🚀 go-pugleaf

Hi.  Here is a small patch to "improve" the timing logic for
pl74.  Apparently I was giving it way too much time and a couple
of versions which ran on ICS was suddenly losing on time.  Sorry
for the problems.

-- 
Chua Kong Sian   <kongsian@nsrc.nus.sg>
National Supercomputing Research Centre
Singapore



*** ../src/search.c	Wed Feb 22 08:46:50 1995
--- search.c	Sat Mar  4 09:41:41 1995
***************
*** 269,275 ****
  	      /* calculate avg time per move remaining */
  
  	      ResponseTime = TimeControl.clock[side] / TimeControl.moves[side];
! 	      ResponseTime += TCadd * 3 / 4;
  	      TCleft = (int) ResponseTime / 5;
  	      if (TimeControl.moves[side] < 5)
  		TCcount = MAXTCCOUNTX - 10;
--- 269,276 ----
  	      /* calculate avg time per move remaining */
  
  	      ResponseTime = TimeControl.clock[side] / TimeControl.moves[side];
! 	      ResponseTime = ResponseTime * 2 / 3;
! 	      ResponseTime += TCadd / 2;
  	      TCleft = (int) ResponseTime / 5;
  	      if (TimeControl.moves[side] < 5)
  		TCcount = MAXTCCOUNTX - 10;
***************
*** 1245,1251 ****
  #endif
  	      ElapsedTime (2);
  	      TCcount++;
- 	      if (et > ResponseTime) flag.timeout = true;
  	      if (!background && Sdepth > 2)
  		{
  		  if (best < alpha)
--- 1246,1251 ----