Summer/Fall 2021 Maintenance!!

[zsunberg]

Alright, I am finally going to put in some effort at long-overdue maintenance on this package! If you are interested in helping with a particular task, please let me know which task and/or comment in the individual issues.

There are four major steps that I plan to take in this order:

1. Make sure that the appropriate github actions are installed (Package Maintenance Github Actions #356)
2. Consolidate the tools packages (Consolidate Packages into POMDPTools.jl #290)
3. Improve Documentation(documentation label)
4. Release POMDPs.jl version 1.0

@bkraske @WhiffleFish @MaximeBouton @Shushman @rejuvyesh @mossr @lassepe thought you might have suggestions for other issues that need to be fixed or be able to help with something. Please tag anyone else who might be interested.

[zsunberg]

@qhho @himanshugupta1009

[zsunberg]

Fixing this is also super important: https://github.com/JuliaPOMDP/quickpomdps I want to make that work well (this will require some python ecosystem knowledge, so help is especially needed here)

[rejuvyesh]

I can work on this. Let me see what's the current best practice.

[rejuvyesh]

One question is what should we pre-install? All the related packages under JuliaPOMDP org. Add a function that does the package management on the julia side from python?

[zsunberg]

@rejuvyesh thanks! I don't think we should install any solvers by default. Rather, we should give instructions (I don't think it will be too hard to do from julia import Pkg; Pkg.add("QMDP") or something like that.

The most important thing to add is testing. Currently QuickPOMDP doesn't work because of some quirks that I added surrounding reward and observation. I don't think this should be difficult to fix with QuickPOMDPs.preprocess(::PyObject).