Almost 2 month ago an issue was reported.
AppDaemon stops after updating to 0.15.0 · Issue #287 · hassio-addons/addon-appdaemon · GitHub
The priority is high but, it has not been fixed and there is no report or response about the prediction of this bug.
I have some automations in AppDaemon and I really like the concept.
But now I don´t dare to spend more time on it, and I wonder if I should look for something else.
Does anybody have some inside information about this tropic?
The problem seems to be that AppDaemon, Pyscript and Node red are separate software, which can get into a state where it is no longer supported or maintained.
It would be nice if a “not yaml” concept was developed by Home Assistant, so this risk is reduced.
I think you misread…
The link is a link to a reported issue in v0.15.0 (and explains how to resolve it).
It is nowhere mentioned that AppDaemon will stop development and in the meantime v0.16.0 is already available:
(time to remove your comment on GitHub?)
But Release v0.16.0 has the same problem
Have you tried changing the path to the new location for the secrets file?
That is all that was needed for me:
The problem is in the update; as far as I understood, starting from scratch works fine 
Node-Red?
It exists completely independently of Home Assistant and is supported by a large community. People use it to automate their homes (without Home Assistant, openHAB, etc). It would be very surprising to see that open-source project become “no longer supported or maintained”.
Ok, but I think it is a bit scary, to uninstall and start from scratch. This is why I wonder about the future of AppDaemon.
Yes, changing the path to tne new location works, but In the filesystem I get duplicated appdaemon folders. So I see two problems, the concrete duplicated folders and that I can not figure out if anybody is working on this issue.
AppDaemon and the AppDaemon addon for HA are two separate projects. The addon is used to bring AD into Home Assistant making it easy for people to use it. AD can be used independently of the add-on.
While the releases/updates for AD are relatively slow, the project is still active. As far as the future goes? Who knows.
Very interesting. So one solution to the problem could be to use AppDaemon with Docker. As I see it, it would remove the concrete problem and remove one depencency (addon for HA).
It does seems like the installation requires more work than the addon.
I guess there is also some work to get visual studio code to work in this setup.
Can anybody see more work?
If I get Appdaemon to work on docker, will everything then work the same way?
Is it likely that the “only” work would be to get Appdaemon to run on Docker, Visual studio code to run on Docker and then use the exact same python and yaml code that I already have?
If we are splitting hairs here, Home Assistant is also open source and could go dark at any moment. Will it? Likely not. It really depends on the scale of the user base that will drive this. More likely is that open source with a large user base could go closed source private software company.
Yes, you could certainly move over to using your own docker container for AD (several of us do). One benefit to this is you can use the development release (if you wish) and obviously you have more control over your AD instance.
Spinning up AD in Docker yourself is more work but it is documented here: Docker tutorial — AppDaemon 4.4.2 documentation
As far as running vscode on docker, I’m not sure about that. I prefer to run vscode locally on my workstation and I push changes via SFTP to the host where my docker container runs. You can also connect vscode to a remote docker instance and make changes. There are a lot of options, just need to find the workflow that you like.
edit: AD works exactly the same way in docker as it does under HA (which is also ran inside of a docker container).