quote:Original post by bishop_pass Kind of a tough question - but I would be inclined to think that as long as you play, the realtime rule is in effect.
But then you''d get to a boring part, and you''d have to turn your game system off and on to skip to the next interesting part. Also, wouldn''t this work just as well with a human-generated storyline?
Requiring someone at the other end to constantly generate something for you doesn''t seem very wise an aspect of a game, not to mention that that person would most likely be effected by things they may have seen during that specific day etc. which may be detrimental to play.
Actually, I have this graduate CS course titled "Knowledge representation and reasoning", and this seems very similar to the kind of stuff that I have to do for the reasoning part. Except that we code in prolog.
Remember 2 years ago when I was just getting started with programming AI in prolog and I couldn''t do tree pruning and posted here? Well, I''ve mastered that art now ^_^
Well if you got to an end of an episode you''d get a screen going, "Congradulations you''ve finished episode: blah, do you wish to continue or save?" or something like that.
I think this sounds like a really good idea, though getting computer generated story lines may be rather hard (I know there''s AI research going on about computer generate story lines atm, I''m sure google will turn up someting).
quote:Original post by Raduprv On a serious note, I was thinking of doing something like that, for the quests in my game (random quests). Then I realised that the quality of computer generated quests doesn''t even coem close to real, human generated quests, so I gave up.
Given the implementation you proposed, I could see how you would believe that. I think one of the goals here is to explore an implementation, but I think the requirement for doing so is to look at the other thread I created about fictional character histories. By creating a custom fiction generator for each type of genre (or specific game) and taking heed of knowledge and symbolic methods as well as looking at AI scripts promoted by Roger Schank (not the scripting techniques used in game programming, which are an entirely different thing altogether), one could come up with an effective generator which could improve with each revision through simple addition.
_______________________________ "To understand the horse you'll find that you're going to be working on yourself. The horse will give you the answers and he will question you to see if you are sure or not."
- Ray Hunt, in Think Harmony With Horses ALU - SHRDLU - WORDNET - CYC - SWALE - AM - CD - J.M. - K.S. | CAA - BCHA - AQHA - APHA - R.H. - T.D. | 395 - SPS - GORDIE - SCMA - R.M. - G.R. - V.C. - C.F.
quote:Original post by bishop_pass Kind of a tough question - but I would be inclined to think that as long as you play, the realtime rule is in effect.
But then you''d get to a boring part, and you''d have to turn your game system off and on to skip to the next interesting part. Also, wouldn''t this work just as well with a human-generated storyline?
I''ve been a longtime proponent of a game intelligent enough to know what a lull is, and to generate something intense to happen. There are all kinds of intense things that a program can have waiting in the wings to throw at you, such as being mugged, a brawl breaking out (even if you''re a spectator), betting opportunities, something interesting to discover, etc. But before you get a preconceived notion of what I''m talking about, you''d have to learn about my ideas of Situation Creators, knowledge representation, etc., unless you already believe in what I''m saying.
_______________________________ "To understand the horse you'll find that you're going to be working on yourself. The horse will give you the answers and he will question you to see if you are sure or not."
- Ray Hunt, in Think Harmony With Horses ALU - SHRDLU - WORDNET - CYC - SWALE - AM - CD - J.M. - K.S. | CAA - BCHA - AQHA - APHA - R.H. - T.D. | 395 - SPS - GORDIE - SCMA - R.M. - G.R. - V.C. - C.F.
This would make a GBA game. You could skimp on the story a bit, and it would allow the characters to be put into interesting situations as soon as the GBA is turned on.
~~~~~ "the best thing about betting on apathy is that even when you lose, you dont care." - nethead. Download and play Slime King I.
No, I''ve actually been reading most of your posts, and think that that''s a good idea. It would be cool to have some methode of figuring out what the player likes, and then adding more of that element into the game.
So, to reinvigorate the subject at hand, the point here, is to devise an implementation that could fabricate a transitional piece of fiction between episodes. I''m an advocate of using Lisp type symbolic list like structures to encode knowledge, so even if you would want to code such a thing in C or C++, in the end, your toplevel view of the problem might be in a set of nested lists. Disagree if you wish, but please, no C++ arrays of RPG stats as we got in the last thread.
As far as specifically tackling such a problem, I believe the best way is to initially restrict the domain heavily, focusing on the methods necessary to reproduce a tiny subset of what you ultimately wish to accomplish. So, for starters, you might take the outcome of a particular episode, think of the setting for a new episode, think of an example of how the character got from episode one to episode two, and then derive with knowledge a methodology to reproduce that.
_______________________________ "To understand the horse you'll find that you're going to be working on yourself. The horse will give you the answers and he will question you to see if you are sure or not."
- Ray Hunt, in Think Harmony With Horses ALU - SHRDLU - WORDNET - CYC - SWALE - AM - CD - J.M. - K.S. | CAA - BCHA - AQHA - APHA - R.H. - T.D. | 395 - SPS - GORDIE - SCMA - R.M. - G.R. - V.C. - C.F.
I think it''s time we further evaluated this subject.
_______________________________ "To understand the horse you'll find that you're going to be working on yourself. The horse will give you the answers and he will question you to see if you are sure or not."
- Ray Hunt, in Think Harmony With Horses ALU - SHRDLU - WORDNET - CYC - SWALE - AM - CD - J.M. - K.S. | CAA - BCHA - AQHA - APHA - R.H. - T.D. | 395 - SPS - GORDIE - SCMA - R.M. - G.R. - V.C. - C.F.
Please take no offense, but I''m wondering what the point of this thread (and the others like it) is..? Are you asking for ideas, suggestions, or maybe help implementing such a thing? Are you just trying to get other people to consider the possibilities, or maybe trying to ''sell'' people on your ideas?
I think that if you could create an entirely ''random''(never the same twice, future depends on what the player did in the past, etc) game it would be pretty much the perfect game, but you need more than just history generation and fiction generation - you need a whole system of generating places, people, events, inventions, etc so that the game can at any time take the previous state and the actions of the player and generate a new scene that can contain anything real life does (if not more).
Imagine playing a game that is bigger than the real world - not difficult when it can generate anything and everything you should see when you should see it.
"Walk not the trodden path, for it has borne it's burden." -John, Flying Monk
