README.md

Pacman, a naive implementation of the classic game in C++

Original document location.

Git Repository

This project’s canonical repositories is hosted on Gothel Software.

Overview

This project demonstrates a complex system written in modern C++ for our computer science class.

We have chosen the classic game for our project to pay homage to Toru Iwatani’s original Puckman.

Besides management of animated sprite graphics, maze environment and tile positioning, animation speed synchronization, the interesting part might be the ghost’s state machine and their movements.

Potential code sections of interest

Coverage

Chase, Scatter and Phantoms – Sisyphus alike Ghost Fate is a little article of ours describing details and includes screenshots and a short video recording.

Previous Work

To implement the original Puckman game behavior like weighted tile collision, ghost algorithm, etc. - we have used the following documents for reference

We have followed the references closely to achieve a most accurate default mode to match the original, see Deviations from the Original below.

Other Implementations

There exist other free software implementations of this classic game, which we have not analyzed yet (2022-04-03):

License and Copyrights

This project is for educational purposes, intended to be distributed in source form only and is licensed under the MIT license.

This project is authored by Sven Gothel and his son Svenson Han Gothel. Its copyright is held by Gothel Software e.K since 2022.

Namco holds the copyright on the original Puckman since 1980.

Supported Platforms

C++17 and better where the SDL2 library is supported.

Building Binaries

The project requires make, g++ >= 8.3 and the libsdl2 for building.

Installing build dependencies on Debian (11 or better):

apt install git build-essential g++ gcc libc-dev libpthread-stubs0-dev make
apt install libsdl2-dev libsdl2-image-dev libsdl2-ttf-dev libsdl2-mixer-dev

For a generic build use:

git clone https://jausoft.com/cgit/cs_class/pacman.git
cd pacman
make

The binary shall be build to bin/pacman.

Usage

Commandline Arguments

Following commandline arguments are supported

  • -2p to enable 2nd player controlling Blinky when chasing, scattering or scared using W, A, S and D for up, left, down and right.
  • -audio to turn on audio effects, i.e. playing the audio samples.
  • -pixqual <int> to increas pixel filtering quality, i.e. 0 = nearest (default), 1 = linear and 2 = anisotropic filtering.
  • -no_vsync to force off hardware enabled vsync, which in turn enables manual fps synchronization
  • -fps <int> to enforce a specific fps value, which will also set -no_vsync naturally
  • -speed <int> to set the 100% player speed in fields per seconds
  • -wwidth <int> to set the initial window width
  • -wheight <int> to set the initial window height
  • -show_fps to show pacman’s speed statistics and periodically show frames per seconds (fps) on the console
  • -show_modes to show all players’ mode changes
  • -show_moves to show all players’ move criteria like distance and collisions incl. speed changes on the console
  • -show_targets to show the ghost’s target position as a ray on the video screen
  • -show_debug_gfx to show all debug gfx
  • -show_all enable -show_modes, -show_moves, -show_targets and -show_debug_gfx
  • -no_ghosts to disable all ghosts (testing)
  • -invincible to make pacman invincible (testing)
  • -bugfix to turn off the original puckman’s behavior (bugs), see Bugfix Mode below.
  • -decision_on_spot to enable ghot’s deciding next turn on the spot with a more current position, otherwise one tile ahead.
  • -dist_manhatten to use the Manhatten distance function instead of the Euclidean default
  • -level <int> to start at given level
  • -record <basename-of-bmp-files> to record each frame as a bmp file at known fps, see Video Recording Example below.
bin/pacman [-2p] [-audio] [-pixqual <int>] [-no_vsync] [-fps <int>] [-speed <int>] [-wwidth <int>] [-wheight <int>] [-show_fps] [-show_modes] [-show_moves] [-show_targets] [-show_debug_gfx] [-show_all] [-no_ghosts] [-bugfix] [-decision_on_spot] [-dist_manhatten] [-level <int>] [-record <basename-of-bmp-files>]

Keyboard Input

  • General Control
    • End programm: Q or ESCAPE
    • Pause: P
    • Reset: R
    • Fullscreen: F
    • Snapshot: F12 (saved as puckman-snap-abcd.bmp>)
  • Player Control
    • Up: UP or W
    • Left: LEFT or A
    • Down: DOWN or S
    • Right: RIGHT or D

Video Recording Example

The basename for the resulting bmp files is passed after the -record option tag and may contain folder names.

