Author Topic: Increase cost with age  (Read 3100 times)

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
Increase cost with age
« on: October 25, 2006, 01:28:00 AM »
I've been wondering about the "increase cost log(age)" function. If I set the cost to 1 starting at 1000 then when a bot reaches 1000 cycles it will lose 1 extra energy each cycle the rest of it's life. Whouldn't it be better that the cost increases the older the bot gets, like when the bot reaches 2000 cycles then it whould lose 2 energy each cycle, at 3000 cycles it loses 3 energy and so on. I compared bots with different age and it doesn't seem to be working that way now.
The internet is corrupt and controlled by criminally minded people.

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
Increase cost with age
« Reply #1 on: October 25, 2006, 11:11:14 AM »
Quote from: Testlund
I've been wondering about the "increase cost log(age)" function. If I set the cost to 1 starting at 1000 then when a bot reaches 1000 cycles it will lose 1 extra energy each cycle the rest of it's life. Whouldn't it be better that the cost increases the older the bot gets, like when the bot reaches 2000 cycles then it whould lose 2 energy each cycle, at 3000 cycles it loses 3 energy and so on. I compared bots with different age and it doesn't seem to be working that way now.
The option increases the cost logrithmicly by multiplying the cost by log(age).  In your example, at age 1000, the cost would be 3.  At 10,000, it would be 4.  At 100,000 it would be 5.

If you are requesting that the cost be increased linearly, that would be a new feature request though at the time I implemented this, the consensus was for logarithmic.
Many beers....

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
Increase cost with age
« Reply #2 on: October 25, 2006, 01:13:19 PM »
I guess that explains it why a 21 million cycles old just loses energy slightly faster than a 40000 cycles old. If there is no objection I whould prefer it as I described.    If one wants a slower acceleration of the ageing one just sets a higher value where the costs should start. Setting it at 10000 cycles for instance whould mean the cost whould get multiplied every 10000 cycles. What do you think?
The internet is corrupt and controlled by criminally minded people.

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
Increase cost with age
« Reply #3 on: October 25, 2006, 01:17:00 PM »
How about I add a new option (in addition to the current log(N) functionality) which would increase the age cost (X) by some fraction of the age cost (lets call this fraction Y) beginning at age Z and you get to set X, Y and Z?

So, for example, if you set the age cost to 1 and choose to increase it by 0.001 of the age cost per cycle beginning at 1000 cycles then you would have the behaviour you site.
« Last Edit: October 25, 2006, 01:28:19 PM by EricL »
Many beers....

Offline Testlund

  • Bot God
  • *****
  • Posts: 1574
    • View Profile
Increase cost with age
« Reply #4 on: October 25, 2006, 01:52:04 PM »
Yeah, that whould be cool I think. It whould also minimizing the risk of overpopulation slowing down the computer some what.  
The internet is corrupt and controlled by criminally minded people.

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
Increase cost with age
« Reply #5 on: October 25, 2006, 03:37:50 PM »
Quote from: Testlund
Yeah, that whould be cool I think.
Your wish is... well, I had a little time this morning.  This feature is implemented as of 2.42.8c (as well as the CostX reinstatement feature).

2.42.8c Buddy Drop Download
Many beers....