As was foretold, we've added advertisements to the forums! If you have questions, or if you encounter any bugs, please visit this thread: https://forums.penny-arcade.com/discussion/240191/forum-advertisement-faq-and-reports-thread/

Let's Make a Game: Platformer Edition [Playable Alpha (Win/Mac)! See page 8]

ZackSchillingZackSchilling Registered User regular
edited September 2010 in Games and Technology
The demo:
http://www.youtube.com/watch?v=twI6I0FUHkw

Old demos:
Old Files (Will be updated soon):
Google Code page is set up: http://code.google.com/p/project906/

This page holds an old version of the game that uses a software rendering engine. The OpenGL rewrite was so extensive that I had to basically recreate the project from scratch. These links are for historical purposes only.

Mac OS X Binary: http://project906.googlecode.com/files/platformer-osx-a3.zip
Windows Binary: http://project906.googlecode.com/files/platformer-win-a3.zip
Data Files (REQUIRED TO RUN): http://project906.googlecode.com/files/platformer-data-a3.zip
Alpha 3 src: http://project906.googlecode.com/files/platformer-src-a3.zip

The back story:

I've always wanted to create a platformer. I've been brewing ideas for the past year or so of how I want my very own game to play, what its gimmick would be, and various other details. It would control like a mix between Mario and Sonic. Gameplay would be a mixture of platforming and beat-em-up combat. The twist would be a built-in rhythm gameplay engine driving combos, attack strength, and jump height.

Last week, I stopped thinking and started doing. I got Pygame installed on my laptop and just started developing. First I wrote up a parallax background scroller, then movement/collision detection. Eventually, I got used to the development environment and wrote up some basic physics, control code, USB pad support.

Alpha 1: I now have the basics of a platformer. There's a little man and there's a level. He can move and jump.
Alpha 2: He doesn't get stuck in the walls anymore.
Alpha 3: He can also do other things, like punch, kick, stomp the ground, hit things with his head, and throw spiked balls. There are enemies that hop around and hunt the player.
Alpha 4: There are spikes and exploding balloons and a trigger system that allows for scripted events in levels. There is a camera that moves, zooms and tilts. The levels are rendered in full 3D

Picture-1.jpg

There's also a rhythm engine hidden somewhere in the back end, just to make sure it all keeps running as a develop things. Who knows if this will ever work out to be the game I imagined. It probably won't at all. I'm ok with that.

What this thread is:

I'd like to gauge interest in the game, get feedback on how it feels at this early stage, and, ideally, recruit some people for this project, no development experience necessary. I'm looking for a few people who would like to design some levels, create some sprites, etc. If someone wants to help with programming, just let me know. I think I'm all set but, who am I kidding, I'd love for someone to swoop in and whip up the perfect collision detection system.

Picture-2.jpg

This is strictly just for fun. If the game ever gets to the point where it would be a compelling experience for others to play, you get your due credit and it goes up in a released form for free. If not, that's just how it goes.

There are always threads that are like: "Hey guys, let's make a game!" and then they never go anywhere. This has already gone somewhere. There is a running engine packaged into executables, ready for new level designs, sprites, and (you'll have to work with me on this one) animations. Sure, it's early and in-progress. You have to start somewhere. Get in on the ground floor.

How to start:

First, download the game and try it. There are no enemies or goals or interface to speak of. It just is. Take it in. If it's running like crap or you'd like to change the demo level or controls, open up the data folder and poke around with "conf.txt" and "controls.txt" Windows users, you may need to use Wordpad to edit the files because notepad could choke on their linebreak format.

Next, if you're interested in levels or sprites, you'll need a graphics editing program that can handle transparent PNGs and, for your sanity, layers. I use Photoshop but the gimp will do in a pinch. Anything that can make a legitimate transparent PNG will do.

Open the data folder, open up levels, and pick one of the three examples (vida, testbig, palevel). There are 6 files that make up each level. "back.png", "front.png" and "rear.png" are just graphics with no bearing on gameplay. Front is the layer of the level over your character. Back is the layer immediately behind. Front and back should have the same pixel dimensions. Rear is the panel that sits in the back and gives the illusion of depth. It should be smaller than front and back but larger than the screen size.

"ground.png" is the ground surface, the places the player can walk. "mask.png" represents the walls the player will bounce off of. Play the game and inspect the examples and you'll get a much better idea.

