When I discovered ShinyProxy a few months ago I could not believe it would actually work. But it does, of course. And it is fantastic! Thanks to Open Analytics for making this available.
One issue I been wondering about: I quickly resorted to putting into an R-script all the necessary commands for building new images and uploading and running my own applications in a container (just using the ssh-package). I also ended up handling most of the installation process (shinyproxy, Docker, nginx) etc. from an R-script, since I had to reinstall a number of times.
So, it seems to me it should be super-easy for somebody to make an R-package handling all the details with installing ShinyProxy on a server and running applications. Then users will not have to spend hours trying to comprehend and follow the detailed instructions found on https://www.shinyproxy.io/ . I think doing this could vastly increase the popularity of ShinyProxy. Many of the potential users (like myself) know much more about statistical analysis than about setting up a server.
Anybody working along these lines?