Open Oleksii-Kshenskyi opened 6 months ago
Thinking about this, these are two range-based genes I can implement relatively easily:
To implement these, I need two features that are not yet implemented:
I already have an issue for line of sight mechanics (#10), need an issue for the back off jump action too. For now, this implementation is blocked by those two.
Both #10 and #31 are implemented. Should start planning of this big one. For now, there should be three "scale" genes:
Also, think about one-two other types of genes to implement. Aaaand also think if it's a good idea to start implementing action pipelines (queues) in this one or maybe wait a bit.
UPD: OUTDATED info below - this task has been reduced to just the genetics because otherwise the scope is too wide and the SLP would obviously get violated.
This is the key feature of the entire project. Effectively, these are two different features, but they go hand in hand because if we don't have the ability to save a generation, we cannot simulate multiple generations without graphics as we have to somehow save the result of the simulation and then load it into the visualizer.
The features are:
Saving/Loading results of the simulation: once the simulation has finished, we have to "breed" players with each other, producing a new "superior" generation of players that (hopefully) is going to know what they're doing a bit better than their parents, and save those genes/brains into a file. Also the visualizer has to be able to then load that saved file and show the results of those "evolved" players playing the game.^^^ For the above, visit issue #37.