Something unique

Well, my bachelor student beeing will begin next morning and I think it's time to do some bits of code before there's no time anymore. Well, I'm still shirking a lot of essential game elements like movement systems, map editors, etc. I always wanted to make a game made like a pro would do it, but a pro would just make it, independent how perfect it will be - it has to work. Thinking about how to make a program makes curious how a pro would program it or which approach he would try. The basic class structure, game code elements, etc... Nothing is perfect and I doubt pro code is perfect. Maybe there's no pro code. Just code. Hm. I have to think about that. However, my problem is that I always want to make something special without coding special things. I mean, I usually avoid writing game mechanics. I just write code snippets and classes without direct connection, creating a chainable system of buffers and values. I always thought game programming is about writing complex and flexible code for every possible kind of game. But now I know that game programming is more about how to extend an existing basic code on and on again - there is no "SUPER" system for games. It's always specialized code for certain game situations. Oh, I hate game programming for that fact. I like universal solution, so, game programming is never universal. Especially if you want to create something... unique.



For some reason I removed my LastFM-Link. I feel it was the right thing to do.

Surprise, surprise, it works!

Exactly! Half-Life 2 works on a T30 with 2.0 Ghz and an ATI Mobility radeon 7500. The interesting thing is how the engines clips too detailed/far away objects depending on how much time they need to render. I think there's some kind of a frame time counter and a special render order. Or it's a pre-calculated "needs much or not" value to determine the complexity and needed render time for each object in the game. Cool thing is: you can see all important things like enemies and level geometry but no other dynamically placed objects. That's interesting, because it reflects how unnecessary all the other objects are. Valve, that's seriously interesting! Most game engines won't do this. Now I like Valve more than before, not that I didn't like Valve. I really like Valve, expecially because I can Valve's games with some an extremely low-end hardware without any problems! Not to say that I LOVE Valve! Ok, time to cool down. Now it's time to test some source mods and multiplayer games! There's an probably good Perfect Dark-themed mod for HL 2: Deathmatch. And I like to try this mod! I haven't played Perfect Dark in multplayer mode since five or four ages! Perfect Dark is fun as hell and entertaining in Multiplayer mode. Not to say that it's the only multiplayer FPS I would like to play with my friends. Ok, ok - I would player Stalker, too. But Perfect Dark is really just perfect. It was the first shooter I played and still the coolest one. No dual control, no boring storyline, a ton of new weapons every level and a fucking cool awesome, horny-making multiplayer mode. YES. It's that awesome. Too bad that Rare's Perfect Dark remake is XBox360 only... Damn you, Rare! I hope somebody will make a similarly awesome shooter for the Wii some day. It doesn't have to Perfect, it doesn't have be Dark, but... it has to be GOOD and PLAYABLE. Not BORING and MULTIPLAYER ONLINE ONLY (Like The Conduit, which sucked pretty much! With tons of different weapons, a cool soundtrack, good levels and overall feeling. There is just no decent real FPS on the Wii! Metroid isn't really that FPS. It's more exploration and adventure-style. But I will wait. I'm waiting for the REAL SHOOTER on the Wii. Like Perfect Dark was it on the N64.

Good times, good times.


Let's simulate post-apocalytic, nuclear worlds!

Yay, finished Fallout 3 in about, 5 days? A well-made and entertaining game, a perfect post-apocalyptic 1950s-based survival simulation! Ok, it's not that perfect, especially because it's too simple to become invincible in the game. I reached the level 20 (maximal) before even finishing the second main quest. It's a shame! And boring after a while. Of course, the quest design is good, Bethesda's engine makes perfect use of streamed content at high performance without beeing unplayable buggy. But it's just not that entertaining if you can beat every enemies with one or two headshots (not even counting overpowered, nuclear weapons!). A reduction of gameplay, nothing left but exploration and quests without rewarding rewards. This game's problems are:

  • too much xp for enemies/quests
  • too low maximum level
  • too few enemy types (it get's really boring after a while)
  • too many non-unique rewards/too much money-only rewards

Well, that's what I encountered after playing this game. After all, it's still a very entertaining game with good graphics, a brutal world and awesome 50s design choices. Hunting for artifacts of the past in Fallout 3's wasteland, I felt like in a combination of Bioshock, Stalker and partly Oblivion (same movement feeling, same streaming). Also, the radio is a cool but extremely un-used feature. Just a few radio channels and only one real channel with charming old music. I found the radio feature beautiful but not that good implemented. A shame. Having a dozen of interesting military/SOS/music signals would have been sooo awesome! Well, It's still good.

