deduplicate models

.cargo/config.toml

@@ -1,2 +0,0 @@
-[registries.strafesnet]
-index = "sparse+https://git.itzana.me/api/packages/strafesnet/cargo/"

.gitignore

@@ -1 +1,2 @@
 /target
+/textures

Cargo.toml

@@ -1,36 +1,30 @@
 [package]
 name = "strafe-client"
-version = "0.10.5"
+version = "0.8.0"
 edition = "2021"
-repository = "https://git.itzana.me/StrafesNET/strafe-client"
-license = "Custom"
-description = "StrafesNET game client for bhop and surf."
-authors = ["Rhys Lloyd <krakow20@gmail.com>"]
 
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
-[features]
-default = ["snf"]
-snf = ["dep:strafesnet_snf"]
-source = ["dep:strafesnet_deferred_loader", "dep:strafesnet_bsp_loader"]
-roblox = ["dep:strafesnet_deferred_loader", "dep:strafesnet_rbx_loader"]
 
 [dependencies]
+async-executor = "1.5.1"
 bytemuck = { version = "1.13.1", features = ["derive"] }
 configparser = "3.0.2"
 ddsfile = "0.5.1"
-glam = "0.29.0"
-id = { version = "0.1.0", registry = "strafesnet" }
+env_logger = "0.10.0"
+glam = "0.24.1"
+lazy-regex = "3.0.2"
+log = "0.4.20"
+obj = "0.10.2"
 parking_lot = "0.12.1"
 pollster = "0.3.0"
-strafesnet_bsp_loader = { version = "0.2.1", registry = "strafesnet", optional = true }
-strafesnet_common = { version = "0.5.2", registry = "strafesnet" }
-strafesnet_deferred_loader = { version = "0.4.0", features = ["legacy"], registry = "strafesnet", optional = true }
-strafesnet_rbx_loader = { version = "0.5.1", registry = "strafesnet", optional = true }
-strafesnet_snf = { version = "0.2.0", registry = "strafesnet", optional = true }
-wgpu = "22.1.0"
-winit = "0.30.5"
+rbx_binary = "0.7.1"
+rbx_dom_weak = "2.5.0"
+rbx_reflection_database = "0.2.7"
+rbx_xml = "0.13.1"
+wgpu = "0.17.0"
+winit = "0.28.6"
 
-[profile.release]
+#[profile.release]
 #lto = true
-strip = true
-codegen-units = 1
+#strip = true
+#codegen-units = 1

LICENSE

@@ -1,5 +1,5 @@
 /*******************************************************
-* Copyright (C) 2023-2024 Rhys Lloyd <krakow20@gmail.com>
+* Copyright (C) 2023 Rhys Lloyd <krakow20@gmail.com>
 *
 * This file is part of the StrafesNET bhop/surf client.
 *

src/aabb.rs

@@ -0,0 +1,89 @@
+use crate::integer::Planar64Vec3;
+
+#[derive(Debug,Clone,Copy,Hash,Eq,PartialEq)]
+pub enum AabbFace{
+	Right,//+X
+	Top,
+	Back,
+	Left,
+	Bottom,
+	Front,
+}
+#[derive(Clone)]
+pub struct Aabb{
+	pub min:Planar64Vec3,
+	pub max:Planar64Vec3,
+}
+
+impl Default for Aabb {
+	fn default()->Self {
+		Self{min:Planar64Vec3::MAX,max:Planar64Vec3::MIN}
+	}
+}
+
+impl Aabb{
+	const VERTEX_DATA:[Planar64Vec3;8]=[
+		Planar64Vec3::int( 1,-1,-1),
+		Planar64Vec3::int( 1, 1,-1),
+		Planar64Vec3::int( 1, 1, 1),
+		Planar64Vec3::int( 1,-1, 1),
+		Planar64Vec3::int(-1,-1, 1),
+		Planar64Vec3::int(-1, 1, 1),
+		Planar64Vec3::int(-1, 1,-1),
+		Planar64Vec3::int(-1,-1,-1),
+	];
+
+	pub fn grow(&mut self,point:Planar64Vec3){
+		self.min=self.min.min(point);
+		self.max=self.max.max(point);
+	}
+	pub fn join(&mut self,aabb:&Aabb){
+		self.min=self.min.min(aabb.min);
+		self.max=self.max.max(aabb.max);
+	}
+	pub fn inflate(&mut self,hs:Planar64Vec3){
+		self.min-=hs;
+		self.max+=hs;
+	}
+	pub fn intersects(&self,aabb:&Aabb)->bool{
+		(self.min.cmplt(aabb.max)&aabb.min.cmplt(self.max)).all()
+	}
+	pub fn normal(face:AabbFace)->Planar64Vec3{
+		match face {
+			AabbFace::Right=>Planar64Vec3::int(1,0,0),
+			AabbFace::Top=>Planar64Vec3::int(0,1,0),
+			AabbFace::Back=>Planar64Vec3::int(0,0,1),
+			AabbFace::Left=>Planar64Vec3::int(-1,0,0),
+			AabbFace::Bottom=>Planar64Vec3::int(0,-1,0),
+			AabbFace::Front=>Planar64Vec3::int(0,0,-1),
+		}
+	}
+	pub fn unit_vertices()->[Planar64Vec3;8] {
+		return Self::VERTEX_DATA;
+	}
+	// pub fn face(&self,face:AabbFace)->Aabb {
+	// 	let mut aabb=self.clone();
+	// 	//in this implementation face = worldspace aabb face
+	// 	match face {
+	// 		AabbFace::Right => aabb.min.x=aabb.max.x,
+	// 		AabbFace::Top => aabb.min.y=aabb.max.y,
+	// 		AabbFace::Back => aabb.min.z=aabb.max.z,
+	// 		AabbFace::Left => aabb.max.x=aabb.min.x,
+	// 		AabbFace::Bottom => aabb.max.y=aabb.min.y,
+	// 		AabbFace::Front => aabb.max.z=aabb.min.z,
+	// 	}
+	// 	return aabb;
+	// }
+	pub fn center(&self)->Planar64Vec3{
+		return self.min.midpoint(self.max)
+	}
+	//probably use floats for area & volume because we don't care about precision
+	// pub fn area_weight(&self)->f32{
+	// 	let d=self.max-self.min;
+	// 	d.x*d.y+d.y*d.z+d.z*d.x
+	// }
+	// pub fn volume(&self)->f32{
+	// 	let d=self.max-self.min;
+	// 	d.x*d.y*d.z
+	// }
+}

src/bvh.rs

@@ -0,0 +1,107 @@
+use crate::aabb::Aabb;
+
+//da algaritum
+//lista boxens
+//sort by {minx,maxx,miny,maxy,minz,maxz} (6 lists)
+//find the sets that minimizes the sum of surface areas
+//splitting is done when the minimum split sum of surface areas is larger than the node's own surface area
+
+//start with bisection into octrees because a bad bvh is still 1000x better than no bvh
+//sort the centerpoints on each axis (3 lists)
+//bv is put into octant based on whether it is upper or lower in each list
+#[derive(Default)]
+pub struct BvhNode{
+	children:Vec<Self>,
+	models:Vec<u32>,
+	aabb:Aabb,
+}
+
+impl BvhNode{
+	pub fn the_tester<F:FnMut(u32)>(&self,aabb:&Aabb,f:&mut F){
+		for &model in &self.models{
+			f(model);
+		}
+		for child in &self.children{
+			if aabb.intersects(&child.aabb){
+				child.the_tester(aabb,f);
+			}
+		}
+	}
+}
+
+pub fn generate_bvh(boxen:Vec<Aabb>)->BvhNode{
+	generate_bvh_node(boxen.into_iter().enumerate().collect())
+}
+
+fn generate_bvh_node(boxen:Vec<(usize,Aabb)>)->BvhNode{
+	let n=boxen.len();
+	if n<20{
+		let mut aabb=Aabb::default();
+		let models=boxen.into_iter().map(|b|{aabb.join(&b.1);b.0 as u32}).collect();
+		BvhNode{
+			children:Vec::new(),
+			models,
+			aabb,
+		}
+	}else{
+		let mut octant=std::collections::HashMap::with_capacity(n);//this ids which octant the boxen is put in
+		let mut sort_x=Vec::with_capacity(n);
+		let mut sort_y=Vec::with_capacity(n);
+		let mut sort_z=Vec::with_capacity(n);
+		for (i,aabb) in boxen.iter(){
+			let center=aabb.center();
+			octant.insert(*i,0);
+			sort_x.push((*i,center.x()));
+			sort_y.push((*i,center.y()));
+			sort_z.push((*i,center.z()));
+		}
+		sort_x.sort_by(|tup0,tup1|tup0.1.partial_cmp(&tup1.1).unwrap());
+		sort_y.sort_by(|tup0,tup1|tup0.1.partial_cmp(&tup1.1).unwrap());
+		sort_z.sort_by(|tup0,tup1|tup0.1.partial_cmp(&tup1.1).unwrap());
+		let h=n/2;
+		let median_x=sort_x[h].1;
+		let median_y=sort_y[h].1;
+		let median_z=sort_z[h].1;
+		for (i,c) in sort_x{
+			if median_x<c{
+				octant.insert(i,octant[&i]+1<<0);
+			}
+		}
+		for (i,c) in sort_y{
+			if median_y<c{
+				octant.insert(i,octant[&i]+1<<1);
+			}
+		}
+		for (i,c) in sort_z{
+			if median_z<c{
+				octant.insert(i,octant[&i]+1<<2);
+			}
+		}
+		//generate lists for unique octant values
+		let mut list_list=Vec::with_capacity(8);
+		let mut octant_list=Vec::with_capacity(8);
+		for (i,aabb) in boxen.into_iter(){
+			let octant_id=octant[&i];
+			let list_id=if let Some(list_id)=octant_list.iter().position(|&id|id==octant_id){
+				list_id
+			}else{
+				let list_id=list_list.len();
+				octant_list.push(octant_id);
+				list_list.push(Vec::new());
+				list_id
+			};
+			list_list[list_id].push((i,aabb));
+		}
+		let mut aabb=Aabb::default();
+		let children=list_list.into_iter().map(|b|{
+			let node=generate_bvh_node(b);
+			aabb.join(&node.aabb);
+			node
+		}).collect();
+		BvhNode{
+			children,
+			models:Vec::new(),
+			aabb,
+		}
+	}
+}

src/compat_worker.rs

@@ -1,21 +0,0 @@
-pub type QNWorker<'a,Task>=CompatNWorker<'a,Task>;
-pub type INWorker<'a,Task>=CompatNWorker<'a,Task>;
-
-pub struct CompatNWorker<'a,Task>{
-	data:std::marker::PhantomData<Task>,
-	f:Box<dyn FnMut(Task)+Send+'a>,
-}
-
-impl<'a,Task> CompatNWorker<'a,Task>{
-	pub fn new(f:impl FnMut(Task)+Send+'a)->CompatNWorker<'a,Task>{
-		Self{
-			data:std::marker::PhantomData,
-			f:Box::new(f),
-		}
-	}
-
-	pub fn send(&mut self,task:Task)->Result<(),()>{
-		(self.f)(task);
-		Ok(())
-	}
-}

src/face_crawler.rs

@@ -1,127 +0,0 @@
-use crate::physics::Body;
-use crate::model_physics::{GigaTime,FEV,MeshQuery,DirectedEdge,MinkowskiMesh,MinkowskiFace,MinkowskiDirectedEdge,MinkowskiVert};
-use strafesnet_common::integer::{Time,Fixed,Ratio};
-
-#[derive(Debug)]
-enum Transition<F,E:DirectedEdge,V>{
-	Miss,
-	Next(FEV<F,E,V>,GigaTime),
-	Hit(F,GigaTime),
-}
-
-type MinkowskiFEV=FEV<MinkowskiFace,MinkowskiDirectedEdge,MinkowskiVert>;
-type MinkowskiTransition=Transition<MinkowskiFace,MinkowskiDirectedEdge,MinkowskiVert>;
-
-	fn next_transition(fev:&MinkowskiFEV,body_time:GigaTime,mesh:&MinkowskiMesh,body:&Body,mut best_time:GigaTime)->MinkowskiTransition{
-		//conflicting derivative means it crosses in the wrong direction.
-		//if the transition time is equal to an already tested transition, do not replace the current best.
-		let mut best_transition=MinkowskiTransition::Miss;
-		match fev{
-			&MinkowskiFEV::Face(face_id)=>{
-				//test own face collision time, ignoring roots with zero or conflicting derivative
-				//n=face.normal d=face.dot
-				//n.a t^2+n.v t+n.p-d==0
-				let (n,d)=mesh.face_nd(face_id);
-				//TODO: use higher precision d value?
-				//use the mesh transform translation instead of baking it into the d value.
-				for dt in Fixed::<4,128>::zeroes2((n.dot(body.position)-d)*2,n.dot(body.velocity)*2,n.dot(body.acceleration)){
-					if body_time.le_ratio(dt)&&dt.lt_ratio(best_time)&&n.dot(body.extrapolated_velocity_ratio_dt(dt)).is_negative(){
-						best_time=dt;
-						best_transition=MinkowskiTransition::Hit(face_id,dt);
-						break;
-					}
-				}
-				//test each edge collision time, ignoring roots with zero or conflicting derivative
-				for &directed_edge_id in mesh.face_edges(face_id).iter(){
-					let edge_n=mesh.directed_edge_n(directed_edge_id);
-					let n=n.cross(edge_n);
-					let verts=mesh.edge_verts(directed_edge_id.as_undirected());
-					//WARNING: d is moved out of the *2 block because of adding two vertices!
-					//WARNING: precision is swept under the rug!
-					for dt in Fixed::<4,128>::zeroes2(n.dot(body.position*2-(mesh.vert(verts[0])+mesh.vert(verts[1]))).fix_4(),n.dot(body.velocity).fix_4()*2,n.dot(body.acceleration).fix_4()){
-						if body_time.le_ratio(dt)&&dt.lt_ratio(best_time)&&n.dot(body.extrapolated_velocity_ratio_dt(dt)).is_negative(){
-							best_time=dt;
-							best_transition=MinkowskiTransition::Next(MinkowskiFEV::Edge(directed_edge_id.as_undirected()),dt);
-							break;
-						}
-					}
-				}
-				//if none:
-			},
-			&MinkowskiFEV::Edge(edge_id)=>{
-				//test each face collision time, ignoring roots with zero or conflicting derivative
-				let edge_n=mesh.edge_n(edge_id);
-				let edge_verts=mesh.edge_verts(edge_id);
-				let delta_pos=body.position*2-(mesh.vert(edge_verts[0])+mesh.vert(edge_verts[1]));
-				for (i,&edge_face_id) in mesh.edge_faces(edge_id).iter().enumerate(){
-					let face_n=mesh.face_nd(edge_face_id).0;
-					//edge_n gets parity from the order of edge_faces
-					let n=face_n.cross(edge_n)*((i as i64)*2-1);
-					//WARNING yada yada d *2
-					for dt in Fixed::<4,128>::zeroes2(n.dot(delta_pos).fix_4(),n.dot(body.velocity).fix_4()*2,n.dot(body.acceleration).fix_4()){
-						if body_time.le_ratio(dt)&&dt.lt_ratio(best_time)&&n.dot(body.extrapolated_velocity_ratio_dt(dt)).is_negative(){
-							best_time=dt;
-							best_transition=MinkowskiTransition::Next(MinkowskiFEV::Face(edge_face_id),dt);
-							break;
-						}
-					}
-				}
-				//test each vertex collision time, ignoring roots with zero or conflicting derivative
-				for (i,&vert_id) in edge_verts.iter().enumerate(){
-					//vertex normal gets parity from vert index
-					let n=edge_n*(1-2*(i as i64));
-					for dt in Fixed::<2,64>::zeroes2((n.dot(body.position-mesh.vert(vert_id)))*2,n.dot(body.velocity)*2,n.dot(body.acceleration)){
-						if body_time.le_ratio(dt)&&dt.lt_ratio(best_time)&&n.dot(body.extrapolated_velocity_ratio_dt(dt)).is_negative(){
-							let dt=Ratio::new(dt.num.fix_4(),dt.den.fix_4());
-							best_time=dt;
-							best_transition=MinkowskiTransition::Next(MinkowskiFEV::Vert(vert_id),dt);
-							break;
-						}
-					}
-				}
-				//if none:
-			},
-			&MinkowskiFEV::Vert(vert_id)=>{
-				//test each edge collision time, ignoring roots with zero or conflicting derivative
-				for &directed_edge_id in mesh.vert_edges(vert_id).iter(){
-					//edge is directed away from vertex, but we want the dot product to turn out negative
-					let n=-mesh.directed_edge_n(directed_edge_id);
-					for dt in Fixed::<2,64>::zeroes2((n.dot(body.position-mesh.vert(vert_id)))*2,n.dot(body.velocity)*2,n.dot(body.acceleration)){
-						if body_time.le_ratio(dt)&&dt.lt_ratio(best_time)&&n.dot(body.extrapolated_velocity_ratio_dt(dt)).is_negative(){
-							let dt=Ratio::new(dt.num.fix_4(),dt.den.fix_4());
-							best_time=dt;
-							best_transition=MinkowskiTransition::Next(MinkowskiFEV::Edge(directed_edge_id.as_undirected()),dt);
-							break;
-						}
-					}
-				}
-				//if none:
-			},
-		}
-		best_transition
-	}
-pub enum CrawlResult<F,E:DirectedEdge,V>{
-	Miss(FEV<F,E,V>),
-	Hit(F,GigaTime),
-}
-type MinkowskiCrawlResult=CrawlResult<MinkowskiFace,MinkowskiDirectedEdge,MinkowskiVert>;
-pub fn crawl_fev(mut fev:MinkowskiFEV,mesh:&MinkowskiMesh,relative_body:&Body,start_time:Time,time_limit:Time)->MinkowskiCrawlResult{
-	let mut body_time={
-		let r=(start_time-relative_body.time).to_ratio();
-		Ratio::new(r.num.fix_4(),r.den.fix_4())
-	};
-	let time_limit={
-		let r=(time_limit-relative_body.time).to_ratio();
-		Ratio::new(r.num.fix_4(),r.den.fix_4())
-	};
-	for _ in 0..20{
-		match next_transition(&fev,body_time,mesh,relative_body,time_limit){
-			Transition::Miss=>return CrawlResult::Miss(fev),
-			Transition::Next(next_fev,next_time)=>(fev,body_time)=(next_fev,next_time),
-			Transition::Hit(face,time)=>return CrawlResult::Hit(face,time),
-		}
-	}
-	//TODO: fix all bugs
-	//println!("Too many iterations!  Using default behaviour instead of crashing...");
-	CrawlResult::Miss(fev)
-}

src/file.rs

@@ -1,144 +0,0 @@
-use std::io::Read;
-
-#[derive(Debug)]
-pub enum ReadError{
-	#[cfg(feature="roblox")]
-	Roblox(strafesnet_rbx_loader::ReadError),
-	#[cfg(feature="source")]
-	Source(strafesnet_bsp_loader::ReadError),
-	#[cfg(feature="snf")]
-	StrafesNET(strafesnet_snf::Error),
-	#[cfg(feature="snf")]
-	StrafesNETMap(strafesnet_snf::map::Error),
-	Io(std::io::Error),
-	UnknownFileFormat,
-}
-impl std::fmt::Display for ReadError{
-	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
-		write!(f,"{self:?}")
-	}
-}
-impl std::error::Error for ReadError{}
-
-pub enum DataStructure{
-	#[cfg(feature="roblox")]
-	Roblox(strafesnet_rbx_loader::Model),
-	#[cfg(feature="source")]
-	Source(strafesnet_bsp_loader::Bsp),
-	#[cfg(feature="snf")]
-	StrafesNET(strafesnet_common::map::CompleteMap),
-}
-
-pub fn read<R:Read+std::io::Seek>(input:R)->Result<DataStructure,ReadError>{
-	let mut buf=std::io::BufReader::new(input);
-	let peek=std::io::BufRead::fill_buf(&mut buf).map_err(ReadError::Io)?;
-	match &peek[0..4]{
-		#[cfg(feature="roblox")]
-		b"<rob"=>Ok(DataStructure::Roblox(strafesnet_rbx_loader::read(buf).map_err(ReadError::Roblox)?)),
-		#[cfg(feature="source")]
-		b"VBSP"=>Ok(DataStructure::Source(strafesnet_bsp_loader::read(buf).map_err(ReadError::Source)?)),
-		#[cfg(feature="snf")]
-		b"SNFM"=>Ok(DataStructure::StrafesNET(
-			strafesnet_snf::read_map(buf).map_err(ReadError::StrafesNET)?
-			.into_complete_map().map_err(ReadError::StrafesNETMap)?
-		)),
-		_=>Err(ReadError::UnknownFileFormat),
-	}
-}
-
-#[derive(Debug)]
-pub enum LoadError{
-	ReadError(ReadError),
-	File(std::io::Error),
-	Io(std::io::Error),
-}
-impl std::fmt::Display for LoadError{
-	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
-		write!(f,"{self:?}")
-	}
-}
-impl std::error::Error for LoadError{}
-
-pub fn load<P:AsRef<std::path::Path>>(path:P)->Result<strafesnet_common::map::CompleteMap,LoadError>{
-	//blocking because it's simpler...
-	let file=std::fs::File::open(path).map_err(LoadError::File)?;
-	match read(file).map_err(LoadError::ReadError)?{
-		#[cfg(feature="snf")]
-		DataStructure::StrafesNET(map)=>Ok(map),
-		#[cfg(feature="roblox")]
-		DataStructure::Roblox(model)=>{
-			let mut place=model.into_place();
-			place.run_scripts();
-
-			let mut loader=strafesnet_deferred_loader::roblox_legacy();
-
-			let (texture_loader,mesh_loader)=loader.get_inner_mut();
-
-			let map_step1=strafesnet_rbx_loader::convert(
-				&place,
-				|name|texture_loader.acquire_render_config_id(name),
-				|name|mesh_loader.acquire_mesh_id(name),
-			);
-
-			let meshpart_meshes=mesh_loader.load_meshes().map_err(LoadError::Io)?;
-
-			let map_step2=map_step1.add_meshpart_meshes_and_calculate_attributes(
-				meshpart_meshes.into_iter().map(|(mesh_id,loader_model)|
-					(mesh_id,strafesnet_rbx_loader::data::RobloxMeshBytes::new(loader_model.get()))
-				)
-			);
-
-			let (textures,render_configs)=loader.into_render_configs().map_err(LoadError::Io)?.consume();
-
-			let map=map_step2.add_render_configs_and_textures(
-				render_configs.into_iter(),
-				textures.into_iter().map(|(texture_id,texture)|
-					(texture_id,match texture{
-						strafesnet_deferred_loader::texture::Texture::ImageDDS(data)=>data,
-					})
-				)
-			);
-
-			Ok(map)
-		},
-		#[cfg(feature="source")]
-		DataStructure::Source(bsp)=>{
-			let mut loader=strafesnet_deferred_loader::source_legacy();
-
-			let (texture_loader,mesh_loader)=loader.get_inner_mut();
-
-			let map_step1=strafesnet_bsp_loader::convert(
-				&bsp,
-				|name|texture_loader.acquire_render_config_id(name),
-				|name|mesh_loader.acquire_mesh_id(name),
-			);
-
-			let prop_meshes=mesh_loader.load_meshes(bsp.as_ref());
-
-			let map_step2=map_step1.add_prop_meshes(
-				//the type conflagulator 9000
-				prop_meshes.into_iter().map(|(mesh_id,loader_model)|
-					(mesh_id,strafesnet_bsp_loader::data::ModelData{
-						mdl:strafesnet_bsp_loader::data::MdlData::new(loader_model.mdl.get()),
-						vtx:strafesnet_bsp_loader::data::VtxData::new(loader_model.vtx.get()),
-						vvd:strafesnet_bsp_loader::data::VvdData::new(loader_model.vvd.get()),
-					})
-				),
-				|name|texture_loader.acquire_render_config_id(name),
-			);
-
-			let (textures,render_configs)=loader.into_render_configs().map_err(LoadError::Io)?.consume();
-
-			let map=map_step2.add_render_configs_and_textures(
-				render_configs.into_iter(),
-				textures.into_iter().map(|(texture_id,texture)|
-					(texture_id,match texture{
-						strafesnet_deferred_loader::texture::Texture::ImageDDS(data)=>data,
-					})
-				),
-			);
-
-			Ok(map)
-		},
-	}
-}