Picture-3.jpg
This picture shows ground and walls overlayed in Photoshop.

"props.txt" determines what song plays and at what coordinates the player starts. Eventually, this will flesh out to do more.

The details:

The Controls
Left/Right arrows: Movement
Up arrow: Jump
Left Ctrl: Run
Tab: Spawn a basic character entity called "Jumper"
Space: Rhythm code interface
Esc: Quit

More pictures:
Picture-4.jpg
Picture-5.jpg
Picture-6.jpg

ghost-robot.jpg
ZackSchilling on
«13456

Posts

  • urahonkyurahonky Registered User regular
    edited January 2009
    Oh man. I may have to mess with this some... Looks fun. :)

    urahonky on
  • UberEvUberEv Registered User regular
    edited January 2009
    I enjoyed it as well. I'm just a bit worried about the file sizes required for these maps. With such large background files, you are going to be eating up a lot of space. Very cool to play though.

    UberEv on
  • ZackSchillingZackSchilling Registered User regular
    edited January 2009
    A quick note: the jittery collision detection on the walls is a temporary fix for a bug I just noticed before posting. It's a real ugly bandage on a fundamental problem with sprite ejection. I'm working on it so for now, just understand that it's not supposed to be like that. (It used to be silky smooth :( )

    ZackSchilling on
    ghost-robot.jpg
  • FremFrem Registered User regular
    edited January 2009
    $ python main.py 
    Traceback (most recent call last):
      File "main.py", line 1388, in <module>
        if __name__ == '__main__': main()
      File "main.py", line 1149, in main
        gameinfo = GameInfo()
      File "main.py", line 757, in __init__
        self.background = ScrollingBG(self)
      File "main.py", line 1060, in __init__
        self.ground.mask = pygame.mask.from_surface(self.ground.image, 127)
    AttributeError: 'module' object has no attribute 'mask'
    

    Running Ubuntu 8.04, Python 2.5.2, PyGame 1.7.1

    Edit: suggestion - the file size would go down and levels would be easier to edit if you either pulled a Sonic the Hedgehog and used a lot of sloped/curved tiles, or if you saved the levels in SVG format or something.

    Frem on
  • ValkunValkun Registered User regular
    edited January 2009
    Interesting demo. Have you put any thought into in game "terrain" creation/destruction effects? Tiles would largely defeat the feel of the game. With the physics cleaned up, I could see this being a pretty fun platformer.

    I guess I should go back to working on my own game so it doesn't become one of those projects that never go anywhere. :lol:

    Valkun on
  • vegeta_666vegeta_666 CanadaRegistered User regular
    edited January 2009
    Shotgun Sunrise would like to have a word with you.

    However, this looks good and colour me interested.

    vegeta_666 on
    Sob24Nm.png
    Steam: abunchofdaftpunk | PSN: noautomobilesgo | Lastfm: sjchszeppelin | Backloggery: colincummings | 3DS FC: 1392-6019-0219 |
  • dolemelangedolemelange Registered User regular
    edited January 2009
    Wow looks good man, very interesting idea - community platformer!

    dolemelange on
  • ZackSchillingZackSchilling Registered User regular
    edited January 2009
    Frem wrote: »
    $ python main.py 
    Traceback (most recent call last):
      File "main.py", line 1388, in <module>
        if __name__ == '__main__': main()
      File "main.py", line 1149, in main
        gameinfo = GameInfo()
      File "main.py", line 757, in __init__
        self.background = ScrollingBG(self)
      File "main.py", line 1060, in __init__
        self.ground.mask = pygame.mask.from_surface(self.ground.image, 127)
    AttributeError: 'module' object has no attribute 'mask'
    

    Running Ubuntu 8.04, Python 2.5.2, PyGame 1.7.1

    Edit: suggestion - the file size would go down and levels would be easier to edit if you either pulled a Sonic the Hedgehog and used a lot of sloped/curved tiles, or if you saved the levels in SVG format or something.

    I'm still thinking about the level format. As impractical as giant PNGs are, it's the simplest way to do what I want. Also, Sonic had a continuous listing of the slope at each pixel.

    You need to upgrade to pygame 1.8.x. 1.7 doesn't support collision masks, a major feature I use.

    Edit: Make that a major feature I don't use anymore because they totally suck.