The resulting bmp files may be converted to video using scripts/bmps_to_mp4.sh.

We assume we are in the project folder having bin/pacman build and created a video folder.

  • Use the first command to start the game while recording bmp snapshots each frame to video/puckman-01-*.bmp.
  • Use the second command to convert same files to video/puckman-01.mp4.
  • Use the third command to delete the bmp files
  • Use the forth command to play the video with mpv
bin/pacman -record video/puckman-01 -wwidth 1044 -wheight 1080 -show_targets -show_debug_gfx
scripts/bmps_to_mp4.sh video/puckman-01
rm video/puckman-01*bmp
mpv video/puckman-01.mp4

Deviations from the Original

While we have focused on implementing the original behavior most accurately, some aspects are not fully in our control and are discussed here.

PRNG Scared Mode

The original uses a known seeding iteration, used as a memory address within the game rom. The last bits of the addressed rom byte represents the PRNG value for the ghost’s direction in scared mode.

Since the roms can’t be used, we use the PRNG value of another algorithm with same seeding sequence to preserver the periodic attrbutes at least.

Frame Sync Timing

This implementation uses the monitor’s frames per second number to approximate the sub-tile step-width for the desired tiles per frame pace.

This is achieved via Keyframe interval for animation, see keyframei_t.

Below we added measurements from pacman via commandline argument -show_fps running along the bottom longest line from collision to collision.

For each level we measured the slower first walk eating pellets and the second walk faster without pellets.

Level 1: Fields/s deviation on longest line measured 7.955 f/s to 8.000 f/s or 0.5635%

bin/pacman -no_ghosts -show_fps -level 1

[    8,471] pacman stats: speed 0.71%, td 3515ms, fields[25.00 walked, actual 7.112/s, requested 7.100/s, diff 0.1743%], fps[draw 60.03/s, tick 56.90/s], frames[draw 211, synced 11], [fps 60, frames/field 8, fields/s 7.500000/7.100000 (diff 0.400000, 3.200001f/s, 0.938969 ms, sync 19/f), center 0.500000], [26.500000/32.500000 26/32, last[dir R, collided 1], [walked[25.000000, 25], center 25, entered 25]]

[   11,872] pacman stats: speed 0.80%, td 3017ms, fields[24.00 walked, actual 7.955/s, requested 8.000/s, diff 0.5635%], fps[draw 59.99/s, tick 56.02/s], frames[draw 181, synced 12], [fps 60, frames/field 7, fields/s 8.571428/8.000000 (diff 0.571428, 3.999998f/s, 1.190477 ms, sync 15/f), center 0.428571], [1.428571/32.428570 1/32, last[dir L, collided 1], [walked[23.999989, 23], center 24, entered 24]]

Level 5: Fields/s deviation on longest line measured 8.772 f/s to 8.700 f/s or 0.8267%

bin/pacman -no_ghosts -show_fps -level 5

[    8,270] pacman stats: speed 0.87%, td 2850ms, fields[25.00 walked, actual 8.772/s, requested 8.700/s, diff 0.8267%], fps[draw 60.00/s, tick 52.63/s], frames[draw 171, synced 21], [fps 60, frames/field 6, fields/s 10.000000/8.700000 (diff 1.300000, 7.800001f/s, 2.490423 ms, sync 8/f), center 0.500000], [26.500000/32.500000 26/32, last[dir R, collided 1], [walked[24.999977, 25], center 25, entered 25]]

[   11,004] pacman stats: speed 1.00%, td 2400ms, fields[23.83 walked, actual 9.931/s, requested 10.000/s, diff 0.6945%], fps[draw 60.00/s, tick 60.00/s], frames[draw 144, synced 0], [fps 60, frames/field 6, fields/s 10.000000/10.000000 (diff 0.000000, 0.000000f/s, 0.000000 ms, sync -1/f), center 0.500000], [1.500000/32.500000 1/32, last[dir L, collided 1], [walked[23.833315, 23], center 24, entered 23]]

Below 3% deviation should be good enough for this game using >= 10% speed differences in modes.