src/framework.rs

@@ -0,0 +1,516 @@
+use std::future::Future;
+#[cfg(target_arch = "wasm32")]
+use std::str::FromStr;
+#[cfg(target_arch = "wasm32")]
+use web_sys::{ImageBitmapRenderingContext, OffscreenCanvas};
+use winit::{
+	event::{self, WindowEvent, DeviceEvent},
+	event_loop::{ControlFlow, EventLoop},
+};
+
+#[allow(dead_code)]
+pub fn cast_slice<T>(data: &[T]) -> &[u8] {
+	use std::{mem::size_of, slice::from_raw_parts};
+
+	unsafe { from_raw_parts(data.as_ptr() as *const u8, data.len() * size_of::<T>()) }
+}
+
+#[allow(dead_code)]
+pub enum ShaderStage {
+	Vertex,
+	Fragment,
+	Compute,
+}
+
+pub trait Example: 'static + Sized {
+	fn optional_features() -> wgpu::Features {
+		wgpu::Features::empty()
+	}
+	fn required_features() -> wgpu::Features {
+		wgpu::Features::empty()
+	}
+	fn required_downlevel_capabilities() -> wgpu::DownlevelCapabilities {
+		wgpu::DownlevelCapabilities {
+			flags: wgpu::DownlevelFlags::empty(),
+			shader_model: wgpu::ShaderModel::Sm5,
+			..wgpu::DownlevelCapabilities::default()
+		}
+	}
+	fn required_limits() -> wgpu::Limits {
+		wgpu::Limits::downlevel_webgl2_defaults() // These downlevel limits will allow the code to run on all possible hardware
+	}
+	fn init(
+		config: &wgpu::SurfaceConfiguration,
+		adapter: &wgpu::Adapter,
+		device: &wgpu::Device,
+		queue: &wgpu::Queue,
+	) -> Self;
+	fn resize(
+		&mut self,
+		config: &wgpu::SurfaceConfiguration,
+		device: &wgpu::Device,
+		queue: &wgpu::Queue,
+	);
+	fn update(&mut self, window: &winit::window::Window, device: &wgpu::Device, queue: &wgpu::Queue, event: WindowEvent);
+	fn device_event(&mut self, window: &winit::window::Window, event: DeviceEvent);
+	fn load_file(&mut self, path:std::path::PathBuf, device: &wgpu::Device, queue: &wgpu::Queue);
+	fn render(
+		&mut self,
+		view: &wgpu::TextureView,
+		device: &wgpu::Device,
+		queue: &wgpu::Queue,
+		spawner: &Spawner,
+	);
+}
+
+struct Setup {
+	window: winit::window::Window,
+	event_loop: EventLoop<()>,
+	instance: wgpu::Instance,
+	size: winit::dpi::PhysicalSize<u32>,
+	surface: wgpu::Surface,
+	adapter: wgpu::Adapter,
+	device: wgpu::Device,
+	queue: wgpu::Queue,
+	#[cfg(target_arch = "wasm32")]
+	offscreen_canvas_setup: Option<OffscreenCanvasSetup>,
+}
+
+#[cfg(target_arch = "wasm32")]
+struct OffscreenCanvasSetup {
+	offscreen_canvas: OffscreenCanvas,
+	bitmap_renderer: ImageBitmapRenderingContext,
+}
+
+async fn setup<E: Example>(title: &str) -> Setup {
+	#[cfg(not(target_arch = "wasm32"))]
+	{
+		env_logger::init();
+	};
+
+	let event_loop = EventLoop::new();
+	let mut builder = winit::window::WindowBuilder::new();
+	builder = builder.with_title(title);
+	#[cfg(windows_OFF)] // TODO
+	{
+		use winit::platform::windows::WindowBuilderExtWindows;
+		builder = builder.with_no_redirection_bitmap(true);
+	}
+	let window = builder.build(&event_loop).unwrap();
+
+	#[cfg(target_arch = "wasm32")]
+	{
+		use winit::platform::web::WindowExtWebSys;
+		let query_string = web_sys::window().unwrap().location().search().unwrap();
+		let level: log::Level = parse_url_query_string(&query_string, "RUST_LOG")
+			.and_then(|x| x.parse().ok())
+			.unwrap_or(log::Level::Error);
+		console_log::init_with_level(level).expect("could not initialize logger");
+		std::panic::set_hook(Box::new(console_error_panic_hook::hook));
+		// On wasm, append the canvas to the document body
+		web_sys::window()
+			.and_then(|win| win.document())
+			.and_then(|doc| doc.body())
+			.and_then(|body| {
+				body.append_child(&web_sys::Element::from(window.canvas()))
+					.ok()
+			})
+			.expect("couldn't append canvas to document body");
+	}
+
+	#[cfg(target_arch = "wasm32")]
+	let mut offscreen_canvas_setup: Option<OffscreenCanvasSetup> = None;
+	#[cfg(target_arch = "wasm32")]
+	{
+		use wasm_bindgen::JsCast;
+		use winit::platform::web::WindowExtWebSys;
+
+		let query_string = web_sys::window().unwrap().location().search().unwrap();
+		if let Some(offscreen_canvas_param) =
+			parse_url_query_string(&query_string, "offscreen_canvas")
+		{
+			if FromStr::from_str(offscreen_canvas_param) == Ok(true) {
+				log::info!("Creating OffscreenCanvasSetup");
+
+				let offscreen_canvas =
+					OffscreenCanvas::new(1024, 768).expect("couldn't create OffscreenCanvas");
+
+				let bitmap_renderer = window
+					.canvas()
+					.get_context("bitmaprenderer")
+					.expect("couldn't create ImageBitmapRenderingContext (Result)")
+					.expect("couldn't create ImageBitmapRenderingContext (Option)")
+					.dyn_into::<ImageBitmapRenderingContext>()
+					.expect("couldn't convert into ImageBitmapRenderingContext");
+
+				offscreen_canvas_setup = Some(OffscreenCanvasSetup {
+					offscreen_canvas,
+					bitmap_renderer,
+				})
+			}
+		}
+	};
+
+	log::info!("Initializing the surface...");
+
+	let backends = wgpu::util::backend_bits_from_env().unwrap_or_else(wgpu::Backends::all);
+	let dx12_shader_compiler = wgpu::util::dx12_shader_compiler_from_env().unwrap_or_default();
+
+	let instance = wgpu::Instance::new(wgpu::InstanceDescriptor {
+		backends,
+		dx12_shader_compiler,
+	});
+	let (size, surface) = unsafe {
+		let size = window.inner_size();
+
+		#[cfg(any(not(target_arch = "wasm32"), target_os = "emscripten"))]
+		let surface = instance.create_surface(&window).unwrap();
+		#[cfg(all(target_arch = "wasm32", not(target_os = "emscripten")))]
+		let surface = {
+			if let Some(offscreen_canvas_setup) = &offscreen_canvas_setup {
+				log::info!("Creating surface from OffscreenCanvas");
+				instance.create_surface_from_offscreen_canvas(
+					offscreen_canvas_setup.offscreen_canvas.clone(),
+				)
+			} else {
+				instance.create_surface(&window)
+			}
+		}
+		.unwrap();
+
+		(size, surface)
+	};
+
+	let adapter;
+
+	let optional_features = E::optional_features();
+	let required_features = E::required_features();
+
+	//no helper function smh gotta write it myself
+	let adapters = instance.enumerate_adapters(backends);
+
+	let mut chosen_adapter = None;
+	let mut chosen_adapter_score=0;
+	for adapter in adapters {
+		if !adapter.is_surface_supported(&surface) {
+			continue;
+		}
+
+		let score=match adapter.get_info().device_type{
+			wgpu::DeviceType::IntegratedGpu=>3,
+			wgpu::DeviceType::DiscreteGpu=>4,
+			wgpu::DeviceType::VirtualGpu=>2,
+			wgpu::DeviceType::Other|wgpu::DeviceType::Cpu=>1,
+		};
+
+		let adapter_features = adapter.features();
+		if chosen_adapter_score<score&&adapter_features.contains(required_features) {
+			chosen_adapter_score=score;
+			chosen_adapter=Some(adapter);
+		}
+	}
+
+	if let Some(maybe_chosen_adapter) = chosen_adapter{
+		adapter=maybe_chosen_adapter;
+	}else{
+		panic!("No suitable GPU adapters found on the system!");
+	}
+
+	#[cfg(not(target_arch = "wasm32"))]
+	{
+		let adapter_info = adapter.get_info();
+		println!("Using {} ({:?})", adapter_info.name, adapter_info.backend);
+	}
+
+	let required_downlevel_capabilities = E::required_downlevel_capabilities();
+	let downlevel_capabilities = adapter.get_downlevel_capabilities();
+	assert!(
+		downlevel_capabilities.shader_model >= required_downlevel_capabilities.shader_model,
+		"Adapter does not support the minimum shader model required to run this example: {:?}",
+		required_downlevel_capabilities.shader_model
+	);
+	assert!(
+		downlevel_capabilities
+			.flags
+			.contains(required_downlevel_capabilities.flags),
+		"Adapter does not support the downlevel capabilities required to run this example: {:?}",
+		required_downlevel_capabilities.flags - downlevel_capabilities.flags
+	);
+
+	// Make sure we use the texture resolution limits from the adapter, so we can support images the size of the surface.
+	let needed_limits = E::required_limits().using_resolution(adapter.limits());
+
+	let trace_dir = std::env::var("WGPU_TRACE");
+	let (device, queue) = adapter
+		.request_device(
+			&wgpu::DeviceDescriptor {
+				label: None,
+				features: (optional_features & adapter.features()) | required_features,
+				limits: needed_limits,
+			},
+			trace_dir.ok().as_ref().map(std::path::Path::new),
+		)
+		.await
+		.expect("Unable to find a suitable GPU adapter!");
+
+	Setup {
+		window,
+		event_loop,
+		instance,
+		size,
+		surface,
+		adapter,
+		device,
+		queue,
+		#[cfg(target_arch = "wasm32")]
+		offscreen_canvas_setup,
+	}
+}
+
+fn start<E: Example>(
+	#[cfg(not(target_arch = "wasm32"))] Setup {
+		window,
+		event_loop,
+		instance,
+		size,
+		surface,
+		adapter,
+		device,
+		queue,
+	}: Setup,
+	#[cfg(target_arch = "wasm32")] Setup {
+		window,
+		event_loop,
+		instance,
+		size,
+		surface,
+		adapter,
+		device,
+		queue,
+		offscreen_canvas_setup,
+	}: Setup,
+) {
+	let spawner = Spawner::new();
+	let mut config = surface
+		.get_default_config(&adapter, size.width, size.height)
+		.expect("Surface isn't supported by the adapter.");
+	let surface_view_format = config.format.add_srgb_suffix();
+	config.view_formats.push(surface_view_format);
+	surface.configure(&device, &config);
+
+	log::info!("Initializing the example...");
+	let mut example = E::init(&config, &adapter, &device, &queue);
+
+	log::info!("Entering render loop...");
+	event_loop.run(move |event, _, control_flow| {
+		let _ = (&instance, &adapter); // force ownership by the closure
+		*control_flow = if cfg!(feature = "metal-auto-capture") {
+			ControlFlow::Exit
+		} else {
+			ControlFlow::Poll
+		};
+		match event {
+			event::Event::RedrawEventsCleared => {
+				#[cfg(not(target_arch = "wasm32"))]
+				spawner.run_until_stalled();
+
+				window.request_redraw();
+			}
+			event::Event::WindowEvent {
+				event:
+					WindowEvent::Resized(size)
+					| WindowEvent::ScaleFactorChanged {
+						new_inner_size: &mut size,
+						..
+					},
+				..
+			} => {
+				// Once winit is fixed, the detection conditions here can be removed.
+				// https://github.com/rust-windowing/winit/issues/2876
+				let max_dimension = adapter.limits().max_texture_dimension_2d;
+				if size.width > max_dimension || size.height > max_dimension {
+					log::warn!(
+						"The resizing size {:?} exceeds the limit of {}.",
+						size,
+						max_dimension
+					);
+				} else {
+					log::info!("Resizing to {:?}", size);
+					config.width = size.width.max(1);
+					config.height = size.height.max(1);
+					example.resize(&config, &device, &queue);
+					surface.configure(&device, &config);
+				}
+			}
+			event::Event::WindowEvent { event, .. } => match event {
+				WindowEvent::KeyboardInput {
+					input:
+						event::KeyboardInput {
+							virtual_keycode: Some(event::VirtualKeyCode::Escape),
+							state: event::ElementState::Pressed,
+							..
+						},
+					..
+				}
+				| WindowEvent::CloseRequested => {
+					*control_flow = ControlFlow::Exit;
+				}
+				#[cfg(not(target_arch = "wasm32"))]
+				WindowEvent::KeyboardInput {
+					input:
+						event::KeyboardInput {
+							virtual_keycode: Some(event::VirtualKeyCode::Scroll),
+							state: event::ElementState::Pressed,
+							..
+						},
+					..
+				} => {
+					println!("{:#?}", instance.generate_report());
+				}
+				_ => {
+					example.update(&window,&device,&queue,event);
+				}
+			},
+			event::Event::DeviceEvent {
+				event,
+				..
+			} => {
+				example.device_event(&window,event);
+			},
+			event::Event::RedrawRequested(_) => {
+
+				let frame = match surface.get_current_texture() {
+					Ok(frame) => frame,
+					Err(_) => {
+						surface.configure(&device, &config);
+						surface
+							.get_current_texture()
+							.expect("Failed to acquire next surface texture!")
+					}
+				};
+				let view = frame.texture.create_view(&wgpu::TextureViewDescriptor {
+					format: Some(surface_view_format),
+					..wgpu::TextureViewDescriptor::default()
+				});
+
+				example.render(&view, &device, &queue, &spawner);
+
+				frame.present();
+
+				#[cfg(target_arch = "wasm32")]
+				{
+					if let Some(offscreen_canvas_setup) = &offscreen_canvas_setup {
+						let image_bitmap = offscreen_canvas_setup
+							.offscreen_canvas
+							.transfer_to_image_bitmap()
+							.expect("couldn't transfer offscreen canvas to image bitmap.");
+						offscreen_canvas_setup
+							.bitmap_renderer
+							.transfer_from_image_bitmap(&image_bitmap);
+
+						log::info!("Transferring OffscreenCanvas to ImageBitmapRenderer");
+					}
+				}
+			}
+			_ => {}
+		}
+	});
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+pub struct Spawner<'a> {
+	executor: async_executor::LocalExecutor<'a>,
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+impl<'a> Spawner<'a> {
+	fn new() -> Self {
+		Self {
+			executor: async_executor::LocalExecutor::new(),
+		}
+	}
+
+	#[allow(dead_code)]
+	pub fn spawn_local(&self, future: impl Future<Output = ()> + 'a) {
+		self.executor.spawn(future).detach();
+	}
+
+	fn run_until_stalled(&self) {
+		while self.executor.try_tick() {}
+	}
+}
+
+#[cfg(target_arch = "wasm32")]
+pub struct Spawner {}
+
+#[cfg(target_arch = "wasm32")]
+impl Spawner {
+	fn new() -> Self {
+		Self {}
+	}
+
+	#[allow(dead_code)]
+	pub fn spawn_local(&self, future: impl Future<Output = ()> + 'static) {
+		wasm_bindgen_futures::spawn_local(future);
+	}
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+pub fn run<E: Example>(title: &str) {
+	let setup = pollster::block_on(setup::<E>(title));
+	start::<E>(setup);
+}
+
+#[cfg(target_arch = "wasm32")]
+pub fn run<E: Example>(title: &str) {
+	use wasm_bindgen::prelude::*;
+
+	let title = title.to_owned();
+	wasm_bindgen_futures::spawn_local(async move {
+		let setup = setup::<E>(&title).await;
+		let start_closure = Closure::once_into_js(move || start::<E>(setup));
+
+		// make sure to handle JS exceptions thrown inside start.
+		// Otherwise wasm_bindgen_futures Queue would break and never handle any tasks again.
+		// This is required, because winit uses JS exception for control flow to escape from `run`.
+		if let Err(error) = call_catch(&start_closure) {
+			let is_control_flow_exception = error.dyn_ref::<js_sys::Error>().map_or(false, |e| {
+				e.message().includes("Using exceptions for control flow", 0)
+			});
+
+			if !is_control_flow_exception {
+				web_sys::console::error_1(&error);
+			}
+		}
+
+		#[wasm_bindgen]
+		extern "C" {
+			#[wasm_bindgen(catch, js_namespace = Function, js_name = "prototype.call.call")]
+			fn call_catch(this: &JsValue) -> Result<(), JsValue>;
+		}
+	});
+}
+
+#[cfg(target_arch = "wasm32")]
+/// Parse the query string as returned by `web_sys::window()?.location().search()?` and get a
+/// specific key out of it.
+pub fn parse_url_query_string<'a>(query: &'a str, search_key: &str) -> Option<&'a str> {
+	let query_string = query.strip_prefix('?')?;
+
+	for pair in query_string.split('&') {
+		let mut pair = pair.split('=');
+		let key = pair.next()?;
+		let value = pair.next()?;
+
+		if key == search_key {
+			return Some(value);
+		}
+	}
+
+	None
+}
+
+// This allows treating the framework as a standalone example,
+// thus avoiding listing the example names in `Cargo.toml`.
+#[allow(dead_code)]
+fn main() {}