    ZackSchilling on
    ghost-robot.jpg
  • ZackSchillingZackSchilling Registered User regular
    edited January 2009
    Update: I spent the day investigating performance issues and potential fixes, including switching the graphics to OpenGL, switching the whole project to C++/GLUT/OpenGL/OpenAL.

    Eventually, I did a performance profile of the game's code and found out that what's causing the most trouble is by far the physics calculations. Specifically the mask-based collision detection. I was surprised at how little the software rendering of the graphics (all that blitting!) was by comparison.

    Since collision needs a rewrite anyway, I installed the NumPy mathematics library for Python and I'm in the process of transforming the level masks into NumPy arrays for lightning-fast collision checks and surface normal calculations. Luckily, this means absolutely nothing changes in terms of data files getting read in. In the end, the code will look neater and run faster, locking the frame rate at 60. So that's what I'm working on.

    ZackSchilling on
    ghost-robot.jpg
  • PeewiPeewi Registered User regular
    edited January 2009
    The way you bump off everything makes it very hard to navigate small spaces, which there are a bunch of on your sample levels.

    The game would slow down a lot when there were four or more balls on the screen.

    Your first post contains "Space: Rhythm code interface", but I don't notice anything happening when pressing space and it doesn't seem to be mentioned in the controls file.

    Peewi on
  • ZackSchillingZackSchilling Registered User regular
    edited January 2009
    Peewi wrote: »
    The way you bump off everything makes it very hard to navigate small spaces, which there are a bunch of on your sample levels.

    The game would slow down a lot when there were four or more balls on the screen.

    Your first post contains "Space: Rhythm code interface", but I don't notice anything happening when pressing space and it doesn't seem to be mentioned in the controls file.

    1.) The way the character bumps off of things is made a lot more severe because of a hack I placed on top of collision detection to stop the player from getting stuck in walls. This does make it hard to navigate small areas. The Testbig level was designed entirely to present odd cases to the collision detection code to make sure we never got stuck in an infinite loop, to make sure that bizarre level shapes do not clip you through walls, etc. The level is not meant to be fun, it's meant to push the limits of the code so I can quickly test to see if the physics are working pixel perfect like they are supposed to.

    I made a design decision that the player would bounce off walls the way he does, so that's not going away unless someone convinces me that killing all momentum is more fun. If after all the collision code is fixed, it's still really hard to navigate tight spaces, the solution is just to avoid tight spaces in level creation.

    2.) I know. You add 4 balls and the frame rate drops to 45. Any more and it takes a nosedive. That's why I added the feature to spawn the balls. I needed to test the physics and collision code on more than just the player to see when I'd encounter slowdown. I was hoping it would be more than 4 balls on most machines, but that just means I'll have to aggressively optimize the physics (which I am doing, see my post above yours) and even then, the game may have to limit the number of enemies onscreen to about 5 or 6. Non-physical sprites (effects, environment animations) will not count toward this total.

    3.) I didn't detail what this does exactly. It's not in the control file because it's not part of the standard control object that unifies keyboard/joystick and passes stuff to the player.

    If you want to know, open the conf file and make sure training is switched to 1. Run the game and forget moving around. Instead, tap the spacebar to a beat. You don't have to tap for the whole song. Press esc when you're done. The game will capture this tapping to the beat as a rhythm track. Now, set in the conf file, training=0. Run the game again. This time try to tap the space bar to the same beat. The game will grade you on how well you are hitting each beat in the upper right.

    This is the beginning of the rhythm code. Once I'm satisfied with the way the platforming elements work, I'll move back to this and start fleshing it out. (Give it a better interface, visual feedback of when beats are coming, tie rhythm to physical variable in the game, etc)

    ZackSchilling on
    ghost-robot.jpg
  • RichardTauberRichardTauber Kvlt Registered User regular
    edited January 2009
    Hey man, my Desert Bus remake would like to have a word with you too.

    Hope this gets done though!

    RichardTauber on
  • urahonkyurahonky Registered User regular
    edited January 2009
    Not taking away from your thread Zack, but does anyone know how well this GDK is?

    http://www.microsoft.com/express/samples/gamecreators/default.aspx

    I am thinking of just playing around with it (making a tower defense game, heh) just on my free time, and since it works with Visual C++ Express it won't cost a dime. Unless you guys think using the OP's software would be easier.

    urahonky on
  • ZackSchillingZackSchilling Registered User regular
    edited January 2009
    Holy shit you guys. I'm rewriting the collision detection to use Numpy, a super fast array crunching python extension written in optimized C.

