v0.3.0: Towards better abstraction

[chaosprint]

• Perhaps I should wrap all the Glicol code as some specific task, and let users run everything exactly the same as gym.
  For example:

env = rf.make("sin-perc-tunner")
observation = env.reset()
action = env.action_space.sample()
print(action)
observation, reward, done, info = env.step(action) # reward is calculated with mean squared error
plt.plot(observation) # make your own import matplotlib
print(reward, done, info)

• On top of that, I will provide APIs for defining an env:
  rf.def_env(code, total_step, step_len, target, action_space)
• Still, to better explain all these, the docs should include Glicol interactive editor.