src/graphics.rs

@@ -1,989 +0,0 @@
-use std::borrow::Cow;
-use std::collections::{HashSet,HashMap};
-use strafesnet_common::map;
-use strafesnet_common::integer;
-use strafesnet_common::model::{self, ColorId, NormalId, PolygonIter, PositionId, RenderConfigId, TextureCoordinateId, VertexId};
-use wgpu::{util::DeviceExt,AstcBlock,AstcChannel};
-use crate::model_graphics::{self,IndexedGraphicsMeshOwnedRenderConfig,IndexedGraphicsMeshOwnedRenderConfigId,GraphicsMeshOwnedRenderConfig,GraphicsModelColor4,GraphicsModelOwned,GraphicsVertex};
-
-struct Indices{
-	count:u32,
-	buf:wgpu::Buffer,
-	format:wgpu::IndexFormat,
-}
-impl Indices{
-	fn new<T:bytemuck::Pod>(device:&wgpu::Device,indices:&Vec<T>,format:wgpu::IndexFormat)->Self{
-		Self{
-			buf:device.create_buffer_init(&wgpu::util::BufferInitDescriptor{
-				label:Some("Index"),
-				contents:bytemuck::cast_slice(indices),
-				usage:wgpu::BufferUsages::INDEX,
-			}),
-			count:indices.len() as u32,
-			format,
-		}
-	}
-}
-struct GraphicsModel{
-	indices:Indices,
-	vertex_buf:wgpu::Buffer,
-	bind_group:wgpu::BindGroup,
-	instance_count:u32,
-}
-
-struct GraphicsSamplers{
-	repeat:wgpu::Sampler,
-}
-
-struct GraphicsBindGroupLayouts{
-	model:wgpu::BindGroupLayout,
-}
-
-struct GraphicsBindGroups{
-	camera:wgpu::BindGroup,
-	skybox_texture:wgpu::BindGroup,
-}
-
-struct GraphicsPipelines{
-	skybox:wgpu::RenderPipeline,
-	model:wgpu::RenderPipeline,
-}
-
-struct GraphicsCamera{
-	screen_size:glam::UVec2,
-	fov:glam::Vec2,//slope
-	//camera angles and such are extrapolated and passed in every time
-}
-
-#[inline]
-fn perspective_rh(fov_x_slope:f32,fov_y_slope:f32,z_near:f32,z_far:f32)->glam::Mat4{
-	//glam_assert!(z_near > 0.0 && z_far > 0.0);
-	let r=z_far/(z_near-z_far);
-	glam::Mat4::from_cols(
-		glam::Vec4::new(1.0/fov_x_slope,0.0,0.0,0.0),
-		glam::Vec4::new(0.0,1.0/fov_y_slope,0.0,0.0),
-		glam::Vec4::new(0.0,0.0,r,-1.0),
-		glam::Vec4::new(0.0,0.0,r*z_near,0.0),
-	)
-}
-impl GraphicsCamera{
-	pub fn proj(&self)->glam::Mat4{
-		perspective_rh(self.fov.x,self.fov.y,0.4,4000.0)
-	}
-	pub fn world(&self,pos:glam::Vec3,angles:glam::Vec2)->glam::Mat4{
-		//f32 good enough for view matrix
-		glam::Mat4::from_translation(pos)*glam::Mat4::from_euler(glam::EulerRot::YXZ,angles.x,angles.y,0f32)
-	}
-
-	pub fn to_uniform_data(&self,pos:glam::Vec3,angles:glam::Vec2)->[f32;16*4]{
-		let proj=self.proj();
-		let proj_inv=proj.inverse();
-		let view_inv=self.world(pos,angles);
-		let view=view_inv.inverse();
-
-		let mut raw=[0f32; 16 * 4];
-		raw[..16].copy_from_slice(&AsRef::<[f32; 16]>::as_ref(&proj)[..]);
-		raw[16..32].copy_from_slice(&AsRef::<[f32; 16]>::as_ref(&proj_inv)[..]);
-		raw[32..48].copy_from_slice(&AsRef::<[f32; 16]>::as_ref(&view)[..]);
-		raw[48..64].copy_from_slice(&AsRef::<[f32; 16]>::as_ref(&view_inv)[..]);
-		raw
-	}
-}
-impl std::default::Default for GraphicsCamera{
-	fn default()->Self{
-		Self{
-			screen_size:glam::UVec2::ONE,
-			fov:glam::Vec2::ONE,
-		}
-	}
-}
-
-pub struct FrameState{
-	pub body:crate::physics::Body,
-	pub camera:crate::physics::PhysicsCamera,
-	pub time:integer::Time,
-}
-
-pub struct GraphicsState{
-	pipelines:GraphicsPipelines,
-	bind_groups:GraphicsBindGroups,
-	bind_group_layouts:GraphicsBindGroupLayouts,
-	samplers:GraphicsSamplers,
-	camera:GraphicsCamera,
-	camera_buf:wgpu::Buffer,
-	temp_squid_texture_view:wgpu::TextureView,
-	models:Vec<GraphicsModel>,
-	depth_view:wgpu::TextureView,
-	staging_belt:wgpu::util::StagingBelt,
-}
-
-impl GraphicsState{
-	const DEPTH_FORMAT:wgpu::TextureFormat=wgpu::TextureFormat::Depth24Plus;
-	fn create_depth_texture(
-		config:&wgpu::SurfaceConfiguration,
-		device:&wgpu::Device,
-	)->wgpu::TextureView{
-		let depth_texture=device.create_texture(&wgpu::TextureDescriptor{
-			size:wgpu::Extent3d{
-				width:config.width,
-				height:config.height,
-				depth_or_array_layers:1,
-			},
-			mip_level_count:1,
-			sample_count:1,
-			dimension:wgpu::TextureDimension::D2,
-			format:Self::DEPTH_FORMAT,
-			usage:wgpu::TextureUsages::RENDER_ATTACHMENT,
-			label:None,
-			view_formats:&[],
-		});
-
-		depth_texture.create_view(&wgpu::TextureViewDescriptor::default())
-	}
-	pub fn clear(&mut self){
-		self.models.clear();
-	}
-	pub fn load_user_settings(&mut self,user_settings:&crate::settings::UserSettings){
-		self.camera.fov=user_settings.calculate_fov(1.0,&self.camera.screen_size).as_vec2();
-	}
-	pub fn generate_models(&mut self,device:&wgpu::Device,queue:&wgpu::Queue,map:&map::CompleteMap){
-		//generate texture view per texture
-		let texture_views:HashMap<strafesnet_common::model::TextureId,wgpu::TextureView>=map.textures.iter().enumerate().filter_map(|(texture_id,texture_data)|{
-			let texture_id=model::TextureId::new(texture_id as u32);
-			let image=match ddsfile::Dds::read(std::io::Cursor::new(texture_data)){
-				Ok(image)=>image,
-				Err(e)=>{
-					println!("Error loading texture: {e}");
-					return None;
-				},
-			};
-
-			let (mut width,mut height)=(image.get_width(),image.get_height());
-
-			let format=match image.header10.unwrap().dxgi_format{
-				ddsfile::DxgiFormat::R8G8B8A8_UNorm_sRGB=>wgpu::TextureFormat::Rgba8UnormSrgb,
-				ddsfile::DxgiFormat::BC7_UNorm_sRGB =>{
-					//floor(w,4),should be ceil(w,4)
-					width=width/4*4;
-					height=height/4*4;
-					wgpu::TextureFormat::Bc7RgbaUnormSrgb
-				},
-				other=>{
-					println!("unsupported texture format{:?}",other);
-					return None;
-				},
-			};
-
-			let size=wgpu::Extent3d{
-				width,
-				height,
-				depth_or_array_layers:1,
-			};
-
-			let layer_size=wgpu::Extent3d{
-				depth_or_array_layers:1,
-				..size
-			};
-			let max_mips=layer_size.max_mips(wgpu::TextureDimension::D2);
-
-			let texture=device.create_texture_with_data(
-				queue,
-				&wgpu::TextureDescriptor{
-					size,
-					mip_level_count:max_mips,
-					sample_count:1,
-					dimension:wgpu::TextureDimension::D2,
-					format,
-					usage:wgpu::TextureUsages::TEXTURE_BINDING|wgpu::TextureUsages::COPY_DST,
-					label:Some(format!("Texture{}",texture_id.get()).as_str()),
-					view_formats:&[],
-				},
-				wgpu::util::TextureDataOrder::LayerMajor,
-				&image.data,
-			);
-			Some((texture_id,texture.create_view(&wgpu::TextureViewDescriptor{
-				label:Some(format!("Texture{} View",texture_id.get()).as_str()),
-				dimension:Some(wgpu::TextureViewDimension::D2),
-				..wgpu::TextureViewDescriptor::default()
-			})))
-		}).collect();
-		let num_textures=texture_views.len();
-
-		//split groups with different textures into separate models
-		//the models received here are supposed to be tightly packed,i.e. no code needs to check if two models are using the same groups.
-		let indexed_models_len=map.models.len();
-		//models split into graphics_group.RenderConfigId
-		let mut owned_mesh_id_from_mesh_id_render_config_id:HashMap<model::MeshId,HashMap<RenderConfigId,IndexedGraphicsMeshOwnedRenderConfigId>>=HashMap::new();
-		let mut unique_render_config_models:Vec<IndexedGraphicsMeshOwnedRenderConfig>=Vec::with_capacity(indexed_models_len);
-		for model in &map.models{
-			//wow
-			let instance=GraphicsModelOwned{
-				transform:model.transform.into(),
-				normal_transform:glam::Mat3::from_cols_array_2d(&model.transform.matrix3.to_array().map(|row|row.map(Into::into))).inverse().transpose(),
-				color:GraphicsModelColor4::new(model.color),
-			};
-			//get or create owned mesh map
-			let owned_mesh_map=owned_mesh_id_from_mesh_id_render_config_id
-			.entry(model.mesh).or_insert_with(||{
-				let mut owned_mesh_map=HashMap::new();
-				//add mesh if renderid never before seen for this model
-				//add instance
-				//convert Model into GraphicsModelOwned
-				//check each group, if it's using a new render config then make a new clone of the model
-				if let Some(mesh)=map.meshes.get(model.mesh.get() as usize){
-					for graphics_group in mesh.graphics_groups.iter(){
-						//get or create owned mesh
-						let owned_mesh_id=owned_mesh_map
-						.entry(graphics_group.render).or_insert_with(||{
-							//create
-							let owned_mesh_id=IndexedGraphicsMeshOwnedRenderConfigId::new(unique_render_config_models.len() as u32);
-							unique_render_config_models.push(IndexedGraphicsMeshOwnedRenderConfig{
-								unique_pos:mesh.unique_pos.iter().map(|v|v.to_array().map(Into::into)).collect(),
-								unique_tex:mesh.unique_tex.iter().map(|v|*v.as_ref()).collect(),
-								unique_normal:mesh.unique_normal.iter().map(|v|v.to_array().map(Into::into)).collect(),
-								unique_color:mesh.unique_color.iter().map(|v|*v.as_ref()).collect(),
-								unique_vertices:mesh.unique_vertices.clone(),
-								render_config:graphics_group.render,
-								polys:model::PolygonGroup::PolygonList(model::PolygonList::new(Vec::new())),
-								instances:Vec::new(),
-							});
-							owned_mesh_id
-						});
-						let owned_mesh=unique_render_config_models.get_mut(owned_mesh_id.get() as usize).unwrap();
-						match &mut owned_mesh.polys{
-							model::PolygonGroup::PolygonList(polygon_list)=>polygon_list.extend(
-								graphics_group.groups.iter().flat_map(|polygon_group_id|{
-									mesh.polygon_groups[polygon_group_id.get() as usize].polys()
-								})
-								.map(|vertex_id_slice|
-									vertex_id_slice.to_vec()
-								)
-							),
-						}
-					}
-				}
-				owned_mesh_map
-			});
-			for owned_mesh_id in owned_mesh_map.values(){
-				let owned_mesh=unique_render_config_models.get_mut(owned_mesh_id.get() as usize).unwrap();
-				let render_config=&map.render_configs[owned_mesh.render_config.get() as usize];
-				if model.color.w==0.0&&render_config.texture.is_none(){
-					continue;
-				}
-				owned_mesh.instances.push(instance.clone());
-			}
-		}
-		//check every model to see if it's using the same (texture,color) but has few instances,if it is combine it into one model
-		//1. collect unique instances of texture and color,note model id
-		//2. for each model id,check if removing it from the pool decreases both the model count and instance count by more than one
-		//3. transpose all models that stay in the set
-
-		//best plan:benchmark set_bind_group,set_vertex_buffer,set_index_buffer and draw_indexed
-		//check if the estimated render performance is better by transposing multiple model instances into one model instance
-
-		//for now:just deduplicate single models...
-		let mut deduplicated_models=Vec::with_capacity(indexed_models_len);//use indexed_models_len because the list will likely get smaller instead of bigger
-		let mut unique_texture_color=HashMap::new();//texture->color->vec![(model_id,instance_id)]
-		for (model_id,model) in unique_render_config_models.iter().enumerate(){
-			//for now:filter out models with more than one instance
-			if 1<model.instances.len(){
-				continue;
-			}
-			//populate hashmap
-			let unique_color=unique_texture_color
-			.entry(model.render_config)
-			.or_insert_with(||HashMap::new());
-			//separate instances by color
-			for (instance_id,instance) in model.instances.iter().enumerate(){
-				let model_instance_list=unique_color
-				.entry(instance.color)
-				.or_insert_with(||Vec::new());
-				//add model instance to list
-				model_instance_list.push((model_id,instance_id));
-			}
-		}
-		//populate a hashset of models selected for transposition
-		//construct transposed models
-		let mut selected_model_instances=HashSet::new();
-		for (render_config,unique_color) in unique_texture_color.into_iter(){
-			for (color,model_instance_list) in unique_color.into_iter(){
-				//world transforming one model does not meet the definition of deduplicaiton
-				if 1<model_instance_list.len(){
-					//create model
-					let mut unique_pos=Vec::new();
-					let mut pos_id_from=HashMap::new();
-					let mut unique_tex=Vec::new();
-					let mut tex_id_from=HashMap::new();
-					let mut unique_normal=Vec::new();
-					let mut normal_id_from=HashMap::new();
-					let mut unique_color=Vec::new();
-					let mut color_id_from=HashMap::new();
-					let mut unique_vertices=Vec::new();
-					let mut vertex_id_from=HashMap::new();
-
-					let mut polys=Vec::new();
-					//transform instance vertices
-					for (model_id,instance_id) in model_instance_list.into_iter(){
-						//populate hashset to prevent these models from being copied
-						selected_model_instances.insert(model_id);
-						//there is only one instance per model
-						let model=&unique_render_config_models[model_id];
-						let instance=&model.instances[instance_id];
-						//just hash word slices LOL
-						let map_pos_id:Vec<PositionId>=model.unique_pos.iter().map(|untransformed_pos|{
-							let pos=instance.transform.transform_point3(glam::Vec3::from_array(untransformed_pos.clone())).to_array();
-							let h=bytemuck::cast::<[f32;3],[u32;3]>(pos);
-							PositionId::new(*pos_id_from.entry(h).or_insert_with(||{
-								let pos_id=unique_pos.len();
-								unique_pos.push(pos);
-								pos_id
-							}) as u32)
-						}).collect();
-						let map_tex_id:Vec<TextureCoordinateId>=model.unique_tex.iter().map(|&tex|{
-							let h=bytemuck::cast::<[f32;2],[u32;2]>(tex);
-							TextureCoordinateId::new(*tex_id_from.entry(h).or_insert_with(||{
-								let tex_id=unique_tex.len();
-								unique_tex.push(tex);
-								tex_id
-							}) as u32)
-						}).collect();
-						let map_normal_id:Vec<NormalId>=model.unique_normal.iter().map(|untransformed_normal|{
-							let normal=(instance.normal_transform*glam::Vec3::from_array(untransformed_normal.clone())).to_array();
-							let h=bytemuck::cast::<[f32;3],[u32;3]>(normal);
-							NormalId::new(*normal_id_from.entry(h).or_insert_with(||{
-								let normal_id=unique_normal.len();
-								unique_normal.push(normal);
-								normal_id
-							}) as u32)
-						}).collect();
-						let map_color_id:Vec<ColorId>=model.unique_color.iter().map(|&color|{
-							let h=bytemuck::cast::<[f32;4],[u32;4]>(color);
-							ColorId::new(*color_id_from.entry(h).or_insert_with(||{
-								let color_id=unique_color.len();
-								unique_color.push(color);
-								color_id
-							}) as u32)
-						}).collect();
-						//map the indexed vertices onto new indices
-						//creating the vertex map is slightly different because the vertices are directly hashable
-						let map_vertex_id:Vec<VertexId>=model.unique_vertices.iter().map(|unmapped_vertex|{
-							let vertex=model::IndexedVertex{
-								pos:map_pos_id[unmapped_vertex.pos.get() as usize],
-								tex:map_tex_id[unmapped_vertex.tex.get() as usize],
-								normal:map_normal_id[unmapped_vertex.normal.get() as usize],
-								color:map_color_id[unmapped_vertex.color.get() as usize],
-							};
-							VertexId::new(*vertex_id_from.entry(vertex.clone()).or_insert_with(||{
-								let vertex_id=unique_vertices.len();
-								unique_vertices.push(vertex);
-								vertex_id
-							}) as u32)
-						}).collect();
-						polys.extend(model.polys.polys().map(|poly|
-							poly.iter().map(|vertex_id|
-								map_vertex_id[vertex_id.get() as usize]
-							).collect()
-						));
-					}
-					//push model into dedup
-					deduplicated_models.push(IndexedGraphicsMeshOwnedRenderConfig{
-						unique_pos,
-						unique_tex,
-						unique_normal,
-						unique_color,
-						unique_vertices,
-						render_config,
-						polys:model::PolygonGroup::PolygonList(model::PolygonList::new(polys)),
-						instances:vec![GraphicsModelOwned{
-							transform:glam::Mat4::IDENTITY,
-							normal_transform:glam::Mat3::IDENTITY,
-							color
-						}],
-					});
-				}
-			}
-		}
-		//fill untouched models
-		for (model_id,model) in unique_render_config_models.into_iter().enumerate(){
-			if !selected_model_instances.contains(&model_id){
-				deduplicated_models.push(model);
-			}
-		}
-
-		//de-index models
-		let deduplicated_models_len=deduplicated_models.len();
-		let models:Vec<GraphicsMeshOwnedRenderConfig>=deduplicated_models.into_iter().map(|model|{
-			let mut vertices=Vec::new();
-			let mut index_from_vertex=HashMap::new();//::<IndexedVertex,usize>
-			//this mut be combined in a more complex way if the models use different render patterns per group
-			let mut indices=Vec::new();
-			for poly in model.polys.polys(){
-				let mut poly_vertices=poly.iter()
-				.map(|&vertex_index|*index_from_vertex.entry(vertex_index).or_insert_with(||{
-					let i=vertices.len();
-					let vertex=&model.unique_vertices[vertex_index.get() as usize];
-					vertices.push(GraphicsVertex{
-						pos:model.unique_pos[vertex.pos.get() as usize],
-						tex:model.unique_tex[vertex.tex.get() as usize],
-						normal:model.unique_normal[vertex.normal.get() as usize],
-						color:model.unique_color[vertex.color.get() as usize],
-					});
-					i
-				}));
-
-				let a=poly_vertices.next().unwrap();
-				let mut b=poly_vertices.next().unwrap();
-
-				poly_vertices.for_each(|c|{
-					indices.extend([a,b,c]);
-					b=c;
-				});
-			}
-			GraphicsMeshOwnedRenderConfig{
-				instances:model.instances,
-				indices:if (u32::MAX as usize)<vertices.len(){
-					panic!("Model has too many vertices!")
-				}else if (u16::MAX as usize)<vertices.len(){
-					model_graphics::Indices::U32(indices.into_iter().map(|vertex_idx|vertex_idx as u32).collect())
-				}else{
-					model_graphics::Indices::U16(indices.into_iter().map(|vertex_idx|vertex_idx as u16).collect())
-				},
-				vertices,
-				render_config:model.render_config,
-			}
-		}).collect();
-		//.into_iter() the modeldata vec so entities can be /moved/ to models.entities
-		let mut model_count=0;
-		let mut instance_count=0;
-		let uniform_buffer_binding_size=crate::setup::required_limits().max_uniform_buffer_binding_size as usize;
-		let chunk_size=uniform_buffer_binding_size/MODEL_BUFFER_SIZE_BYTES;
-		self.models.reserve(models.len());
-		for model in models.into_iter(){
-			instance_count+=model.instances.len();
-			for instances_chunk in model.instances.rchunks(chunk_size){
-				model_count+=1;
-				let mut model_uniforms=get_instances_buffer_data(instances_chunk);
-				//TEMP: fill with zeroes to pass validation
-				model_uniforms.resize(MODEL_BUFFER_SIZE*512,0.0f32);
-				let model_buf=device.create_buffer_init(&wgpu::util::BufferInitDescriptor{
-					label:Some(format!("Model{} Buf",model_count).as_str()),
-					contents:bytemuck::cast_slice(&model_uniforms),
-					usage:wgpu::BufferUsages::UNIFORM|wgpu::BufferUsages::COPY_DST,
-				});
-				let render_config=&map.render_configs[model.render_config.get() as usize];
-				let texture_view=render_config.texture.and_then(|texture_id|
-					texture_views.get(&texture_id)
-				).unwrap_or(&self.temp_squid_texture_view);
-				let bind_group=device.create_bind_group(&wgpu::BindGroupDescriptor{
-					layout:&self.bind_group_layouts.model,
-					entries:&[
-						wgpu::BindGroupEntry{
-							binding:0,
-							resource:model_buf.as_entire_binding(),
-						},
-						wgpu::BindGroupEntry{
-							binding:1,
-							resource:wgpu::BindingResource::TextureView(texture_view),
-						},
-						wgpu::BindGroupEntry{
-							binding:2,
-							resource:wgpu::BindingResource::Sampler(&self.samplers.repeat),
-						},
-					],
-					label:Some(format!("Model{} Bind Group",model_count).as_str()),
-				});
-				let vertex_buf=device.create_buffer_init(&wgpu::util::BufferInitDescriptor{
-					label:Some("Vertex"),
-					contents:bytemuck::cast_slice(&model.vertices),
-					usage:wgpu::BufferUsages::VERTEX,
-				});
-				//all of these are being moved here
-				self.models.push(GraphicsModel{
-					instance_count:instances_chunk.len() as u32,
-					vertex_buf,
-					indices:match &model.indices{
-						model_graphics::Indices::U32(indices)=>Indices::new(device,indices,wgpu::IndexFormat::Uint32),
-						model_graphics::Indices::U16(indices)=>Indices::new(device,indices,wgpu::IndexFormat::Uint16),
-					},
-					bind_group,
-				});
-			}
-		}
-		println!("Texture References={}",num_textures);
-		println!("Textures Loaded={}",texture_views.len());
-		println!("Indexed Models={}",indexed_models_len);
-		println!("Deduplicated Models={}",deduplicated_models_len);
-		println!("Graphics Objects:{}",self.models.len());
-		println!("Graphics Instances:{}",instance_count);
-	}
-
-	pub fn new(
-		device:&wgpu::Device,
-		queue:&wgpu::Queue,
-		config:&wgpu::SurfaceConfiguration,
-	)->Self{
-		let camera_bind_group_layout=device.create_bind_group_layout(&wgpu::BindGroupLayoutDescriptor{
-			label:None,
-			entries:&[
-				wgpu::BindGroupLayoutEntry{
-					binding:0,
-					visibility:wgpu::ShaderStages::VERTEX,
-					ty:wgpu::BindingType::Buffer{
-						ty:wgpu::BufferBindingType::Uniform,
-						has_dynamic_offset:false,
-						min_binding_size:None,
-					},
-					count:None,
-				},
-			],
-		});
-		let skybox_texture_bind_group_layout=device.create_bind_group_layout(&wgpu::BindGroupLayoutDescriptor{
-			label:Some("Skybox Texture Bind Group Layout"),
-			entries:&[
-				wgpu::BindGroupLayoutEntry{
-					binding:0,
-					visibility:wgpu::ShaderStages::FRAGMENT,
-					ty:wgpu::BindingType::Texture{
-						sample_type:wgpu::TextureSampleType::Float{filterable:true},
-						multisampled:false,
-						view_dimension:wgpu::TextureViewDimension::Cube,
-					},
-					count:None,
-				},
-				wgpu::BindGroupLayoutEntry{
-					binding:1,
-					visibility:wgpu::ShaderStages::FRAGMENT,
-					ty:wgpu::BindingType::Sampler(wgpu::SamplerBindingType::Filtering),
-					count:None,
-				},
-			],
-		});
-		let model_bind_group_layout=device.create_bind_group_layout(&wgpu::BindGroupLayoutDescriptor{
-			label:Some("Model Bind Group Layout"),
-			entries:&[
-				wgpu::BindGroupLayoutEntry{
-					binding:0,
-					visibility:wgpu::ShaderStages::VERTEX,
-					ty:wgpu::BindingType::Buffer{
-						ty:wgpu::BufferBindingType::Uniform,
-						has_dynamic_offset:false,
-						min_binding_size:None,
-					},
-					count:None,
-				},
-				wgpu::BindGroupLayoutEntry{
-					binding:1,
-					visibility:wgpu::ShaderStages::FRAGMENT,
-					ty:wgpu::BindingType::Texture{
-						sample_type:wgpu::TextureSampleType::Float{filterable:true},
-						multisampled:false,
-						view_dimension:wgpu::TextureViewDimension::D2,
-					},
-					count:None,
-				},
-				wgpu::BindGroupLayoutEntry{
-					binding:2,
-					visibility:wgpu::ShaderStages::FRAGMENT,
-					ty:wgpu::BindingType::Sampler(wgpu::SamplerBindingType::Filtering),
-					count:None,
-				},
-			],
-		});
-
-		let clamp_sampler=device.create_sampler(&wgpu::SamplerDescriptor{
-			label:Some("Clamp Sampler"),
-			address_mode_u:wgpu::AddressMode::ClampToEdge,
-			address_mode_v:wgpu::AddressMode::ClampToEdge,
-			address_mode_w:wgpu::AddressMode::ClampToEdge,
-			mag_filter:wgpu::FilterMode::Linear,
-			min_filter:wgpu::FilterMode::Linear,
-			mipmap_filter:wgpu::FilterMode::Linear,
-			..Default::default()
-		});
-		let repeat_sampler=device.create_sampler(&wgpu::SamplerDescriptor{
-			label:Some("Repeat Sampler"),
-			address_mode_u:wgpu::AddressMode::Repeat,
-			address_mode_v:wgpu::AddressMode::Repeat,
-			address_mode_w:wgpu::AddressMode::Repeat,
-			mag_filter:wgpu::FilterMode::Linear,
-			min_filter:wgpu::FilterMode::Linear,
-			mipmap_filter:wgpu::FilterMode::Linear,
-			anisotropy_clamp:16,
-			..Default::default()
-		});
-
-		// Create the render pipeline
-		let shader=device.create_shader_module(wgpu::ShaderModuleDescriptor{
-			label:None,
-			source:wgpu::ShaderSource::Wgsl(Cow::Borrowed(include_str!("shader.wgsl"))),
-		});
-
-		//load textures
-		let device_features=device.features();
-
-		let skybox_texture_view={
-			let skybox_format=if device_features.contains(wgpu::Features::TEXTURE_COMPRESSION_ASTC){
-				println!("Using ASTC");
-				wgpu::TextureFormat::Astc{
-					block:AstcBlock::B4x4,
-					channel:AstcChannel::UnormSrgb,
-				}
-			}else if device_features.contains(wgpu::Features::TEXTURE_COMPRESSION_ETC2){
-				println!("Using ETC2");
-				wgpu::TextureFormat::Etc2Rgb8UnormSrgb
-			}else if device_features.contains(wgpu::Features::TEXTURE_COMPRESSION_BC){
-				println!("Using BC");
-				wgpu::TextureFormat::Bc1RgbaUnormSrgb
-			}else{
-				println!("Using plain");
-				wgpu::TextureFormat::Bgra8UnormSrgb
-			};
-
-			let bytes=match skybox_format{
-				wgpu::TextureFormat::Astc{
-					block:AstcBlock::B4x4,
-					channel:AstcChannel::UnormSrgb,
-				}=>&include_bytes!("../images/astc.dds")[..],
-				wgpu::TextureFormat::Etc2Rgb8UnormSrgb=>&include_bytes!("../images/etc2.dds")[..],
-				wgpu::TextureFormat::Bc1RgbaUnormSrgb=>&include_bytes!("../images/bc1.dds")[..],
-				wgpu::TextureFormat::Bgra8UnormSrgb=>&include_bytes!("../images/bgra.dds")[..],
-				_=>unreachable!(),
-			};
-
-			let skybox_image=ddsfile::Dds::read(&mut std::io::Cursor::new(bytes)).unwrap();
-
-			let size=wgpu::Extent3d{
-				width:skybox_image.get_width(),
-				height:skybox_image.get_height(),
-				depth_or_array_layers:6,
-			};
-
-			let layer_size=wgpu::Extent3d{
-				depth_or_array_layers:1,
-				..size
-			};
-			let max_mips=layer_size.max_mips(wgpu::TextureDimension::D2);
-
-			let skybox_texture=device.create_texture_with_data(
-				queue,
-				&wgpu::TextureDescriptor{
-					size,
-					mip_level_count:max_mips,
-					sample_count:1,
-					dimension:wgpu::TextureDimension::D2,
-					format:skybox_format,
-					usage:wgpu::TextureUsages::TEXTURE_BINDING|wgpu::TextureUsages::COPY_DST,
-					label:Some("Skybox Texture"),
-					view_formats:&[],
-				},
-				wgpu::util::TextureDataOrder::LayerMajor,
-				&skybox_image.data,
-			);
-
-			skybox_texture.create_view(&wgpu::TextureViewDescriptor{
-				label:Some("Skybox Texture View"),
-				dimension:Some(wgpu::TextureViewDimension::Cube),
-				..wgpu::TextureViewDescriptor::default()
-			})
-		};
-
-		//squid
-		let squid_texture_view={
-			let bytes=include_bytes!("../images/squid.dds");
-
-			let image=ddsfile::Dds::read(&mut std::io::Cursor::new(bytes)).unwrap();
-
-			let size=wgpu::Extent3d{
-				width:image.get_width(),
-				height:image.get_height(),
-				depth_or_array_layers:1,
-			};
-
-			let layer_size=wgpu::Extent3d{
-				depth_or_array_layers:1,
-				..size
-			};
-			let max_mips=layer_size.max_mips(wgpu::TextureDimension::D2);
-
-			let texture=device.create_texture_with_data(
-				queue,
-				&wgpu::TextureDescriptor{
-					size,
-					mip_level_count:max_mips,
-					sample_count:1,
-					dimension:wgpu::TextureDimension::D2,
-					format:wgpu::TextureFormat::Bc7RgbaUnorm,
-					usage:wgpu::TextureUsages::TEXTURE_BINDING|wgpu::TextureUsages::COPY_DST,
-					label:Some("Squid Texture"),
-					view_formats:&[],
-				},
-				wgpu::util::TextureDataOrder::LayerMajor,
-				&image.data,
-			);
-
-			texture.create_view(&wgpu::TextureViewDescriptor{
-				label:Some("Squid Texture View"),
-				dimension:Some(wgpu::TextureViewDimension::D2),
-				..wgpu::TextureViewDescriptor::default()
-			})
-		};
-
-		let model_pipeline_layout=device.create_pipeline_layout(&wgpu::PipelineLayoutDescriptor{
-			label:None,
-			bind_group_layouts:&[
-				&camera_bind_group_layout,
-				&skybox_texture_bind_group_layout,
-				&model_bind_group_layout,
-			],
-			push_constant_ranges:&[],
-		});
-		let sky_pipeline_layout=device.create_pipeline_layout(&wgpu::PipelineLayoutDescriptor{
-			label:None,
-			bind_group_layouts:&[
-				&camera_bind_group_layout,
-				&skybox_texture_bind_group_layout,
-			],
-			push_constant_ranges:&[],
-		});
-
-		// Create the render pipelines
-		let sky_pipeline=device.create_render_pipeline(&wgpu::RenderPipelineDescriptor{
-			label:Some("Sky Pipeline"),
-			layout:Some(&sky_pipeline_layout),
-			vertex:wgpu::VertexState{
-				module:&shader,
-				entry_point:"vs_sky",
-				buffers:&[],
-				compilation_options:wgpu::PipelineCompilationOptions::default(),
-			},
-			fragment:Some(wgpu::FragmentState{
-				module:&shader,
-				entry_point:"fs_sky",
-				targets:&[Some(config.view_formats[0].into())],
-				compilation_options:wgpu::PipelineCompilationOptions::default(),
-			}),
-			primitive:wgpu::PrimitiveState{
-				front_face:wgpu::FrontFace::Cw,
-				..Default::default()
-			},
-			depth_stencil:Some(wgpu::DepthStencilState{
-				format:Self::DEPTH_FORMAT,
-				depth_write_enabled:false,
-				depth_compare:wgpu::CompareFunction::LessEqual,
-				stencil:wgpu::StencilState::default(),
-				bias:wgpu::DepthBiasState::default(),
-			}),
-			multisample:wgpu::MultisampleState::default(),
-			multiview:None,
-			cache:None,
-		});
-		let model_pipeline=device.create_render_pipeline(&wgpu::RenderPipelineDescriptor{
-			label:Some("Model Pipeline"),
-			layout:Some(&model_pipeline_layout),
-			vertex:wgpu::VertexState{
-				module:&shader,
-				entry_point:"vs_entity_texture",
-				buffers:&[wgpu::VertexBufferLayout{
-					array_stride:std::mem::size_of::<GraphicsVertex>() as wgpu::BufferAddress,
-					step_mode:wgpu::VertexStepMode::Vertex,
-					attributes:&wgpu::vertex_attr_array![0=>Float32x3,1=>Float32x2,2=>Float32x3,3=>Float32x4],
-				}],
-				compilation_options:wgpu::PipelineCompilationOptions::default(),
-			},
-			fragment:Some(wgpu::FragmentState{
-				module:&shader,
-				entry_point:"fs_entity_texture",
-				targets:&[Some(config.view_formats[0].into())],
-				compilation_options:wgpu::PipelineCompilationOptions::default(),
-			}),
-			primitive:wgpu::PrimitiveState{
-				front_face:wgpu::FrontFace::Cw,
-				cull_mode:Some(wgpu::Face::Front),
-				..Default::default()
-			},
-			depth_stencil:Some(wgpu::DepthStencilState{
-				format:Self::DEPTH_FORMAT,
-				depth_write_enabled:true,
-				depth_compare:wgpu::CompareFunction::LessEqual,
-				stencil:wgpu::StencilState::default(),
-				bias:wgpu::DepthBiasState::default(),
-			}),
-			multisample:wgpu::MultisampleState::default(),
-			multiview:None,
-			cache:None,
-		});
-
-		let camera=GraphicsCamera::default();
-		let camera_uniforms=camera.to_uniform_data(glam::Vec3::ZERO,glam::Vec2::ZERO);
-		let camera_buf=device.create_buffer_init(&wgpu::util::BufferInitDescriptor{
-			label:Some("Camera"),
-			contents:bytemuck::cast_slice(&camera_uniforms),
-			usage:wgpu::BufferUsages::UNIFORM|wgpu::BufferUsages::COPY_DST,
-		});
-		let camera_bind_group=device.create_bind_group(&wgpu::BindGroupDescriptor{
-			layout:&camera_bind_group_layout,
-			entries:&[
-				wgpu::BindGroupEntry{
-					binding:0,
-					resource:camera_buf.as_entire_binding(),
-				},
-			],
-			label:Some("Camera"),
-		});
-
-		let skybox_texture_bind_group=device.create_bind_group(&wgpu::BindGroupDescriptor{
-			layout:&skybox_texture_bind_group_layout,
-			entries:&[
-				wgpu::BindGroupEntry{
-					binding:0,
-					resource:wgpu::BindingResource::TextureView(&skybox_texture_view),
-				},
-				wgpu::BindGroupEntry{
-					binding:1,
-					resource:wgpu::BindingResource::Sampler(&clamp_sampler),
-				},
-			],
-			label:Some("Sky Texture"),
-		});
-
-		let depth_view=Self::create_depth_texture(config,device);
-
-		Self{
-			pipelines:GraphicsPipelines{
-				skybox:sky_pipeline,
-				model:model_pipeline
-			},
-			bind_groups:GraphicsBindGroups{
-				camera:camera_bind_group,
-				skybox_texture:skybox_texture_bind_group,
-			},
-			camera,
-			camera_buf,
-			models:Vec::new(),
-			depth_view,
-			staging_belt:wgpu::util::StagingBelt::new(0x100),
-			bind_group_layouts:GraphicsBindGroupLayouts{model:model_bind_group_layout},
-			samplers:GraphicsSamplers{repeat:repeat_sampler},
-			temp_squid_texture_view:squid_texture_view,
-		}
-	}
-	pub fn resize(
-		&mut self,
-		device:&wgpu::Device,
-		config:&wgpu::SurfaceConfiguration,
-		user_settings:&crate::settings::UserSettings,
-	){
-		self.depth_view=Self::create_depth_texture(config,device);
-		self.camera.screen_size=glam::uvec2(config.width,config.height);
-		self.load_user_settings(user_settings);
-	}
-	pub fn render(
-		&mut self,
-		view:&wgpu::TextureView,
-		device:&wgpu::Device,
-		queue:&wgpu::Queue,
-		frame_state:FrameState,
-	){
-		//TODO:use scheduled frame times to create beautiful smoothing simulation physics extrapolation assuming no input
-
-		let mut encoder=device.create_command_encoder(&wgpu::CommandEncoderDescriptor{label:None});
-
-		// update rotation
-		let camera_uniforms=self.camera.to_uniform_data(
-			frame_state.body.extrapolated_position(frame_state.time).map(Into::<f32>::into).to_array().into(),
-			frame_state.camera.simulate_move_angles(glam::IVec2::ZERO)
-		);
-		self.staging_belt
-			.write_buffer(
-				&mut encoder,
-				&self.camera_buf,
-				0,
-				wgpu::BufferSize::new((camera_uniforms.len() * 4) as wgpu::BufferAddress).unwrap(),
-				device,
-			)
-			.copy_from_slice(bytemuck::cast_slice(&camera_uniforms));
-		//This code only needs to run when the uniforms change
-		/*
-		for model in self.models.iter(){
-			let model_uniforms=get_instances_buffer_data(&model.instances);
-			self.staging_belt
-				.write_buffer(
-					&mut encoder,
-					&model.model_buf,//description of where data will be written when command is executed
-					0,//offset in staging belt?
-					wgpu::BufferSize::new((model_uniforms.len() * 4) as wgpu::BufferAddress).unwrap(),
-					device,
-				)
-				.copy_from_slice(bytemuck::cast_slice(&model_uniforms));
-		}
-		*/
-		self.staging_belt.finish();
-
-		{
-			let mut rpass=encoder.begin_render_pass(&wgpu::RenderPassDescriptor{
-				label:None,
-				color_attachments:&[Some(wgpu::RenderPassColorAttachment{
-					view,
-					resolve_target:None,
-					ops:wgpu::Operations{
-						load:wgpu::LoadOp::Clear(wgpu::Color{
-							r:0.1,
-							g:0.2,
-							b:0.3,
-							a:1.0,
-						}),
-						store:wgpu::StoreOp::Store,
-					},
-				})],
-				depth_stencil_attachment:Some(wgpu::RenderPassDepthStencilAttachment{
-					view:&self.depth_view,
-					depth_ops:Some(wgpu::Operations{
-						load:wgpu::LoadOp::Clear(1.0),
-						store:wgpu::StoreOp::Discard,
-					}),
-					stencil_ops:None,
-				}),
-				timestamp_writes:Default::default(),
-				occlusion_query_set:Default::default(),
-			});
-
-			rpass.set_bind_group(0,&self.bind_groups.camera,&[]);
-			rpass.set_bind_group(1,&self.bind_groups.skybox_texture,&[]);
-
-			rpass.set_pipeline(&self.pipelines.model);
-			for model in &self.models{
-				rpass.set_bind_group(2,&model.bind_group,&[]);
-				rpass.set_vertex_buffer(0,model.vertex_buf.slice(..));
-				rpass.set_index_buffer(model.indices.buf.slice(..),model.indices.format);
-				//TODO: loop over triangle strips
-				rpass.draw_indexed(0..model.indices.count,0,0..model.instance_count);
-			}
-
-			rpass.set_pipeline(&self.pipelines.skybox);
-			rpass.draw(0..3,0..1);
-		}
-
-		queue.submit(std::iter::once(encoder.finish()));
-
-		self.staging_belt.recall();
-	}
-}
-const MODEL_BUFFER_SIZE:usize=4*4 + 12 + 4;//let size=std::mem::size_of::<ModelInstance>();
-const MODEL_BUFFER_SIZE_BYTES:usize=MODEL_BUFFER_SIZE*4;
-fn get_instances_buffer_data(instances:&[GraphicsModelOwned])->Vec<f32>{
-	let mut raw=Vec::with_capacity(MODEL_BUFFER_SIZE*instances.len());
-	for mi in instances{
-		//model transform
-		raw.extend_from_slice(&AsRef::<[f32; 4*4]>::as_ref(&mi.transform)[..]);
-		//normal transform
-		raw.extend_from_slice(AsRef::<[f32; 3]>::as_ref(&mi.normal_transform.x_axis));
-		raw.extend_from_slice(&[0.0]);
-		raw.extend_from_slice(AsRef::<[f32; 3]>::as_ref(&mi.normal_transform.y_axis));
-		raw.extend_from_slice(&[0.0]);
-		raw.extend_from_slice(AsRef::<[f32; 3]>::as_ref(&mi.normal_transform.z_axis));
-		raw.extend_from_slice(&[0.0]);
-		//color
-		raw.extend_from_slice(AsRef::<[f32; 4]>::as_ref(&mi.color.get()));
-	}
-	raw
-}