I'm looking forward to playing the new S.T.A.L.K.E.R., which will be released in November, here in Germany. Although the release date isn't available, Precursors will probably be a nice experience. Hhmmmmm.... I have a better idea. I'll try to get HL2 working on my laptop. An ATI Mobility Radeon 7500 awaits my torture!



Well, let's be honest: the banner SUCKED. Text has also a lot of beauty.



Well, this is more a reskin than a redesign, but I think it looks OK atm. I switched to a monochrome theme instead of the blue/orange stuff. Looks less smooth but fits the general style of what I'm doing. Also, I like the banner. But it's a bit too big and I'll probably changed it in the future.

Stuff to look at


Ok, ok. Complaining about usually non-realistic light models/calculations in video game (which aren't even close to reality) is pretty lame. I changed a lot of silly light issues and I think working well now. You may not see what's happening on these screenshots, but it has a lot of potential! Alpha blending works right now, even if you won't notice it that much I have a lot of ideas and most of them are small graphics related tweaks of the existing code. Using a second map-sized buffer for fog and other special effects could solve problems I will encounter when implementing special effects. Well, this game is huge, memory-wasting shitload of CPU consumption. But it's so cool! My own depth-less raytracer! And it's even possible to make perspective version of, I just need to hack the light ray directions. Wooo. I mean, I feel like always after creating something cool-looking. But this time I think it looks better. Much better than before. Well, if I had more motivation, I would starting making a map editor.

Glass, transparency and light color

A few days ago I implemented alpha blending for transparent characters and it happened that I had to rethink my light alghorythms. The more features you add, the more changes you have to make if it's not a modular environment. But who can say his light engine is modular? Beside this slightly boring question, I also noticed that the typical apha blending doesn't work with what I'm doing in my game. I mean alpha blending is fine but not for a realistic lightning system in a raytracer! Reality is different, and it's no wonder that alpha blending + light reflection doesn't go well together. So... I tried to find some more informations about how transparency, glass color and light do work together. Odyssey! Seriously! I'm not into light physics, but I doubt somebody more into it could help me here. That always bugged me about physics and math: Everybody is saying it's soooo important, but if it comes to exact real-life-simulating things it doesn't help much. It's alsways approximation, nothing exact. Seems I have to make some light experiments again.



I plan major blog redesign. Different colors, deco pixels and a section for my music. A while ago I did my music in mini albums formats on LastFM, but that's not how I like at the moment. I think it's the best to release single tracks from time to with a little text and some informations about it. I always like to talk a lot about music and what's behind it, so this will please my musician mind.

I already used different music "hosting" sites, but I'm somehow not satisfied with them. A personal, totally customizable place for my creations is what they deserve. Not such a big, ugly and ad-filled music site... No. Just music and text. That's all what I need at the moment.


Spread the night light

Since I discovered that some personal problems which keep you away from beeing happy doesn't exist at night time, there's a lot of creative potential to support! Yes, indeed. The night is my new time for creative/energetic outputs in every format. Especially because everything is silent at night, nobody will disturb and hear you! Awesome for flying through spacey mind castles of shattered visions!


Moving sound

Sometimes I hear soundtracks which immediately trigger a special mood. Soundtracks like from Super Metroid, Donkey Kong Country, etc... Soundtracks from a specific synth era, using every possible bit of sound they can get for the best atmosphere they ever thought of. Not that it is like that, but they do a very moody job especially for low-frequency sounds. Listening to Donkey Kong Country, I think it's because they mostly have a rythmic and ambientish base. I rarely hear real themes or melodies. They're are silent or not memorable. They only part surviving is the combination and you'll never get the chance to reproduce the feeling... Strange. And Super Metroid does by repeating special parts over and over again. I just say Maridia or the Oil Swamp thingy.


