Re: Even MORE galaxy suggestions...

From: desj@ccr-p.ida.org (David desJardins)

Date: Thu, 06 Aug 1992 23:56:57 +0000

Michael Hall <hall@kronos.arc.nasa.gov> writes:
>The offensive player adapts by breaking his attack into N pieces.  The
>largest mass (but low tech) force is sent *away* from the real targets,
>diverting the intercepting defensive forces.  Then the N-1 attack groups
>pound the defender's planets senseless just like there were no interception
>command at all.  The offensive player loses his decoy fleet, but the
>defender loses N-1 worlds.

I think an improvement on the Intercept command would be to intercept
*all* forces which leave the target planet for destinations within
range.  The intercept group could be divided so that it intercepts each
enemy force with a fraction of its ships proportional to the size of the
enemy force (but at least one in every case, if possible).

Thus, suppose I have a group with 13 ships, and I order that group to
intercept enemy ships from planet X.  The enemy sends 50 mass units from
X to Y, and 30 mass units from X to Z, and 10 mass units stay at X.
Then 2 of my ships would be sent to X, and 7 to Y, and 4 to Z.

I'll even contribute some code to this effect.  All of the changes are
in interceptphase().  (This is to version 2.92, so I don't know how much
use it is, but it wasn't too much work.)

David desJardins

void interceptphase (void)
{
  int i,j,jj,k,*n;
  double maxratio,minmass,maxdist,*v;
  player *P,*P2;
  group *g,*g2;
  planet *p,*p2;

i = 1;
  for (p=planets; p; p=p->next)
    i++;
  v = alloc (i * sizeof (double));
  n = alloc (i * sizeof (int));
  for (P=players; P; P=P->next)
    for (g=P->groups; g; g=g->next)
      if (g->dist < 0)
	{
	  maxdist = 2 * g->type->drive * g->tech.drive * DRIVEMAGIC / shipmass (g);
	  p = g->from;
	  memset (v,0,i * sizeof (double));
	  for (P2=players; P2; P2=P2->next)
	    if (P2 != P)
	      for (g2=P2->groups; g2; g2=g2->next)
		if (g2->from == p && g2->where != p &&
		    g2->dist == dist (p,g2->where) &&
		    dist (g->where,g2->where) < maxdist)
		  v[ptonum (planets,g2->where)] +=
		    g2->ships * shipmass (g2);
	  memset (n,0,i * sizeof (int));
	  k = g->ships;
	  for (j=1; j<i; j++)
	    if (v[j] > 0.)
	      {
		n[j]++;
		k--;
	      }
	  if (k == g->ships)
	    continue;
	  while (k > 0)
	    {
	      maxratio = 0.;
	      for (j=1; j<i; j++)
		if (v[j] > n[j]*maxratio)
		  {
		    maxratio = v[j]/n[j];
		    jj = j;
		  }
	      n[jj]++;
	      k--;
	    }
	  while (k < 0)
	    {
	      minmass = 0.;
	      for (j=1; j<i; j++)
		if (v[j] < minmass)
		  {
		    minmass = v[j];
		    jj = j;
		  }
	      n[jj]--;
	      k++;
	    }
	  for (j=1,p2=planets; j!=i; j++,p2=p2->next)
	    if (n[j] > 0)
	      {
		if (n[j] != g->ships)
		  {
		    g2 = alloc (sizeof (group));
		    *g2 = *g;
		    addlist (&P->groups,g2);
		    g->ships -= n[j];
		    g2->ships = n[j];
		    send (g2,p2);
		  }
		else
		  send (g,p2);
	      }
	}
  free (v);
  free (n);
}

References

Even MORE galaxy suggestions...
d8ilvja@dtek.chalmers.se (Jakob Ilves)
Wed, 05 Aug 1992 21:41:09 +0000
Re: Even MORE galaxy suggestions...
hall@kronos.arc.nasa.gov (Michael Hall)
Thu, 06 Aug 1992 20:03:25 +0000