TUER: Truly Unusual Experience of Revolution, FPS using JOGL

WIP games, tools & toy projects
#121

We are here to learn things too. That is not a problem.

The next version won’t be updated now. I will update only the source code tomorrow because it is not stable enough. This new version has some improvements :

  • 2 times faster ;D
  • the bots are visible even in the accelerated version!!
  • the alpha test is less used to allow people with an old card to have a quite good experience
  • the fullscreen mode might work on any Linux systems this time
  • “xset -r off” at the start, “xset -r on” at the end in a shutdown hook for the users under Linux
    However, there are some regression :
  • the bots run too quick!
  • the keyboard answers hardly
  • the bots respawn even in a cleared area when you restart
  • the game uses too much memory and crashes at the start sometimes (java heap space ???)

When the game is stable, I will update the java archive too.

#122

http://membres.lycos.fr/javalution/projects/snapshot_3.png

This is a screenshot of the next version ;D under AMD Sempron 2600+ 2 GB DDRAM 400 Mhz ATI X1950 AGP 8x 256 MB Mandriva Linux 2007 with the proprietary driver.

#123

NEW UPDATE OF THE SOURCE CODE :
new features :

  • draw several pieces of a huge VBO without rebinding
  • the objects can be broken (they only disappear) in the experimental mode

coming soon :

  • some methods to perform multiple draws with the same VBO without rebinding
  • impacts, rockets and explosions
  • disappearance of the slow mode which will cause a huge decrease of the memory usage and a noticeable decrease of the loading time
  • no more java.lang.outOfMemoryError

WARNING! The method called “glBindBufferARB” must be used as rarely as possible. If you draw the same objects with the same VBO at different positions for the same frame, do your best to bind the VBO one time only or your performance will hugely decrease. For example, if I bind the VBO which contains the coordinates of all the objects each time I want to draw an object, I get only 33 FPS whereas I get 142 FPS when I bind the VBO once for all objects. Some 3D cards use an extension allowing multiDraw. It might be interesting but you should plan to find an alternative solution when a 3D card isn’t compatible with this extension.

the source code :
Link removed

the game itself (not updated for a week) :
Link removed
Link removed

#124

NEW UPDATE OF THE JAVA EXECUTABLE ARCHIVE :

new features :

  • draw several pieces of a huge VBO without rebinding it and the same VBO at different positions (may use glMultiDrawArrays if available)
  • the objects can be broken (they only disappear) in the experimental mode
  • the bots appear in the experimental mode too ;D
  • TWICE quicker than the former update ;D
  • less bugs
  • some contradictions between the slow mode and the experimental mode :frowning:
  • the bots are still too quick
  • might work under Microsoft Vista

I included a cheat key to become invincible to test if everything works. I’m sorry for the delay, my internet service provider doesn’t want to repair my own connection.

the source code :
Link removed

the game itself (currently updated) :
Link removed
Link removed

#125

IMPORTANT INFORMATION :

erikd might be interested by this. The declaration is available at this address :
Link removed

I think that some people might laugh about me after reading this document but it is my attempt to suggest another way of living from gaming without charging directly the users and some money has been saved to drive this project possible in some years. I hope that some motivated computer engineer will join the Foundation. The document isn’t finalised.

#126

You seem to be saying that the primary source of funding would be subsidies (called ‘grants’ in the UK) from government, corporate or charitable sources. This is far from impossible but you’d need to show what benefit would be gained from the money. Writing and playing games is fun but does it actually benefit anyone? I think you’d need to ‘add value’ somehow ie; provide employment to disadvantaged people, have educational content in the games, &c before you’d get any serious money. Otherwise you’d just be asking people to pay you to do what you want to do!

#127

Some organisations are ready to pay, it is not only a dream, it isn’t something I will never realize. As I said, I consider that everyone should have an equal access to culture and art, I consider that the video games are works of art then it benefits to the people. Money is going to arrive, no need to add value, it is a serious project and the organisations which agree with helping me have been easily convinced, that’s not the problem. They only want me not to stay alone. They expect from me that I find some very motivated people to take part into my project. It is gonna be alright ;D I need to be patient, that’s all.

#128

NEW UPDATE :

  • only minor improvements to prevent the game from crashing

I have an internet conection now ;D.
Please help me, my 3D card doesn’t work anymore, I had to install the older one. Can someone test my game again? (erase the caches before)
I’m going to implement a lossly algorithm in order to reduce the count of textures passed through the bus by a factor of about 100 as a temporary solution. I hope it will work.