src/graphics_worker.rs

@@ -1,65 +0,0 @@
-pub enum Instruction{
-	Render(crate::graphics::FrameState),
-	//UpdateModel(crate::graphics::GraphicsModelUpdate),
-	Resize(winit::dpi::PhysicalSize<u32>,crate::settings::UserSettings),
-	ChangeMap(strafesnet_common::map::CompleteMap),
-}
-
-//Ideally the graphics thread worker description is:
-/*
-WorkerDescription{
-	input:Immediate,
-	output:Realtime(PoolOrdering::Ordered(3)),
-}
-*/
-//up to three frames in flight, dropping new frame requests when all three are busy, and dropping output frames when one renders out of order
-
-pub fn new<'a>(
-	mut graphics:crate::graphics::GraphicsState,
-	mut config:wgpu::SurfaceConfiguration,
-	surface:wgpu::Surface<'a>,
-	device:wgpu::Device,
-	queue:wgpu::Queue,
-)->crate::compat_worker::INWorker<'a,Instruction>{
-	let mut resize=None;
-	crate::compat_worker::INWorker::new(move |ins:Instruction|{
-		match ins{
-			Instruction::ChangeMap(map)=>{
-				graphics.clear();
-				graphics.generate_models(&device,&queue,&map);
-			},
-			Instruction::Resize(size,user_settings)=>{
-				resize=Some((size,user_settings));
-			}
-			Instruction::Render(frame_state)=>{
-				if let Some((size,user_settings))=resize.take(){
-					println!("Resizing to {:?}",size);
-					let t0=std::time::Instant::now();
-					config.width=size.width.max(1);
-					config.height=size.height.max(1);
-					surface.configure(&device,&config);
-					graphics.resize(&device,&config,&user_settings);
-					println!("Resize took {:?}",t0.elapsed());
-				}
-				//this has to go deeper somehow
-				let frame=match surface.get_current_texture(){
-					Ok(frame)=>frame,
-					Err(_)=>{
-						surface.configure(&device,&config);
-						surface
-							.get_current_texture()
-							.expect("Failed to acquire next surface texture!")
-					}
-				};
-				let view=frame.texture.create_view(&wgpu::TextureViewDescriptor{
-					format:Some(config.view_formats[0]),
-					..wgpu::TextureViewDescriptor::default()
-				});
-
-				graphics.render(&view,&device,&queue,frame_state);
-
-				frame.present();
-			}
-		}
-	})
-}

src/instruction.rs

@@ -0,0 +1,50 @@
+use crate::integer::Time;
+
+#[derive(Debug)]
+pub struct TimedInstruction<I>{
+	pub time:Time,
+	pub instruction:I,
+}
+
+pub trait InstructionEmitter<I>{
+	fn next_instruction(&self,time_limit:Time)->Option<TimedInstruction<I>>;
+}
+pub trait InstructionConsumer<I>{
+	fn process_instruction(&mut self, instruction:TimedInstruction<I>);
+}
+
+//PROPER PRIVATE FIELDS!!!
+pub struct InstructionCollector<I>{
+	time:Time,
+	instruction:Option<I>,
+}
+impl<I> InstructionCollector<I>{
+	pub fn new(time:Time)->Self{
+		Self{
+			time,
+			instruction:None
+		}
+	}
+
+	pub fn collect(&mut self,instruction:Option<TimedInstruction<I>>){
+		match instruction{
+			Some(unwrap_instruction)=>{
+				if unwrap_instruction.time<self.time {
+					self.time=unwrap_instruction.time;
+					self.instruction=Some(unwrap_instruction.instruction);
+				}
+			},
+			None=>(),
+		}
+	}
+	pub fn instruction(self)->Option<TimedInstruction<I>>{
+		//STEAL INSTRUCTION AND DESTROY INSTRUCTIONCOLLECTOR
+		match self.instruction{
+			Some(instruction)=>Some(TimedInstruction{
+				time:self.time,
+				instruction
+			}),
+			None=>None,
+		}
+	}
+}

src/integer.rs

@@ -0,0 +1,925 @@
+//integer units
+#[derive(Clone,Copy,Hash,PartialEq,PartialOrd,Debug)]
+pub struct Time(i64);
+impl Time{
+	pub const ZERO:Self=Self(0);
+	pub const ONE_SECOND:Self=Self(1_000_000_000);
+	pub const ONE_MILLISECOND:Self=Self(1_000_000);
+	pub const ONE_MICROSECOND:Self=Self(1_000);
+	pub const ONE_NANOSECOND:Self=Self(1);
+	#[inline]
+	pub fn from_secs(num:i64)->Self{
+		Self(Self::ONE_SECOND.0*num)
+	}
+	#[inline]
+	pub fn from_millis(num:i64)->Self{
+		Self(Self::ONE_MILLISECOND.0*num)
+	}
+	#[inline]
+	pub fn from_micros(num:i64)->Self{
+		Self(Self::ONE_MICROSECOND.0*num)
+	}
+	#[inline]
+	pub fn from_nanos(num:i64)->Self{
+		Self(Self::ONE_NANOSECOND.0*num)
+	}
+	//should I have checked subtraction? force all time variables to be positive?
+	#[inline]
+	pub fn nanos(&self)->i64{
+		self.0
+	}
+}
+impl From<Planar64> for Time{
+	#[inline]
+	fn from(value:Planar64)->Self{
+		Time((((value.0 as i128)*1_000_000_000)>>32) as i64)
+	}
+}
+impl std::fmt::Display for Time{
+	#[inline]
+	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
+		write!(f,"{}s+{:09}ns",self.0/Self::ONE_SECOND.0,self.0%Self::ONE_SECOND.0)
+	}
+}
+impl std::ops::Neg for Time{
+	type Output=Time;
+	#[inline]
+	fn neg(self)->Self::Output {
+		Time(-self.0)
+	}
+}
+impl std::ops::Add<Time> for Time{
+	type Output=Time;
+	#[inline]
+	fn add(self,rhs:Self)->Self::Output {
+		Time(self.0+rhs.0)
+	}
+}
+impl std::ops::Sub<Time> for Time{
+	type Output=Time;
+	#[inline]
+	fn sub(self,rhs:Self)->Self::Output {
+		Time(self.0-rhs.0)
+	}
+}
+impl std::ops::Mul<Time> for Time{
+	type Output=Time;
+	#[inline]
+	fn mul(self,rhs:Time)->Self::Output{
+		Self((((self.0 as i128)*(rhs.0 as i128))/1_000_000_000) as i64)
+	}
+}
+impl std::ops::Div<i64> for Time{
+	type Output=Time;
+	#[inline]
+	fn div(self,rhs:i64)->Self::Output {
+		Time(self.0/rhs)
+	}
+}
+
+#[inline]
+const fn gcd(mut a:u64,mut b:u64)->u64{
+	while b!=0{
+		(a,b)=(b,a.rem_euclid(b));
+	};
+	a
+}
+#[derive(Clone,Hash)]
+pub struct Ratio64{
+	num:i64,
+	den:u64,
+}
+impl Ratio64{
+	pub const ZERO:Self=Ratio64{num:0,den:1};
+	pub const ONE:Self=Ratio64{num:1,den:1};
+	#[inline]
+	pub const fn new(num:i64,den:u64)->Option<Ratio64>{
+		if den==0{
+			None
+		}else{
+			let d=gcd(num.unsigned_abs(),den);
+			Some(Self{num:num/d as i64,den:den/d})
+		}
+	}
+	#[inline]
+	pub fn mul_int(&self,rhs:i64)->i64{
+		rhs*self.num/self.den as i64
+	}
+	#[inline]
+	pub fn rhs_div_int(&self,rhs:i64)->i64{
+		rhs*self.den as i64/self.num
+	}
+	#[inline]
+	pub fn mul_ref(&self,rhs:&Ratio64)->Ratio64{
+		let (num,den)=(self.num*rhs.num,self.den*rhs.den);
+		let d=gcd(num.unsigned_abs(),den);
+		Self{
+			num:num/d as i64,
+			den:den/d,
+		}
+	}
+}
+//from num_traits crate
+#[inline]
+fn integer_decode_f32(f: f32) -> (u64, i16, i8) {
+	let bits: u32 = f.to_bits();
+	let sign: i8 = if bits >> 31 == 0 { 1 } else { -1 };
+	let mut exponent: i16 = ((bits >> 23) & 0xff) as i16;
+	let mantissa = if exponent == 0 {
+		(bits & 0x7fffff) << 1
+	} else {
+		(bits & 0x7fffff) | 0x800000
+	};
+	// Exponent bias + mantissa shift
+	exponent -= 127 + 23;
+	(mantissa as u64, exponent, sign)
+}
+#[inline]
+fn integer_decode_f64(f: f64) -> (u64, i16, i8) {
+	let bits: u64 = f.to_bits();
+	let sign: i8 = if bits >> 63 == 0 { 1 } else { -1 };
+	let mut exponent: i16 = ((bits >> 52) & 0x7ff) as i16;
+	let mantissa = if exponent == 0 {
+		(bits & 0xfffffffffffff) << 1
+	} else {
+		(bits & 0xfffffffffffff) | 0x10000000000000
+	};
+	// Exponent bias + mantissa shift
+	exponent -= 1023 + 52;
+	(mantissa, exponent, sign)
+}
+#[derive(Debug)]
+pub enum Ratio64TryFromFloatError{
+	Nan,
+	Infinite,
+	Subnormal,
+	HighlyNegativeExponent(i16),
+	HighlyPositiveExponent(i16),
+}
+const MAX_DENOMINATOR:u128=u64::MAX as u128;
+#[inline]
+fn ratio64_from_mes((m,e,s):(u64,i16,i8))->Result<Ratio64,Ratio64TryFromFloatError>{
+	if e< -127{
+		//this can also just be zero
+		Err(Ratio64TryFromFloatError::HighlyNegativeExponent(e))
+	}else if e< -63{
+		//approximate input ratio within denominator limit
+		let mut target_num=m as u128;
+		let mut target_den=1u128<<-e;
+
+		let mut num=1;
+		let mut den=0;
+		let mut prev_num=0;
+		let mut prev_den=1;
+
+		while target_den!=0{
+			let whole=target_num/target_den;
+			(target_num,target_den)=(target_den,target_num-whole*target_den);
+			let new_num=whole*num+prev_num;
+			let new_den=whole*den+prev_den;
+			if MAX_DENOMINATOR<new_den{
+				break;
+			}else{
+				(prev_num,prev_den)=(num,den);
+				(num,den)=(new_num,new_den);
+			}
+		}
+
+		Ok(Ratio64::new(num as i64,den as u64).unwrap())
+	}else if e<0{
+		Ok(Ratio64::new((m as i64)*(s as i64),1<<-e).unwrap())
+	}else if (64-m.leading_zeros() as i16)+e<64{
+		Ok(Ratio64::new((m as i64)*(s as i64)*(1<<e),1).unwrap())
+	}else{
+		Err(Ratio64TryFromFloatError::HighlyPositiveExponent(e))
+	}
+}
+impl TryFrom<f32> for Ratio64{
+	type Error=Ratio64TryFromFloatError;
+	#[inline]
+	fn try_from(value:f32)->Result<Self,Self::Error>{
+		match value.classify(){
+			std::num::FpCategory::Nan=>Err(Self::Error::Nan),
+			std::num::FpCategory::Infinite=>Err(Self::Error::Infinite),
+			std::num::FpCategory::Zero=>Ok(Self::ZERO),
+			std::num::FpCategory::Subnormal=>Err(Self::Error::Subnormal),
+			std::num::FpCategory::Normal=>ratio64_from_mes(integer_decode_f32(value)),
+		}
+	}
+}
+impl TryFrom<f64> for Ratio64{
+	type Error=Ratio64TryFromFloatError;
+	#[inline]
+	fn try_from(value:f64)->Result<Self,Self::Error>{
+		match value.classify(){
+			std::num::FpCategory::Nan=>Err(Self::Error::Nan),
+			std::num::FpCategory::Infinite=>Err(Self::Error::Infinite),
+			std::num::FpCategory::Zero=>Ok(Self::ZERO),
+			std::num::FpCategory::Subnormal=>Err(Self::Error::Subnormal),
+			std::num::FpCategory::Normal=>ratio64_from_mes(integer_decode_f64(value)),
+		}
+	}
+}
+impl std::ops::Mul<Ratio64> for Ratio64{
+	type Output=Ratio64;
+	#[inline]
+	fn mul(self,rhs:Ratio64)->Self::Output{
+		let (num,den)=(self.num*rhs.num,self.den*rhs.den);
+		let d=gcd(num.unsigned_abs(),den);
+		Self{
+			num:num/d as i64,
+			den:den/d,
+		}
+	}
+}
+impl std::ops::Mul<i64> for Ratio64{
+	type Output=Ratio64;
+	#[inline]
+	fn mul(self,rhs:i64)->Self::Output {
+		Self{
+			num:self.num*rhs,
+			den:self.den,
+		}
+	}
+}
+impl std::ops::Div<u64> for Ratio64{
+	type Output=Ratio64;
+	#[inline]
+	fn div(self,rhs:u64)->Self::Output {
+		Self{
+			num:self.num,
+			den:self.den*rhs,
+		}
+	}
+}
+#[derive(Clone,Hash)]
+pub struct Ratio64Vec2{
+	pub x:Ratio64,
+	pub y:Ratio64,
+}
+impl Ratio64Vec2{
+	pub const ONE:Self=Self{x:Ratio64::ONE,y:Ratio64::ONE};
+	#[inline]
+	pub fn new(x:Ratio64,y:Ratio64)->Self{
+		Self{x,y}
+	}
+	#[inline]
+	pub fn mul_int(&self,rhs:glam::I64Vec2)->glam::I64Vec2{
+		glam::i64vec2(
+			self.x.mul_int(rhs.x),
+			self.y.mul_int(rhs.y),
+		)
+	}
+}
+impl std::ops::Mul<i64> for Ratio64Vec2{
+	type Output=Ratio64Vec2;
+	#[inline]
+	fn mul(self,rhs:i64)->Self::Output {
+		Self{
+			x:self.x*rhs,
+			y:self.y*rhs,
+		}
+	}
+}
+
+///[-pi,pi) = [-2^31,2^31-1]
+#[derive(Clone,Copy,Hash)]
+pub struct Angle32(i32);
+impl Angle32{
+	pub const FRAC_PI_2:Self=Self(1<<30);
+	pub const PI:Self=Self(-1<<31);
+	#[inline]
+	pub fn wrap_from_i64(theta:i64)->Self{
+		//take lower bits
+		//note: this was checked on compiler explorer and compiles to 1 instruction!
+		Self(i32::from_ne_bytes(((theta&((1<<32)-1)) as u32).to_ne_bytes()))
+	}
+	#[inline]
+	pub fn clamp_from_i64(theta:i64)->Self{
+		//the assembly is a bit confusing for this, I thought it was checking the same thing twice
+		//but it's just checking and then overwriting the value for both upper and lower bounds.
+		Self(theta.clamp(i32::MIN as i64,i32::MAX as i64) as i32)
+	}
+	#[inline]
+	pub fn get(&self)->i32{
+		self.0
+	}
+	/// Clamps the value towards the midpoint of the range.
+	/// Note that theta_min can be larger than theta_max and it will wrap clamp the other way around
+	#[inline]
+	pub fn clamp(&self,theta_min:Self,theta_max:Self)->Self{
+		//((max-min as u32)/2 as i32)+min
+		let midpoint=((
+			u32::from_ne_bytes(theta_max.0.to_ne_bytes())
+			.wrapping_sub(u32::from_ne_bytes(theta_min.0.to_ne_bytes()))
+			/2
+		) as i32)//(u32::MAX/2) as i32 ALWAYS works
+		.wrapping_add(theta_min.0);
+		//(theta-mid).clamp(max-mid,min-mid)+mid
+		Self(
+			self.0.wrapping_sub(midpoint)
+			.max(theta_min.0.wrapping_sub(midpoint))
+			.min(theta_max.0.wrapping_sub(midpoint))
+			.wrapping_add(midpoint)
+		)
+	}
+	/*
+	#[inline]
+	pub fn cos(&self)->Unit32{
+		//TODO: fix this rounding towards 0
+		Unit32(unsafe{((self.0 as f64*ANGLE32_TO_FLOAT64_RADIANS).cos()*UNIT32_ONE_FLOAT64).to_int_unchecked()})
+	}
+	#[inline]
+	pub fn sin(&self)->Unit32{
+		//TODO: fix this rounding towards 0
+		Unit32(unsafe{((self.0 as f64*ANGLE32_TO_FLOAT64_RADIANS).sin()*UNIT32_ONE_FLOAT64).to_int_unchecked()})
+	}
+	*/
+}
+const ANGLE32_TO_FLOAT64_RADIANS:f64=std::f64::consts::PI/((1i64<<31) as f64);
+impl Into<f32> for Angle32{
+	#[inline]
+	fn into(self)->f32{
+		(self.0 as f64*ANGLE32_TO_FLOAT64_RADIANS) as f32
+	}
+}
+impl std::ops::Neg for Angle32{
+	type Output=Angle32;
+	#[inline]
+	fn neg(self)->Self::Output{
+		Angle32(self.0.wrapping_neg())
+	}
+}
+impl std::ops::Add<Angle32> for Angle32{
+	type Output=Angle32;
+	#[inline]
+	fn add(self,rhs:Self)->Self::Output {
+		Angle32(self.0.wrapping_add(rhs.0))
+	}
+}
+impl std::ops::Sub<Angle32> for Angle32{
+	type Output=Angle32;
+	#[inline]
+	fn sub(self,rhs:Self)->Self::Output {
+		Angle32(self.0.wrapping_sub(rhs.0))
+	}
+}
+impl std::ops::Mul<i32> for Angle32{
+	type Output=Angle32;
+	#[inline]
+	fn mul(self,rhs:i32)->Self::Output {
+		Angle32(self.0.wrapping_mul(rhs))
+	}
+}
+impl std::ops::Mul<Angle32> for Angle32{
+	type Output=Angle32;
+	#[inline]
+	fn mul(self,rhs:Self)->Self::Output {
+		Angle32(self.0.wrapping_mul(rhs.0))
+	}
+}
+
+/* Unit type unused for now, may revive it for map files
+///[-1.0,1.0] = [-2^30,2^30]
+pub struct Unit32(i32);
+impl Unit32{
+	#[inline]
+	pub fn as_planar64(&self) -> Planar64{
+		Planar64(4*(self.0 as i64))
+	}
+}
+const UNIT32_ONE_FLOAT64=((1<<30) as f64);
+///[-1.0,1.0] = [-2^30,2^30]
+pub struct Unit32Vec3(glam::IVec3);
+impl TryFrom<[f32;3]> for Unit32Vec3{
+	type Error=Unit32TryFromFloatError;
+	fn try_from(value:[f32;3])->Result<Self,Self::Error>{
+		Ok(Self(glam::ivec3(
+			Unit32::try_from(Planar64::try_from(value[0])?)?.0,
+			Unit32::try_from(Planar64::try_from(value[1])?)?.0,
+			Unit32::try_from(Planar64::try_from(value[2])?)?.0,
+		)))
+	}
+}
+*/
+
+///[-1.0,1.0] = [-2^32,2^32]
+#[derive(Clone,Copy,Hash,Eq,Ord,PartialEq,PartialOrd)]
+pub struct Planar64(i64);
+impl Planar64{
+	pub const ZERO:Self=Self(0);
+	pub const ONE:Self=Self(1<<32);
+	#[inline]
+	pub const fn int(num:i32)->Self{
+		Self(Self::ONE.0*num as i64)
+	}
+	#[inline]
+	pub const fn raw(num:i64)->Self{
+		Self(num)
+	}
+	#[inline]
+	pub const fn get(&self)->i64{
+		self.0
+	}
+	pub fn sqrt(&self)->Self{
+		Planar64(unsafe{(((self.0 as i128)<<32) as f64).sqrt().to_int_unchecked()})
+	}
+}
+const PLANAR64_ONE_FLOAT32:f32=(1u64<<32) as f32;
+const PLANAR64_CONVERT_TO_FLOAT32:f32=1.0/PLANAR64_ONE_FLOAT32;
+const PLANAR64_ONE_FLOAT64:f64=(1u64<<32) as f64;
+impl Into<f32> for Planar64{
+	#[inline]
+	fn into(self)->f32{
+		self.0 as f32*PLANAR64_CONVERT_TO_FLOAT32
+	}
+}
+impl From<Ratio64> for Planar64{
+	#[inline]
+	fn from(ratio:Ratio64)->Self{
+		Self((((ratio.num as i128)<<32)/ratio.den as i128) as i64)
+	}
+}
+#[derive(Debug)]
+pub enum Planar64TryFromFloatError{
+	Nan,
+	Infinite,
+	Subnormal,
+	HighlyNegativeExponent(i16),
+	HighlyPositiveExponent(i16),
+}
+#[inline]
+fn planar64_from_mes((m,e,s):(u64,i16,i8))->Result<Planar64,Planar64TryFromFloatError>{
+	let e32=e+32;
+	if e32<0&&(m>>-e32)==0{//shifting m will underflow to 0
+		Ok(Planar64::ZERO)
+		// println!("m{} e{} s{}",m,e,s);
+		// println!("f={}",(m as f64)*(2.0f64.powf(e as f64))*(s as f64));
+		// Err(Planar64TryFromFloatError::HighlyNegativeExponent(e))
+	}else if (64-m.leading_zeros() as i16)+e32<64{//shifting m will not overflow
+		if e32<0{
+			Ok(Planar64((m as i64)*(s as i64)>>-e32))
+		}else{
+			Ok(Planar64((m as i64)*(s as i64)<<e32))
+		}
+	}else{//if shifting m will overflow (prev check failed)
+		Err(Planar64TryFromFloatError::HighlyPositiveExponent(e))
+	}
+}
+impl TryFrom<f32> for Planar64{
+	type Error=Planar64TryFromFloatError;
+	#[inline]
+	fn try_from(value:f32)->Result<Self,Self::Error>{
+		match value.classify(){
+			std::num::FpCategory::Nan=>Err(Self::Error::Nan),
+			std::num::FpCategory::Infinite=>Err(Self::Error::Infinite),
+			std::num::FpCategory::Zero=>Ok(Self::ZERO),
+			std::num::FpCategory::Subnormal=>Err(Self::Error::Subnormal),
+			std::num::FpCategory::Normal=>planar64_from_mes(integer_decode_f32(value)),
+		}
+	}
+}
+impl TryFrom<f64> for Planar64{
+	type Error=Planar64TryFromFloatError;
+	#[inline]
+	fn try_from(value:f64)->Result<Self,Self::Error>{
+		match value.classify(){
+			std::num::FpCategory::Nan=>Err(Self::Error::Nan),
+			std::num::FpCategory::Infinite=>Err(Self::Error::Infinite),
+			std::num::FpCategory::Zero=>Ok(Self::ZERO),
+			std::num::FpCategory::Subnormal=>Err(Self::Error::Subnormal),
+			std::num::FpCategory::Normal=>planar64_from_mes(integer_decode_f64(value)),
+		}
+	}
+}
+impl std::fmt::Display for Planar64{
+	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
+		write!(f,"{:.3}",
+			Into::<f32>::into(*self),
+		)
+	}
+}
+impl std::ops::Neg for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn neg(self)->Self::Output{
+		Planar64(-self.0)
+	}
+}
+impl std::ops::Add<Planar64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn add(self, rhs: Self) -> Self::Output {
+		Planar64(self.0+rhs.0)
+	}
+}
+impl std::ops::Sub<Planar64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn sub(self, rhs: Self) -> Self::Output {
+		Planar64(self.0-rhs.0)
+	}
+}
+impl std::ops::Mul<i64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn mul(self, rhs: i64) -> Self::Output {
+		Planar64(self.0*rhs)
+	}
+}
+impl std::ops::Mul<Planar64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn mul(self, rhs: Self) -> Self::Output {
+		Planar64((((self.0 as i128)*(rhs.0 as i128))>>32) as i64)
+	}
+}
+impl std::ops::Div<i64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn div(self, rhs: i64) -> Self::Output {
+		Planar64(self.0/rhs)
+	}
+}
+impl std::ops::Div<Planar64> for Planar64{
+	type Output=Planar64;
+	#[inline]
+	fn div(self, rhs: Planar64) -> Self::Output {
+		Planar64((((self.0 as i128)<<32)/rhs.0 as i128) as i64)
+	}
+}
+// impl PartialOrd<i64> for Planar64{
+// 	fn partial_cmp(&self, other: &i64) -> Option<std::cmp::Ordering> {
+// 		self.0.partial_cmp(other)
+// 	}
+// }
+
+
+///[-1.0,1.0] = [-2^32,2^32]
+#[derive(Clone,Copy,Default,Hash,Eq,PartialEq)]
+pub struct Planar64Vec3(glam::I64Vec3);
+impl Planar64Vec3{
+	pub const ZERO:Self=Planar64Vec3(glam::I64Vec3::ZERO);
+	pub const ONE:Self=Self::int(1,1,1);
+	pub const X:Self=Self::int(1,0,0);
+	pub const Y:Self=Self::int(0,1,0);
+	pub const Z:Self=Self::int(0,0,1);
+	pub const NEG_X:Self=Self::int(-1,0,0);
+	pub const NEG_Y:Self=Self::int(0,-1,0);
+	pub const NEG_Z:Self=Self::int(0,0,-1);
+	pub const MIN:Self=Planar64Vec3(glam::I64Vec3::MIN);
+	pub const MAX:Self=Planar64Vec3(glam::I64Vec3::MAX);
+	#[inline]
+	pub const fn int(x:i32,y:i32,z:i32)->Self{
+		Self(glam::i64vec3((x as i64)<<32,(y as i64)<<32,(z as i64)<<32))
+	}
+	#[inline]
+	pub fn x(&self)->Planar64{
+		Planar64(self.0.x)
+	}
+	#[inline]
+	pub fn y(&self)->Planar64{
+		Planar64(self.0.y)
+	}
+	#[inline]
+	pub fn z(&self)->Planar64{
+		Planar64(self.0.z)
+	}
+	#[inline]
+	pub fn min(&self,rhs:Self)->Self{
+		Self(glam::i64vec3(
+			self.0.x.min(rhs.0.x),
+			self.0.y.min(rhs.0.y),
+			self.0.z.min(rhs.0.z),
+		))
+	}
+	#[inline]
+	pub fn max(&self,rhs:Self)->Self{
+		Self(glam::i64vec3(
+			self.0.x.max(rhs.0.x),
+			self.0.y.max(rhs.0.y),
+			self.0.z.max(rhs.0.z),
+		))
+	}
+	#[inline]
+	pub fn midpoint(&self,rhs:Self)->Self{
+		Self((self.0+rhs.0)/2)
+	}
+	#[inline]
+	pub fn cmplt(&self,rhs:Self)->glam::BVec3{
+		self.0.cmplt(rhs.0)
+	}
+	#[inline]
+	pub fn dot(&self,rhs:Self)->Planar64{
+		Planar64(((
+			(self.0.x as i128)*(rhs.0.x as i128)+
+			(self.0.y as i128)*(rhs.0.y as i128)+
+			(self.0.z as i128)*(rhs.0.z as i128)
+		)>>32) as i64)
+	}
+	#[inline]
+	pub fn length(&self)->Planar64{
+		let radicand=(self.0.x as i128)*(self.0.x as i128)+(self.0.y as i128)*(self.0.y as i128)+(self.0.z as i128)*(self.0.z as i128);
+		Planar64(unsafe{(radicand as f64).sqrt().to_int_unchecked()})
+	}
+	#[inline]
+	pub fn with_length(&self,length:Planar64)->Self{
+		let radicand=(self.0.x as i128)*(self.0.x as i128)+(self.0.y as i128)*(self.0.y as i128)+(self.0.z as i128)*(self.0.z as i128);
+		let self_length:i128=unsafe{(radicand as f64).sqrt().to_int_unchecked()};
+		//self.0*length/self_length
+		Planar64Vec3(
+			glam::i64vec3(
+				((self.0.x as i128)*(length.0 as i128)/self_length) as i64,
+				((self.0.y as i128)*(length.0 as i128)/self_length) as i64,
+				((self.0.z as i128)*(length.0 as i128)/self_length) as i64,
+			)
+		)
+	}
+}
+impl Into<glam::Vec3> for Planar64Vec3{
+	#[inline]
+	fn into(self)->glam::Vec3{
+		glam::vec3(
+			self.0.x as f32,
+			self.0.y as f32,
+			self.0.z as f32,
+		)*PLANAR64_CONVERT_TO_FLOAT32
+	}
+}
+impl TryFrom<[f32;3]> for Planar64Vec3{
+	type Error=Planar64TryFromFloatError;
+	#[inline]
+	fn try_from(value:[f32;3])->Result<Self,Self::Error>{
+		Ok(Self(glam::i64vec3(
+			Planar64::try_from(value[0])?.0,
+			Planar64::try_from(value[1])?.0,
+			Planar64::try_from(value[2])?.0,
+		)))
+	}
+}
+impl TryFrom<glam::Vec3A> for Planar64Vec3{
+	type Error=Planar64TryFromFloatError;
+	#[inline]
+	fn try_from(value:glam::Vec3A)->Result<Self,Self::Error>{
+		Ok(Self(glam::i64vec3(
+			Planar64::try_from(value.x)?.0,
+			Planar64::try_from(value.y)?.0,
+			Planar64::try_from(value.z)?.0,
+		)))
+	}
+}
+impl std::fmt::Display for Planar64Vec3{
+	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
+		write!(f,"{:.3},{:.3},{:.3}",
+			Into::<f32>::into(self.x()),Into::<f32>::into(self.y()),Into::<f32>::into(self.z()),
+		)
+	}
+}
+impl std::ops::Neg for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn neg(self)->Self::Output{
+		Planar64Vec3(-self.0)
+	}
+}
+impl std::ops::Add<Planar64Vec3> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn add(self,rhs:Planar64Vec3) -> Self::Output {
+		Planar64Vec3(self.0+rhs.0)
+	}
+}
+impl std::ops::AddAssign<Planar64Vec3> for Planar64Vec3{
+	#[inline]
+	fn add_assign(&mut self,rhs:Planar64Vec3){
+		*self=*self+rhs
+	}
+}
+impl std::ops::Sub<Planar64Vec3> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn sub(self,rhs:Planar64Vec3) -> Self::Output {
+		Planar64Vec3(self.0-rhs.0)
+	}
+}
+impl std::ops::SubAssign<Planar64Vec3> for Planar64Vec3{
+	#[inline]
+	fn sub_assign(&mut self,rhs:Planar64Vec3){
+		*self=*self-rhs
+	}
+}
+impl std::ops::Mul<Planar64Vec3> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn mul(self, rhs: Planar64Vec3) -> Self::Output {
+		Planar64Vec3(glam::i64vec3(
+			(((self.0.x as i128)*(rhs.0.x as i128))>>32) as i64,
+			(((self.0.y as i128)*(rhs.0.y as i128))>>32) as i64,
+			(((self.0.z as i128)*(rhs.0.z as i128))>>32) as i64
+		))
+	}
+}
+impl std::ops::Mul<Planar64> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn mul(self, rhs: Planar64) -> Self::Output {
+		Planar64Vec3(glam::i64vec3(
+			(((self.0.x as i128)*(rhs.0 as i128))>>32) as i64,
+			(((self.0.y as i128)*(rhs.0 as i128))>>32) as i64,
+			(((self.0.z as i128)*(rhs.0 as i128))>>32) as i64
+		))
+	}
+}
+impl std::ops::Mul<i64> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn mul(self,rhs:i64)->Self::Output {
+		Planar64Vec3(glam::i64vec3(
+			self.0.x*rhs,
+			self.0.y*rhs,
+			self.0.z*rhs
+		))
+	}
+}
+impl std::ops::Mul<Time> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn mul(self,rhs:Time)->Self::Output{
+		Planar64Vec3(glam::i64vec3(
+			(((self.0.x as i128)*(rhs.0 as i128))/1_000_000_000) as i64,
+			(((self.0.y as i128)*(rhs.0 as i128))/1_000_000_000) as i64,
+			(((self.0.z as i128)*(rhs.0 as i128))/1_000_000_000) as i64
+		))
+	}
+}
+impl std::ops::Div<i64> for Planar64Vec3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn div(self,rhs:i64)->Self::Output{
+		Planar64Vec3(glam::i64vec3(
+			self.0.x/rhs,
+			self.0.y/rhs,
+			self.0.z/rhs,
+		))
+	}
+}
+
+///[-1.0,1.0] = [-2^32,2^32]
+#[derive(Clone,Copy)]
+pub struct Planar64Mat3{
+	x_axis:Planar64Vec3,
+	y_axis:Planar64Vec3,
+	z_axis:Planar64Vec3,
+}
+impl Default for Planar64Mat3{
+	#[inline]
+	fn default() -> Self {
+		Self{
+			x_axis:Planar64Vec3::X,
+			y_axis:Planar64Vec3::Y,
+			z_axis:Planar64Vec3::Z,
+		}
+	}
+}
+impl std::ops::Mul<Planar64Vec3> for Planar64Mat3{
+	type Output=Planar64Vec3;
+	#[inline]
+	fn mul(self,rhs:Planar64Vec3) -> Self::Output {
+		self.x_axis*rhs.x()
+		+self.y_axis*rhs.y()
+		+self.z_axis*rhs.z()
+	}
+}
+
+impl Planar64Mat3{
+	#[inline]
+	pub fn from_cols(x_axis:Planar64Vec3,y_axis:Planar64Vec3,z_axis:Planar64Vec3)->Self{
+		Self{
+			x_axis,
+			y_axis,
+			z_axis,
+		}
+	}
+	pub const fn int_from_cols_array(array:[i32;9])->Self{
+		Self{
+			x_axis:Planar64Vec3::int(array[0],array[1],array[2]),
+			y_axis:Planar64Vec3::int(array[3],array[4],array[5]),
+			z_axis:Planar64Vec3::int(array[6],array[7],array[8]),
+		}
+	}
+	#[inline]
+	pub fn from_rotation_y(angle:Angle32)->Self{
+		let theta=angle.0 as f64*ANGLE32_TO_FLOAT64_RADIANS;
+		let (s,c)=theta.sin_cos();
+		let (c,s)=(c*PLANAR64_ONE_FLOAT64,s*PLANAR64_ONE_FLOAT64);
+		//TODO: fix this rounding towards 0
+		let (c,s):(i64,i64)=(unsafe{c.to_int_unchecked()},unsafe{s.to_int_unchecked()});
+		Self::from_cols(
+			Planar64Vec3(glam::i64vec3(c,0,-s)),
+			Planar64Vec3::Y,
+			Planar64Vec3(glam::i64vec3(s,0,c)),
+		)
+	}
+}
+impl Into<glam::Mat3> for Planar64Mat3{
+	#[inline]
+	fn into(self)->glam::Mat3{
+		glam::Mat3::from_cols(
+			self.x_axis.into(),
+			self.y_axis.into(),
+			self.z_axis.into(),
+		)
+	}
+}
+impl TryFrom<glam::Mat3A> for Planar64Mat3{
+	type Error=Planar64TryFromFloatError;
+	#[inline]
+	fn try_from(value:glam::Mat3A)->Result<Self,Self::Error>{
+		Ok(Self{
+			x_axis:Planar64Vec3::try_from(value.x_axis)?,
+			y_axis:Planar64Vec3::try_from(value.y_axis)?,
+			z_axis:Planar64Vec3::try_from(value.z_axis)?,
+		})
+	}
+}
+impl std::fmt::Display for Planar64Mat3{
+	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
+		write!(f,"\n{:.3},{:.3},{:.3}\n{:.3},{:.3},{:.3}\n{:.3},{:.3},{:.3}",
+			Into::<f32>::into(self.x_axis.x()),Into::<f32>::into(self.x_axis.y()),Into::<f32>::into(self.x_axis.z()),
+			Into::<f32>::into(self.y_axis.x()),Into::<f32>::into(self.y_axis.y()),Into::<f32>::into(self.y_axis.z()),
+			Into::<f32>::into(self.z_axis.x()),Into::<f32>::into(self.z_axis.y()),Into::<f32>::into(self.z_axis.z()),
+		)
+	}
+}
+impl std::ops::Div<i64> for Planar64Mat3{
+	type Output=Planar64Mat3;
+	#[inline]
+	fn div(self,rhs:i64)->Self::Output{
+		Planar64Mat3{
+			x_axis:self.x_axis/rhs,
+			y_axis:self.y_axis/rhs,
+			z_axis:self.z_axis/rhs,
+		}
+	}
+}
+
+///[-1.0,1.0] = [-2^32,2^32]
+#[derive(Clone,Copy,Default)]
+pub struct Planar64Affine3{
+	pub matrix3:Planar64Mat3,//includes scale above 1
+	pub translation:Planar64Vec3,
+}
+
+impl Planar64Affine3{
+	#[inline]
+	pub fn new(matrix3:Planar64Mat3,translation:Planar64Vec3)->Self{
+		Self{matrix3,translation}
+	}
+	#[inline]
+	pub fn transform_point3(&self,point:Planar64Vec3) -> Planar64Vec3{
+		Planar64Vec3(
+			self.translation.0
+			+(self.matrix3.x_axis*point.x()).0
+			+(self.matrix3.y_axis*point.y()).0
+			+(self.matrix3.z_axis*point.z()).0
+		)
+	}
+}
+impl Into<glam::Mat4> for Planar64Affine3{
+	#[inline]
+	fn into(self)->glam::Mat4{
+		glam::Mat4::from_cols_array(&[
+			self.matrix3.x_axis.0.x as f32,self.matrix3.x_axis.0.y as f32,self.matrix3.x_axis.0.z as f32,0.0,
+			self.matrix3.y_axis.0.x as f32,self.matrix3.y_axis.0.y as f32,self.matrix3.y_axis.0.z as f32,0.0,
+			self.matrix3.z_axis.0.x as f32,self.matrix3.z_axis.0.y as f32,self.matrix3.z_axis.0.z as f32,0.0,
+			self.translation.0.x as f32,self.translation.0.y as f32,self.translation.0.z as f32,PLANAR64_ONE_FLOAT32
+		])*PLANAR64_CONVERT_TO_FLOAT32
+	}
+}
+impl TryFrom<glam::Affine3A> for Planar64Affine3{
+	type Error=Planar64TryFromFloatError;
+	fn try_from(value: glam::Affine3A)->Result<Self, Self::Error> {
+		Ok(Self{
+			matrix3:Planar64Mat3::try_from(value.matrix3)?,
+			translation:Planar64Vec3::try_from(value.translation)?
+		})
+	}
+}
+impl std::fmt::Display for Planar64Affine3{
+	fn fmt(&self,f:&mut std::fmt::Formatter<'_>)->std::fmt::Result{
+		write!(f,"translation: {:.3},{:.3},{:.3}\nmatrix3:\n{:.3},{:.3},{:.3}\n{:.3},{:.3},{:.3}\n{:.3},{:.3},{:.3}",
+			Into::<f32>::into(self.translation.x()),Into::<f32>::into(self.translation.y()),Into::<f32>::into(self.translation.z()),
+			Into::<f32>::into(self.matrix3.x_axis.x()),Into::<f32>::into(self.matrix3.x_axis.y()),Into::<f32>::into(self.matrix3.x_axis.z()),
+			Into::<f32>::into(self.matrix3.y_axis.x()),Into::<f32>::into(self.matrix3.y_axis.y()),Into::<f32>::into(self.matrix3.y_axis.z()),
+			Into::<f32>::into(self.matrix3.z_axis.x()),Into::<f32>::into(self.matrix3.z_axis.y()),Into::<f32>::into(self.matrix3.z_axis.z()),
+		)
+	}
+}
+
+#[test]
+fn test_sqrt(){
+	let r=Planar64::int(400);
+	println!("r{}",r.get());
+	let s=r.sqrt();
+	println!("s{}",s.get());
+}

