short term notes.txt

//////////////////
HIGH PRIORITY
//////////////////

- research why is there so much ping?
	- netplayjs demos have similar ping spikes (only tested fps game!!! pls test simpler game)
	- peerjs has bad routing?
		- I see little talk of this online
		- use our own peerjs signaling server for STUN/TURN stuff? Would this help in any way
	- because packet size big?
		- for the love of god send the world snapshot in binary form or something more compressed
		- Currently, serialized packets is plain text of json, including numbers in string form wtf.
			- 22kb/packet, could be much less
		- Tried & failed to use TextEncoder/TextDecoder. It should work!!!! try on codesandbox
	- because packet frequency so high?
		- No evidence of this, but its a theory
	- ! I think the ping spikes could be tied to performance
		- try chrome performance tab
		- Abilities, stages with more stuff
		- react may be to blame

- matchmaking server w/ rooms
	- I think I want it obfuscated behind an automatic 'Find match' button
		- automatically find games by testing ping with various opponents
		- If no existing opponent hosts are suitble, become a host and wait in the queue for others to join
		- probably be transparent about:
			- '0 people currently looking for matches'
			- 'Found opponent, but the ping was too high (200ms)'

- more stages
	- bouncy red surfaces
	- curved walls
		- just jack up the border-radius on a hollow rectangle? is that possible?
		- polylines or https://rapier.rs/docs/user_guides/javascript/colliders/#heightfields
		- research: HTML canvas drawing methods for this

- more characters
	- paddle thing
		- flexible sportsfriends paddle
		- maybe with the ability to 'stiffen up' ;) and grow longer omg that is vile but would make for good gameplay I swaer to god
	- tank
		- can lay mines
		- & probably shoot? but then we probably want to differentitate its shot from the ship'screen
		- what about a rotating cannon arm that is controllable with stick/rotation?
	- car
	- sticky

//////////////////
MEDIUM PRIORITY
//////////////////

- fps stats:
	- how do? in netplay, we tick WAY MORE than fps

- Got feedback? Something broken? Email me 

- replace all instances of 60 with an fps constant duhdoy

- ship gamepad controls

- crunchy animations:
	- contacts should probably glow (lighting on walls & objects) & spark (particle effects)
	- screen shake?

- combine control stick inputs and keyboard WASD inputs more automatically?

- control config menu
	- can maybe be extended to be a tutorial as well
	- encourage gampepad
	- eventually should have specialized instructions for each bindable control for each character.
	- localStorage??

- think about name
	- tab title https://nextjs.org/docs/api-reference/next/head

- research: deal with stalling better
	- for one, we could display netplay errors & prompt a restart
	- game should not stall... at all. ie the clock should just keep ticking while stalled
	- ^ maybe only possible with authoratative server..
		-BUT what would it even look like to NOT stall on a server?
			- I promised myself it would be an easy p2p->sever port
				- therefore it must be easy to do p2p stall handling assuming host's machine doesn't stall..

- cpu players

- sentry reporting

//////////////////
LOW PRIORITY
//////////////////

- game end first-to-x ui & implementation

- convert to simpler configged inputs BEFORE input serialization
	- no reason to at the moment? How costly is it.
	- is there actually any advantage?
		- yeah cuz rn we need both clients to know each others' input config & apply it after deserialization
			- that's whack.

- research: how bad is p2p.
	- is it exploitable in any way? Does the host have an advantage?
		- theoretically. But it would be hard to do in a subtle way
		- redux dev config would be a funny backdoor 
	- are there any actual advantages to the server-of-truth approach?
	- how difficult would it be to port to server-of-truth?
	- what is the tech debt of sticking with p2p? methinks not very high?