I have heard of it, but never looked closely. Being that I have limited bandwidth at the moment, I am just applying the simplest measures to solve the issue I have - how to identify what packages from the package manager in Ubuntu, in this case, are required to run some software that I am developing. With the chroot set up, I can then just copy in my project, and attempt to run it, and it will flag all the missing packages from a base install, then I install everything that gets rid of these errors, and I have a log in the shell scrollback history that I can use to add this list to the documentation for my project. From what I understand of Docker, its application is not really directly relevant to this, as it is more or less like a revamp of package management systems, and it handles dependencies, so either way I would still be in the same situation, not able to know for certain exactly which packages my project requires exactly.

Once I do know, it probably will mostly apply to any project I do based on the same set of libraries imported in the code. Anything new after that, will probably be obvious. The Piston library has a particular set of requirements, and my code may not directly require them but will still require them to be present for Piston to work. Later on maybe I will add other elements, perhaps I will create chart renderers, for example, which will involve Cairo and maybe OpenGL.

But I probably only need to solve this main problem once. I am just doing it so that when someone goes and looks at my github repositories, they have enough information to get it running and see it in action. This is important from a marketing perspective, since personal testimony is good for sales, as is for organisation, the ability for an interested person to get up and running quickly to start contributing.

Marketing is still important, even if you are not trying to sell something. Finding collaborators requires marketing. Gaining votes for ideas and projects is marketing.