    I just got the code to a spot where it compiles. I haven't optimized it yet. It doesn't even totally work right. But...

    I opened the game, walked out into the level. 60FPS. I spawn a jumper guy. 60FPS. I spawn 10 jumper guys. 60 FPS. I mash the tab key as many times as I can, spawning about 100 jumper guys.

    SIXTY FRAMES PER SECOND. HOLY SHIT. I expected a speedup but nothing this dramatic. The game engine can now support hundreds and hundreds of enemies onscreen at once. More than even makes sense from a design perspective.

    I've arrived. I'll get this version up once I finish the physics rewrite, hopefully very soon. Numpy is the goddamn greatest thing ever.

    Edit: This screenshot, busy as all hell, was taken running at 60FPS on my laptop. I haven't even started optimizing.

    Picture-7.jpg

    ZackSchilling on
    ghost-robot.jpg
  • SeeksSeeks Registered User regular
    edited January 2009
    I gotta say, I'm impressed with this so far.

    Seeks on
    userbar.jpg
    desura_Userbar.png
  • Zilla360Zilla360 21st Century. |She/Her| Trans* Woman In Aviators Firing A Bazooka. ⚛️Registered User regular
    edited January 2009
    Holy shit you guys. I'm rewriting the collision detection to use Numpy, a super fast array crunching python extension written in optimized C.
    Awesome. Also how about if we go this way gameplay-wise? :
    emnmnme wrote: »
    Oh yeah, remember Voodoo Vince? It was a mediocre platformer on the XBOX but I imagined it to be so much more when I watched the trailer.

    http://www.youtube.com/watch?v=vl9VlCLc7qg&feature=PlayList&p=785E773D16201CE2&playnext=1&index=92

    I had it in my head that the doll named Vince couldn't attack enemies directly but had to creatively kill himself to progress. There could have been plenty of great brain teasers a la Warioland or Sphinx and the Cursed Mummy but noooo. Reality was much harsher and the final product ended up being a Crash Bandicoot clone at best. Instead of looking for trouble, you simply pressed a button and an anvil or a chainsaw appeared above you. Lazy.
    ^Taken from another thread. A puzzle-platformer (braid-like) would be pretty cool. :)

    Zilla360 on
    |Ko-Fi Me! ☕😎|NH844lc.png | PSN | chi-logo-only-favicon.png(C.H.I) Ltd. |🏳️⚧️♥️
  • mastriusmastrius Registered User regular
    edited January 2009
    I am awful at trying to do anything with this. Of course I only just decided to try today...I read up on things and have tried to do things and so far IM ABLE TO CREATE FUCKING WORDS! Oh thats right. All hail me. *sniff* This isnt easy to learn....This thread has inspired me to keep at it though as I usually just lazily decided to say fuck it and give up. But not this time! Now I can make WOOOOORRRDDDSSS!

    (Its a start....right? :( )

    mastrius on
    "You're like a kitten! A kitten who doesn't speak Japanese." ~ Juliet Starling
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    If you have any questions about Pygame, let me know. There was a ton of stuff I wished I had someone to ask, but it was impossible to talk to anyone about. Googling was rarely helpful.

    On the game front, the rewrite was a success. The new collision system is 99% complete. I'm now tracking down the last little bug that caused sticky walls (if you were falling, you could hold the direction of the wall and actually come to a full stop because the friction and elasticity of the collisions would bleed away all your force). There would also be certain walls where the normal calculation would go a little wonky and your bounce would send you back into the wall, getting you completely stuck. I'm currently fixing that.

    ZackSchilling on
    ghost-robot.jpg
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    Zilla360 wrote: »
    Holy shit you guys. I'm rewriting the collision detection to use Numpy, a super fast array crunching python extension written in optimized C.
    Awesome. Also how about if we go this way gameplay-wise? :
    emnmnme wrote: »
    Oh yeah, remember Voodoo Vince? It was a mediocre platformer on the XBOX but I imagined it to be so much more when I watched the trailer.

    http://www.youtube.com/watch?v=vl9VlCLc7qg&feature=PlayList&p=785E773D16201CE2&playnext=1&index=92

    I had it in my head that the doll named Vince couldn't attack enemies directly but had to creatively kill himself to progress. There could have been plenty of great brain teasers a la Warioland or Sphinx and the Cursed Mummy but noooo. Reality was much harsher and the final product ended up being a Crash Bandicoot clone at best. Instead of looking for trouble, you simply pressed a button and an anvil or a chainsaw appeared above you. Lazy.
    ^Taken from another thread. A puzzle-platformer (braid-like) would be pretty cool. :)

    I'm not sure how well that would jive with the most basic elements I had planned (platforming, rhythm/combat).