#129

Works OK (4fps & 15fps slow & fast modes) but the bullets aren’t showing in fast mode!
Aiming is v. tricky too - it seems to move in discrete steps rather than continuously so it keeps swinging back and forth past the target as you try to aim…

[quote=“gouessej,post:128,topic:29428”]
See you in Wolfenstein ET! ;D

#130

Thank you very much. The microprocessor seems to be too busy to handle the keyboard properly, that is a serious problem and I don’t know how to solve it.

#131

For Linux users :

I have found a solution for those who had a problem with the taskbar drawn onto the screen! Unfortunately, you have to modify your configuration through the menu of KDE : system -> settings -> KDE -> appearance -> “instrument panel” -> masking -> allow the other windows to recover the instrument panel (taskbar?)
This is translated from French, sorry if it is not exactly the good words. Nevertheless, the performances are not stable even with this, the game turns at between 20 and 120 FPS, I don’t know what to do. If you have a solution, it will help me a lot.

#132

I just had another go with this. It worked fine, but I have a few remarks:

  • Movement is way too fast!
  • You might want to smooth out mouse movement as well. Apart from the mouse being too sensitive, it’s also hard to aim because of the large steps.
  • I tried the game twice, and the millisecond I saw a bad guy, I was instantly killed. Make it easier at the start.
  • The collision sphere (or whatever it is in your game) is a bit too small; It seems the player is able to get too close to the wall, as if you’re a really thin stick figure. Besides that it seems unnatural, it’s sometimes hard to see where you’re going when you’re close to a wall looking straight at it. (maybe this has also something to do with the movement being too sensitive and fast).
#133

Excellent remarks! Yes, movement is too fast, that’s why I try to implement time-based animations and time-based movements. I will use some equations for the uniformly accelerated rectilinear movements something like : x(t)=1/2a0t²+v0*t+x0

Yes the step is too large now as the game works quicker and as I will show you, the next version will have a better framerate ;D to allow people with an old 3D card to have fun too. I will compute the width of the step by using the average framerate.

The collision system needs to be improved. It uses voxel-based detection except for the rockets and the bots.

the source code :
Link removed

the game itself :
Link removed
Link removed

#134

Now look at the next version, it is not finished but the framerate is twice bigger on my old ATI radeon 9250 Pro :
Link removed

My driver is very bad. It should be better for you with a proprietary driver.

#135

Now, the walls are ok, the ceil and the floor are missing :

Link removed

The source code and the game will be updated when everything appear.

the source code :
Link removed

the game itself :
Link removed
Link removed

#136

The walls, the floor and the ceiling are complete in the next version, the impact appear too! However, I need to improve one of the algorithm to choose which coordinates really need to be given to the 3D card. Now, I’m going to try to transmit only the coordinates in front of the player.

the source code :
Link removed

the game itself :
Link removed
Link removed

#137

NEW UPDATE :

  • impacts on the walls when you shoot
  • a very small increase of speed but the displayed area is temporarly less deep

I will improve the accuracy of the impacts and implement a cell-and-portal algorithm with some other visibility culling methods to limit the amount of data on the bus. My old ATI Radeon 9250 Pro displays only 20 FPS, it is not enough, I’m very angry. It’s really frustrating and the open source driver is very bad. The display of the walls is consuming the most part of the execution time. When I don’t display anything, the game turns at between 222 FPS and 500 FPS. When I display only the walls, even just the walls close to me, the frame rate is close to 22 FPS. I hope I will get 40 FPS in a few weeks.

The other problem comes from the alpha test. My driver is so bad that the objects are often “flickering”, they appear and they disappear. I didn’t have this problem with the proprietary driver on the ATI X1950 Pro.

The implementation of the cell-and-portal algorithm might take me a long time. Therefore, I will try to implement other interesting visual effects to allow you to see something new, rockets and explosions at first. When it is ok, I will put the slow mode out of my game to decrease the memory usage.

the source code :
Link removed

the game itself :
Link removed
Link removed

#138

erikd, please, as your configuration is the following : WinXP, 512MB, AMD 2200+, GeForce 4 Ti4200, IE6, Java 1.6 update 2
can you give me the frame rate you get with the experimental mode?

#139

I got a steady 60-ish framerate, I suppose it’s vsync’ed.
There were some glitches with objects close to walls (those big pots with plants in your screenshots) that sometimes disappear.

#140

Oh, and this time it took like forever to start (to even get past the JWS ‘downloading’ screen).
Maybe this is the delay you were talking about? Maybe JWS and your host don’t get along for some strange reason (something like your host limiting the number of requests coming from JWS, maybe)…