I know the feel about taking some unnecessary risk, but still you will certainly not be the first to try to clean the floor:
A lot of people are actively using Docker right now. For instance I'm building https://reesd.com on it (i.e. I don't have a prior system that I replace). Getting a few bugs from Docker is not different to me than getting a few bugs from my own systems. I write my integration tests and move along.
Other people are further along the road building a lot of different things, trying a lot of different ideas. I would think that unless you're doing something radically different, a lot of ground has already been covered (and will be covered until you have completely switched to Docker).
Another reason I was not afraid of Docker is that it is built around existing pieces. For instance some people say "Docker is just a script around the lxc tools, what's the point". That, in my opinion, is reassuring. If I need to, I will be able to drop to the next layer, or another similar project will take over.
A lot of people are actively using Docker right now. For instance I'm building https://reesd.com on it (i.e. I don't have a prior system that I replace). Getting a few bugs from Docker is not different to me than getting a few bugs from my own systems. I write my integration tests and move along.
Other people are further along the road building a lot of different things, trying a lot of different ideas. I would think that unless you're doing something radically different, a lot of ground has already been covered (and will be covered until you have completely switched to Docker).
Another reason I was not afraid of Docker is that it is built around existing pieces. For instance some people say "Docker is just a script around the lxc tools, what's the point". That, in my opinion, is reassuring. If I need to, I will be able to drop to the next layer, or another similar project will take over.
So I think it is already/almost no longer true.