Posted by Cash Car Star on Thu Jan 20th at 8:58pm 2005
Cash Car Star
member
1260 posts
301 snarkmarks
Registered: Apr 7th 2002
Location: Connecticut (sigh)
Occupation: post-student
Posted by DrGlass on Fri Jan 21st at 12:12am 2005
DrGlass
member
1825 posts
293 snarkmarks
Registered: Dec 12th 2004
Location: USA
Occupation: 2D/3D digital artist
Posted by Foxpup on Fri Jan 21st at 12:19am 2005
Foxpup
member
380 posts
38 snarkmarks
Registered: Nov 26th 2004
Location: the Land of Oz

Occupation: Student
Posted by satchmo on Fri Jan 21st at 1:01am 2005
It's entirely possible. In fact, most of my single player maps are non-linear, including my latest one, Justice. (http://www.snarkpit.com/maps.php?map=1643)
But one of my first maps ever made (for the original Half-Life) was non-linear as well. (http://www.snarkpit.com/maps.php?map=1325). It's not easy to anticipate all the possible route the player might take, but that's part of the fun and challenge. It's the reason why I wrote an extensive walkthrough for this map (Escape from Black Mesa).
There are always more than one ways to beat these maps, and that adds to their re-play value.
[addsig]satchmo
member
2077 posts
396 snarkmarks
Registered: Nov 24th 2004
Location: Los Angeles, U.S.

Occupation: pediatrician
Posted by ReNo on Fri Jan 21st at 1:03am 2005
[addsig]
ReNo
member
5457 posts
933 snarkmarks
Registered: Aug 22nd 2001
Location: Scotland
Occupation: Level Designer
Posted by satchmo on Fri Jan 21st at 4:58am 2005
Hey, I live in L.A. Self-pimping is virtually expected. Some people (esp. actors) here have developed it into an art form.
[addsig]
satchmo
member
2077 posts
396 snarkmarks
Registered: Nov 24th 2004
Location: Los Angeles, U.S.

Occupation: pediatrician
Posted by Foxpup on Fri Jan 21st at 5:13am 2005
Foxpup
member
380 posts
38 snarkmarks
Registered: Nov 26th 2004
Location: the Land of Oz

Occupation: Student
Posted by Crono on Fri Jan 21st at 5:29am 2005
You're not suggesting a ridiculous amount of possibilities from each and every action are you? [addsig]
Posted by Foxpup on Fri Jan 21st at 5:47am 2005
Yes actually I am. [addsig]
Foxpup
member
380 posts
38 snarkmarks
Registered: Nov 26th 2004
Location: the Land of Oz

Occupation: Student
Posted by Crono on Fri Jan 21st at 6:05am 2005
Yes actually I am.
Then you have no idea how miserably it would fail, no offense.
Not to mention, it would take years of studying to understand that level of AI development.
HL2 (and any game with scripting for that matter) can do this "kind of" no where near the level you're thinking, though.
This wouldn't run on a normal workstation or gaming computer, to be blunt. [addsig]
Posted by Gorbachev on Fri Jan 21st at 7:25am 2005
[addsig]
Posted by Campaignjunkie on Fri Jan 21st at 7:50am 2005
[addsig]
Campaignjunkie
member
1309 posts
291 snarkmarks
Registered: Feb 12th 2002
Location: West Coast, USA
Occupation: Student
Posted by Crono on Fri Jan 21st at 8:08am 2005
You'd use fuzzy logic. It's a type of AI programming "thought process", or so you could call it. It's pretty damn complicated stuff.
Basically, you make a tree of choices. Then each node of that tree goes into sub-tree of sub-choices and so on until you get to the last node in the series (which you use logical statements to decide which to choose in each level) and that is the final "decision, action, reaction? etc.
It's a doable concept, even in a game. The only problem is this is for one aspect, for example, a personality. He's talking about for EVERY SINGLE ELEMENT in the game. THAT will not run (nor would it be anything a development team could realistically finish).
Fuzzy logic is the way to go for this type of thinking. But, yes, NOTHING in computing is random, not even a random number generator. Also, usually if something says "random" when speaking about computers, it generally means the fastest choice.
Complex stuff. What I was saying though, is something that is 10% of this level of complexity could be simulated in something like HL2, but I don't think it'd be worth it.
I'd mark this under "crazy game ideas", along with Kojima's idea to have the game disc RUINED when you die in the game, forcing you to play cautiously. [addsig]
Posted by Cash Car Star on Fri Jan 21st at 8:08pm 2005
I haven't programmed in quite some time, but I remember trying to create random numbers for dice games in QBasic. If I just used the typical random number command, the same sequence would always occur. If I used the Randomize Timer command before generating random numbers, they would be truly random (or at least I never detected any patterns). I always assumed this worked by taking a random seed at the time the command was entered, which to a computer would be far more precise than a human could ever be, and this time would be based upon when the user pressed a button to run the program. Therefore, the exact time at which it went off would be completely unpredictable. It's kind of like how if you have a precise machine toss a die, you could theoretically calculate all the forces on the die and accurately predict the result. When a human tosses a die, there's no way to maintain such precision with the spin, velocity and altitude and as such, it is completely random.
So basically I've always considered the sole source of true randomness to be results based on human action more precise than can be controlled. Anything that taps into this source can become truely random.
Cash Car Star
member
1260 posts
301 snarkmarks
Registered: Apr 7th 2002
Location: Connecticut (sigh)
Occupation: post-student
Posted by Foxpup on Wed Jan 26th at 12:03am 2005
Foxpup
member
380 posts
38 snarkmarks
Registered: Nov 26th 2004
Location: the Land of Oz

Occupation: Student
Posted by Rof on Wed Jan 26th at 12:07am 2005
Man, 20 years ago my Speccy only had a 3.5 MHz Z80 CPU. I should've overclocked it.
[addsig]
Posted by Crono on Wed Jan 26th at 1:05am 2005
For your idea to "work" we would need computers with connections that are 100% efficient (meaning, no heat is given off) and all forms of storage would need to be instantaneous. This is impossible, since we can't make a connection that is 100% efficient (which is what, when you get down to it, computers rely on for speed).
This isn't even taking into account all the flaws with every computational design ever. A true simulation of real life, which is what you're talking about, not only will never have proper hardware, but would never be completed on a software level. It's too complex. [addsig]
Posted by BlisTer on Wed Jan 26th at 2:44am 2005
it's not 100% certain that a technology which seems impossible now, will still be impossible in the future. Technology advances exponentially, and many boundaries that were predicted to be unbreakable, have already been surpassed. I'll always remember that my geography teacher told us 7 years ago that fusion would be impossible. Well guess what mr. Maes, they're almost in the construction stage of ITER.
So when you say efficient connections, i think upgraded superconductors. Could be it turns out a whole other technology will be used, but the point is that there is a chance.
[addsig]Posted by Crono on Wed Jan 26th at 2:47am 2005
Snarkpit v6.1.0 created this page in 1.0761 seconds.


