gamedata.md

Game data and visualization process

Purpose

This document describes a high level overview of the collection of data about the game and how it should be exposed.

What is the objective?:

We want to finely-grained visualize what happens during a game, down to the memory reads and writes that were performed by an instruction.

What do we need for it?:

We need to collect the following data in the game:

• The player that is currently running something
• The program counter of the currently executing process / The process that will start executing next
• The program counters of the players other programs
• The memory reads that were performed while running the instruction
• The memory writes that were performed while running the instruction
• Whether the process died

Rough Process:

Frontend / API

• Player(s) join the game and submit code
• The round / game is started

Interpreter

• The interpreter is provided with the data about players (Name, Code, etc.)
• The interpreter runs the game.
  • The game data is collected by the interpreter (duh).
    • The Setup (placing of the instructions etc.) will also be included.
    • The game data will be exposed as a sequential list of game data objects, each containing the data about a single execution of a process, ordered by the order of execution.
  • The interpreter exposes some kind of way to get the game data after a / the round is done.

API

• After the interpreter is done simulating, the API retrieves the game data from the interpreter via the method exposed by the interpreter.
• The API exposes some kind of way to retrieve the data about a / the round.

Frontend

• The Frontend requests that data from the API and visualizes it.

Extras

• The frontend aims to achieve ~30 TPS for the visualization, but it remains to be seen how viable that is in regard to speed and or practicality (how fast is too fast to view etc.)