Pushing for Pushing's sake

I had two fabulous days while playing the first game in the Jak and Daxter series. It was great! It was an awesome and fair game, I collected 98% items in just two days and I'm still motivated. That's a point, 'cause I tend to drop games after getting bored. So, the second part isn't as motiving as the first, which leads me to a more productive activity: my game. I did some major redesigns and class-ifications, next part making an editor gui and the level editor itself. It's not that the engine is finished, but I it could be nice to have the possibility for quicking editing or so. Hardcoded levels suck. Saving is another part. But I think the tile structure is ready to export at the moment. It's not necessary, but some new graphic effects get better when using special editors.

A problem is that I like to make more eye candy for it instead of boring GUI stuff. I had that awesome bumpmapping idea where every tile has informations about it's rotation (which I can use for physics, too!). But bumpmapping needs a lot of rotation/height information and I don't like to add another hardcoded level (SUCKS ASS). I also had an idea how to display transparent characters using monochrome fonts in SDL, which requires a major rewrite whereas bumpmapping is pretty minor. Adding some kind of fake specular light (I think it was called like that) will be possible too. Yeah, the rotation could be really useful for me. And I will take the time to think about a non-quick'n-dirty solution for it. A system where you can specify angle to angle ranges for color fades, fancy shading... or other dynamic appereances. I like the idea and I think I'm trying to compensate my lack of 3D knowledge with it. You know... I don't have a 3D engine (display-wise) and I'm just pushing it away, ignoring a necessary pull. Otherwise, I came far enough to continue my engine. I have a lot of ideas, mostly graphical stuff. The physics part of it is different and I don't want to work on it until I finished the graphical base. The physics will be simple, small one-tile objects only. Pretty basic. Physics is a base for system and my AI system. AI without physics is stupid, because the AI has to USE and KNOW the physics before it can walk and interact with the world properly. I don't want to have a game where enemies shoot and walk like there's no climbable wall between it and the hero.

Eye candy first. I like eye candy. It's something I can code over and over again without getting bored. Sparkles! Oh, no. I hate particles. My game won't have so many particles. It will have fx layers, lightning and shadows. I even had an idea for a special cloud layer using the depth buffer... A real cloud information on the map would be VERY hard to calculate. I know there's a possibility to reach volumetric clouds when combining a 3D version of dirty rectangles (not for graphics but for map tiles) where you update every cloud tile within a bounding box. The same goes for fire and other stuff and I hope it will work good. Performance may be a problem for the the (hopefully no too non-soon) completely unleashed game, I think. Atleast for my 2.0 GHz Intel laptop.

Oh nose, there's so much to do and to think. I will start thinking now, like always.


F5 your code

I'm currently refreshing and reworking existing code for my game. That means:

- cleaning up messy prototype code
- turn multiple equal codelines into useful classes
- improve definition/declaration structure
- make classes out of everything to ensure minimally copied data

All in all I'm not in the mood for coding ground-breaking new technologies, but maintaining code structure and the code itself is the perfect work atm. Some days are full of amazing code ideas (the "OMG" days) where I'm just coding a couple of necessary lines needed to get the new feature working. Days like today are generally more productive or let's say effective. A solid code in a game is a solid code and shouldn't be changed until you really need to. A messy and ugly prototype code can only get worse, exploiting non-OMG days to clean up is always a good idea. And I think non-OMG days are followed by OMG days and vice versa. A chessboard of work. Checkmate.


Quaterniated Matrix

I just read this and I'm stunned. Totally. I never knew that 3D math was so simple behind. It's nothing more than some fancy matrix operations! OK, OK, I know that 3D math is a bit more, but always asked myself what the magic behind matrices and quaternions and so on is. Now I know it! Not exactly and entirely, but I have a hunch how it works and that's also an inportant thing. Maybe it's like that I just need a giant list how everything works and can apply it after reading for a while. A phenomenon I came across after finishing pre-university school. But that's a good thing - most people understand it far better when learning it day-by-day. I like the full program, but only if it's theoretical stuff where you have to understand and combine, not to use. Actually using all these calculations is another story and will usually result in weird experimention cycles. So, what I understood is something that will be helpful for the rest of my life! Atleast scientifically. And I'm looking forward using it in programming projects. Maybe I will even be able to write a complete software 3D renderer. My dream could become reality!