I am interested and have been playing around with TinkerPop & Docker, particularly for my Gremlin training.
For the training I did at DataDay Texas back in January, I had students clone this and run it locally:
https://github.com/experoinc/gremlin-lang-intro. It is a bit of a hack, but was sufficient for my purposes. Note that it loads 5 different data sets (classic, modern, the crew, grateful dead, and a "citations" one I got from Ted).
For training purposes my requirements are:
- Each student has own graph(s) which are separate from the other students.
- Students have ability to "reset" their environment back to a known state.
- Ability to load known data sets for examples and practice.
- Ability to mutate the graph and have mutations persist from one command to the next
- Overall solution should be as lightweight as possible.
I would be curious to hear what other requirements are out there. I suspect that there are Proof of Concept use cases where this would be valuable.
I'm also interested in having an image which has a "notebook"-style interface (e.g. Jupyter, Apache Zeppelin). There's a PR from a fellow Experonaut which adds Jupyter to the repo above, but I have been out of the office for the past couple of weeks and haven't evaluated it yet.