Open Source HTML5 Puzzle Game Engine
License: MIT License
Open Source HTML5 Puzzle Game Engine
Try it out at https://www.puzzlescript.net.
If you're interested in recompiling/modifing/hacking the engine, there is development setup info here. If you're just interested in learning how to use the engine/make games in it, the documentation is here.
E.g. [ Eyeball | Player ] -> [ ... | Corpse ]
It's interpreted by PuzzleScript as [ Eyeball | Player ] -> [ | Corpse ]
But looks like it might mean [ Eyeball | Player ] -> [ Eyeball | Corpse ]
Test project:
https://github.com/increpare/PuzzleScript/blob/master/js/engine.js#L1810
This line
if (loopCount > 00) {
should probably be
if (loopCount > 200) {
In puzzlescript capital letters aren't working with the MESSAGE command (for rules)
Is this intentional? Or is it a bug?
But why is it case insensitive in rules when upper case works in the messages between levels?
Not a bug but would be nice.
Unsure if it's easily doable with the way music is sourced from Youtube though.
fix before upploading
Here's another example of win/again weirdness:
http://www.puzzlescript.net/editor.html?hack=9232397
"again" seems to take priority over "win"!
Probably the behaviour I'd expect is:
When predicting a turn to see if again needs to fire, the win command is ignored (this is my theory for why there's no delay in my first example)
If win is fired, again is ignored even if it would trigger another turn.
from alan
This is a pretty stand-alone task, maybe someone else would be interested in having a look -
right now for the index page it pulls the last twenty and selects 3 at random. i don't know if it's possible but it'd be neat if someone could make it trawl 3 completely at random from its archives.
https://github.com/increpare/PuzzleScript/blob/master/index.html
Friend reported this, does seem to be the case that you get a blank page when trying to run a PuzzleScript game with cookies disabled.
The line "Late vertical [Crate | ... | Player] -> restart" seems to cause an uncaught type error in the rulesToMask function in compiler.js, specifically when handling "...".
This causes the rebuild and run buttons to not work at all, and the reason for the error is also not apparent at first due to there being no obvious error message.
I don't have a good repro for this yet. The whole editor page will lock up and no longer respond to mouse or keyboard input. It generally seems to happen while typing. I've seen it happen when typing an equals sign in a Legend line, and when typing a hash sign on a CollisionLayers line (I was trying to guess the comment syntax). The last time, I was lucky enough to have the debugging console open and could pause it and step through. The minified Javascript is a bit hard to decipher, but comparing it to the source I believe it was stuck in the loop in runMode. In this case, the content of the stream was just "#", and pos never moved beyond 0.
I'll follow-up if I find a reliable repro.
Here's a minimal repro. If you try to compile this buggy code, the compiler dies with an uncaught exception. It should probably have reported an error that Bug hasn't been assigned to a layer, but as far as I can tell it's gotten to the point where it's trying to report that Bug and Background can't appear on the same tile, but it's tripping up because it assumes that Bug has a layer and it doesn't.
title Uncaught TypeError compiling buggy code
author Weeble
========
OBJECTS
========
Background1
LIGHTGREEN
Player
Black
Bug
Red
=======
LEGEND
=======
P = Player
b = Bug and Background1 (This is bad. They shouldn't be on same tile.)
Background = Bug or Background1
=======
SOUNDS
=======
================
COLLISIONLAYERS
================
Background1 (Oops, we forgot to assign Bug to a layer.)
Player
======
RULES
======
==============
WINCONDITIONS
==============
=======
LEVELS
=======
Pb
You can't define an object called Moving, but you can define a synonym Moving = MovingL or MovingR and it won't error.
It will of course give you errors if you try using that definition in a rule.
If a message is extremely long and the first message of the game, it doesn't display at all; it just waits for "x". To see this in action, put the following at the start of the levels section:
message Hello there, good reader. I don't know what the maximum message length is, so I figured I'd test it by typing this great big message. That way, I could get an idea for what the maximum message length is. Luckily, it seems to be pretty big, but ultimately I wonder if it splits long messages across pages, or more likely, if long messages just don't work properly.
message test
Also, if a message is just a little bit too long, the message will overwrite the word "continue" a bit and then following long messages can behave weirdly. For an example:
message Hello there, good reader. I don't know what the maximum message length is, so I figured I'd test it by typing this great big message. That way, I could get an idea for what the maximum message length is. Luckily, it seems to be pretty big, but ultimately I wonder if it splits long messages across pages, or more likely, if long
message Hello there, good reader. I don't know what the maximum message length is, so I figured I'd test it by typing this great big message. That way, I could get an idea for what the maximum message length is. Luckily, it seems to be pretty big, but ultimately I wonder if it splits long messages across pages, or more likely, if long messages just don't work properly.
<div class="footer" > <a href="http://">www.puzzlescript.net</a>
If you undo/reset to a state when the level was differently sized, it gets corrupted.
If you try to export a game which uses beta features, you don't get a warning. This is probably fine since the exported code wouldn't ever change?
