rightfuture
Emperor
- Joined
- Mar 11, 2008
- Messages
- 1,649
AIAndy - Thank you kindly for your feedback and discussion. I always appreciate it.
I actually do think this way with no apologies.
I feel strongly about you guys with good and specific reasons, and can accurately and specifically defend most of you guys' individual exceptionality in different ways.
Sometimes I may come off a little positive and optimistic (yeah, I know), but I do feel strongly about C2C, and your work. I also feel strongly about C2C's progress and potential, and I have no illusions about it's limitations or reality. Things are what you make of them, and I'm neither blowing 'smoke' nor being careless with my praise.
Just because it's glowing or descriptive doesn't mean it's not accurate nor deserved.
When I meet new people, they can somewhat frequently think that I'm exaggerating because I make some claims that seem unlikely, but almost everyone who eventually gets to know me, Will pull me aside at some point and tell me that they were wrong about that. It is a matter of perspective what you believe ATM. I would rather you find out for yourself than take my word for it or try to change it.
I concede that I should probably tailor my conversations for a disbelieving and skeptical audience, but I've discovered that some people won't believe you anyway, and if it's accurate then people will be able to see for themselves, even if it takes a long time. I neither claim that I am right about everything I say, or that my desire to be positive and appreciate in my obviously 'PR' fashion is not a little 'enthusiastic'. Please factor-in that I fight pretty severe exhaustion from sleep apnea, and that I have had to overcome some terrible odds more times than I like to count.
I hope you all take me at the truth you can see for yourself, and give me a chance to prove that I believe it.
As far as my content I will be more specific.
As far as the links I'm not sure if you were referring to the real time infinite detail graphic rendering video, the elite procedural generation video or my statements on the ideas in general. I did mean what I said about everybody and C2C.
My purpose was to draw everyone attention to the ability to 'vary up' formulas to produce complex behavior and content.
The links I used above were there to illustrate in general, the relatively complex behavior(and content) can be simulated with simple formulas.
To simplify even further, imagine a fixed fractal formula for a tree leaf/stem being applied to a procedural generation formula (basically a random appearing -{actually non-random} seed/formula that can generate/re-generate content or patterns on the fly). It is basically a linear formula that repeatedly iterates(writes over itself) to produce a much more complex texture.
It can be used to draw a leaf, a tree, and even a forest using techniques to apply that fractal texture with slight variations. This can be done using very similiar formulas to apply to map and terrain generation, (like the mapscripts being mentioned in this thread). With slight variations fractals can create complicated.
Simple behavior AI, can likewise be generated using certain non-linear formulas which can exhibit slightly varied iterations(basically loops) that can simulate basic behaviors.
Simple variations (with non-random seed formulas like procedural generation and fractals) can tweak the formulas direction so that it settles into a slightly different behavior. ( I know you understand most of this, just illustrating for beginers)
For example, Imagine the scouting behavior where a unit cycles between 2 points.
- This is usually a linear pathing formula. Go from point A to point B. And reverse it to form a loop. The formula basically outputs x,y coordinates between the x/y map coordinate of A and B.
- A looped non-linear formula is basically a circle that iterates (or changes slightly) in a known pattern (with some potential variations). They can create constrained behavior that follows some interesting rules and appears natural and potentially intelligent. This and some very basic learning algorithms were probably the explanation for the intelligent animal behavior in the LionHead video game Black and White. It doesn't take much to create complex behavior but you have to understand it, to make it work. Imagine scouting patterns that are not always predictable, but can deviate slightly and keep the game interesting. I am always a fan of 'living world' type behavior. Sim Earth, Sim City like games that use behavior that can be approximately modeled with patterns and statistic trending.
- Like in the video on procedural generation, applying simple fractal formulas can make non-random patterns seem quite random and differentiated. Think of the difference in programs like MS Paint and Photoshop. A brush in MS Paint from a single x,y data point would probably be a dot or a point. A brush in Photoshop can be textured with complex 'fractal' patterns that easily make 'paint' an extremely large variety of textures with simple formulas that are still based on a single x,y data point. To make things appear more complex,varied, and detailed fractals can be very useful.
Used correctly, any of these formulas can produce and re-produce relatively complex data relatively little management or resources. All you have to do is look at higher resolution terrain generation software of most types to see it working faster than you suspect.
(as you said AIANDY some of those links were intended as advertising and somewhat suspect in their benefit. Most people do not spend money to advertise vaporware, though they will talk it up with their purposes in mind. All in all they can be potentially useful. I think their underlying concepts were sound or I would not have shared them. And I also think the discussion was worthy of helping this thread and more.)
I hope I have made myself more clear. Please excuse my tendency to invariably and inconsistently deal my words and thoughts. Severe exhaustion will do that to anyone, I just have to fight it more than most.
My point here is that I am pretty sure that what I'm saying has more value than you may consider.
I do have a tendency to over simplify or make things more complex then they have to be. Mostly because I'm distracted.
So please consider my statements and arguments because I believe strongly that they could help C2C. Please ask me to clarify if you think I'm approaching this too simply. I may only know so much, but I am usually a lot more knowledgeable that people think.
We all benefit from feedback.
Edit would a text version of the overlays be enough? Something like a trace print that comes out with some of the map scripts identifying the terrain in each plot as a character?