src/load_roblox.rs

@@ -0,0 +1,484 @@
+use crate::primitives;
+use crate::integer::{Planar64,Planar64Vec3,Planar64Mat3,Planar64Affine3};
+
+fn class_is_a(class: &str, superclass: &str) -> bool {
+	if class==superclass {
+		return true
+	}
+	let class_descriptor=rbx_reflection_database::get().classes.get(class);
+	if let Some(descriptor) = &class_descriptor {
+		if let Some(class_super) = &descriptor.superclass {
+			return class_is_a(&class_super, superclass)
+		}
+	}
+	return false
+}
+fn recursive_collect_superclass(objects: &mut std::vec::Vec<rbx_dom_weak::types::Ref>,dom: &rbx_dom_weak::WeakDom, instance: &rbx_dom_weak::Instance, superclass: &str){
+	for &referent in instance.children() {
+		if let Some(c) = dom.get_by_ref(referent) {
+			if class_is_a(c.class.as_str(), superclass) {
+				objects.push(c.referent());//copy ref
+			}
+			recursive_collect_superclass(objects,dom,c,superclass);
+		}
+	}
+}
+fn get_texture_refs(dom:&rbx_dom_weak::WeakDom) -> Vec<rbx_dom_weak::types::Ref>{
+	let mut objects = std::vec::Vec::new();
+	recursive_collect_superclass(&mut objects, dom, dom.root(),"Decal");
+	//get ids
+	//clear vec
+	//next class
+	objects
+}
+fn planar64_affine3_from_roblox(cf:&rbx_dom_weak::types::CFrame,size:&rbx_dom_weak::types::Vector3)->Planar64Affine3{
+	Planar64Affine3::new(
+		Planar64Mat3::from_cols(
+			Planar64Vec3::try_from([cf.orientation.x.x,cf.orientation.y.x,cf.orientation.z.x]).unwrap()
+			*Planar64::try_from(size.x/2.0).unwrap(),
+			Planar64Vec3::try_from([cf.orientation.x.y,cf.orientation.y.y,cf.orientation.z.y]).unwrap()
+			*Planar64::try_from(size.y/2.0).unwrap(),
+			Planar64Vec3::try_from([cf.orientation.x.z,cf.orientation.y.z,cf.orientation.z.z]).unwrap()
+			*Planar64::try_from(size.z/2.0).unwrap(),
+		),
+		Planar64Vec3::try_from([cf.position.x,cf.position.y,cf.position.z]).unwrap()
+	)
+}
+fn get_attributes(name:&str,can_collide:bool,velocity:Planar64Vec3,force_intersecting:bool)->crate::model::CollisionAttributes{
+	let mut general=crate::model::GameMechanicAttributes::default();
+	let mut intersecting=crate::model::IntersectingAttributes::default();
+	let mut contacting=crate::model::ContactingAttributes::default();
+	let mut force_can_collide=can_collide;
+	match name{
+		"Water"=>intersecting.water=Some(crate::model::IntersectingWater{density:Planar64::ONE,viscosity:Planar64::ONE/10,current:velocity}),
+		"Accelerator"=>{force_can_collide=false;intersecting.accelerator=Some(crate::model::IntersectingAccelerator{acceleration:velocity})},
+		"MapFinish"=>{force_can_collide=false;general.zone=Some(crate::model::GameMechanicZone{mode_id:0,behaviour:crate::model::ZoneBehaviour::Finish})},
+		"MapAnticheat"=>{force_can_collide=false;general.zone=Some(crate::model::GameMechanicZone{mode_id:0,behaviour:crate::model::ZoneBehaviour::Anitcheat})},
+		"Platform"=>general.teleport_behaviour=Some(crate::model::TeleportBehaviour::StageElement(crate::model::GameMechanicStageElement{
+			mode_id:0,
+			stage_id:0,
+			force:false,
+			behaviour:crate::model::StageElementBehaviour::Platform,
+		})),
+		other=>{
+			if let Some(captures)=lazy_regex::regex!(r"^(Force)?(Spawn|SpawnAt|Trigger|Teleport|Platform)(\d+)$")
+			.captures(other){
+				general.teleport_behaviour=Some(crate::model::TeleportBehaviour::StageElement(crate::model::GameMechanicStageElement{
+					mode_id:0,
+					stage_id:captures[3].parse::<u32>().unwrap(),
+					force:match captures.get(1){
+						Some(m)=>m.as_str()=="Force",
+						None=>false,
+					},
+					behaviour:match &captures[2]{
+						"Spawn"|"SpawnAt"=>crate::model::StageElementBehaviour::SpawnAt,
+						"Trigger"=>{force_can_collide=false;crate::model::StageElementBehaviour::Trigger},
+						"Teleport"=>{force_can_collide=false;crate::model::StageElementBehaviour::Teleport},
+						"Platform"=>crate::model::StageElementBehaviour::Platform,
+						_=>panic!("regex1[2] messed up bad"),
+					}
+				}));
+			}else if let Some(captures)=lazy_regex::regex!(r"^Bonus(Finish|Anticheat)(\d+)$")
+			.captures(other){
+				force_can_collide=false;
+				match &captures[1]{
+					"Finish"=>general.zone=Some(crate::model::GameMechanicZone{mode_id:captures[2].parse::<u32>().unwrap(),behaviour:crate::model::ZoneBehaviour::Finish}),
+					"Anticheat"=>general.zone=Some(crate::model::GameMechanicZone{mode_id:captures[2].parse::<u32>().unwrap(),behaviour:crate::model::ZoneBehaviour::Anitcheat}),
+					_=>panic!("regex2[1] messed up bad"),
+				}
+			}
+		}
+	}
+	//need some way to skip this
+	if velocity!=Planar64Vec3::ZERO{
+		general.booster=Some(crate::model::GameMechanicBooster{velocity});
+	}
+	match force_can_collide{
+		true=>{
+			match name{
+				"Bounce"=>contacting.elasticity=Some(u32::MAX),
+				"Surf"=>contacting.surf=Some(crate::model::ContactingSurf{}),
+				"Ladder"=>contacting.ladder=Some(crate::model::ContactingLadder{sticky:true}),
+				other=>{
+					if let Some(captures)=lazy_regex::regex!(r"^(Jump|WormholeIn)(\d+)$")
+					.captures(other){
+						match &captures[1]{
+							"Jump"=>general.jump_limit=Some(crate::model::GameMechanicJumpLimit{count:captures[2].parse::<u32>().unwrap()}),
+							"WormholeIn"=>general.teleport_behaviour=Some(crate::model::TeleportBehaviour::Wormhole(crate::model::GameMechanicWormhole{destination_model_id:captures[2].parse::<u32>().unwrap()})),
+							_=>panic!("regex3[1] messed up bad"),
+						}
+					}
+				}
+			}
+			crate::model::CollisionAttributes::Contact{contacting,general}
+		},
+		false=>if force_intersecting
+		||general.jump_limit.is_some()
+		||general.booster.is_some()
+		||general.zone.is_some()
+		||general.teleport_behaviour.is_some()
+		||intersecting.water.is_some()
+		||intersecting.accelerator.is_some()
+		{
+			crate::model::CollisionAttributes::Intersect{intersecting,general}
+		}else{
+			crate::model::CollisionAttributes::Decoration
+		},
+	}
+}
+
+struct RobloxAssetId(u64);
+struct RobloxAssetIdParseErr;
+impl std::str::FromStr for RobloxAssetId {
+	type Err=RobloxAssetIdParseErr;
+	fn from_str(s: &str) -> Result<Self, Self::Err>{
+		let regman=lazy_regex::regex!(r"(\d+)$");
+		if let Some(captures) = regman.captures(s) {
+			if captures.len()==2{//captures[0] is all captures concatenated, and then each individual capture
+				if let Ok(id) = captures[0].parse::<u64>() {
+					return Ok(Self(id));
+				}
+			}
+		}
+		Err(RobloxAssetIdParseErr)
+	}
+}
+#[derive(Clone,Copy,PartialEq)]
+struct RobloxTextureTransform{
+	offset_u:f32,
+	offset_v:f32,
+	scale_u:f32,
+	scale_v:f32,
+}
+impl std::cmp::Eq for RobloxTextureTransform{}//????
+impl std::default::Default for RobloxTextureTransform{
+    fn default() -> Self {
+        Self{offset_u:0.0,offset_v:0.0,scale_u:1.0,scale_v:1.0}
+    }
+}
+impl std::hash::Hash for RobloxTextureTransform {
+	fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+		self.offset_u.to_ne_bytes().hash(state);
+		self.offset_v.to_ne_bytes().hash(state);
+		self.scale_u.to_ne_bytes().hash(state);
+		self.scale_v.to_ne_bytes().hash(state);
+	}
+}
+#[derive(Clone,PartialEq)]
+struct RobloxFaceTextureDescription{
+	texture:u32,
+	color:glam::Vec4,
+	transform:RobloxTextureTransform,
+}
+impl std::cmp::Eq for RobloxFaceTextureDescription{}//????
+impl std::hash::Hash for RobloxFaceTextureDescription {
+	fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+		self.texture.hash(state);
+		self.transform.hash(state);
+        for &el in self.color.as_ref().iter() {
+            el.to_ne_bytes().hash(state);
+        }
+    }
+}
+impl RobloxFaceTextureDescription{
+	fn to_face_description(&self)->primitives::FaceDescription{
+		primitives::FaceDescription{
+			texture:Some(self.texture),
+			transform:glam::Affine2::from_translation(
+				glam::vec2(self.transform.offset_u,self.transform.offset_v)
+			)
+			*glam::Affine2::from_scale(
+				glam::vec2(self.transform.scale_u,self.transform.scale_v)
+			),
+			color:self.color,
+		}
+	}
+}
+type RobloxPartDescription=[Option<RobloxFaceTextureDescription>;6];
+type RobloxWedgeDescription=[Option<RobloxFaceTextureDescription>;5];
+type RobloxCornerWedgeDescription=[Option<RobloxFaceTextureDescription>;5];
+#[derive(Clone,Eq,Hash,PartialEq)]
+enum RobloxBasePartDescription{
+	Sphere,
+	Part(RobloxPartDescription),
+	Cylinder,
+	Wedge(RobloxWedgeDescription),
+	CornerWedge(RobloxCornerWedgeDescription),
+}
+pub fn generate_indexed_models(dom:rbx_dom_weak::WeakDom) -> crate::model::IndexedModelInstances{
+	//IndexedModelInstances includes textures
+	let mut spawn_point=Planar64Vec3::ZERO;
+
+	let mut indexed_models=Vec::new();
+	let mut model_id_from_description=std::collections::HashMap::<RobloxBasePartDescription,usize>::new();
+
+	let mut texture_id_from_asset_id=std::collections::HashMap::<u64,u32>::new();
+	let mut asset_id_from_texture_id=Vec::new();
+
+	let mut object_refs=Vec::new();
+	let mut temp_objects=Vec::new();
+	recursive_collect_superclass(&mut object_refs, &dom, dom.root(),"BasePart");
+	for object_ref in object_refs {
+		if let Some(object)=dom.get_by_ref(object_ref){
+			if let (
+					Some(rbx_dom_weak::types::Variant::CFrame(cf)),
+					Some(rbx_dom_weak::types::Variant::Vector3(size)),
+					Some(rbx_dom_weak::types::Variant::Vector3(velocity)),
+					Some(rbx_dom_weak::types::Variant::Float32(transparency)),
+					Some(rbx_dom_weak::types::Variant::Color3uint8(color3)),
+					Some(rbx_dom_weak::types::Variant::Bool(can_collide)),
+				) = (
+					object.properties.get("CFrame"),
+					object.properties.get("Size"),
+					object.properties.get("Velocity"),
+					object.properties.get("Transparency"),
+					object.properties.get("Color"),
+					object.properties.get("CanCollide"),
+				)
+			{
+				let model_transform=planar64_affine3_from_roblox(cf,size);
+
+				//push TempIndexedAttributes
+				let mut force_intersecting=false;
+				let mut temp_indexing_attributes=Vec::new();
+				if let Some(attr)=match &object.name[..]{
+					"MapStart"=>{
+						spawn_point=model_transform.transform_point3(Planar64Vec3::ZERO)+Planar64Vec3::Y*5/2;
+						Some(crate::model::TempIndexedAttributes::Start{mode_id:0})
+					},
+					"UnorderedCheckpoint"=>Some(crate::model::TempIndexedAttributes::UnorderedCheckpoint{mode_id:0}),
+					other=>{
+						let regman=lazy_regex::regex!(r"^(BonusStart|Spawn|ForceSpawn|OrderedCheckpoint)(\d+)$");
+						if let Some(captures) = regman.captures(other) {
+							match &captures[1]{
+								"BonusStart"=>Some(crate::model::TempIndexedAttributes::Start{mode_id:captures[2].parse::<u32>().unwrap()}),
+								"Spawn"|"ForceSpawn"=>Some(crate::model::TempIndexedAttributes::Spawn{mode_id:0,stage_id:captures[2].parse::<u32>().unwrap()}),
+								"OrderedCheckpoint"=>Some(crate::model::TempIndexedAttributes::OrderedCheckpoint{mode_id:0,checkpoint_id:captures[2].parse::<u32>().unwrap()}),
+								_=>None,
+							}
+						}else{
+							None
+						}
+					}
+				}{
+					force_intersecting=true;
+					temp_indexing_attributes.push(attr);
+				}
+
+				//TODO: also detect "CylinderMesh" etc here
+				let shape=match &object.class[..]{
+					"Part"=>{
+						if let Some(rbx_dom_weak::types::Variant::Enum(shape))=object.properties.get("Shape"){
+							match shape.to_u32(){
+								0=>primitives::Primitives::Sphere,
+								1=>primitives::Primitives::Cube,
+								2=>primitives::Primitives::Cylinder,
+								3=>primitives::Primitives::Wedge,
+								4=>primitives::Primitives::CornerWedge,
+								_=>panic!("Funky roblox PartType={};",shape.to_u32()),
+							}
+						}else{
+							panic!("Part has no Shape!");
+						}
+					},
+					"WedgePart"=>primitives::Primitives::Wedge,
+					"CornerWedgePart"=>primitives::Primitives::CornerWedge,
+					_=>{
+						println!("Unsupported BasePart ClassName={}; defaulting to cube",object.class);
+						primitives::Primitives::Cube
+					}
+				};
+
+				//use the biggest one and cut it down later...
+				let mut part_texture_description:RobloxPartDescription=[None,None,None,None,None,None];
+				temp_objects.clear();
+				recursive_collect_superclass(&mut temp_objects, &dom, object,"Decal");
+				for &decal_ref in &temp_objects{
+					if let Some(decal)=dom.get_by_ref(decal_ref){
+						if let (
+							Some(rbx_dom_weak::types::Variant::Content(content)),
+							Some(rbx_dom_weak::types::Variant::Enum(normalid)),
+							Some(rbx_dom_weak::types::Variant::Color3(decal_color3)),
+							Some(rbx_dom_weak::types::Variant::Float32(decal_transparency)),
+						) = (
+							decal.properties.get("Texture"),
+							decal.properties.get("Face"),
+							decal.properties.get("Color3"),
+							decal.properties.get("Transparency"),
+						) {
+							if let Ok(asset_id)=content.clone().into_string().parse::<RobloxAssetId>(){
+								let texture_id=if let Some(&texture_id)=texture_id_from_asset_id.get(&asset_id.0){
+									texture_id
+								}else{
+									let texture_id=asset_id_from_texture_id.len() as u32;
+									texture_id_from_asset_id.insert(asset_id.0,texture_id);
+									asset_id_from_texture_id.push(asset_id.0);
+									texture_id
+								};
+								let normal_id=normalid.to_u32();
+								if normal_id<6{
+									let (roblox_texture_color,roblox_texture_transform)=if decal.class=="Texture"{
+										//generate tranform
+										if let (
+												Some(rbx_dom_weak::types::Variant::Float32(ox)),
+												Some(rbx_dom_weak::types::Variant::Float32(oy)),
+												Some(rbx_dom_weak::types::Variant::Float32(sx)),
+												Some(rbx_dom_weak::types::Variant::Float32(sy)),
+											) = (
+												decal.properties.get("OffsetStudsU"),
+												decal.properties.get("OffsetStudsV"),
+												decal.properties.get("StudsPerTileU"),
+												decal.properties.get("StudsPerTileV"),
+											)
+										{
+											let (size_u,size_v)=match normal_id{
+												0=>(size.z,size.y),//right
+												1=>(size.x,size.z),//top
+												2=>(size.x,size.y),//back
+												3=>(size.z,size.y),//left
+												4=>(size.x,size.z),//bottom
+												5=>(size.x,size.y),//front
+												_=>panic!("unreachable"),
+											};
+											(
+												glam::vec4(decal_color3.r,decal_color3.g,decal_color3.b,1.0-*decal_transparency),
+												RobloxTextureTransform{
+													offset_u:*ox/(*sx),offset_v:*oy/(*sy),
+													scale_u:size_u/(*sx),scale_v:size_v/(*sy),
+												}
+											)
+										}else{
+											(glam::Vec4::ONE,RobloxTextureTransform::default())
+										}
+									}else{
+										(glam::Vec4::ONE,RobloxTextureTransform::default())
+									};
+									part_texture_description[normal_id as usize]=Some(RobloxFaceTextureDescription{
+										texture:texture_id,
+										color:roblox_texture_color,
+										transform:roblox_texture_transform,
+									});
+								}else{
+									println!("NormalId={} unsupported for shape={:?}",normal_id,shape);
+								}
+							}
+						}
+					}
+				}
+				//obscure rust syntax "slice pattern"
+				let [
+					f0,//Cube::Right
+					f1,//Cube::Top
+					f2,//Cube::Back
+					f3,//Cube::Left
+					f4,//Cube::Bottom
+					f5,//Cube::Front
+				]=part_texture_description;
+				let basepart_texture_description=match shape{
+					primitives::Primitives::Sphere=>RobloxBasePartDescription::Sphere,
+					primitives::Primitives::Cube=>RobloxBasePartDescription::Part([f0,f1,f2,f3,f4,f5]),
+					primitives::Primitives::Cylinder=>RobloxBasePartDescription::Cylinder,
+					//use front face texture first and use top face texture as a fallback
+					primitives::Primitives::Wedge=>RobloxBasePartDescription::Wedge([
+						f0,//Cube::Right->Wedge::Right
+						if f5.is_some(){f5}else{f1},//Cube::Front|Cube::Top->Wedge::TopFront
+						f2,//Cube::Back->Wedge::Back
+						f3,//Cube::Left->Wedge::Left
+						f4,//Cube::Bottom->Wedge::Bottom
+					]),
+					//TODO: fix Left+Back texture coordinates to match roblox when not overwridden by Top
+					primitives::Primitives::CornerWedge=>RobloxBasePartDescription::CornerWedge([
+						f0,//Cube::Right->CornerWedge::Right
+						if f2.is_some(){f2}else{f1.clone()},//Cube::Back|Cube::Top->CornerWedge::TopBack
+						if f3.is_some(){f3}else{f1},//Cube::Left|Cube::Top->CornerWedge::TopLeft
+						f4,//Cube::Bottom->CornerWedge::Bottom
+						f5,//Cube::Front->CornerWedge::Front
+					]),
+				};
+				//make new model if unit cube has not been created before
+				let model_id=if let Some(&model_id)=model_id_from_description.get(&basepart_texture_description){
+					//push to existing texture model
+					model_id
+				}else{
+					let model_id=indexed_models.len();
+					model_id_from_description.insert(basepart_texture_description.clone(),model_id);//borrow checker going crazy
+					indexed_models.push(match basepart_texture_description{
+						RobloxBasePartDescription::Sphere=>primitives::unit_sphere(),
+						RobloxBasePartDescription::Part(part_texture_description)=>{
+							let mut cube_face_description=primitives::CubeFaceDescription::new();
+							for (face_id,roblox_face_description) in part_texture_description.iter().enumerate(){
+								cube_face_description.insert(
+								match face_id{
+									0=>primitives::CubeFace::Right,
+									1=>primitives::CubeFace::Top,
+									2=>primitives::CubeFace::Back,
+									3=>primitives::CubeFace::Left,
+									4=>primitives::CubeFace::Bottom,
+									5=>primitives::CubeFace::Front,
+									_=>panic!("unreachable"),
+								},
+								match roblox_face_description{
+									Some(roblox_texture_transform)=>roblox_texture_transform.to_face_description(),
+									None=>primitives::FaceDescription::default(),
+								});
+							}
+							primitives::generate_partial_unit_cube(cube_face_description)
+						},
+						RobloxBasePartDescription::Cylinder=>primitives::unit_cylinder(),
+						RobloxBasePartDescription::Wedge(wedge_texture_description)=>{
+							let mut wedge_face_description=primitives::WedgeFaceDescription::new();
+							for (face_id,roblox_face_description) in wedge_texture_description.iter().enumerate(){
+								wedge_face_description.insert(
+								match face_id{
+									0=>primitives::WedgeFace::Right,
+									1=>primitives::WedgeFace::TopFront,
+									2=>primitives::WedgeFace::Back,
+									3=>primitives::WedgeFace::Left,
+									4=>primitives::WedgeFace::Bottom,
+									_=>panic!("unreachable"),
+								},
+								match roblox_face_description{
+									Some(roblox_texture_transform)=>roblox_texture_transform.to_face_description(),
+									None=>primitives::FaceDescription::default(),
+								});
+							}
+							primitives::generate_partial_unit_wedge(wedge_face_description)
+						},
+						RobloxBasePartDescription::CornerWedge(cornerwedge_texture_description)=>{
+							let mut cornerwedge_face_description=primitives::CornerWedgeFaceDescription::new();
+							for (face_id,roblox_face_description) in cornerwedge_texture_description.iter().enumerate(){
+								cornerwedge_face_description.insert(
+								match face_id{
+									0=>primitives::CornerWedgeFace::Right,
+									1=>primitives::CornerWedgeFace::TopBack,
+									2=>primitives::CornerWedgeFace::TopLeft,
+									3=>primitives::CornerWedgeFace::Bottom,
+									4=>primitives::CornerWedgeFace::Front,
+									_=>panic!("unreachable"),
+								},
+								match roblox_face_description{
+									Some(roblox_texture_transform)=>roblox_texture_transform.to_face_description(),
+									None=>primitives::FaceDescription::default(),
+								});
+							}
+							primitives::generate_partial_unit_cornerwedge(cornerwedge_face_description)
+						},
+					});
+					model_id
+				};
+				indexed_models[model_id].instances.push(crate::model::ModelInstance {
+					transform:model_transform,
+					color:glam::vec4(color3.r as f32/255f32, color3.g as f32/255f32, color3.b as f32/255f32, 1.0-*transparency),
+					attributes:get_attributes(&object.name,*can_collide,Planar64Vec3::try_from([velocity.x,velocity.y,velocity.z]).unwrap(),force_intersecting),
+					temp_indexing:temp_indexing_attributes,
+				});
+			}
+		}
+	}
+	crate::model::IndexedModelInstances{
+		textures:asset_id_from_texture_id.iter().map(|t|t.to_string()).collect(),
+		models:indexed_models,
+		spawn_point,
+		modes:Vec::new(),
+	}
+}