However, the exported game is still trying to display a warning message, by looking for an element with id "errormessage". There is no such element on the standalone page, so the javascript code errors, with the effect that it appears to be stuck on the loading screen.
(It's not actually stuck though, it's just not showing the title screen. If you press X it will start the game)
http://www.puzzlescript.net/editor.html?hack=6893137
Open the level editor. Save. I see this output:
""""""undefinedundefinedundefined"""
"h""""undefinedundefinedundefined"$"
""""""undefinedundefinedundefined"""
""""""undefinedundefinedundefined"""
iundefinedundefinedundefinediundefinedundefinedundefinedundefinedundefinedundefinedundefined
"undefinedundefinedundefined"undefinedundefinedundefinedundefinedundefinedundefinedundefined
i"""iundefinedundefinedundefinedundefinedundefinedundefinedundefined
"undefinedundefinedundefined"undefinedundefinedundefinedundefinedundefinedundefinedundefined
iundefined@undefinediundefinedundefinedundefinedundefinedundefinedundefinedundefined
http://www.draknek.org/games/puzzlescript/8.68-HACK.php
Every time you start a game, it makes a really quiet squeaking sound that's not anything like one of the sounds I've defined (none of which should be playing in any case)
Sorry I haven't made a nice simple test-case here!
when I share my puzzlescript game this is the most commonly requested thing - no idea how feasible it is, but thought I'd suggest it at least.
I appreciate that the following may well fall into the category of "won't fix" :) as the general case-insensitivity of PuzzleScript is clearly a feature.
I'm working on a new game involving different animals. There's a marker to show the player which animal they are currently allowed to move. An extract from my current legend is this:
D = Dog and Floor and Marker
C = Cat and Floor and Marker
M = Mouse and Floor and Marker
d = Dog and Floor
c = Cat and Floor
m = Mouse and Floor
Sadly, this doesn't work, as 'm' and 'M' are treated the same, so, for example, on the cat and mouse level, where I want a controllable cat and uncontrollable mice, all my mice still end up with markers even if I've specified them with 'm'.
If legend characters were case-insensitive, it would be easier to create these levels directly, as I could use 'M' to denote 'Mouse and Floor and Marker' (controllable mouse) and 'm' to denote 'Mouse and Floor' (uncontrollable mouse).
Even if legends will always be case-insensitive, though, specifying eg both 'M' and 'm' in a legend ought to trigger some kind of 'Things Will Not Behave As You Expect' error message, since they won't.
Cheers,
Wayne
Attempting to pair the "cancel" keyword with a "sfx" keyword gives the following error: "The CANCEL command can only appear by itself on the right hand side of the arrow."
Since "cancel" is often used to block illegal moves, it seems like it would be useful to allow some kind of feedback sound to alert the player. Why not allow "sfx" (and maybe even "message") to be the exception to this rule?
Thanks!
Looked into the specifics of this, it's because editing the console's innerHTML several hundred times in quick succession is slow. If the log entries are queued up and added all at once at the end of every turn, it should be totally useable.
http://www.puzzlescript.net/editor.html?hack=9075300
[ Player ] -> [ ]
[ Spawn ] -> [ Spawn Crate ]
Open the example in the editor then start the level editor. Select the white object (Background2) and paint some of it onto the map. Press the S button to save the level.
The regions you painted should be saved as the symbol "o", but they are in fact saved as the symbol "x". It's definitely related to having multiple backgrounds, but it doesn't always seem to happen solely from having multiple backgrounds.
My suspicion is that it's a bug in levelEditorClick. It looks like it tries to mix in the default background tile with every tile it saves, regardless of whether that tile already has a background tile. But I'm not entirely sure I understand the binary encoding used for tiles, nor how backgroundid relates to the objects you've declared as being background, so take that with a pinch of salt.
title Level editor background bug
author Weeble
homepage www.puzzlescript.net
Background1
DARKGREY
Background2
WHITE
Player
RED
X = Background1
O = Background2
P = Player
Background = Background1 or Background2
Background
XXXOOP
XXXOOO
XXXOOO
I tried to use the Action action as a shortcut to skip levels (useful for prototyping), so have the rule:
[Action Player] -> WIN
However, since I also have require_player_movement as a prelude, when I hit the Action key, the win condition is not triggered as the Action action is cancelled (due to non-movement). Logs:
require_player_movement set, but no player movement detected, so cancelling turn.
An ugly work-around is to change the rule to include movement, eg:
[Action Player ] -> [> Player] WIN
I think it would be nice if Action could be considered a 'movement' or have another prelude that will encompass this (eg. require_player_movement_or_action)
If you have a bunch of rules which could cancel a turn, you can't tell which one it is
In Firefox, hack links work fine. In Chrome the link just goes to "#", and I get this error in the console:
Uncaught ReferenceError: Mobile is not defined
[ Red ] -> [ down A up B ]
[ up A up B ] -> [ Blue ]
An example of the resulting malformed link: <a href="http://">www.puzzlescript.net</a>
It looks like the relevant portion from standalone.html isn't making it into the standalone_inlined template, as can be seen from the snippets below:
standalone.html: <a href="http://"+'"__HOMEPAGE__"'>"__HOMEPAGE__"</a>
standalone_inlined.txt: <a href="http://">"__HOMEPAGE__"</a>
I'd offer to post a fix, but I have no idea how the inlining works.
