# bowntz

audio-visual pseudo-physical simulation of colliding circles https://code.mathr.co.uk/bowntz

Latest on Hackage: | 1 |

This package is not currently in any snapshots. If you're interested in using it, we recommend adding it to Stackage Nightly. Doing so will make builds more reliable, and allow stackage.org to host generated Haddocks.

GPL-3 licensed by

**Claude Heiland-Allen**Maintained by

**claude@mathr.co.uk**Bowntz -- an audio-visual pseudo-physical simulation of colliding circles

Copyright (C) 2010,2013,2015,2016 Claude-Heiland-Allen <claude@mathr.co.uk>

Build

-----

$ ghc -O2 -Wall --make bowntz.hs

Bowntz requires hsc3 >= 0.8 (newer than Hackage, as of 2010-03-12).

Usage

-----

$ scsynth -u 57110 &

-- wait for scsynth to start, connect its JACK ports --

$ ./bowntz

Bowntz requires a "clean" SC3 server to be running on its usual port.

Bowntz leaves this SC3 server in an "unclean" state when it exits.

This situation is far from ideal, but it's the current state of the code.

Notes

-----

The world is a collection of N circles which each move at a constant speed

in a straight line (in the absence of collisions). Given two such circles

it is possible to calculate the precise time of their next collision. For

N circles, O(N^2) such "next" collisions may occur.

Pick the soonest next collision between circles A and B. This collision

will change the velocities of A and B, but no other circle. Update the

collection of collisions, by removing all the other "next" collisions

involving either A or B, and computing the new collisions of A and B with

all the other circles (O(N) collisions). For maintaining a consistent

time-base for all circles in the world, it makes sense to advance in a

straight line each circle to its position at the time of the collision.

Given a world at time T, to find the world at time T+dT simply step

through the future collisions one by one until the soonest collision is

after T+dT, at which point one can just advance the circles in straight

lines to T+dT.

T T+1 T+2 T+3 T+4 T+5 T+6 T+7

... | | | | | | | | ...

----*---*---*---------*---------------**-*--*-*-------*------*--

Care needs to be taken in case of numerical imprecision: one can get an

infinite number of collisions without time increasing.

To do

-----

General code clean-up and documentation.

Optimizations based on collision possibility tree, example:

A 7 possible

| A~*~B

+---B A~*~E 5 impossible

| +---C B~*~C !A~*~C

| +---D B~*~D !A~*~D

| C~*~D !C~*~F

+---E B~*~E !D~*~F

+---F E~*~F !A~*~F

data Node = Node ID [Node]

possibles n@(Node _ ns) = possible n `map` ns ++ liftM2 possible ns

Infinite zoom into circles within circles within circles within circles?

When (if ever) would it be safe to prune "outside"?

Some kind of loop where circles contain themselves (I did try this in

another unreleased version but it resulted in hyperexponential kinetic

energy increase, which naturally caused problems).

Non-flat space.

Angular momentum affecting collisions.

Remove all sources of non-determinism to enable two-pass rendering:

1 pass to record audio from SC3 in realtime

2 pass to record video from OpenGL in non-realtime

Or just record ball positions in the first pass for later drawing (because

callback races and/or timing instability might affect the reproducibility

unless very precise care is taken at all stages..)

Interactivity.

More impressive audio-visuals.

Won't

-----

Anything other than circles (ok, maybe infinite lines would be ok, but

asymmetry with angular momentum would quickly make collision detection

get horribly complicated).

Legal

-----

License: GPL

This program is free software: you can redistribute it and/or modify

it under the terms of the GNU General Public License as published by

the Free Software Foundation, either version 3 of the License, or

(at your option) any later version.

This program is distributed in the hope that it will be useful,

but WITHOUT ANY WARRANTY; without even the implied warranty of

MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

GNU General Public License for more details.

You should have received a copy of the GNU General Public License

along with this program. If not, see <http://www.gnu.org/licenses/>.

Copyright (C) 2010,2013,2015,2016 Claude-Heiland-Allen <claude@mathr.co.uk>

Build

-----

$ ghc -O2 -Wall --make bowntz.hs

Bowntz requires hsc3 >= 0.8 (newer than Hackage, as of 2010-03-12).

Usage

-----

$ scsynth -u 57110 &

-- wait for scsynth to start, connect its JACK ports --

$ ./bowntz

Bowntz requires a "clean" SC3 server to be running on its usual port.

Bowntz leaves this SC3 server in an "unclean" state when it exits.

This situation is far from ideal, but it's the current state of the code.

Notes

-----

The world is a collection of N circles which each move at a constant speed

in a straight line (in the absence of collisions). Given two such circles

it is possible to calculate the precise time of their next collision. For

N circles, O(N^2) such "next" collisions may occur.

Pick the soonest next collision between circles A and B. This collision

will change the velocities of A and B, but no other circle. Update the

collection of collisions, by removing all the other "next" collisions

involving either A or B, and computing the new collisions of A and B with

all the other circles (O(N) collisions). For maintaining a consistent

time-base for all circles in the world, it makes sense to advance in a

straight line each circle to its position at the time of the collision.

Given a world at time T, to find the world at time T+dT simply step

through the future collisions one by one until the soonest collision is

after T+dT, at which point one can just advance the circles in straight

lines to T+dT.

T T+1 T+2 T+3 T+4 T+5 T+6 T+7

... | | | | | | | | ...

----*---*---*---------*---------------**-*--*-*-------*------*--

Care needs to be taken in case of numerical imprecision: one can get an

infinite number of collisions without time increasing.

To do

-----

General code clean-up and documentation.

Optimizations based on collision possibility tree, example:

A 7 possible

| A~*~B

+---B A~*~E 5 impossible

| +---C B~*~C !A~*~C

| +---D B~*~D !A~*~D

| C~*~D !C~*~F

+---E B~*~E !D~*~F

+---F E~*~F !A~*~F

data Node = Node ID [Node]

possibles n@(Node _ ns) = possible n `map` ns ++ liftM2 possible ns

Infinite zoom into circles within circles within circles within circles?

When (if ever) would it be safe to prune "outside"?

Some kind of loop where circles contain themselves (I did try this in

another unreleased version but it resulted in hyperexponential kinetic

energy increase, which naturally caused problems).

Non-flat space.

Angular momentum affecting collisions.

Remove all sources of non-determinism to enable two-pass rendering:

1 pass to record audio from SC3 in realtime

2 pass to record video from OpenGL in non-realtime

Or just record ball positions in the first pass for later drawing (because

callback races and/or timing instability might affect the reproducibility

unless very precise care is taken at all stages..)

Interactivity.

More impressive audio-visuals.

Won't

-----

Anything other than circles (ok, maybe infinite lines would be ok, but

asymmetry with angular momentum would quickly make collision detection

get horribly complicated).

Legal

-----

License: GPL

This program is free software: you can redistribute it and/or modify

it under the terms of the GNU General Public License as published by

the Free Software Foundation, either version 3 of the License, or

(at your option) any later version.

This program is distributed in the hope that it will be useful,

but WITHOUT ANY WARRANTY; without even the implied warranty of

MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

GNU General Public License for more details.

You should have received a copy of the GNU General Public License

along with this program. If not, see <http://www.gnu.org/licenses/>.

comments powered byDisqus

A service provided by
FP Complete