src/model.rs

@@ -0,0 +1,246 @@
+use crate::integer::{Planar64,Planar64Vec3,Planar64Affine3};
+pub type TextureCoordinate=glam::Vec2;
+pub type Color4=glam::Vec4;
+#[derive(Clone,Hash,PartialEq,Eq)]
+pub struct IndexedVertex{
+	pub pos:u32,
+	pub tex:u32,
+	pub normal:u32,
+	pub color:u32,
+}
+pub struct IndexedPolygon{
+	pub vertices:Vec<u32>,
+}
+pub struct IndexedGroup{
+	pub texture:Option<u32>,//RenderPattern? material/texture/shader/flat color
+	pub polys:Vec<IndexedPolygon>,
+}
+pub struct IndexedModel{
+	pub unique_pos:Vec<Planar64Vec3>,
+	pub unique_normal:Vec<Planar64Vec3>,
+	pub unique_tex:Vec<TextureCoordinate>,
+	pub unique_color:Vec<Color4>,
+	pub unique_vertices:Vec<IndexedVertex>,
+	pub groups: Vec<IndexedGroup>,
+	pub instances:Vec<ModelInstance>,
+}
+pub struct ModelInstance{
+	//pub id:u64,//this does not actually help with map fixes resimulating bots, they must always be resimulated
+	pub transform:Planar64Affine3,
+	pub color:Color4,//transparency is in here
+	pub attributes:CollisionAttributes,
+	pub temp_indexing:Vec<TempIndexedAttributes>,
+}
+impl std::default::Default for ModelInstance{
+	fn default() -> Self {
+		Self{
+			color:Color4::ONE,
+			transform:Default::default(),
+			attributes:Default::default(),
+			temp_indexing:Default::default(),
+		}
+	}
+}
+pub struct IndexedModelInstances{
+	pub textures:Vec<String>,//RenderPattern
+	pub models:Vec<IndexedModel>,
+	//may make this into an object later.
+	pub modes:Vec<ModeDescription>,
+	pub spawn_point:Planar64Vec3,
+}
+//stage description referencing flattened ids is spooky, but the map loading is meant to be deterministic.
+pub struct ModeDescription{
+	pub start:u32,//start=model_id
+	pub spawns:Vec<u32>,//spawns[spawn_id]=model_id
+	pub ordered_checkpoints:Vec<u32>,//ordered_checkpoints[checkpoint_id]=model_id
+	pub unordered_checkpoints:Vec<u32>,//unordered_checkpoints[checkpoint_id]=model_id
+	pub spawn_from_stage_id:std::collections::HashMap::<u32,usize>,
+	pub ordered_checkpoint_from_checkpoint_id:std::collections::HashMap::<u32,usize>,
+}
+impl ModeDescription{
+	pub fn get_spawn_model_id(&self,stage_id:u32)->Option<&u32>{
+		if let Some(&spawn)=self.spawn_from_stage_id.get(&stage_id){
+			self.spawns.get(spawn)
+		}else{
+			None
+		}
+	}
+	pub fn get_ordered_checkpoint_model_id(&self,checkpoint_id:u32)->Option<&u32>{
+		if let Some(&checkpoint)=self.ordered_checkpoint_from_checkpoint_id.get(&checkpoint_id){
+			self.ordered_checkpoints.get(checkpoint)
+		}else{
+			None
+		}
+	}
+}
+pub enum TempIndexedAttributes{
+	Start{
+		mode_id:u32,
+	},
+	Spawn{
+		mode_id:u32,
+		stage_id:u32,
+	},
+	OrderedCheckpoint{
+		mode_id:u32,
+		checkpoint_id:u32,
+	},
+	UnorderedCheckpoint{
+		mode_id:u32,
+	},
+}
+
+//you have this effect while in contact
+#[derive(Clone)]
+pub struct ContactingSurf{}
+#[derive(Clone)]
+pub struct ContactingLadder{
+	pub sticky:bool
+}
+//you have this effect while intersecting
+#[derive(Clone)]
+pub struct IntersectingWater{
+	pub viscosity:Planar64,
+	pub density:Planar64,
+	pub current:Planar64Vec3,
+}
+#[derive(Clone)]
+pub struct IntersectingAccelerator{
+	pub acceleration:Planar64Vec3
+}
+//All models can be given these attributes
+#[derive(Clone)]
+pub struct GameMechanicJumpLimit{
+	pub count:u32,
+}
+#[derive(Clone)]
+pub struct GameMechanicBooster{
+	pub velocity:Planar64Vec3,
+}
+#[derive(Clone)]
+pub enum ZoneBehaviour{
+	//Start is indexed
+	//Checkpoints are indexed
+	Finish,
+	Anitcheat,
+}
+#[derive(Clone)]
+pub struct GameMechanicZone{
+	pub mode_id:u32,
+	pub behaviour:ZoneBehaviour,
+}
+// enum TrapCondition{
+// 	FasterThan(i64),
+// 	SlowerThan(i64),
+// 	InRange(i64,i64),
+// 	OutsideRange(i64,i64),
+// }
+#[derive(Clone)]
+pub enum StageElementBehaviour{
+ 	//Spawn,//The behaviour of stepping on a spawn setting the spawnid
+ 	SpawnAt,
+ 	Trigger,
+ 	Teleport,
+ 	Platform,
+ 	//Speedtrap(TrapCondition),//Acts as a trigger with a speed condition
+}
+#[derive(Clone)]
+pub struct GameMechanicStageElement{
+	pub mode_id:u32,
+	pub stage_id:u32,//which spawn to send to
+	pub force:bool,//allow setting to lower spawn id i.e. 7->3
+	pub behaviour:StageElementBehaviour
+}
+#[derive(Clone)]
+pub struct GameMechanicWormhole{
+	//destination does not need to be another wormhole
+	//this defines a one way portal to a destination model transform
+	//two of these can create a two way wormhole
+	pub destination_model_id:u32,
+	//(position,angles)*=origin.transform.inverse()*destination.transform
+}
+#[derive(Clone)]
+pub enum TeleportBehaviour{
+	StageElement(GameMechanicStageElement),
+	Wormhole(GameMechanicWormhole),
+}
+#[derive(Default,Clone)]
+pub struct GameMechanicAttributes{
+	pub jump_limit:Option<GameMechanicJumpLimit>,
+	pub booster:Option<GameMechanicBooster>,
+	pub zone:Option<GameMechanicZone>,
+	pub teleport_behaviour:Option<TeleportBehaviour>,
+}
+#[derive(Default,Clone)]
+pub struct ContactingAttributes{
+	pub elasticity:Option<u32>,//[1/2^32,1] 0=None (elasticity+1)/2^32
+	//friction?
+	pub surf:Option<ContactingSurf>,
+	pub ladder:Option<ContactingLadder>,
+}
+#[derive(Default,Clone)]
+pub struct IntersectingAttributes{
+	pub water:Option<IntersectingWater>,
+	pub accelerator:Option<IntersectingAccelerator>,
+}
+//Spawn(u32) NO! spawns are indexed in the map header instead of marked with attibutes
+pub enum CollisionAttributes{
+	Decoration,//visual only
+	Contact{//track whether you are contacting the object
+		contacting:ContactingAttributes,
+		general:GameMechanicAttributes,
+	},
+	Intersect{//track whether you are intersecting the object
+		intersecting:IntersectingAttributes,
+		general:GameMechanicAttributes,
+	},
+}
+impl std::default::Default for CollisionAttributes{
+	fn default() -> Self {
+		Self::Contact{
+			contacting:ContactingAttributes::default(),
+			general:GameMechanicAttributes::default()
+		}
+	}
+}
+
+pub fn generate_indexed_model_list_from_obj(data:obj::ObjData,color:Color4)->Vec<IndexedModel>{
+	let mut unique_vertex_index = std::collections::HashMap::<obj::IndexTuple,u32>::new();
+	return data.objects.iter().map(|object|{
+		unique_vertex_index.clear();
+		let mut unique_vertices = Vec::new();
+		let groups = object.groups.iter().map(|group|{
+			IndexedGroup{
+				texture:None,
+				polys:group.polys.iter().map(|poly|{
+					IndexedPolygon{
+						vertices:poly.0.iter().map(|&tup|{
+							if let Some(&i)=unique_vertex_index.get(&tup){
+								i
+							}else{
+								let i=unique_vertices.len() as u32;
+								unique_vertices.push(IndexedVertex{
+									pos: tup.0 as u32,
+									tex: tup.1.unwrap() as u32,
+									normal: tup.2.unwrap() as u32,
+									color: 0,
+								});
+								unique_vertex_index.insert(tup,i);
+								i
+							}
+						}).collect()
+					}
+				}).collect()
+			}
+		}).collect();
+		IndexedModel{
+			unique_pos: data.position.iter().map(|&v|Planar64Vec3::try_from(v).unwrap()).collect(),
+			unique_tex: data.texture.iter().map(|&v|TextureCoordinate::from_array(v)).collect(),
+			unique_normal: data.normal.iter().map(|&v|Planar64Vec3::try_from(v).unwrap()).collect(),
+			unique_color: vec![color],
+			unique_vertices,
+			groups,
+			instances:Vec::new(),
+		}
+	}).collect()
+}

src/model_graphics.rs

@@ -1,48 +1,55 @@
-use bytemuck::{Pod,Zeroable};
-use strafesnet_common::model::{IndexedVertex,PolygonGroup,RenderConfigId};
-#[derive(Clone,Copy,Pod,Zeroable)]
+use bytemuck::{Pod, Zeroable};
+use crate::model::{IndexedVertex,IndexedPolygon};
+#[derive(Clone, Copy, Pod, Zeroable)]
 #[repr(C)]
-pub struct GraphicsVertex{
-	pub pos:[f32;3],
-	pub tex:[f32;2],
-	pub normal:[f32;3],
-	pub color:[f32;4],
+pub struct GraphicsVertex {
+	pub pos: [f32; 3],
+	pub tex: [f32; 2],
+	pub normal: [f32; 3],
+	pub color: [f32; 4],
 }
-#[derive(Clone,Copy,id::Id)]
-pub struct IndexedGraphicsMeshOwnedRenderConfigId(u32);
-pub struct IndexedGraphicsMeshOwnedRenderConfig{
-	pub unique_pos:Vec<[f32;3]>,
-	pub unique_tex:Vec<[f32;2]>,
-	pub unique_normal:Vec<[f32;3]>,
-	pub unique_color:Vec<[f32;4]>,
+pub struct IndexedGroupFixedTexture{
+	pub polys:Vec<IndexedPolygon>,
+}
+pub struct IndexedModelGraphicsSingleTexture{
+	pub unique_pos:Vec<[f32; 3]>,
+	pub unique_tex:Vec<[f32; 2]>,
+	pub unique_normal:Vec<[f32; 3]>,
+	pub unique_color:Vec<[f32; 4]>,
 	pub unique_vertices:Vec<IndexedVertex>,
-	pub render_config:RenderConfigId,
-	pub polys:PolygonGroup,
-	pub instances:Vec<GraphicsModelOwned>,
+	pub texture:Option<u32>,//RenderPattern? material/texture/shader/flat color
+	pub groups: Vec<IndexedGroupFixedTexture>,
+	pub instances:Vec<ModelGraphicsInstance>,
 }
-pub enum Indices{
-	U32(Vec<u32>),
-	U16(Vec<u16>),
+pub struct ModelGraphicsSingleTexture{
+	pub instances: Vec<ModelGraphicsInstance>,
+	pub vertices: Vec<GraphicsVertex>,
+	pub entities: Vec<Vec<u16>>,
+	pub texture: Option<u32>,
 }
-pub struct GraphicsMeshOwnedRenderConfig{
-	pub vertices:Vec<GraphicsVertex>,
-	pub indices:Indices,
-	pub render_config:RenderConfigId,
-	pub instances:Vec<GraphicsModelOwned>,
+#[derive(Clone,PartialEq)]
+pub struct ModelGraphicsColor4(glam::Vec4);
+impl ModelGraphicsColor4{
+	pub const fn get(&self)->glam::Vec4{
+		self.0
+	}
 }
-#[derive(Clone,Copy,PartialEq,id::Id)]
-pub struct GraphicsModelColor4(glam::Vec4);
-impl std::hash::Hash for GraphicsModelColor4{
-	fn hash<H:std::hash::Hasher>(&self,state:&mut H) {
+impl From<glam::Vec4> for ModelGraphicsColor4{
+	fn from(value:glam::Vec4)->Self{
+		Self(value)
+	}
+}
+impl std::hash::Hash for ModelGraphicsColor4{
+	fn hash<H: std::hash::Hasher>(&self,state:&mut H) {
 		for &f in self.0.as_ref(){
-			bytemuck::cast::<f32,u32>(f).hash(state);
+			u32::from_ne_bytes(f.to_ne_bytes()).hash(state);
 		}
 	}
 }
-impl Eq for GraphicsModelColor4{}
+impl Eq for ModelGraphicsColor4{}
 #[derive(Clone)]
-pub struct GraphicsModelOwned{
+pub struct ModelGraphicsInstance{
 	pub transform:glam::Mat4,
 	pub normal_transform:glam::Mat3,
-	pub color:GraphicsModelColor4,
-}
+	pub color:ModelGraphicsColor4,
+}

src/physics_worker.rs

@@ -1,242 +0,0 @@
-use strafesnet_common::mouse::MouseState;
-use strafesnet_common::physics::Instruction as PhysicsInputInstruction;
-use strafesnet_common::integer::Time;
-use strafesnet_common::instruction::TimedInstruction;
-use strafesnet_common::timer::{Scaled,Timer,TimerState};
-use mouse_interpolator::MouseInterpolator;
-
-#[derive(Debug)]
-pub enum InputInstruction{
-	MoveMouse(glam::IVec2),
-	MoveRight(bool),
-	MoveUp(bool),
-	MoveBack(bool),
-	MoveLeft(bool),
-	MoveDown(bool),
-	MoveForward(bool),
-	Jump(bool),
-	Zoom(bool),
-	ResetAndRestart,
-	ResetAndSpawn(strafesnet_common::gameplay_modes::ModeId,strafesnet_common::gameplay_modes::StageId),
-	PracticeFly,
-}
-pub enum Instruction{
-	Input(InputInstruction),
-	Render,
-	Resize(winit::dpi::PhysicalSize<u32>),
-	ChangeMap(strafesnet_common::map::CompleteMap),
-	//SetPaused is not an InputInstruction: the physics doesn't know that it's paused.
-	SetPaused(bool),
-	//Graphics(crate::graphics_worker::Instruction),
-}
-mod mouse_interpolator{
-	use super::*;
-	//TODO: move this or tab
-pub struct MouseInterpolator{
-	//"PlayerController"
-	user_settings:crate::settings::UserSettings,
-	//"MouseInterpolator"
-	timeline:std::collections::VecDeque<TimedInstruction<PhysicsInputInstruction>>,
-	last_mouse_time:Time,//this value is pre-transformed to simulation time
-	mouse_blocking:bool,
-	//"Simulation"
-	timer:Timer<Scaled>,
-	physics:crate::physics::PhysicsContext,
-
-}
-impl MouseInterpolator{
-	pub fn new(
-		physics:crate::physics::PhysicsContext,
-		user_settings:crate::settings::UserSettings,
-	)->MouseInterpolator{
-		MouseInterpolator{
-			mouse_blocking:true,
-			last_mouse_time:physics.get_next_mouse().time,
-			timeline:std::collections::VecDeque::new(),
-			timer:Timer::from_state(Scaled::identity(),false),
-			physics,
-			user_settings,
-		}
-	}
-	fn push_mouse_instruction(&mut self,ins:&TimedInstruction<Instruction>,m:glam::IVec2){
-		if self.mouse_blocking{
-			//tell the game state which is living in the past about its future
-			self.timeline.push_front(TimedInstruction{
-				time:self.last_mouse_time,
-				instruction:PhysicsInputInstruction::SetNextMouse(MouseState{time:self.timer.time(ins.time),pos:m}),
-			});
-		}else{
-			//mouse has just started moving again after being still for longer than 10ms.
-			//replace the entire mouse interpolation state to avoid an intermediate state with identical m0.t m1.t timestamps which will divide by zero
-			self.timeline.push_front(TimedInstruction{
-				time:self.last_mouse_time,
-				instruction:PhysicsInputInstruction::ReplaceMouse(
-					MouseState{time:self.last_mouse_time,pos:self.physics.get_next_mouse().pos},
-					MouseState{time:self.timer.time(ins.time),pos:m}
-				),
-			});
-			//delay physics execution until we have an interpolation target
-			self.mouse_blocking=true;
-		}
-		self.last_mouse_time=self.timer.time(ins.time);
-	}
-	fn push(&mut self,time:Time,phys_input:PhysicsInputInstruction){
-		//This is always a non-mouse event
-		self.timeline.push_back(TimedInstruction{
-			time:self.timer.time(time),
-			instruction:phys_input,
-		});
-	}
-	/// returns should_empty_queue
-	/// may or may not mutate internal state XD!
-	fn map_instruction(&mut self,ins:&TimedInstruction<Instruction>)->bool{
-		let mut update_mouse_blocking=true;
-		match &ins.instruction{
-			Instruction::Input(input_instruction)=>match input_instruction{
-				&InputInstruction::MoveMouse(m)=>{
-					if !self.timer.is_paused(){
-						self.push_mouse_instruction(ins,m);
-					}
-					update_mouse_blocking=false;
-				},
-				&InputInstruction::MoveForward(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveForward(s)),
-				&InputInstruction::MoveLeft(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveLeft(s)),
-				&InputInstruction::MoveBack(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveBack(s)),
-				&InputInstruction::MoveRight(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveRight(s)),
-				&InputInstruction::MoveUp(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveUp(s)),
-				&InputInstruction::MoveDown(s)=>self.push(ins.time,PhysicsInputInstruction::SetMoveDown(s)),
-				&InputInstruction::Jump(s)=>self.push(ins.time,PhysicsInputInstruction::SetJump(s)),
-				&InputInstruction::Zoom(s)=>self.push(ins.time,PhysicsInputInstruction::SetZoom(s)),
-				&InputInstruction::ResetAndSpawn(mode_id,stage_id)=>{
-					self.push(ins.time,PhysicsInputInstruction::Reset);
-					self.push(ins.time,PhysicsInputInstruction::SetSensitivity(self.user_settings.calculate_sensitivity()));
-					self.push(ins.time,PhysicsInputInstruction::Spawn(mode_id,stage_id));
-				},
-				InputInstruction::ResetAndRestart=>{
-					self.push(ins.time,PhysicsInputInstruction::Reset);
-					self.push(ins.time,PhysicsInputInstruction::SetSensitivity(self.user_settings.calculate_sensitivity()));
-					self.push(ins.time,PhysicsInputInstruction::Restart);
-				},
-				InputInstruction::PracticeFly=>self.push(ins.time,PhysicsInputInstruction::PracticeFly),
-			},
-			//do these really need to idle the physics?
-			//sending None dumps the instruction queue
-			Instruction::ChangeMap(_)=>self.push(ins.time,PhysicsInputInstruction::Idle),
-			Instruction::Resize(_)=>self.push(ins.time,PhysicsInputInstruction::Idle),
-			Instruction::Render=>self.push(ins.time,PhysicsInputInstruction::Idle),
-			&Instruction::SetPaused(paused)=>{
-				if let Err(e)=self.timer.set_paused(ins.time,paused){
-					println!("Cannot pause: {e}");
-				}
-				self.push(ins.time,PhysicsInputInstruction::Idle);
-			},
-		}
-		if update_mouse_blocking{
-			//this returns the bool for us
-			self.update_mouse_blocking(ins.time)
-		}else{
-			//do flush that queue
-			true
-		}
-	}
-	/// must check if self.mouse_blocking==true before calling!
-	fn unblock_mouse(&mut self,time:Time){
-		//push an event to extrapolate no movement from
-		self.timeline.push_front(TimedInstruction{
-			time:self.last_mouse_time,
-			instruction:PhysicsInputInstruction::SetNextMouse(MouseState{time:self.timer.time(time),pos:self.physics.get_next_mouse().pos}),
-		});
-		self.last_mouse_time=self.timer.time(time);
-		//stop blocking. the mouse is not moving so the physics does not need to live in the past and wait for interpolation targets.
-		self.mouse_blocking=false;
-	}
-	fn update_mouse_blocking(&mut self,time:Time)->bool{
-		if self.mouse_blocking{
-			//assume the mouse has stopped moving after 10ms.
-			//shitty mice are 125Hz which is 8ms so this should cover that.
-			//setting this to 100us still doesn't print even though it's 10x lower than the polling rate,
-			//so mouse events are probably not handled separately from drawing and fire right before it :(
-			if Time::from_millis(10)<self.timer.time(time)-self.physics.get_next_mouse().time{
-				self.unblock_mouse(time);
-				true
-			}else{
-				false
-			}
-		}else{
-			//keep this up to date so that it can be used as a known-timestamp
-			//that the mouse was not moving when the mouse starts moving again
-			self.last_mouse_time=self.timer.time(time);
-			true
-		}
-	}
-	fn empty_queue(&mut self){
-		while let Some(instruction)=self.timeline.pop_front(){
-			self.physics.run_input_instruction(instruction);
-		}
-	}
-	pub fn handle_instruction(&mut self,ins:&TimedInstruction<Instruction>){
-		let should_empty_queue=self.map_instruction(ins);
-		if should_empty_queue{
-			self.empty_queue();
-		}
-	}
-	pub fn get_frame_state(&self,time:Time)->crate::graphics::FrameState{
-		crate::graphics::FrameState{
-			body:self.physics.camera_body(),
-			camera:self.physics.camera(),
-			time:self.timer.time(time),
-		}
-	}
-	pub fn change_map(&mut self,time:Time,map:&strafesnet_common::map::CompleteMap){
-		//dump any pending interpolation state
-		if self.mouse_blocking{
-			self.unblock_mouse(time);
-		}
-		self.empty_queue();
-
-		//doing it like this to avoid doing PhysicsInstruction::ChangeMap(Rc<CompleteMap>)
-		self.physics.generate_models(&map);
-
-		//use the standard input interface so the instructions are written out to bots
-		self.handle_instruction(&TimedInstruction{
-			time:self.timer.time(time),
-			instruction:Instruction::Input(InputInstruction::ResetAndSpawn(
-				strafesnet_common::gameplay_modes::ModeId::MAIN,
-				strafesnet_common::gameplay_modes::StageId::FIRST,
-			)),
-		});
-	}
-	pub const fn user_settings(&self)->&crate::settings::UserSettings{
-		&self.user_settings
-	}
-}
-}
-
-pub fn new<'a>(
-	mut graphics_worker:crate::compat_worker::INWorker<'a,crate::graphics_worker::Instruction>,
-	user_settings:crate::settings::UserSettings,
-)->crate::compat_worker::QNWorker<'a,TimedInstruction<Instruction>>{
-	let physics=crate::physics::PhysicsContext::default();
-	let mut interpolator=MouseInterpolator::new(
-		physics,
-		user_settings
-	);
-	crate::compat_worker::QNWorker::new(move |ins:TimedInstruction<Instruction>|{
-		interpolator.handle_instruction(&ins);
-		match ins.instruction{
-			Instruction::Render=>{
-				let frame_state=interpolator.get_frame_state(ins.time);
-				graphics_worker.send(crate::graphics_worker::Instruction::Render(frame_state)).unwrap();
-			},
-			Instruction::Resize(size)=>{
-				graphics_worker.send(crate::graphics_worker::Instruction::Resize(size,interpolator.user_settings().clone())).unwrap();
-			},
-			Instruction::ChangeMap(map)=>{
-				interpolator.change_map(ins.time,&map);
-				graphics_worker.send(crate::graphics_worker::Instruction::ChangeMap(map)).unwrap();
-			},
-			Instruction::Input(_)=>(),
-			Instruction::SetPaused(_)=>(),
-		}
-	})
-}

