display a message when daemon closes due to an error in your procfile?

[thdaraujo]

Thanks for working on this tool, it's really great!

This is more of a question or maybe an opportunity to improve user-experience or documentation.

I was running overmind as a daemon, and one of the apps listed in the Procfile had a problem to launch. Then the daemon would quickly shutdown. If I tried to connect to one of the apps or run echo, the socket wouldn't be available anymore (naturally!), so you'd see the error about the socket not being available:

overmind: dial unix ./.overmind.sock: connect: no such file or directory

That all makes sense after you understand what's really going on. But this behaviour can be a bit confusing, because I thought that overmind stopped working, or that there was an issue creating the socket. In reality, it just shutdown really quickly.

Now, if you try to echo or connect, it won't work either, so it's hard to know what's going on.

But if you then go and launch overmind in the foreground, you'll be able to see that the problem is your procfile, not overmind.

Do you think we should document this? Or maybe have a way to show that overmind has shutdown successfully, and capture a bit of the output log somewhere? That would help with debugging situations like this.

I'd be happy to contribute, just need to know your thoughts on this. Thanks!