TUER: Truly Unusual Experience of Revolution, FPS using JOGL

There is something wrong with my FTP client. The files are too big (4.8 MB). If you give me your email adress, I will send it to you as fast as I can.

Ok I solved the problem.

Link removed

Hi! My exams will be finished in a few days. I will begin designing the full hardware accelerated mode, hoping that it will work better than now. I wonā€™t change the keyboard config now but I will build a menu to change the mouse sensibility and the keys.

The screenshot feature is going to be repaired in a few hours. Could someone test my game under Microsoft Vista? I need to check if it works fine.

Does anybody know how to create a directory without having a problem of permission in Java? The screenshot feature doesnā€™t work. Iā€™m looking for a way to detect keyboard type in Java as some users of the game have got a AZERTY keyboard and the others have got a QWERTY keyboard.

;D The screenshot feature works under Linux. I will change the configuration of the keyboard in a few days.

No one here seems to be interested in my projectā€¦ :ā€™(

I have just updated the java archive. The configuration has been changed as someone asked me for QWERTY and AZERTY keyboards. For AZERTY :
Z to go forth
S to go back
Q and D for left and right strafe
The arrows allow to turn, to go forth and to go back.
SPACE and left mouse button to shoot
The mouse allow to turn too.
F1 to perform a screen capture! (sometimes it doesnā€™t work under Microsoft Windows, check the default directory and its user create directory permissions).

If there is something wrong during the capture, you will see a small message very quickly.

Under Microsoft Windows, the problem appears only with firefox, I donā€™t know why.

[quote]No one here seems to be interested in my projectā€¦
[/quote]
Iā€™m waiting for the hw accellerated version :wink:

Ok! No problem! I have been working on this major improvement for some weeks. It will be ready at September, maybe before. I am going to try to draw the crosshairs at first, the weapons, the walls and then the objects and the ennemies. I will use multipass rendering or if possible multitexturing and blending to handle the marks of the walls after an explosion. The portals wonā€™t be implemented directly. I want to remove raycasting at first and then improve the way to render the graphics, step by step. Thank you for your honest remark. Be patient ;D

The screenshot feature works only if the directory chosen by your browser as the ā€œhomeā€ for my game allows the creation of a directory. Under Linux, it works perfectly, the directory called ā€œsnapā€ is created in your home.
Under Windows, this directory is created on your desktop with Internet Explorer and in ā€œProgram Files/Mozilla Firefox/ā€ with Firefox.

About the hardware accelerated version, I worked this weekend on the computation of the vertices and the textures coordinates of the walls by using the world map. I hope that I will be able to show you some screenshots in a few days.

Hello! Push ā€œF3ā€ during the game and you will be able to test the EXPERIMENTAL fully hardware accelerated mode! Now, I only draw the wall in this mode, I will complete it step by step. There is no space subdivision. Then, the performance may be increased in the future. I plan to use a cells-and-portals algorithm. The game might crash under Microsoft Windows with PCs using Intel graphic chipset above all. I added a FPS counter at the bottom of the screen. Iā€™m waiting for your first impression. I have a very little time to work on the game but I donā€™t give up! ;D

Link removed

There is a problem with Intel 82810 chipset. The bug has been corrected, I will update the java archive Monday. Sorry for the delay.

Cool, 63-70 FPS (I think it locked that as the cap) on my computer. If anything it runs too fast now o_O. Of course I have a 7900 gs in here, so that helps. But no serious performance issues so good job so far! Now just work on filling in those objects so I donā€™t keep running into invisible barriersā€¦

Oh also, complements on the smooth switch between hardware and software. Nice setup for comparison and testing.

Thank you very much! I want to improve the objects. These are only zoomed bitmaps. It is not locked by a cap. The laptop computer of my best friend turns at 81 FPS on my game. The performance may be increased by spatial subdivisions. I will use cell-and-portal algorithms. It will become faster. These invisible barriers match with objects these are not drawn now in the experimental mode. I have a lot of things to do. The lighting system is primitive. At first, I will use some simple light sources and after that, I might use pixel shading for dynamic lighting. The GUI is too much poor and I need a multiplayer modeā€¦ I would like to add a crosshair and other weapons. The extern module for the display of MD3 models is not fully optimised.

Finally, I appreciate to have some feedbacks. Sometimes, Iā€™m sad because only a very few people seem to be interested in my project and there is still no link from https://jogl.dev.java.net/ to my website.

Exp-mode works fine on Core2 Duo@3Ghz, Vista, HD2900XT at around 91fps. However, it doesnā€™t work on an old Intel onboard chipset (all whiteā€¦).

Athlon 64 X2 3800+,XP, X800Pro works fine too. The fps counter goes crazy and shows something like 926363636272819 fps most of the time, but apart from that it works fine.
BTW: Software mode runs with 5fps on that machine while it outputs 15fps on the Core2.

Iā€™d like to play it but everytime I tried I get 0 fps on software mode.

Athlon 2500+, 9800pro, 1GB ram.

60 fps hardware accelerate mode.

Thanks for the feedback! This bug is solved, I am updating the java archive right now. I donā€™t have internet at home, then it complicates many things. It was all white because I had forgotten a glBegin and a glEnd in one of my methods :stuck_out_tongue:

It is the first time that the FPS counter goes crazy! Maybe the computation become false when the interval between two frames become very very small. I will study that point.

;D Now the experimental version shows the picture and uses backface culling. It does not increase performance now because the zbuffer spends too much time in excluding vertices. It is my fault, I have not implemented any spacial subdivisions algorithm. :o