src/primitives.rs

@@ -0,0 +1,510 @@
+use crate::model::{Color4,TextureCoordinate,IndexedModel,IndexedPolygon,IndexedGroup,IndexedVertex};
+use crate::integer::Planar64Vec3;
+
+#[derive(Debug)]
+pub enum Primitives{
+	Sphere,
+	Cube,
+	Cylinder,
+	Wedge,
+	CornerWedge,
+}
+#[derive(Hash,PartialEq,Eq)]
+pub enum CubeFace{
+	Right,
+	Top,
+	Back,
+	Left,
+	Bottom,
+	Front,
+}
+const CUBE_DEFAULT_TEXTURE_COORDS:[TextureCoordinate;4]=[
+	TextureCoordinate::new(0.0,0.0),
+	TextureCoordinate::new(1.0,0.0),
+	TextureCoordinate::new(1.0,1.0),
+	TextureCoordinate::new(0.0,1.0),
+];
+const CUBE_DEFAULT_VERTICES:[Planar64Vec3;8]=[
+	Planar64Vec3::int(-1,-1, 1),//0 left bottom back
+	Planar64Vec3::int( 1,-1, 1),//1 right bottom back
+	Planar64Vec3::int( 1, 1, 1),//2 right top back
+	Planar64Vec3::int(-1, 1, 1),//3 left top back
+	Planar64Vec3::int(-1, 1,-1),//4 left top front
+	Planar64Vec3::int( 1, 1,-1),//5 right top front
+	Planar64Vec3::int( 1,-1,-1),//6 right bottom front
+	Planar64Vec3::int(-1,-1,-1),//7 left bottom front
+];
+const CUBE_DEFAULT_NORMALS:[Planar64Vec3;6]=[
+	Planar64Vec3::int( 1, 0, 0),//CubeFace::Right
+	Planar64Vec3::int( 0, 1, 0),//CubeFace::Top
+	Planar64Vec3::int( 0, 0, 1),//CubeFace::Back
+	Planar64Vec3::int(-1, 0, 0),//CubeFace::Left
+	Planar64Vec3::int( 0,-1, 0),//CubeFace::Bottom
+	Planar64Vec3::int( 0, 0,-1),//CubeFace::Front
+];
+const CUBE_DEFAULT_POLYS:[[[u32;3];4];6]=[
+	// right (1, 0, 0)
+	[
+		[6,2,0],//[vertex,tex,norm]
+		[5,1,0],
+		[2,0,0],
+		[1,3,0],
+	],
+	// top (0, 1, 0)
+	[
+		[5,3,1],
+		[4,2,1],
+		[3,1,1],
+		[2,0,1],
+	],
+	// back (0, 0, 1)
+	[
+		[0,3,2],
+		[1,2,2],
+		[2,1,2],
+		[3,0,2],
+	],
+	// left (-1, 0, 0)
+	[
+		[0,2,3],
+		[3,1,3],
+		[4,0,3],
+		[7,3,3],
+	],
+	// bottom (0,-1, 0)
+	[
+		[1,1,4],
+		[0,0,4],
+		[7,3,4],
+		[6,2,4],
+	],
+	// front (0, 0,-1)
+	[
+		[4,1,5],
+		[5,0,5],
+		[6,3,5],
+		[7,2,5],
+	],
+];
+
+#[derive(Hash,PartialEq,Eq)]
+pub enum WedgeFace{
+	Right,
+	TopFront,
+	Back,
+	Left,
+	Bottom,
+}
+const WEDGE_DEFAULT_NORMALS:[Planar64Vec3;5]=[
+	Planar64Vec3::int( 1, 0, 0),//Wedge::Right
+	Planar64Vec3::int( 0, 1,-1),//Wedge::TopFront
+	Planar64Vec3::int( 0, 0, 1),//Wedge::Back
+	Planar64Vec3::int(-1, 0, 0),//Wedge::Left
+	Planar64Vec3::int( 0,-1, 0),//Wedge::Bottom
+];
+/*
+local cornerWedgeVerticies = {
+	Vector3.new(-1/2,-1/2,-1/2),7
+	Vector3.new(-1/2,-1/2, 1/2),0
+	Vector3.new( 1/2,-1/2,-1/2),6
+	Vector3.new( 1/2,-1/2, 1/2),1
+	Vector3.new( 1/2, 1/2,-1/2),5
+}
+*/
+#[derive(Hash,PartialEq,Eq)]
+pub enum CornerWedgeFace{
+	Right,
+	TopBack,
+	TopLeft,
+	Bottom,
+	Front,
+}
+const CORNERWEDGE_DEFAULT_NORMALS:[Planar64Vec3;5]=[
+	Planar64Vec3::int( 1, 0, 0),//CornerWedge::Right
+	Planar64Vec3::int( 0, 1, 1),//CornerWedge::BackTop
+	Planar64Vec3::int(-1, 1, 0),//CornerWedge::LeftTop
+	Planar64Vec3::int( 0,-1, 0),//CornerWedge::Bottom
+	Planar64Vec3::int( 0, 0,-1),//CornerWedge::Front
+];
+//HashMap fits this use case perfectly but feels like using a sledgehammer to drive a nail
+pub fn unit_sphere()->crate::model::IndexedModel{
+	let mut indexed_model=crate::model::generate_indexed_model_list_from_obj(obj::ObjData::load_buf(&include_bytes!("../models/suzanne.obj")[..]).unwrap(),Color4::ONE).remove(0);
+	for pos in indexed_model.unique_pos.iter_mut(){
+		*pos=*pos/2;
+	}
+	indexed_model
+}
+pub type CubeFaceDescription=std::collections::HashMap::<CubeFace,FaceDescription>;
+pub fn unit_cube()->crate::model::IndexedModel{
+	let mut t=CubeFaceDescription::new();
+	t.insert(CubeFace::Right,FaceDescription::default());
+	t.insert(CubeFace::Top,FaceDescription::default());
+	t.insert(CubeFace::Back,FaceDescription::default());
+	t.insert(CubeFace::Left,FaceDescription::default());
+	t.insert(CubeFace::Bottom,FaceDescription::default());
+	t.insert(CubeFace::Front,FaceDescription::default());
+	generate_partial_unit_cube(t)
+}
+const TEAPOT_TRANSFORM:crate::integer::Planar64Mat3=crate::integer::Planar64Mat3::int_from_cols_array([0,1,0, -1,0,0, 0,0,1]);
+pub fn unit_cylinder()->crate::model::IndexedModel{
+	let mut indexed_model=crate::model::generate_indexed_model_list_from_obj(obj::ObjData::load_buf(&include_bytes!("../models/teapot.obj")[..]).unwrap(),Color4::ONE).remove(0);
+	for pos in indexed_model.unique_pos.iter_mut(){
+		*pos=TEAPOT_TRANSFORM*(*pos)/10;
+	}
+	indexed_model
+}
+pub type WedgeFaceDescription=std::collections::HashMap::<WedgeFace,FaceDescription>;
+pub fn unit_wedge()->crate::model::IndexedModel{
+	let mut t=WedgeFaceDescription::new();
+	t.insert(WedgeFace::Right,FaceDescription::default());
+	t.insert(WedgeFace::TopFront,FaceDescription::default());
+	t.insert(WedgeFace::Back,FaceDescription::default());
+	t.insert(WedgeFace::Left,FaceDescription::default());
+	t.insert(WedgeFace::Bottom,FaceDescription::default());
+	generate_partial_unit_wedge(t)
+}
+pub type CornerWedgeFaceDescription=std::collections::HashMap::<CornerWedgeFace,FaceDescription>;
+pub fn unit_cornerwedge()->crate::model::IndexedModel{
+	let mut t=CornerWedgeFaceDescription::new();
+	t.insert(CornerWedgeFace::Right,FaceDescription::default());
+	t.insert(CornerWedgeFace::TopBack,FaceDescription::default());
+	t.insert(CornerWedgeFace::TopLeft,FaceDescription::default());
+	t.insert(CornerWedgeFace::Bottom,FaceDescription::default());
+	t.insert(CornerWedgeFace::Front,FaceDescription::default());
+	generate_partial_unit_cornerwedge(t)
+}
+
+#[derive(Copy,Clone)]
+pub struct FaceDescription{
+	pub texture:Option<u32>,
+	pub transform:glam::Affine2,
+	pub color:Color4,
+}
+impl std::default::Default for FaceDescription{
+	fn default()->Self {
+		Self{
+			texture:None,
+			transform:glam::Affine2::IDENTITY,
+			color:Color4::new(1.0,1.0,1.0,0.0),//zero alpha to hide the default texture
+		}
+	}
+}
+impl FaceDescription{
+	pub fn new(texture:u32,transform:glam::Affine2,color:Color4)->Self{
+		Self{texture:Some(texture),transform,color}
+	}
+	pub fn from_texture(texture:u32)->Self{
+		Self{
+			texture:Some(texture),
+			transform:glam::Affine2::IDENTITY,
+			color:Color4::ONE,
+		}
+	}
+}
+//TODO: it's probably better to use a shared vertex buffer between all primitives and use indexed rendering instead of generating a unique vertex buffer for each primitive.
+//implementation: put all roblox primitives into one model.groups <- this won't work but I forget why
+pub fn generate_partial_unit_cube(face_descriptions:CubeFaceDescription)->crate::model::IndexedModel{
+	let mut generated_pos=Vec::new();
+	let mut generated_tex=Vec::new();
+	let mut generated_normal=Vec::new();
+	let mut generated_color=Vec::new();
+	let mut generated_vertices=Vec::new();
+	let mut groups=Vec::new();
+	let mut transforms=Vec::new();
+	//note that on a cube every vertex is guaranteed to be unique, so there's no need to hash them against existing vertices.
+	for (face,face_description) in face_descriptions.into_iter(){
+		//assume that scanning short lists is faster than hashing.
+		let transform_index=if let Some(transform_index)=transforms.iter().position(|&transform|transform==face_description.transform){
+			transform_index
+		}else{
+			//create new transform_index
+			let transform_index=transforms.len();
+			transforms.push(face_description.transform);
+			for tex in CUBE_DEFAULT_TEXTURE_COORDS{
+				generated_tex.push(face_description.transform.transform_point2(tex));
+			}
+			transform_index
+		} as u32;
+		let color_index=if let Some(color_index)=generated_color.iter().position(|&color|color==face_description.color){
+			color_index
+		}else{
+			//create new color_index
+			let color_index=generated_color.len();
+			generated_color.push(face_description.color);
+			color_index
+		} as u32;
+		let face_id=match face{
+			CubeFace::Right => 0,
+			CubeFace::Top => 1,
+			CubeFace::Back => 2,
+			CubeFace::Left => 3,
+			CubeFace::Bottom => 4,
+			CubeFace::Front => 5,
+		};
+		//always push normal
+		let normal_index=generated_normal.len() as u32;
+		generated_normal.push(CUBE_DEFAULT_NORMALS[face_id]);
+		//push vertices as they are needed
+		groups.push(IndexedGroup{
+			texture:face_description.texture,
+			polys:vec![IndexedPolygon{
+				vertices:CUBE_DEFAULT_POLYS[face_id].map(|tup|{
+					let pos=CUBE_DEFAULT_VERTICES[tup[0] as usize];
+					let pos_index=if let Some(pos_index)=generated_pos.iter().position(|&p|p==pos){
+						pos_index
+					}else{
+						//create new pos_index
+						let pos_index=generated_pos.len();
+						generated_pos.push(pos);
+						pos_index
+					} as u32;
+					//always push vertex
+					let vertex=IndexedVertex{
+						pos:pos_index,
+						tex:tup[1]+4*transform_index,
+						normal:normal_index,
+						color:color_index,
+					};
+					let vert_index=generated_vertices.len();
+					generated_vertices.push(vertex);
+					vert_index as u32
+				}).to_vec(),
+			}],
+		});
+	}
+	IndexedModel{
+		unique_pos:generated_pos,
+		unique_tex:generated_tex,
+		unique_normal:generated_normal,
+		unique_color:generated_color,
+		unique_vertices:generated_vertices,
+		groups,
+		instances:Vec::new(),
+	}
+}
+//don't think too hard about the copy paste because this is all going into the map tool eventually...
+pub fn generate_partial_unit_wedge(face_descriptions:WedgeFaceDescription)->crate::model::IndexedModel{
+	let wedge_default_polys=vec![
+		// right (1, 0, 0)
+		vec![
+			[6,2,0],//[vertex,tex,norm]
+			[2,0,0],
+			[1,3,0],
+		],
+		// FrontTop (0, 1, -1)
+		vec![
+			[3,1,1],
+			[2,0,1],
+			[6,3,1],
+			[7,2,1],
+		],
+		// back (0, 0, 1)
+		vec![
+			[0,3,2],
+			[1,2,2],
+			[2,1,2],
+			[3,0,2],
+		],
+		// left (-1, 0, 0)
+		vec![
+			[0,2,3],
+			[3,1,3],
+			[7,3,3],
+		],
+		// bottom (0,-1, 0)
+		vec![
+			[1,1,4],
+			[0,0,4],
+			[7,3,4],
+			[6,2,4],
+		],
+	];
+	let mut generated_pos=Vec::new();
+	let mut generated_tex=Vec::new();
+	let mut generated_normal=Vec::new();
+	let mut generated_color=Vec::new();
+	let mut generated_vertices=Vec::new();
+	let mut groups=Vec::new();
+	let mut transforms=Vec::new();
+	//note that on a cube every vertex is guaranteed to be unique, so there's no need to hash them against existing vertices.
+	for (face,face_description) in face_descriptions.into_iter(){
+		//assume that scanning short lists is faster than hashing.
+		let transform_index=if let Some(transform_index)=transforms.iter().position(|&transform|transform==face_description.transform){
+			transform_index
+		}else{
+			//create new transform_index
+			let transform_index=transforms.len();
+			transforms.push(face_description.transform);
+			for tex in CUBE_DEFAULT_TEXTURE_COORDS{
+				generated_tex.push(face_description.transform.transform_point2(tex));
+			}
+			transform_index
+		} as u32;
+		let color_index=if let Some(color_index)=generated_color.iter().position(|&color|color==face_description.color){
+			color_index
+		}else{
+			//create new color_index
+			let color_index=generated_color.len();
+			generated_color.push(face_description.color);
+			color_index
+		} as u32;
+		let face_id=match face{
+			WedgeFace::Right => 0,
+			WedgeFace::TopFront => 1,
+			WedgeFace::Back => 2,
+			WedgeFace::Left => 3,
+			WedgeFace::Bottom => 4,
+		};
+		//always push normal
+		let normal_index=generated_normal.len() as u32;
+		generated_normal.push(WEDGE_DEFAULT_NORMALS[face_id]);
+		//push vertices as they are needed
+		groups.push(IndexedGroup{
+			texture:face_description.texture,
+			polys:vec![IndexedPolygon{
+				vertices:wedge_default_polys[face_id].iter().map(|tup|{
+					let pos=CUBE_DEFAULT_VERTICES[tup[0] as usize];
+					let pos_index=if let Some(pos_index)=generated_pos.iter().position(|&p|p==pos){
+						pos_index
+					}else{
+						//create new pos_index
+						let pos_index=generated_pos.len();
+						generated_pos.push(pos);
+						pos_index
+					} as u32;
+					//always push vertex
+					let vertex=IndexedVertex{
+						pos:pos_index,
+						tex:tup[1]+4*transform_index,
+						normal:normal_index,
+						color:color_index,
+					};
+					let vert_index=generated_vertices.len();
+					generated_vertices.push(vertex);
+					vert_index as u32
+				}).collect(),
+			}],
+		});
+	}
+	IndexedModel{
+		unique_pos:generated_pos,
+		unique_tex:generated_tex,
+		unique_normal:generated_normal,
+		unique_color:generated_color,
+		unique_vertices:generated_vertices,
+		groups,
+		instances:Vec::new(),
+	}
+}
+
+pub fn generate_partial_unit_cornerwedge(face_descriptions:CornerWedgeFaceDescription)->crate::model::IndexedModel{
+	let cornerwedge_default_polys=vec![
+		// right (1, 0, 0)
+		vec![
+			[6,2,0],//[vertex,tex,norm]
+			[5,1,0],
+			[1,3,0],
+		],
+		// BackTop (0, 1, 1)
+		vec![
+			[5,3,1],
+			[0,1,1],
+			[1,0,1],
+		],
+		// LeftTop (-1, 1, 0)
+		vec![
+			[5,3,2],
+			[7,2,2],
+			[0,1,2],
+		],
+		// bottom (0,-1, 0)
+		vec![
+			[1,1,3],
+			[0,0,3],
+			[7,3,3],
+			[6,2,3],
+		],
+		// front (0, 0,-1)
+		vec![
+			[5,0,4],
+			[6,3,4],
+			[7,2,4],
+		],
+	];
+	let mut generated_pos=Vec::new();
+	let mut generated_tex=Vec::new();
+	let mut generated_normal=Vec::new();
+	let mut generated_color=Vec::new();
+	let mut generated_vertices=Vec::new();
+	let mut groups=Vec::new();
+	let mut transforms=Vec::new();
+	//note that on a cube every vertex is guaranteed to be unique, so there's no need to hash them against existing vertices.
+	for (face,face_description) in face_descriptions.into_iter(){
+		//assume that scanning short lists is faster than hashing.
+		let transform_index=if let Some(transform_index)=transforms.iter().position(|&transform|transform==face_description.transform){
+			transform_index
+		}else{
+			//create new transform_index
+			let transform_index=transforms.len();
+			transforms.push(face_description.transform);
+			for tex in CUBE_DEFAULT_TEXTURE_COORDS{
+				generated_tex.push(face_description.transform.transform_point2(tex));
+			}
+			transform_index
+		} as u32;
+		let color_index=if let Some(color_index)=generated_color.iter().position(|&color|color==face_description.color){
+			color_index
+		}else{
+			//create new color_index
+			let color_index=generated_color.len();
+			generated_color.push(face_description.color);
+			color_index
+		} as u32;
+		let face_id=match face{
+			CornerWedgeFace::Right => 0,
+			CornerWedgeFace::TopBack => 1,
+			CornerWedgeFace::TopLeft => 2,
+			CornerWedgeFace::Bottom => 3,
+			CornerWedgeFace::Front => 4,
+		};
+		//always push normal
+		let normal_index=generated_normal.len() as u32;
+		generated_normal.push(CORNERWEDGE_DEFAULT_NORMALS[face_id]);
+		//push vertices as they are needed
+		groups.push(IndexedGroup{
+			texture:face_description.texture,
+			polys:vec![IndexedPolygon{
+				vertices:cornerwedge_default_polys[face_id].iter().map(|tup|{
+					let pos=CUBE_DEFAULT_VERTICES[tup[0] as usize];
+					let pos_index=if let Some(pos_index)=generated_pos.iter().position(|&p|p==pos){
+						pos_index
+					}else{
+						//create new pos_index
+						let pos_index=generated_pos.len();
+						generated_pos.push(pos);
+						pos_index
+					} as u32;
+					//always push vertex
+					let vertex=IndexedVertex{
+						pos:pos_index,
+						tex:tup[1]+4*transform_index,
+						normal:normal_index,
+						color:color_index,
+					};
+					let vert_index=generated_vertices.len();
+					generated_vertices.push(vertex);
+					vert_index as u32
+				}).collect(),
+			}],
+		});
+	}
+	IndexedModel{
+		unique_pos:generated_pos,
+		unique_tex:generated_tex,
+		unique_normal:generated_normal,
+		unique_color:generated_color,
+		unique_vertices:generated_vertices,
+		groups,
+		instances:Vec::new(),
+	}
+}

src/settings.rs

@@ -1,14 +1,11 @@
-use strafesnet_common::integer::{Ratio64,Ratio64Vec2};
-#[derive(Clone)]
+use crate::integer::{Ratio64,Ratio64Vec2};
 struct Ratio{
 	ratio:f64,
 }
-#[derive(Clone)]
 enum DerivedFov{
 	FromScreenAspect,
 	FromAspect(Ratio),
 }
-#[derive(Clone)]
 enum Fov{
 	Exactly{x:f64,y:f64},
 	SpecifyXDeriveY{x:f64,y:DerivedFov},
@@ -19,11 +16,9 @@ impl Default for Fov{
 		Fov::SpecifyYDeriveX{x:DerivedFov::FromScreenAspect,y:1.0}
 	}
 }
-#[derive(Clone)]
 enum DerivedSensitivity{
 	FromRatio(Ratio64),
 }
-#[derive(Clone)]
 enum Sensitivity{
 	Exactly{x:Ratio64,y:Ratio64},
 	SpecifyXDeriveY{x:Ratio64,y:DerivedSensitivity},
@@ -35,7 +30,7 @@ impl Default for Sensitivity{
 	}
 }
 
-#[derive(Default,Clone)]
+#[derive(Default)]
 pub struct UserSettings{
 	fov:Fov,
 	sensitivity:Sensitivity,

