Advertisement

Levels of Abstraction

Started by March 29, 2009 08:02 PM
50 comments, last by Platinum_Dragon 15 years, 10 months ago
In your system, I could create a character that steals everyone's stuff, have the omniscient government increase my bounty, hide, and rendervouz with my alt to earn the bounty I worked using my thief.
Maybe I need a system to track down so that only one client comes from one computer? Another thing is the bounty/honor becomes 0 when you die.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:

  • move
  • mask
  • shift
  • move
  • mask
  • shift
  • or

    So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
Advertisement
It wouldn't matter if you restrict one client per computer. Say A and B are friends, A robs B, B kills A to get bounty, they they reverse roles. The problem is that your "bounty" money comes from nowhere. [Edit: actually that wasn't the problem. The problem is that game characters are disposible.]
You would rather have the bounty restricted to the player whom lost something? Well, that way would make it so that stronger player has an advantage, because weaker players cannot even give out bounties because they cannot pay them all. Inflation is, by all ends of the economic scale, a neccessity for the economy to grow. Money in, money out, but inflation is needed. The inflation just have to be minimized to near an ideal value for the economy to flow well. Not really, because the mechanics is that all equipments are lost when they die. But that was too general. Maybe I should change it so that they lose everything in possesion of the charater, that is, only items that are stored away, and money in the bank are safe from lost when they die.

Maybe to stop the abuse from occuring,
1) I should limit it so that if someone rob you then you cannot get that bounty.
2) I should limit it so that you cannot get a bounty that came from you earlier, ie the bounty is trace to its source.
3) You cannot get Honor that came from you, the honor is trace from its source

2 also cover the situation in 1. In this way, they are limit to how they can abuse the situation.

#2 and #3 are essentially the same thing, but they are both there to stop others from arguing about the wordings.








I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:

  • move
  • mask
  • shift
  • move
  • mask
  • shift
  • or

    So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
What is the purpose of the bounty system in your game?

Are you trying to implement a sort of business model or some way to distinguish between honorable killing and dishonorable killing?

Not much, except that there are just a few mechanics that shows you how much you have accomplished before your character dies. Virtually, it is a way to represent how life is not that long and to make the players "graduate" from games. I just want a game that can bring "life" experience to them, and remind them about their "life," and not just playing games.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:

  • move
  • mask
  • shift
  • move
  • mask
  • shift
  • or

    So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
Advertisement
What are some life experiences you want to depict in the game?



I don't want to appear to be asking random questions, so the following is my perspective: I don't think that it is efficient to simulate large system to depict an experience. So if you have list of specific experiences you want to present, you could design specific modules about them. If you want to protray life as a cadet, then you could code specific episode about typical of distinctive features of that life experience. In my perspective the designer's role is to simply bring the player those experiences. If those experiences are meaningful to the player, they would be part of the player's life. So in some sense I am talking about a reversed strategy (read: reversed relation between the game and the player):

The game is not trying to reflect the abilities of the player, to model the player, or to measure the player's potential in any way. The game is only providing lessons for the player, who could directly benefit from it in real life. The only purpose of the game is to augment the player's potential.

In this philosophy, the designer is not trying to make the player work hard to get something in the game world. The designer is trying to provide something that is directly useful for the player in the real world.



What do you think about a game where the player plays a cadet and the game is about the player learning leadership? That could a type of life experience that a player would find interesting to learn by playing a game, and would return to your original focus for the game(?)











Maybe I'm too broad on my ideas. If I were to have the game as creating lessons that can be implemented in life, than it seems that some of my designs are going against my goal. Being a game about learning leardership is one thing, but we all know that not everyone will become a leader. I first have to teach them to be a follower before I can teach them to be a leader. Even if they know how to be a leader, this knowledge will only improve on how they can be a follower. To know how to be a follower and to know how to be a leader are both needed for someone to be a leader.

I really need to somehow be able to implement both aspect of being a leader and being a follower.

I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:

  • move
  • mask
  • shift
  • move
  • mask
  • shift
  • or

    So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?
I think that is fine. Your game could maintain the sequence that the player first learns to be a follower, then learns to be a leader.

There are half a dozen types/styles of leaders. Since almost anyone is in some kind of hierarchy, most players would be in a leading position in some settings. But I think that an interesting point is that some people have a role as a leader but they don't realize that they have such a role. For instances, all parents have roles as leaders. Any person that influences the objective of another person is a leader. So a salesman, when speaking to a protential customer, is a leader. Advertising is leading. Blogging and posting are also leading. Some posts have no effect because the don't lead well. So leadership is applicable in many situations.



I agree, and I will say that my goal is for them to see that every position is a leader and follower at the same time.
I use QueryPerformanceFrequency(), and the result averages to 8 nanoseconds or about 13 cpu cycles (1.66GHz CPU). Is that reasonable?
I though that the assembly equivalent to accessing unaligned data would be something similar to this order:

  • move
  • mask
  • shift
  • move
  • mask
  • shift
  • or

    So it seems reasonable to say that it takes 14 cycles for unaligned data since we'll have to do the series of instructions once to access and once to assign?

This topic is closed to new replies.

Advertisement