Pages

Trilight Demo


This is a demonstration of a project I am working on. Maybe some of you heard about the project Trilight. I liked the idea and was looking forward to playing the game, but some time ago I found out the project was cancelled. So I decided that I could at least try to create a similar game, and after receiving an explicit permission from the original developer Michiel De Mey I started working on it. It is just a very basic demo, but I will do my best to make this a decent game.

Download the demo here

Recommended configuration: Windows XP with Service Pack 3 or higher (if an error pops up about missing "msvcp100.dll", you will need to install Visual C++ binaries)
Direct X 9.0b or higher

Also confirmed that it runs under Wine (tested on Ubuntu 12.10)

Installation: Just unpack the ZIP archive anywhere on your disk and run Trilight.exe

Comments (17)

Loading... Logging you in...
  • Logged in as
Can't play your demo. It says I need some d3dx10_43.dll or something.
2 replies · active 604 weeks ago
mine also said that (today is the 28 of august) and i have windows 8 do i need to download anything
May I recommend packaging the .zip in a folder? I.e. instead of having /Trilight.exe and /data, have /Trilight/data and /Trilight/Trilight.exe ... that way it's easier for people to extract (they don't have to create another folder, they can literally extract it anywhere).

Gonna try the game soon, it looks awesome =)

EDIT: Can't play it... it hates my (standard, 1280x1024) resolution (and it screws up my screen badly). Could you _please_ provide an option to make it windowed when the full version comes out?
10 replies · active 608 weeks ago
"Did you manage to play the game or did it crash?" It does work, but I really don't like playing fullscreen games, because it takes a long time for me to fix the resolution once the game has ended (sometimes I even have to reboot). This is NOT a problem with your game, as it's with every single fullscreen game that changes the resolution to something smaller than my screen (there is an odd bug where you can scale the resolution down, but not up, so even if your game tries to restore the resolution, it won't work).

So if you can make it windowed, I'll love you forever =) Okay, I didn't mean it in THAT way, but, well, you know what I mean =P
I see, that's why it sent me that message. Was wondering what it was about =P

I'm using linux (arch linux). And if you're wondering, I'm using wine 1.6-80-gb6fc57e (latest git, as of 2 weeks ago) =)
Nope, it seemed normal (unless your game was supposed to be at half speed or something =P). For both ShowFPS.exe and SHOWFPS_LINUX.exe, I got a steady 60-62FPS, though that might also be because my graphics card has vsync turned on.

Also, I noticed that in the game, you seemed to have an object sorting issue (for example, if I'm on the ground, and there's a wall right beside me, and I move up and right, it will get stuck, instead of hitting the wall and sliding up)... are you using Box2D for the physics and CR? or using a custom-built one?
Glad I could help! This game looks very promising already =D
"How did you know it was Box2D?" It was kind of a wild guess, but it's the most popular 2D physics engine (by far). "Is it known for these issues?" It's a standard issue with CR, but I'm not sure why Box2D, with the popularity it has, hasn't fixed it yet (I'm guessing it's intentional, for some odd reason).

If this can help you, I've written a tutorial on how to fix it, but it's written for a custom-built AABB CR (i.e. game CR that is not in a library). I'm not sure if you'll find anything useful from it, but, well, if you can, awesome! http://mijyn.github.io/tutorial/2013/06/05/platfo... (the part in question is at the end, but you'll probably want to read most of it to understand how it works).

But since that only works for AABB (though it will work for non-AABB, it'll just be a bit buggy, in the same way as using an AABB algorithm to do CR on non-box objects would be buggy), what you'd have to do if you're not using AABB (which you're obviously not), is to replace all of the functions to their SAT equivalent (but since you're using an engine, you'll probably only have to code in `collides_by_axis`, if that), and, of course, use a loop instead of a simple `if (x && y)`.

Sorry, I know I'm being super-vague here, but it's kind of late, and I can't think properly. I hope this can help you though!

Post a new comment

Comments by