src/setup.rs

@@ -1,292 +0,0 @@
-use crate::window::WindowInstruction;
-use strafesnet_common::instruction::TimedInstruction;
-use strafesnet_common::integer;
-
-fn optional_features()->wgpu::Features{
-	wgpu::Features::TEXTURE_COMPRESSION_ASTC
-	|wgpu::Features::TEXTURE_COMPRESSION_ETC2
-}
-fn required_features()->wgpu::Features{
-	wgpu::Features::TEXTURE_COMPRESSION_BC
-}
-fn required_downlevel_capabilities()->wgpu::DownlevelCapabilities{
-	wgpu::DownlevelCapabilities{
-		flags:wgpu::DownlevelFlags::empty(),
-		shader_model:wgpu::ShaderModel::Sm5,
-		..wgpu::DownlevelCapabilities::default()
-	}
-}
-pub fn required_limits()->wgpu::Limits{
-	wgpu::Limits::default()
-}
-
-struct SetupContextPartial1{
-	backends:wgpu::Backends,
-	instance:wgpu::Instance,
-}
-fn create_window(title:&str,event_loop:&winit::event_loop::EventLoop<()>)->Result<winit::window::Window,winit::error::OsError>{
-	let mut attr=winit::window::WindowAttributes::default();
-	attr=attr.with_title(title);
-	#[cfg(windows_OFF)] // TODO
-	{
-		use winit::platform::windows::WindowBuilderExtWindows;
-		builder=builder.with_no_redirection_bitmap(true);
-	}
-	event_loop.create_window(attr)
-}
-fn create_instance()->SetupContextPartial1{
-	let backends=wgpu::util::backend_bits_from_env().unwrap_or_else(wgpu::Backends::all);
-	let dx12_shader_compiler=wgpu::util::dx12_shader_compiler_from_env().unwrap_or_default();
-	SetupContextPartial1{
-		backends,
-		instance:wgpu::Instance::new(wgpu::InstanceDescriptor{
-			backends,
-			dx12_shader_compiler,
-			..Default::default()
-		}),
-	}
-}
-impl SetupContextPartial1{
-	fn create_surface<'a>(self,window:&'a winit::window::Window)->Result<SetupContextPartial2<'a>,wgpu::CreateSurfaceError>{
-		Ok(SetupContextPartial2{
-			backends:self.backends,
-			surface:self.instance.create_surface(window)?,
-			instance:self.instance,
-		})
-	}
-}
-struct SetupContextPartial2<'a>{
-	backends:wgpu::Backends,
-	instance:wgpu::Instance,
-	surface:wgpu::Surface<'a>,
-}
-impl<'a> SetupContextPartial2<'a>{
-	fn pick_adapter(self)->SetupContextPartial3<'a>{
-		let adapter;
-
-		//TODO: prefer adapter that implements optional features
-		//let optional_features=optional_features();
-		let required_features=required_features();
-
-		//no helper function smh gotta write it myself
-		let adapters=self.instance.enumerate_adapters(self.backends);
-
-		let mut chosen_adapter=None;
-		let mut chosen_adapter_score=0;
-		for adapter in adapters {
-			if !adapter.is_surface_supported(&self.surface) {
-				continue;
-			}
-
-			let score=match adapter.get_info().device_type{
-				wgpu::DeviceType::IntegratedGpu=>3,
-				wgpu::DeviceType::DiscreteGpu=>4,
-				wgpu::DeviceType::VirtualGpu=>2,
-				wgpu::DeviceType::Other|wgpu::DeviceType::Cpu=>1,
-			};
-
-			let adapter_features=adapter.features();
-			if chosen_adapter_score<score&&adapter_features.contains(required_features) {
-				chosen_adapter_score=score;
-				chosen_adapter=Some(adapter);
-			}
-		}
-
-		if let Some(maybe_chosen_adapter)=chosen_adapter{
-			adapter=maybe_chosen_adapter;
-		}else{
-			panic!("No suitable GPU adapters found on the system!");
-		}
-
-
-		let adapter_info=adapter.get_info();
-		println!("Using {} ({:?})", adapter_info.name, adapter_info.backend);
-
-		let required_downlevel_capabilities=required_downlevel_capabilities();
-		let downlevel_capabilities=adapter.get_downlevel_capabilities();
-		assert!(
-			downlevel_capabilities.shader_model >= required_downlevel_capabilities.shader_model,
-			"Adapter does not support the minimum shader model required to run this example: {:?}",
-			required_downlevel_capabilities.shader_model
-		);
-		assert!(
-			downlevel_capabilities
-				.flags
-				.contains(required_downlevel_capabilities.flags),
-			"Adapter does not support the downlevel capabilities required to run this example: {:?}",
-			required_downlevel_capabilities.flags - downlevel_capabilities.flags
-		);
-		SetupContextPartial3{
-			instance:self.instance,
-			surface:self.surface,
-			adapter,
-		}
-	}
-}
-struct SetupContextPartial3<'a>{
-	instance:wgpu::Instance,
-	surface:wgpu::Surface<'a>,
-	adapter:wgpu::Adapter,
-}
-impl<'a> SetupContextPartial3<'a>{
-	fn request_device(self)->SetupContextPartial4<'a>{
-		let optional_features=optional_features();
-		let required_features=required_features();
-
-		// Make sure we use the texture resolution limits from the adapter, so we can support images the size of the surface.
-		let needed_limits=required_limits().using_resolution(self.adapter.limits());
-
-		let trace_dir=std::env::var("WGPU_TRACE");
-		let (device, queue)=pollster::block_on(self.adapter
-			.request_device(
-				&wgpu::DeviceDescriptor {
-					label: None,
-					required_features: (optional_features & self.adapter.features()) | required_features,
-					required_limits: needed_limits,
-					memory_hints:wgpu::MemoryHints::Performance,
-				},
-				trace_dir.ok().as_ref().map(std::path::Path::new),
-			))
-			.expect("Unable to find a suitable GPU adapter!");
-
-		SetupContextPartial4{
-			instance:self.instance,
-			surface:self.surface,
-			adapter:self.adapter,
-			device,
-			queue,
-		}
-	}
-}
-struct SetupContextPartial4<'a>{
-	instance:wgpu::Instance,
-	surface:wgpu::Surface<'a>,
-	adapter:wgpu::Adapter,
-	device:wgpu::Device,
-	queue:wgpu::Queue,
-}
-impl<'a> SetupContextPartial4<'a>{
-	fn configure_surface(self,size:&'a winit::dpi::PhysicalSize<u32>)->SetupContext<'a>{
-		let mut config=self.surface
-			.get_default_config(&self.adapter, size.width, size.height)
-			.expect("Surface isn't supported by the adapter.");
-		let surface_view_format=config.format.add_srgb_suffix();
-		config.view_formats.push(surface_view_format);
-		config.present_mode=wgpu::PresentMode::AutoNoVsync;
-		self.surface.configure(&self.device, &config);
-
-		SetupContext{
-			instance:self.instance,
-			surface:self.surface,
-			device:self.device,
-			queue:self.queue,
-			config,
-		}
-	}
-}
-pub struct SetupContext<'a>{
-	pub instance:wgpu::Instance,
-	pub surface:wgpu::Surface<'a>,
-	pub device:wgpu::Device,
-	pub queue:wgpu::Queue,
-	pub config:wgpu::SurfaceConfiguration,
-}
-
-pub fn setup_and_start(title:String){
-	let event_loop=winit::event_loop::EventLoop::new().unwrap();
-
-	println!("Initializing the surface...");
-
-	let partial_1=create_instance();
-
-	let window=create_window(title.as_str(),&event_loop).unwrap();
-
-	let partial_2=partial_1.create_surface(&window).unwrap();
-
-	let partial_3=partial_2.pick_adapter();
-
-	let partial_4=partial_3.request_device();
-
-	let size=window.inner_size();
-
-	let setup_context=partial_4.configure_surface(&size);
-
-	//dedicated thread to ping request redraw back and resize the window doesn't seem logical
-
-	//the thread that spawns the physics thread
-	let mut window_thread=crate::window::worker(
-		&window,
-		setup_context,
-	);
-
-	if let Some(arg)=std::env::args().nth(1){
-		let path=std::path::PathBuf::from(arg);
-		window_thread.send(TimedInstruction{
-			time:integer::Time::ZERO,
-			instruction:WindowInstruction::WindowEvent(winit::event::WindowEvent::DroppedFile(path)),
-		}).unwrap();
-	};
-
-	println!("Entering event loop...");
-	let root_time=std::time::Instant::now();
-	run_event_loop(event_loop,window_thread,root_time).unwrap();
-}
-
-fn run_event_loop(
-	event_loop:winit::event_loop::EventLoop<()>,
-	mut window_thread:crate::compat_worker::QNWorker<TimedInstruction<WindowInstruction>>,
-	root_time:std::time::Instant
-	)->Result<(),winit::error::EventLoopError>{
-		event_loop.run(move |event,elwt|{
-			let time=integer::Time::from_nanos(root_time.elapsed().as_nanos() as i64);
-			// *control_flow=if cfg!(feature="metal-auto-capture"){
-			// 	winit::event_loop::ControlFlow::Exit
-			// }else{
-			// 	winit::event_loop::ControlFlow::Poll
-			// };
-			match event{
-				winit::event::Event::AboutToWait=>{
-					window_thread.send(TimedInstruction{time,instruction:WindowInstruction::RequestRedraw}).unwrap();
-				}
-				winit::event::Event::WindowEvent {
-					event:
-						// WindowEvent::Resized(size)
-						// | WindowEvent::ScaleFactorChanged {
-						// 	new_inner_size: &mut size,
-						// 	..
-						// },
-						winit::event::WindowEvent::Resized(size),//ignoring scale factor changed for now because mutex bruh
-					window_id:_,
-				} => {
-					window_thread.send(TimedInstruction{time,instruction:WindowInstruction::Resize(size)}).unwrap();
-				}
-				winit::event::Event::WindowEvent{event,..}=>match event{
-					winit::event::WindowEvent::KeyboardInput{
-						event:
-							winit::event::KeyEvent {
-							logical_key: winit::keyboard::Key::Named(winit::keyboard::NamedKey::Escape),
-								state: winit::event::ElementState::Pressed,
-								..
-							},
-						..
-					}
-					|winit::event::WindowEvent::CloseRequested=>{
-						elwt.exit();
-					}
-					winit::event::WindowEvent::RedrawRequested=>{
-						window_thread.send(TimedInstruction{time,instruction:WindowInstruction::Render}).unwrap();
-					}
-					_=>{
-						window_thread.send(TimedInstruction{time,instruction:WindowInstruction::WindowEvent(event)}).unwrap();
-					}
-				},
-				winit::event::Event::DeviceEvent{
-					event,
-					..
-				} => {
-					window_thread.send(TimedInstruction{time,instruction:WindowInstruction::DeviceEvent(event)}).unwrap();
-				},
-				_=>{}
-			}
-		})
-}

src/shader.wgsl

@@ -48,7 +48,7 @@ struct ModelInstance{
 //my fancy idea is to create a megatexture for each model that includes all the textures each intance will need
 //the texture transform then maps the texture coordinates to the location of the specific texture
 //group 1 is the model
-const MAX_MODEL_INSTANCES=512;
+const MAX_MODEL_INSTANCES=4096;
 @group(2)
 @binding(0)
 var<uniform> model_instances: array<ModelInstance, MAX_MODEL_INSTANCES>;

src/sweep.rs

@@ -0,0 +1,8 @@
+
+//something that implements body + hitbox + transform can predict collision
+impl crate::sweep::PredictCollision for Model {
+	fn predict_collision(&self,other:&Model) -> Option<crate::event::EventStruct> {
+		//math!
+		None
+	}
+}

src/window.rs

@@ -1,223 +0,0 @@
-use crate::physics_worker::InputInstruction;
-use strafesnet_common::integer;
-use strafesnet_common::instruction::TimedInstruction;
-
-pub enum WindowInstruction{
-	Resize(winit::dpi::PhysicalSize<u32>),
-	WindowEvent(winit::event::WindowEvent),
-	DeviceEvent(winit::event::DeviceEvent),
-	RequestRedraw,
-	Render,
-}
-
-//holds thread handles to dispatch to
-struct WindowContext<'a>{
-	manual_mouse_lock:bool,
-	mouse:strafesnet_common::mouse::MouseState,//std::sync::Arc<std::sync::Mutex<>>
-	screen_size:glam::UVec2,
-	window:&'a winit::window::Window,
-	physics_thread:crate::compat_worker::QNWorker<'a, TimedInstruction<crate::physics_worker::Instruction>>,
-}
-
-impl WindowContext<'_>{
-	fn get_middle_of_screen(&self)->winit::dpi::PhysicalPosition<f32>{
-		winit::dpi::PhysicalPosition::new(self.screen_size.x as f32/2.0,self.screen_size.y as f32/2.0)
-	}
-	fn window_event(&mut self,time:integer::Time,event: winit::event::WindowEvent) {
-		match event {
-			winit::event::WindowEvent::DroppedFile(path)=>{
-				match crate::file::load(path.as_path()){
-					Ok(map)=>self.physics_thread.send(TimedInstruction{time,instruction:crate::physics_worker::Instruction::ChangeMap(map)}).unwrap(),
-					Err(e)=>println!("Failed to load map: {e}"),
-				}
-			},
-			winit::event::WindowEvent::Focused(state)=>{
-				//pause unpause
-				self.physics_thread.send(TimedInstruction{
-					time,
-					instruction:crate::physics_worker::Instruction::SetPaused(!state),
-				}).unwrap();
-				//recalculate pressed keys on focus
-			},
-			winit::event::WindowEvent::KeyboardInput{
-				event:winit::event::KeyEvent{state,logical_key,repeat:false,..},
-				..
-			}=>{
-				let s=match state{
-					winit::event::ElementState::Pressed=>true,
-					winit::event::ElementState::Released=>false,
-				};
-				match logical_key{
-					winit::keyboard::Key::Named(winit::keyboard::NamedKey::Tab)=>{
-						if s{
-							self.manual_mouse_lock=false;
-							match self.window.set_cursor_position(self.get_middle_of_screen()){
-								Ok(())=>(),
-								Err(e)=>println!("Could not set cursor position: {:?}",e),
-							}
-							match self.window.set_cursor_grab(winit::window::CursorGrabMode::None){
-								Ok(())=>(),
-								Err(e)=>println!("Could not release cursor: {:?}",e),
-							}
-						}else{
-							//if cursor is outside window don't lock but apparently there's no get pos function
-							//let pos=window.get_cursor_pos();
-							match self.window.set_cursor_grab(winit::window::CursorGrabMode::Locked){
-								Ok(())=>(),
-								Err(_)=>{
-									match self.window.set_cursor_grab(winit::window::CursorGrabMode::Confined){
-										Ok(())=>(),
-										Err(e)=>{
-											self.manual_mouse_lock=true;
-											println!("Could not confine cursor: {:?}",e)
-										},
-									}
-								}
-							}
-						}
-						self.window.set_cursor_visible(s);
-					},
-					winit::keyboard::Key::Named(winit::keyboard::NamedKey::F11)=>{
-						if s{
-							if self.window.fullscreen().is_some(){
-								self.window.set_fullscreen(None);
-							}else{
-								self.window.set_fullscreen(Some(winit::window::Fullscreen::Borderless(None)));
-							}
-						}
-					},
-					winit::keyboard::Key::Named(winit::keyboard::NamedKey::Escape)=>{
-						if s{
-							self.manual_mouse_lock=false;
-							match self.window.set_cursor_grab(winit::window::CursorGrabMode::None){
-								Ok(())=>(),
-								Err(e)=>println!("Could not release cursor: {:?}",e),
-							}
-							self.window.set_cursor_visible(true);
-						}
-					},
-					keycode=>{
-						if let Some(input_instruction)=match keycode{
-							winit::keyboard::Key::Named(winit::keyboard::NamedKey::Space)=>Some(InputInstruction::Jump(s)),
-							winit::keyboard::Key::Character(key)=>match key.as_str(){
-								"w"=>Some(InputInstruction::MoveForward(s)),
-								"a"=>Some(InputInstruction::MoveLeft(s)),
-								"s"=>Some(InputInstruction::MoveBack(s)),
-								"d"=>Some(InputInstruction::MoveRight(s)),
-								"e"=>Some(InputInstruction::MoveUp(s)),
-								"q"=>Some(InputInstruction::MoveDown(s)),
-								"z"=>Some(InputInstruction::Zoom(s)),
-								"r"=>if s{
-									//mouse needs to be reset since the position is absolute
-									self.mouse=strafesnet_common::mouse::MouseState::default();
-									Some(InputInstruction::ResetAndRestart)
-								}else{None},
-								"f"=>if s{Some(InputInstruction::PracticeFly)}else{None},
-								_=>None,
-							},
-							_=>None,
-						}{
-							self.physics_thread.send(TimedInstruction{
-								time,
-								instruction:crate::physics_worker::Instruction::Input(input_instruction),
-							}).unwrap();
-						}
-					},
-				}
-			},
-			_=>(),
-		}
-	}
-
-	fn device_event(&mut self,time:integer::Time,event: winit::event::DeviceEvent) {
-		match event {
-			winit::event::DeviceEvent::MouseMotion {
-			    delta,//these (f64,f64) are integers on my machine
-			} => {
-				if self.manual_mouse_lock{
-					match self.window.set_cursor_position(self.get_middle_of_screen()){
-						Ok(())=>(),
-						Err(e)=>println!("Could not set cursor position: {:?}",e),
-					}
-				}
-				//do not step the physics because the mouse polling rate is higher than the physics can run.
-				//essentially the previous input will be overwritten until a true step runs
-				//which is fine because they run all the time.
-				let delta=glam::ivec2(delta.0 as i32,delta.1 as i32);
-				self.mouse.pos+=delta;
-				self.physics_thread.send(TimedInstruction{
-					time,
-					instruction:crate::physics_worker::Instruction::Input(InputInstruction::MoveMouse(self.mouse.pos)),
-				}).unwrap();
-			},
-			winit::event::DeviceEvent::MouseWheel {
-			    delta,
-			} => {
-				println!("mousewheel {:?}",delta);
-				if false{//self.physics.style.use_scroll{
-					self.physics_thread.send(TimedInstruction{
-						time,
-						instruction:crate::physics_worker::Instruction::Input(InputInstruction::Jump(true)),//activates the immediate jump path, but the style modifier prevents controls&CONTROL_JUMP bit from being set to auto jump
-					}).unwrap();
-				}
-			}
-			_=>(),
-		}
-	}
-}
-pub fn worker<'a>(
-	window:&'a winit::window::Window,
-	setup_context:crate::setup::SetupContext<'a>,
-)->crate::compat_worker::QNWorker<'a,TimedInstruction<WindowInstruction>>{
-	// WindowContextSetup::new
-		let user_settings=crate::settings::read_user_settings();
-
-		let mut graphics=crate::graphics::GraphicsState::new(&setup_context.device,&setup_context.queue,&setup_context.config);
-		graphics.load_user_settings(&user_settings);
-
-	//WindowContextSetup::into_context
-		let screen_size=glam::uvec2(setup_context.config.width,setup_context.config.height);
-		let graphics_thread=crate::graphics_worker::new(graphics,setup_context.config,setup_context.surface,setup_context.device,setup_context.queue);
-		let mut window_context=WindowContext{
-			manual_mouse_lock:false,
-			mouse:strafesnet_common::mouse::MouseState::default(),
-			//make sure to update this!!!!!
-			screen_size,
-			window,
-			physics_thread:crate::physics_worker::new(
-				graphics_thread,
-				user_settings,
-			),
-		};
-
-	//WindowContextSetup::into_worker
-		crate::compat_worker::QNWorker::new(move |ins:TimedInstruction<WindowInstruction>|{
-			match ins.instruction{
-				WindowInstruction::RequestRedraw=>{
-					window_context.window.request_redraw();
-				}
-				WindowInstruction::WindowEvent(window_event)=>{
-					window_context.window_event(ins.time,window_event);
-				},
-				WindowInstruction::DeviceEvent(device_event)=>{
-					window_context.device_event(ins.time,device_event);
-				},
-				WindowInstruction::Resize(size)=>{
-					window_context.physics_thread.send(
-						TimedInstruction{
-							time:ins.time,
-							instruction:crate::physics_worker::Instruction::Resize(size)
-						}
-					).unwrap();
-				}
-				WindowInstruction::Render=>{
-					window_context.physics_thread.send(
-						TimedInstruction{
-							time:ins.time,
-							instruction:crate::physics_worker::Instruction::Render
-						}
-					).unwrap();
-				}
-			}
-		})
-}

src/worker.rs

@@ -2,68 +2,16 @@ use std::thread;
 use std::sync::{mpsc,Arc};
 use parking_lot::Mutex;
 
-//WorkerPool
-struct Pool(u32);
-enum PoolOrdering{
-	Single,//single thread cannot get out of order
-	Ordered(u32),//order matters and should be buffered/dropped according to ControlFlow
-	Unordered(u32),//order does not matter
-}
-//WorkerInput
-enum Input{
-	//no input, workers have everything needed at creation
-	None,
-	//Immediate input to any available worker, dropped if they are overflowing (all workers are busy)
-	Immediate,
-	//Queued input is ordered, but serial jobs that mutate state (such as running physics) can only be done with a single worker
-	Queued,//"Fifo"
-	//Query a function to get next input when a thread becomes available
-	//worker stops querying when Query function returns None and dies after all threads complete
-	//lifetimes sound crazy on this one
-	Query,
-	//Queue of length one, the input is replaced if it is submitted twice before the current work finishes
-	Mailbox,
-}
-//WorkerOutput
-enum Output{
-	None(Pool),
-	Realtime(PoolOrdering),//outputs are dropped if they are out of order and order is demanded
-	Buffered(PoolOrdering),//outputs are held back internally if they are out of order and order is demanded
-}
-
-//It would be possible to implement all variants
-//with a query input function and callback output function but I'm not sure if that's worth it.
-//Immediate = Condvar
-//Queued = receiver.recv()
-//a callback function would need to use an async runtime!
-
-//realtime output is an arc mutex of the output value that is assigned every time a worker completes a job
-//buffered output produces a receiver object that can be passed to the creation of another worker
-//when ordering is requested, output is ordered by the order each thread is run
-//which is the same as the order that the input data is processed except for Input::None which has no input data
-//WorkerDescription
-struct Description{
-	input:Input,
-	output:Output,
-}
-
 //The goal here is to have a worker thread that parks itself when it runs out of work.
 //The worker thread publishes the result of its work back to the worker object for every item in the work queue.
-//Previous values do not matter as soon as a new value is produced, which is why it's called "Realtime"
 //The physics (target use case) knows when it has not changed the body, so not updating the value is also an option.
 
-/*
-QR = WorkerDescription{
-	input:Queued,
-	output:Realtime(Single),
-}
-*/
-pub struct QRWorker<Task:Send,Value:Clone>{
+pub struct Worker<Task:Send,Value:Clone> {
 	sender: mpsc::Sender<Task>,
 	value:Arc<Mutex<Value>>,
 }
 
-impl<Task:Send+'static,Value:Clone+Send+'static> QRWorker<Task,Value>{
+impl<Task:Send+'static,Value:Clone+Send+'static> Worker<Task,Value> {
 	pub fn new<F:FnMut(Task)->Value+Send+'static>(value:Value,mut f:F) -> Self {
 		let (sender, receiver) = mpsc::channel::<Task>();
 		let ret=Self {
@@ -97,120 +45,63 @@ impl<Task:Send+'static,Value:Clone+Send+'static> QRWorker<Task,Value>{
 	}
 }
 
-/*
-QN = WorkerDescription{
-	input:Queued,
-	output:None(Single),
-}
-*/
-//None Output Worker does all its work internally from the perspective of the work submitter
-pub struct QNWorker<'a,Task:Send>{
-	sender: mpsc::Sender<Task>,
-	handle:thread::ScopedJoinHandle<'a,()>,
+pub struct CompatWorker<Task,Value:Clone,F>{
+	data:std::marker::PhantomData<Task>,
+	f:F,
+	value:Value,
 }
 
-impl<'a,Task:Send+'a> QNWorker<'a,Task>{
-	pub fn new<F:FnMut(Task)+Send+'a>(scope:&'a thread::Scope<'a,'_>,mut f:F)->QNWorker<'a,Task>{
-		let (sender,receiver)=mpsc::channel::<Task>();
-		let handle=scope.spawn(move ||{
-			loop {
-				match receiver.recv() {
-					Ok(task)=>f(task),
-					Err(_)=>{
-						println!("Worker stopping.",);
-						break;
-					}
-				}
-			}
-		});
-		Self{
-			sender,
-			handle,
+impl<Task,Value:Clone,F:FnMut(Task)->Value> CompatWorker<Task,Value,F> {
+	pub fn new(value:Value,f:F) -> Self {
+		Self {
+			f,
+			value,
+			data:std::marker::PhantomData,
 		}
 	}
-	pub fn send(&self,task:Task)->Result<(),mpsc::SendError<Task>>{
-		self.sender.send(task)
+
+	pub fn send(&mut self,task:Task)->Result<(),()>{
+		self.value=(self.f)(task);
+		Ok(())
+	}
+
+	pub fn grab_clone(&self)->Value{
+		self.value.clone()
 	}
 }
 
-/*
-IN = WorkerDescription{
-	input:Immediate,
-	output:None(Single),
-}
-*/
-//Inputs are dropped if the worker is busy
-pub struct INWorker<'a,Task:Send>{
-	sender: mpsc::SyncSender<Task>,
-	handle:thread::ScopedJoinHandle<'a,()>,
-}
+#[test]//How to run this test with printing: cargo test --release -- --nocapture
+fn test_worker() {
+	println!("hiiiii");
+	// Create the worker thread
+	let worker = Worker::new(crate::physics::Body::with_pva(crate::integer::Planar64Vec3::ZERO,crate::integer::Planar64Vec3::ZERO,crate::integer::Planar64Vec3::ZERO),
+		|_|crate::physics::Body::with_pva(crate::integer::Planar64Vec3::ONE,crate::integer::Planar64Vec3::ONE,crate::integer::Planar64Vec3::ONE)
+	);
 
-impl<'a,Task:Send+'a> INWorker<'a,Task>{
-	pub fn new<F:FnMut(Task)+Send+'a>(scope:&'a thread::Scope<'a,'_>,mut f:F)->INWorker<'a,Task>{
-		let (sender,receiver)=mpsc::sync_channel::<Task>(1);
-		let handle=scope.spawn(move ||{
-			loop {
-				match receiver.recv() {
-					Ok(task)=>f(task),
-					Err(_)=>{
-						println!("Worker stopping.",);
-						break;
-					}
-				}
-			}
-		});
-		Self{
-			sender,
-			handle,
-		}
-	}
-	//blocking!
-	pub fn blocking_send(&self,task:Task)->Result<(), mpsc::SendError<Task>>{
-		self.sender.send(task)
-	}
-	pub fn send(&self,task:Task)->Result<(), mpsc::TrySendError<Task>>{
-		self.sender.try_send(task)
-	}
-}
-
-#[cfg(test)]
-mod test{
-	use super::{thread,QRWorker};
-	use crate::physics;
-	use strafesnet_common::{integer,instruction};
-	#[test]//How to run this test with printing: cargo test --release -- --nocapture
-	fn test_worker() {
-		// Create the worker thread
-		let test_body=physics::Body::new(integer::vec3::ONE,integer::vec3::ONE,integer::vec3::ONE,integer::Time::ZERO);
-		let worker=QRWorker::new(physics::Body::ZERO,
-			|_|physics::Body::new(integer::vec3::ONE,integer::vec3::ONE,integer::vec3::ONE,integer::Time::ZERO)
-		);
-
-		// Send tasks to the worker
-		for _ in 0..5 {
-			let task = instruction::TimedInstruction{
-				time:integer::Time::ZERO,
-				instruction:strafesnet_common::physics::Instruction::Idle,
-			};
-			worker.send(task).unwrap();
-		}
-
-		// Optional: Signal the worker to stop (in a real-world scenario)
-		// sender.send("STOP".to_string()).unwrap();
-
-		// Sleep to allow the worker thread to finish processing
-		thread::sleep(std::time::Duration::from_millis(10));
-
-		// Send a new task
-		let task = instruction::TimedInstruction{
-			time:integer::Time::ZERO,
-			instruction:strafesnet_common::physics::Instruction::Idle,
+	// Send tasks to the worker
+	for _ in 0..5 {
+		let task = crate::instruction::TimedInstruction{
+			time:crate::integer::Time::ZERO,
+			instruction:crate::physics::PhysicsInstruction::StrafeTick,
 		};
 		worker.send(task).unwrap();
-
-		//assert_eq!(test_body,worker.grab_clone());
-
-		// wait long enough to see print from final task
-		thread::sleep(std::time::Duration::from_millis(10));
 	}
+
+	// Optional: Signal the worker to stop (in a real-world scenario)
+	// sender.send("STOP".to_string()).unwrap();
+
+	// Sleep to allow the worker thread to finish processing
+	thread::sleep(std::time::Duration::from_secs(2));
+
+	// Send a new task
+	let task = crate::instruction::TimedInstruction{
+		time:crate::integer::Time::ZERO,
+		instruction:crate::physics::PhysicsInstruction::StrafeTick,
+	};
+	worker.send(task).unwrap();
+
+	println!("value={}",worker.grab_clone());
+
+	// wait long enough to see print from final task
+	thread::sleep(std::time::Duration::from_secs(1));
 }

src/zeroes.rs

@@ -0,0 +1,32 @@
+//find roots of polynomials
+use crate::integer::Planar64;
+
+#[inline]
+pub fn zeroes2(a0:Planar64,a1:Planar64,a2:Planar64) -> Vec<Planar64>{
+	if a2==Planar64::ZERO{
+		return zeroes1(a0, a1);
+	}
+	let mut radicand=a1.get() as i128*a1.get() as i128-a2.get() as i128*a0.get() as i128*4;
+	if 0<radicand {
+		//start with f64 sqrt
+		let planar_radicand=Planar64::raw(unsafe{(radicand as f64).sqrt().to_int_unchecked()});
+		//TODO: one or two newtons
+		if Planar64::ZERO<a2 {
+			return vec![(-a1-planar_radicand)/(a2*2),(-a1+planar_radicand)/(a2*2)];
+		} else {
+			return vec![(-a1+planar_radicand)/(a2*2),(-a1-planar_radicand)/(a2*2)];
+		}
+	} else if radicand==0 {
+		return vec![a1/(a2*-2)];
+	} else {
+		return vec![];
+	}
+}
+#[inline]
+pub fn zeroes1(a0:Planar64,a1:Planar64) -> Vec<Planar64> {
+	if a1==Planar64::ZERO{
+		return vec![];
+	} else {
+		return vec![-a0/a1];
+	}
+}

tools/arcane

@@ -1 +0,0 @@
-mangohud ../target/release/strafe-client bhop_maps/5692113331.snfm

tools/bhop_maps

@@ -1 +0,0 @@
-/run/media/quat/Files/Documents/map-files/verify-scripts/maps/bhop_snfm

tools/cross-compile.sh

@@ -1 +0,0 @@
-cargo build --release --target x86_64-pc-windows-gnu --all-features

tools/iso

@@ -1 +0,0 @@
-mangohud ../target/release/strafe-client bhop_maps/5692124338.snfm

tools/make-demo.sh

@@ -1,4 +0,0 @@
-mkdir -p ../target/demo
-mv ../target/x86_64-pc-windows-gnu/release/strafe-client.exe ../target/demo/strafe-client.exe
-rm ../target/demo.7z
-7z a -t7z -mx=9 -mfb=273 -ms -md=31 -myx=9 -mtm=- -mmt -mmtf -md=1536m -mmf=bt3 -mmc=10000 -mpb=0 -mlc=0 ../target/demo.7z ../target/demo

tools/run

@@ -1 +0,0 @@
-mangohud ../target/release/strafe-client "$1"

tools/settings.conf

@@ -1,4 +0,0 @@
-[camera]
-sensitivity_x=98384
-fov_y=1.0
-#fov_x_from_y_ratio=1.33333333333333333333333333333333

tools/surf_maps

@@ -1 +0,0 @@
-/run/media/quat/Files/Documents/map-files/verify-scripts/maps/surf_snfm

tools/toc

@@ -1 +0,0 @@
-mangohud ../target/release/strafe-client bhop_maps/5692152916.snfm

tools/utopia

@@ -1 +0,0 @@
-mangohud ../target/release/strafe-client surf_maps/5692145408.snfm