postiz won't log out

[nebulon]

Are you seeing any errors in the webinspector or backend of the app?

[humptydumpty]

There are no visible errors in the app. I only see the env file in the filemanager (no logs to check).

[nebulon]

Those errors come from a webextension you have installed in your browser. It does look like bitwarden (which has a bug with latest release, which looks like that) So this has nothing to do with the logout.

Do you even see any network requests happening on logout? If not please report this upstream, who knows maybe they haven't even implemented it.

[humptydumpty]

@nebulon I'm not sure how to figure that out when looking at the network tab. I typed "logout" in the search bar and tried to log out, but no results ever showed up, so I assume the answer is no? In any case, I'll report upstream.

Oh, and that was pretty cool how you figured out my errors are Bitwarden related. I didn't see any reference to it until I clicked on it and searched for the keyword. Thanks Johannes

[humptydumpty]

reported upstream here: https://github.com/gitroomhq/postiz-app/issues/671

[nebulon]

I only recognized the bitwarden issue, since I am seeing the same

[robi]

Has there ever been an app so broken from the getgo?

[nebulon]

Overall we might have been too excited about the app without vetting the selfhosting state of upstream properly.

[nichu42]

@nebulon Still excited. This is the way!

[CptPlastic]

I don't mind them being like this they are labeled as unstable after-all. I would rather work through issues with cloudron.

[humptydumpty]

@nebulon The dev asked the following and I'm not sure how to answer that:

Are you using the non secured option?

https://github.com/gitroomhq/postiz-app/issues/671#issuecomment-2798419551

[robi]

@nebulon another reason to implement docker image sandbox with sysbox runtime In one docker command.

And you can even make a nice dash app icon for it with the proxy feature.

All neatly tucked into a separate non privileged secure container.

[nebulon]

With the hint from the upstream user, we were able to fix the packaging and this should be resolved now.