    ZackSchilling on
    ghost-robot.jpg
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    Quick update: The project is not dead. In fact, quite the opposite. Instead of posting, I've been working. Unfortunately, something has gone awry in the standalone executable-building process, so I can't make neat little exe and app files for you guys like I did in the OP. I'm working on it. I've even joined mailing lists!

    New features:

    -New physics engine based on Numpy array manipulation offers greatly improved performance and no more jerky or shaking wall collisions. Eliminates many bad clipping cases that could trap or incorrectly warp the player.

    -Tiled graphics: You can specify a PNG, a game layer, and a position in the level properties file and the game will stencil the PNG on to the layer before starting the level. This should help keep file sizes down and allow for some image recycling, especially for simpler levels.

    -Animated tiles: You can also specify a layer, position, and an animation name and have the game load up a series of PNGs and animate them. This allows for moving waterfalls, windblown trees, swaying grass, moving clouds in the background. It's all done with PNGs and text files to specify animation order and rate. No coding required. No apparent performance hit either.

    This waterfall looks pretty nice in motion. Screenshot taken with game running at a silky 60fps:
    Picture-8.jpg

    The level config file, showing how I spawned a few repeating animate waterfall tiles to create that effect.
    Picture-9.png

    Picture-10.png

    ZackSchilling on
    ghost-robot.jpg
  • FremFrem Registered User regular
    edited February 2009
    Could you throw this up on SVN at Assembla or Google Code or something? It would make it a lot easier for all of us to stay on the same page (not to mention save bandwidth ;-) )

    Frem on
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    Once the code base and file structure settle down a bit, I'll get more organized and throw the whole thing up on Google Code.

    I need to write a better cross-platform build script too. You know, once I figure out how to get it building again. (The Numpy - pygame interface runs fine on the command line but segfaults if you try to build a package.)

    ZackSchilling on
    ghost-robot.jpg
  • UberEvUberEv Registered User regular
    edited February 2009
    I spoke with the OP to make sure he was cool with this, and he said he was so here I go.

    Since a few people mentioned wanting to contribute to Zack's game, I may have a tool that will help you test ideas before you spend the time coding them to perfection. I'm creating a game development toolkit called Venatio Creo, which is a WYSIWYG toolkit that requires no programming to make fairly complex games. One thing it seems particularly useful for is creating platformers (quickly too, I'm talking in minutes). You should be able to jump right into the program with few problems, but I recommend playing a few of the games that are included with it to see how everything flows.

    It might be a great sandbox for you guys to use to test ideas you have for Zack's game, so give it a shot if you are just itching to develop. Any graphics or sound that you make to try ideas in VC will port just fine to Zach's engine, so you don't need to worry about wasting time and remaking assets. You can get a release candidate from Venatio Creo's Website. Let me know how it goes if you decide to try out the program. It's currently Win32/64 (XP/VISTA) only right now.

    Thanks for letting me mention this Zack.

    UberEv on
  • MiserableMirthMiserableMirth Registered User regular
    edited February 2009
    I would like to help. Not sure what I can do though.

    I'm curious about your rhythm based combat system. Do enemies have their own unique rhythm? How many buttons are involved in making the rhythms?

    MiserableMirth on
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    I would like to help. Not sure what I can do though.

    I'm curious about your rhythm based combat system. Do enemies have their own unique rhythm? How many buttons are involved in making the rhythms?

    Nothing is coded aside from a basic 1-track rhythm system and a little code to compare when you hit the space bar to when the beats in the rhythm track are. Just a little proof-of-concept I can expand on later.

    In terms of what I'm planning, I've got 4 assignable attack (kick and punch, high and low) buttons and a charge button. What I hope to do is tie the power behind punches, kicks and jumps to how on-beat the player is. As you play, each successful move boosts the charge meter. (This could include tapping the charge button right as an enemy attacks to block and absorb). Once some power is in the charge meter, you hold down the charge button and hit an attack or jump to the beat and it gives you double the strength. Hit the beat poorly or miss with the charge active, it hurts you or knocks you around. I'm not sure just yet.

