The authorisation method I'm using is 'Proxy, should I change it to 'Cookie' ?
I've no idea - this is just what I use and it works for me
Thanks - I've changed to the 'cookie' method - seems to be working now.
Mine has gone back to failing after 30 mins 
This is a fun game 
I don't want to tempt fate - but at the moment mine is still working.
Mine works when testing - I go and make a cup of coffee - I come back and test again and it's failed
Or I look out of the window and it fails
edit - of course its not failed at all since I posted that it's failing
I'm going to uninstall applestrudel from my main NR instance running on a Pi and install it on my Win11 machine and see if its stable on that one instead
edit: Up and running on my Win11 machine - going to log anytime Alexa hears a command and see how long it lasts
I've been using the Cookie method to authorise my Alexa node for a few hours. Every time I 'Deploy' a flow on the Pi that includes the Alexa node, I have to re-visit the URL:3458 to re-authorise again.
Is there a way to retain the Cookie?
Applestrudel V5.0.41 has worked fine on my Win11 machine since yesterday afternoon without failure
One thing I've noticed is that on Win 11 I'm running node v18.16.0 and on my Pi I'm running v16.19
So I'm going to update to V18 on my Pi and switch back to using that one and see how it goes
[edit]Pi now at v18.17.1 so lets see how it goes
[edit] Its failed pretty much straight away - moving back to my Win11 instance and will just use MQTT to send/receive Alexa stuff to/from the Pi instance
5.0.42 released about an hour ago - but looks like there is still and issue.
Current state - I stop receiving events for a minute or so then they restart for a minute or so
Rinses and repeats
It seems that whatever option I select when I Deploy the Node-RED flow (for some reason) the Alexa node goes back to asking me to authorise again using URL:3458 - which sometimes fails with Amazon saying... "there was a problem at our end".
I find I need to turn around 3 times, use my left-hand to type in the url and sacrifice a smal animal before I can get it work!
The force is strong with the gremlins when trying to do this - usually takes me about 15 minutes of repeatably trying it using localhost 127.0.0.1 and the actual IP address
Inc logging out and loggin back into my amazon account
but eventually it works
2 Likes
I would go back to using the Proxy setting if that was what has worked for you in the past
It's a real pain as I have some many things in my home that are controlled by capturing the response from BluePrints via 'OnDeviceActivity'. Every thing was working just fine until a week ago. Why does Amazon have to keep "fiddling" with settings or code that ends-up break things??
1 Like
We are just piggy-backing on the functionality of Alexa
If I was running a smart speaker business - I wouldn't give two hoots either for what happens to someone' elses unauthorised software
I'd just make sure it works with my own app
2 Likes
And this morning - it is back to working every time I test it
Strange game - I'd like to play tic-tac-toe instead!
1 Like
I reverted to using the Proxy method to perform the authorisation at 06:00am this morning.
So far (touch wood) 'OnDeviceAct' has been working fine.
PS: I'm using node.js v18.17.1 and Node-RED v3.1.0 on a RPi-4B.
2 Likes
Finally managed to delete my cookie.txt file on my Pi and re-auth with a new file
It now works again on the Pi so I'm back to running it on there
