Author Topic: Proposed changes to .fixang  (Read 2302 times)

Offline EricL

  • Administrator
  • Bot God
  • *****
  • Posts: 2266
    • View Profile
Proposed changes to .fixang
« on: April 18, 2008, 11:42:48 AM »
Right now (2.43.1i) there is no way for a bot to indicate that the angle for a hardened tie it created should rotate freely.  Additionally, the current code behind .fixang does not distribute the range of values evenly across the sysvar mutation space, resulting in the vast majority of mutations fixing the tie angle at 628.  Currently, the code pegs angle requests below -628 or above 628 at -628 or 628 respectivly, which are equilvalent.  As such, fewer than 2% of fixang values result in angles other than 628.

I propsoe the following change to the way .fixang works:

Any negative value allows the tie to rotate freely.
Any postiive (or zero) value fixes the angle at the position given by the value mod 1256.

Thus, half of all values allow for free rotation, half fix the tie angle and the values that do fix the tie angle are evenly distributed with equal probability for each angle.

Similar changes would be made for the .tieangX sysvars (temporarily until a better multi-tie-per-cycle control paradym obviates these).

The behaviour for mutlibots which have a tie harden without them explicitly specifying a tie angle could either be left unchanged (the tie angle becomes fix at the current angle on the cycle it hardens) or coudl be modified so that there is a 50/50 probability of it fixing at the current angle or remainign unfixed.  I prefer the latter.

Comments?
« Last Edit: April 18, 2008, 11:44:00 AM by EricL »
Many beers....