Below 1% deviation is a great match.

Bugs

This software may have bugs :)

Optional Deviations from the Original

Decision on the Spot

As stated in The Pac-Man Dossier, the ghosts select their next direction one tile ahead of an intersection.

With the -decision_on_spot mode enabled, see Commandline Arguments above, the ghosts decide their next direction for each field using a more current pacman position.

The default setting is to use the original one tile ahead of an intersection.

Bugfix Mode

With the -bugfix mode enabled, see Commandline Arguments above, the original puckman behavior (bugs) are disabled.

The list below shall be updated in case we further the implementation and kept in sync with include/pacman/globals.hpp.

By default the original pacman behavior is being implemented:

  • weighted (round) tile position for collision tests
  • Pinky’s up-target not 4 ahead, but 4 ahead and 4 to the left

If false, a more accurate implementation, the pacman bugfix, is used:

  • pixel accurate tile position for collision tests
  • Pinky’s up-target to be 4 ahead as intended

Manhatten Distance

With the -dist_manhatten mode enabled, see Commandline Arguments above, the Manhatten distance function is used instead of the Euclidean default.

Second Player Mode

With the ‘-2p’ mode enabled, see Commandline Arguments above, a 2nd player can control Blinky when chasing, scattering or scared.

Testing Options

The following testing options are available

  • -no_ghosts to disable all ghosts (testing)
  • -invincible to make pacman invincible (testing)

Implementation Status

Done

  • Maze
    • Read maze specification from text file
  • Sprites
  • Speed per tile accurate key-frame animation
    • Weighted tile position from floating position
    • Renderer fps derived step width, i.e. sub-tiles
    • Sync speed by dropping tick, every n-frames
  • Ghost AI
    • Scatter, chase and phantom targets
    • Scared RNG target
      • PRNG with an identical seed value every new level and life for predictable results
    • Next direction_t algo
      • No turning up in Red-Zones if chasing or scattering
      • Optionally use alternative Manhatten distance function instead of the Euclidean default.
    • Grouped wave switch of scatter, chase and frightened
    • Exit Home
      • Using local and global pellet timer
      • Additional maximum time when no pellets are eaten
    • Speed
      • Adjust tunnel speed
      • Elroy 1+2 mode
  • Pacman
    • show eaten ghosts and fruit score onscreen (FREEZE)
    • Freeze pacman only for 3 frames after eating power pellet
  • Level specification (per level)
    • Timings
      • scatter and chase duration per phase
      • frightening/powered duration
    • Speed
      • pacman normal on empty tile or eating
      • pacman powered on empty tile or eating
      • ghost speed normal, frightening or in tunnel
    • Score
  • Score
    • ghost 1-4 per power pellet
    • Fruits / Bonus
  • Sound
    • Using chunks (wav/mp3), mixed from seperated channels
    • Use lossy formats where applicable
  • Persistent game state
    • Snapshot (screenshot)
    • Video recording via bmp snapshots each frame
  • Extension
    • Second player controls Blinky when chasing, scattering or scared

To Do

  • Pacman
    • Lives
  • Sound
    • Complete samples
  • Maze
    • Render maze itself from maze-spec file
  • Persistent game state
    • Save/load game state

Media Data

The pixel data in media/playfield_pacman.png and media/tiles_all.png are copied from Arcade - Pac-Man - General Sprites.png, which were submitted and created by Superjustinbros and assumed to be in the public domain.

Audio samples have been edited using any of the following sources

The included audio samples are incomplete by design, only demonstrating utilizing sound from a coding perspective using loops and concurrently mixed channels.

Changes

1.0.0

  • Reached clean code demo stage with functional implementation.

0.2.0

  • Stability, next_target bug fix and alternative distance function

0.1.0

  • Usable working state with most ghost algorithms for further analysis and demonstration

0.0.1

  • Initial commit with working status and prelim ghost algorithm.
Конвейеры
0 успешных
0 с ошибкой