    Ideally, I'd have 4 or 5 tracks for the player, each would work for any move, but only one, say bass, would give the most power to low kicks and punches. This would all be represented on a scrolling track along the HUD. At no point would it be required to use a charged jump to complete a level, but maybe for shortcuts or secret areas. Uncharged and offbeat moves still do damage, just not as much. Maybe they'll go off with a quarter of the power. The multi-track rhythms and HUD are next on my to do list.

    As for enemies, I don't know yet. It would be neat to tie their movement and attacks to the player's rhythm track too, but I haven't started any AI. It's definitely possible and something I'm thinking about. Maybe give them their own track with just a steady beat at first and work up to reading directly off the player's track as the game progresses.

    ZackSchilling on
    ghost-robot.jpg
  • ImpersonatorImpersonator Registered User regular
    edited February 2009
    UberEv wrote: »
    I spoke with the OP to make sure he was cool with this, and he said he was so here I go.

    Since a few people mentioned wanting to contribute to Zack's game, I may have a tool that will help you test ideas before you spend the time coding them to perfection. I'm creating a game development toolkit called Venatio Creo, which is a WYSIWYG toolkit that requires no programming to make fairly complex games. One thing it seems particularly useful for is creating platformers (quickly too, I'm talking in minutes). You should be able to jump right into the program with few problems, but I recommend playing a few of the games that are included with it to see how everything flows.

    It might be a great sandbox for you guys to use to test ideas you have for Zack's game, so give it a shot if you are just itching to develop. Any graphics or sound that you make to try ideas in VC will port just fine to Zach's engine, so you don't need to worry about wasting time and remaking assets. You can get a release candidate from Venatio Creo's Website. Let me know how it goes if you decide to try out the program. It's currently Win32/64 (XP/VISTA) only right now.

    Thanks for letting me mention this Zack.

    WYSIWYG should be implemented in pretty much every program. Loved the program, I'll download it one of these days and try it (I like to play around with GM).

    Impersonator on
  • MiserableMirthMiserableMirth Registered User regular
    edited February 2009
    Nothing is coded aside from a basic 1-track rhythm system and a little code to compare when you hit the space bar to when the beats in the rhythm track are. Just a little proof-of-concept I can expand on later.

    In terms of what I'm planning, I've got 4 assignable attack (kick and punch, high and low) buttons and a charge button. What I hope to do is tie the power behind punches, kicks and jumps to how on-beat the player is. As you play, each successful move boosts the charge meter. (This could include tapping the charge button right as an enemy attacks to block and absorb). Once some power is in the charge meter, you hold down the charge button and hit an attack or jump to the beat and it gives you double the strength. Hit the beat poorly or miss with the charge active, it hurts you or knocks you around. I'm not sure just yet.

    Ideally, I'd have 4 or 5 tracks for the player, each would work for any move, but only one, say bass, would give the most power to low kicks and punches. This would all be represented on a scrolling track along the HUD. At no point would it be required to use a charged jump to complete a level, but maybe for shortcuts or secret areas. Uncharged and offbeat moves still do damage, just not as much. Maybe they'll go off with a quarter of the power. The multi-track rhythms and HUD are next on my to do list.

    As for enemies, I don't know yet. It would be neat to tie their movement and attacks to the player's rhythm track too, but I haven't started any AI. It's definitely possible and something I'm thinking about. Maybe give them their own track with just a steady beat at first and work up to reading directly off the player's track as the game progresses.
    I think I get the gist of it. I like it a lot. I'd like draw up some design documents just to brainstorm ideas if that is cool with you. I'm not sure how much of the design aspect you have drawn up.

    Also, will different types of platforms react to certain parts of the song? I mean, if you're going to code all those separate tracks, might as well get the most out of it.

    MiserableMirth on
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    Either finding or making royalty-free music that's good enough to listen to is something I'll need help with. An alternative is to make the game with commercial tracks in mind and distribute it without music. The user would then have to go buy the songs on iTunes or something and we'd convert them. I don't like that solution so much because it lends its self to someone pirating all the music and making an unauthorized complete distribution. Then I could be in hot water because of what someone else has done, not to mention I lose control over the distribution of the game.