Attempting to use 'V' as a single-character descriptor in a legend results in the following error message:
"names cannot end with the letter "v", because it's is used as a direction."
I can vaguely see why this might be reasonable behaviour in object names, which must be parsed in conjunction with 'v' as direction in rules.
Harder to see why the ban on 'V' extends to the legend single-character descriptors used to construct the levels.
The workaround is of course simple - don't use 'V' in legends.
It would be nice if there was a way to disable collisions on objects, or allow more layers. I'm probably pushing it, but I'm trying to implement a lot of mechanics in hidden objects on different layers, I don't want these objects to bump into each other.
I don't know what the engine looks like internally, but would it be hard to have one "no-collide" layer?
When using realtime_interval it would be convenient if player movement was buffered until the next automatic turn rather than forcing a new independent turn.
I know this is possible to emulate by [ stationary Player ] in front of all your rules, but it's quite a pain, and I can't imagine a use case that'd be broken by changing this behaviour.
Exported version of a game does not behave in the same way as the online version.
See eg http://www.puzzlescript.net/play.html?p=6850739
That works fine, both at the URL above and in the editor.
However, when I hit export, save the file locally, then run it from my local server, the local version works fine for the first couple of levels, which do not contain any moving NPCs, but seems to ignore most keypresses on the third level (and probably beyond), which does.
The first time you open the editor, the list of tiles at the top doesn't appear. You have to close and re-open edit mode to see them.
This also happens when you resize the level and that makes the game's scale change.
Try this simple level:
http://www.puzzlescript.net/editor.html?hack=6943974
The rules should always draw the hat on top of the player, but it doesn't appear a lot of the time, and it seems to be related to where the player has previously moved.
I've already remarked this on the forum, but seems to make sense to create an issue for it while I'm here.
This message is normally a sign that your code is doing something foolish, but it's often hard to work out exactly what's causing it. If the warning mentioned the rule id that was looping too much it would be very convenient!
doesn't look like key_repeat_interval's being reset after the last again
There seems to be a bug with sound and again. With verbose_logging enabled, it seems to be undoing moves spontaneously and replaying them again, causing the sounds to occur twice in some circumstances. I've mitigated this to a degree by using SFXn instead.
When an object with movement is removed, another object placed at the same position will inherit the removed object's movement.
Here's an example to demonstrate:
http://www.puzzlescript.net/editor.html?hack=7078110
The replacement doesn't have to happen in the same rule – the orphaned movement will stick around until the end of the movement phase unless it's explicitly removed, which can make debugging quite confusing.
I think it's an overall improvement to always remove movement along with objects. Thus:
[ OldObject ] -> [ ]
should behave as if it were:
[ > OldObject ] -> [ ]
[ OldObject ] -> [ ]
If you really want the existing behaviour instead, you can acheive that with:
[ > OldObject ] -> [ > NewObject ]
(which in my opinion is a little nicer anyway, since it's more explicit about what's happening).
see if you can make it happen at a rule group level rather than a per-rule level
If I start a PuzzleScript game and hold an arrow key and alt-tab to a different window, depress the key, then alt-tab back, PuzzleScript thinks the key is still pressed.
It will keep triggering events until you press the arrow again, so PuzzleScript receives the "keyup". Unfortunately that doesn't always work.
I've tested this with Firefox and Chrome on several games, e.g. http://www.puzzlescript.net/play.html?p=6851495 (Just start the game and press the right arrow)
It appears that the norestart command doesn't successfully prevent the user from restarting when they press R.
The W key gets stuck when I hit Command-W to close the tab in Firefox on Mac OS X. Not only does the tab not close, but also W (up) is continuously sent to the game until I press and release W again. Try it out when playing a game – the character in that game will continuously jump when you hit Command-W.
This bug does not happen in Chrome for Mac. I have not tried other browsers or other keyboard shortcuts.
Related to #38.
I've found a problem where the level is ending without a win condition. It seems to occur if the Player is removed on the same turn a message is displayed and an again is triggered. No "Win Condition Satisfied" message is displayed but the game moves to the next level.
Here's an example:
http://www.puzzlescript.net/editor.html?hack=9124914
The win condition is "no Trigger". There is no rule that ever removes the Trigger, but the level ends when you stand on it.
http://www.puzzlescript.net/play.html?p=6963661
music doesn't play on safari, but does on chrome (on osx)
on safari, I can't play the youtube frame at all, even loading the url directly
https://youtube.googleapis.com/v/guG9cVs3ms4?autoplay=1&loop=1&playlist=guG9cVs3ms4
i don't know if there's a way to get it to work without breaking everything else...
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.