April 2012
16 posts
When implementing the ability for the camera to get into places (water/slime/lava/portals) with translucent surfaces, I’ve remembered that portals uses the same pointcontents value as water.
This means that in the case of portals, the chase camera will treat them as water, which means that when water is translucent the chase camera will go through portal surfaces, even if the portals...
Better code design sometimes makes things behave even better than we expected. I just found out that by using a separate list for translucent entities, the chase camera doesn’t get blocked by them anymore.
On the other side, since lava and portals can optionally be translucent now, and water and slime can optionally be opaque, the code needs some adjustments to allow or deny the camera to...
With the correct translucency settings (opaque portals and opaque lava), the opening scene of the start map runs at 17 fps, one above the previous benchmarch.
However, this is not all. I’m running it at 640x240 now, on the Dreamcast, and the game is running way faster than I expected. Watching the demos, the main bottleneck seems to be the dynamic lights now. In any non-underwater scene...
I decided to check it deeper, and the 1.3.1 version still have an old bug that makes the framerate counter unreliable, so the framerate differences between the version 1.3.1 and the current code aren’t as big as indicated in the previous post. Oh well.
The clipping bug is proving to be as insidious as I expected it to be. It goes away when blended BSP entities are turned off, but it only...
I’ve just got the Dreamcast version of the current code to compile. Here are the results so far, measured from the starting point of the start map:
Makaqu 1.3.1:
9 fps when there’s no lava ball
8 fps when the lava ball shows up
Makaqu 1.5 (still unfinished):
16 fps when there’s no lava ball
12 fps when the lava ball shows up
The speed increase is definitely noticeable. The...
Some more bugs fixed. There still is a weird BSP clipping bug to fix, but all the other BSP bugs introduced in the 1.4 alpha seems to have been gone. MDL and SPR entities also needs some more testing, just in case.
I’m also thinking about lowering the threshold of the particle color shading map a little, so they can go a little darker in completely dark places.
Several more bugs fixed. Stability wise, things are now back to how they were in the 1.4 alpha. This means that most of the critical bugs now are related to turbulent (water, slime, lava and portal) surfaces, which aren’t displayed correctly.
However, the entity management is now more robust, and will allow me to finally fix those bugs and finish implementing some new features.
Support for...
Particle lighting is in. It can be turned off, and I’ve also implemented a console variable to disable fullbrights on the color shading map of particles.
The minimum level of light for them is the same as for the viewmodel, which is just enough to distinguish their color in the dark. If I ever see a good reason to allow particles to go fully black, I may make this optional.
I just remembered that particles aren’t lit. This is a feature I really want to get in, just for the sake of behavior completeness.
SPR models also aren’t lit, and I want to make them to be, so they’re getting next on my list. I’ll fix the entity management enough to get them to display again, and then I’m going to work on them.
The config.cfg file included with Makaqu 1.3.1 is outdated.
To fix this, open it in a text editor, and change these lines:
bind DC_AXIS_L +left bind DC_AXIS_R +right bind DC_AXIS_U +lookdown bind DC_AXIS_D +lookup
To these:
bind JOY1_AXIS_X- +left bind JOY1_AXIS_X+ +right bind JOY1_AXIS_Y- +lookdown bind JOY1_AXIS_Y+ +lookup
You can also set the analog stick manually...
Some news for the next release, which will be Makaqu 1.5.
Particle translucency is back, and the particle drawing code has been even more optimized. I thought this translucent rendering would be slower, but for some reason it doesn’t seem to be impacting the framerate.
I’m quite satisfied with the current particle code. There doesn’t seem to be any room for improvement anymore,...
Makaqu 1.3.1 released →
There you go. This has been tested on a real Dreamcast and the analog stick bug is gone.
It was really tempting to add some of the improvements of Makaqu 1.4 alpha into this release, but such a thing would only create confusion. So, there’s nothing else new in it since the version 1.3.
Due to this, the executable of the Windows version didn’t need any changes, so I didn’t even...
Anonymous asked: 2) After researching some total conversions such as QRally and Gold Cup, I'm excited about the possibility of using the Q1 engine for a vehicle simulation. Do you know of any resources/tutorials along this end? I checked the resources on inside3d to no avail.
Anonymous asked: Hey, two questions: 1) I'm completely new to variations of the Quake 1 engine. Do I need to do anything special to install makaqu for windows? I dragged the "quake" directory out of the archive for makaqu 1.3, but running the makaqu executable displays an error saying "Draw_cachepic: failed to load gfx/box_tl.lmp". Question 2 will follow.