    As of right now, there are no "different kinds of platforms." I'm thinking of adding some basic support for moving platforms but I don't know how well my physics engine will behave. I'm sad to admit that it was not coded with relative velocities in mind. Y-axis motion will be ok because the routines that cause the player to hug the ground will stay in effect. X-axis motion will very likely leave the player behind without some kind of ugly hack or system overhaul.

    Lastly, Lenard Lindstrom on the pygame mailing list has identified my obscure problem with building standalone apps. Thank you sir, you are a gentleman and a scholar. I am, as we speak, setting up a google code page. I'll post it and links to the new binaries once they are built and up.

    ZackSchilling on
    ghost-robot.jpg
  • anoffdayanoffday To be changed whenever Anoffday gets around to it. Registered User regular
    edited February 2009
    I've been messing around with creating a platformer as well. It's something I've always wanted to create too. Unfortunately, I know slim to none about programming. I've been messing with Game Maker which people have actually created some pretty sweet games with. Who knows if I'll be able too, though.

    Anyway, yours is coming along nicely. I like the character.

    anoffday on
    Steam: offday
  • UberEvUberEv Registered User regular
    edited February 2009
    anoffday wrote: »
    I've been messing around with creating a platformer as well. It's something I've always wanted to create too. Unfortunately, I know slim to none about programming. I've been messing with Game Maker which people have actually created some pretty sweet games with. Who knows if I'll be able too, though.

    Anyway, yours is coming along nicely. I like the character.

    Well make sure you give Venatio Creo a shot if you are still open to engine ideas. All the details are a few posts above yours. At the very least it can get you started on a simple platformer really quickly. You can also wait until Zack has some of his source code available and try modifying little stuff here and there.

    UberEv on
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    New Release: Alpha 2

    Windows standalone exe
    Mac OS X standalone app
    Platform Independent Source (requires Python, Pygame, and NumPy)

    Data Files (Unzip and place in the same folder as the executable (or on OS X, in the same folder as the .app bundle, NOT inside the bundle)

    Brief FAQ:
    Q: Why are these files so huge?
    A: The standalone executables include an entire Python distribution, Pygame, Numpy, SDL, and various other libraries to ensure they run on any system. The data folder is mostly uncompressed PNGs and music. With tiling, I should be able to shrink it a little going forward. For now, it's ok, my host gives unlimited bandwidth.

    Q: What are the controls?
    A: See the OP

    Q: Are you supposed to be able to cling to the walls?
    A: Yes. It'll make more sense once the air control animations are in place.

    Q: It crashes right away. What can I do?
    A: Make sure you have the data folder along side the executable or app.

    Q: Didn't you say you'd set up a Google Code page?
    A: Coming soon.

    ZackSchilling on
    ghost-robot.jpg
  • ZetxZetx 🐧 Registered User regular
    edited February 2009
    Cool stuff

    Great stuff, Zack. I was going to ask if you had done any tiling stuff, but I guess you haven't yet. I want to make a top-down kinda game and tiling seemed pretty essential.

    I can't remember why I stopped trying to make my game in python, but I guess I can start from scratch again. I can't seem to get very far. :|

    I love these kinda threads, they always make me want to try again.

    Zetx on
    76561197966849367.png
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    There is support for both animated and static tiles over the level images. I have not made full use of the feature in the test levels yet because I just implemented it this morning. Once I, say, replace all the grass in the vida level with tiles, things should get better. It'll keep file sizes down and make it easier to create graphically rich levels as the library of tiles grows.

    ZackSchilling on
    ghost-robot.jpg
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    I put together a quick demo of how things are working now. It should be easier than downloading 50MB of files and editing a config file to play around. Click through and watch in higher quality.

    http://www.youtube.com/watch?v=ypS1AG_Tbx0

    ZackSchilling on
    ghost-robot.jpg
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    Dilemma: I've implemented a little scrolling bar that shows the beats in the song so the player can glance at it. It's just no fun to use at all. It's not easy to watch while you play and it doesn't help much even if you watch it. The scrolling is either uselessly slow (clear but indicates no rhythm) or fast and useless (stuff flies by and you have no time to process while running and jumping). I know it's ugly too but I'm looking for something fun to use. Something unobtrusive while playing that conveys the same information.

    Picture-11.jpg

    The Power indicator is looking pretty sweet though. It animates too, draining and filling over time if you add or remove a large chunk of power.

