An example repo that shows a simple game of Pong -- with all the tools needed to read and modify its inner workings.
Install LÖVE. It's just a 5MB download, open-source and extremely well-behaved.
Download this repo:
git clone https://git.sr.ht/~akkartik/pong.love pong
Run the game:
love pong
When you want to read or modify the game, open up its editor by pressing
ctrl+e
. (Check out more keyboard shortcuts below to see what you can do.)
When you're done reading or making changes, press ctrl+e
again to play Pong
with any changes you made.
At the start, press any key to start a game. Once a game starts:
a
/z
to move the left-hand paddle
up
/down
to move the right-hand paddle
space
to start a new game
ctrl+e
to stop playing and browse the code for the game
ctrl+e
to go back to playing Pong, using any changes you made
ctrl+g
to switch to a different file
ctrl+l
to toggle a second editor on the right for recent logs. Useful when
debugging.
ctrl+=
to zoom in, ctrl+-
to zoom out, ctrl+0
to reset zoom
On the left (source code) side:
ctrl+f
to find patterns within the current filectrl+z
to undo, ctrl+y
to redoalt+right
/alt+left
to jump to the next/previous word, respectivelyshift
+ movement to select text, ctrl+a
to select allctrl+e
to modify the sourcesExclusively tested so far with a US keyboard layout. If you use a different layout, please let me know if things worked, or if you found anything amiss: http://akkartik.name/contact
The primary method here to understand what Pong is doing (or indeed any other
program you choose to turn this repo into) is the log. To emit objects to the
log, use the log
function:
log(2, "log message")
log
takes exactly 2 arguments: a stack frame index to report the file and
line number for, and an object to log.
The stack frame index should be 2
if you call log
directly. If you create
higher levels of abstraction around log
, increment the stack frame
appropriately. Helper f
calling log
should use a stack frame index of 3
to report line numbers from its caller. Helper g
calling f
calling log
should use a stack frame index of 4
, and so on.
The log
function can only emit a single object at a time. However, the
object can be of any type.
Since each entry/line in the log contains a single object (after the filename
and line number), it's easy to specify how different objects should be
rendered. Just make sure the objects contain a field called name
, and create
a function log_render.name
to render it. In this repo, the function
log_render.state
is an example of such a function, specifying how to render
a Pong state consisting of a ball position, ball velocity and paddle
positions.
Rendering functions must all have the same interface as
log_render.state
:
function log_render.state(state, x,y, w)
...
end
The first argument is the object that was logged, the remaining arguments specify the starting position from which to render it, and the width on screen available to it. After the function finishes rendering the object, it should return the y coordinate it drew until, so that the log browser knows where to start rendering the next object.
While pong.love lets you make changes to its source and shows how your changes affect its behavior, the changes don't live in the same place as the original sources. They live in the app's save directory (love.filesystem.getSaveDirectory()).
To distribute your version you'll need to copy the modified files over into the app, either its directory or its .love zip archive. This process is [unfortunately not automated yet](See https://love2d.org/wiki/Game_Distribution).
This repo is a fork of lines.love, an editor for plain text where you can also seamlessly insert line drawings. Its immediate upstream is text.love, a version without support for line drawings. Updates to it can be downloaded from:
Further forks are encouraged. If you show me your fork, I'll link to it here.
Most appreciated. Messages, PRs, patches, forks, it's all good.