Decouple Engine and Simulation #55

New Issue

2018-10-18T08:54:00Z

marty-oehme commented

2018-10-18 08:54:00 +00:00

(Migrated from gitlab.com)

Currently, engine and simulation are tightly coupled to one another. Any operation on the internal simulation (resetting the grid, adding cell patterns, toggling the grid rendering) has to go through the engine right now.

What should be done instead is to have the engine be able to run independently of having a simulation assigned. When a simulation is injected, it will simply run that simulation, calling update and render on it. It should never know more about the simulation than those two simple methods.

Methods that should be kept and reference able: update(), render(), step() [as a single call to update()], toggleEdgeRendering()

Methods that Engine should not know about: reset(), clear(), addPattern(), gridSize

Currently, engine and simulation are tightly coupled to one another. Any operation on the internal simulation (resetting the grid, adding cell patterns, toggling the grid rendering) has to go through the engine right now. What should be done instead is to have the engine be able to run independently of having a simulation assigned. When a simulation is injected, it will simply run that simulation, calling update and render on it. It should never know more about the simulation than those two simple methods. Methods that should be kept and reference able: update(), render(), step() [as a single call to update()], toggleEdgeRendering() Methods that Engine should *not* know about: reset(), clear(), addPattern(), gridSize

marty-oehme commented

2018-10-18 09:54:02 +00:00

(Migrated from gitlab.com)

created branch 55-decouple-engine-and-simulation

created branch [`55-decouple-engine-and-simulation`](https://gitlab.com/marty.oehme/cellular-automata/compare/master...55-decouple-engine-and-simulation)

marty-oehme commented

2018-10-18 09:54:03 +00:00

(Migrated from gitlab.com)

mentioned in merge request !15