Evaluate a dot-language to image mechanism

[niklasfink]

We need a client- or serverside mechanism to translate the generated dot-language to an image (diagram).
A clientside mechanism would reduce serverload, a serverside mechanism would reduce the footprint of the website (only relevant on mobile side).

[niklasfink]

There is a Graphviz Javascript port (viz.js), which is quite good working: https://github.com/mdaines/viz.js/
The Javascript lib size is 3.6mb.

There is a xdot to image generator (Canviz): https://code.google.com/archive/p/canviz/
It's very small, but we need to generate xdot-code out of the dot-code for what reason we need Graphviz. This could be a solution for the mobile version of our project, where the server then generates xdot code after the ruby processing, which should be possible with good performance. Canviz also has some other functions for interacting with the drawn graph.

Roadmap: Use viz.js for further development. If the mobile development gets into process, we will think about a Graphiz server side.

[niklasfink]

Stakeholder @hggross: if satisfied, please confirm roadmap.