4 Likes

@tonesto7 - on another note, is there a reason you just "speak" notifications rather than using the Alexa Device Notification functionality (that persists with the yellow ring)?

The API which is available does not allow us to send device notifications like you are mentioning. Unless something has changed in the last year.