    ZackSchilling on
    ghost-robot.jpg
  • DagrabbitDagrabbit Registered User regular
    edited February 2009
    Couple thoughts:

    In New Super Mario Brothers, the bad guys would dance along to the music. Maybe something like that, swaying trees, mountains, or something to show the rhythm.

    Or, the main character that your staring at all the time could snap his fingers or something, and you have to hit space bar or whatever to beat he's snapping.

    You're right, in an action platformer, people will always be looking at the MC or the environment. Some kind of neat animation could make it clear to the user, and make the whole game visually interesting.

    Dagrabbit on
  • TrusTrus Registered User regular
    edited February 2009
    Dagrabbit wrote: »
    Couple thoughts:

    In New Super Mario Brothers, the bad guys would dance along to the music. Maybe something like that, swaying trees, mountains, or something to show the rhythm.

    Or, the main character that your staring at all the time could snap his fingers or something, and you have to hit space bar or whatever to beat he's snapping.

    You're right, in an action platformer, people will always be looking at the MC or the environment. Some kind of neat animation could make it clear to the user, and make the whole game visually interesting.

    bobbing his head up and down perhaps?

    Trus on
    qFN53.png
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    I thought maybe having all the blue jets pulse would be good too, but I really want to give advance notice of the beats, not just an indication while they're playing.

    One thing I tried is eliminating the white strip shortening the bar, and centering it around the stomach of the player at all times. It's not too distracting and it's easy to watch while you play. Then I was thinking I'd open up the player's sprite, add some kind of hole or receiver in its chest and reskin the beats into a little smaller blue energy orbs, similar in color and style to the arms and legs. Then you get the sort of visual impression that when you hit buttons to the beat, the orb gets sucked in and bursts out the arms/legs depending on your action.

    I'm also considering scrapping the rhythm aspects and focusing more on the platforming and combat, because the music is posing quite a few more problems. For example: where to get decent free tracks and finding the delicate balance between making the levels fun with but not depend on the boosts from following the rhythm. I'd definitely like to finish prototyping a few aspects (combat, basic rhythm) before seriously visiting that idea.

    ZackSchilling on
    ghost-robot.jpg
  • KrisKris Registered User regular
    edited February 2009
    I downloaded the Mac .app and tried running it, but it crashed on startup. Here is the log output:
    07/02/09 10:01:55 PM Platformer[2016] Platformer Error
    07/02/09 10:01:55 PM Platformer[2016] Platformer Error
    An unexpected error has occurred during execution of the main script

    IOError: unable to read font filename

    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] Traceback (most recent call last):
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] File "/Users/kristofernelson/Downloads/Platformer.app/Contents/Resources/__boot__.py", line 137, in <module>
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] _run('Platformer.py')
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] File "/Users/kristofernelson/Downloads/Platformer.app/Contents/Resources/__boot__.py", line 134, in _run
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] execfile(path, globals(), globals())
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] File "/Users/kristofernelson/Downloads/Platformer.app/Contents/Resources/Platformer.py", line 1519, in <module>
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] if __name__ == '__main__': main()
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] File "/Users/kristofernelson/Downloads/Platformer.app/Contents/Resources/Platformer.py", line 1266, in main
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] gameinfo = GameInfo()
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] File "/Users/kristofernelson/Downloads/Platformer.app/Contents/Resources/Platformer.py", line 699, in __init__
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] self.font = pygame.font.Font(os.path.join('data','freesansbold.ttf'), 36)
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] IOError:
    07/02/09 10:01:55 PM [0x0-0x121121].org.pythonmac.unspecified.Platformer[2016] unable to read font filename
    07/02/09 10:02:06 PM com.apple.launchd[106] ([0x0-0x121121].org.pythonmac.unspecified.Platformer[2016]) Exited with exit code: 255
    07/02/09 10:02:06 PM com.apple.launchd[106] ([0x0-0x121121].org.pythonmac.unspecified.Platformer[2016]) Exited with exit code: 255

    Kris on
  • ZackSchillingZackSchilling Registered User regular
    edited February 2009
    You didn't download the data folder. The first thing the game loads is its font and game is saying it can't find it. Go to the OP and get the data folder and throw it in the same folder with the app.

    ZackSchilling on
    ghost-robot.jpg
Sign In or Register to comment.