POD-bot back into shape.

Dedicated..

[Pierre-Marie Baty]

@cooley

This is IMPOSSSSSSSSSSSIBBBBBBBBBBLLLLEEEEEEEEEEEEEEEEEEEH HHHHHHHH
it just DEFIES the LOGIC
man it's just as if I was doing

int i = 1;
assert (i != 0)

and the assertion failed!!!!!!!!!!!!!
I am PUZZLED
if some coder here (Wei Mingzhi?) has an explanation to this bug I WANT to know it

http://racc.bots-united.com/podbot_mm_i386.so
Smoke escaping from my ears.


@walten: what is your metamod version

I know its working somewhat.

I have meta 1.17

here is a screen shot of the text..


http://forums.bots-united.com/attach...ntid=106&stc=1

[Pierre-Marie Baty]

@walten: no matter, something is wrong with your setup. Try deleting (or moving temporarily) all your .cfg files (server.cfg, listenserver.cfg, game.cfg, autoexec.cfg, etc.) reinstall the bot from the ZIP file (use the default config), edit your plugins.ini file so as to have only one line for loading the podbot DLL *WITHOUT* any other plugin, run a dedicated server on developer mode and try again.

Using a STOCK metamod setup and a STOCK bot setup it cannot fail. Once it works try adding your other plugins and config one by one to see which one is causing the problem.

I dont have the full meta mod installed. Only the stuff that came with admin mod. Would that cause a problem?

i have a monster 750 btw

[sPlOrYgOn]

metamod is only metamod.dll and the plugins.ini
adminmod gives both so you have all of metamod unless you mean you don't have the optional metamod file like exec.cfg or config.ini

btw, nice job fixing podbot PMB.
and is there a way to show their skill level?
I have detailed names on but it only shows [POD] infront.

PMB


This must be conflict with either adminmod or statsme or you have fixed the problem. I will not be able to tease this out until tomorrow but the newest biuld worked great with adminmod and statsme off.

Cooley

[Pierre-Marie Baty]

I have not "fixed" the problem, I have removed the failing assert(). But it's an ugly fix because the assertion still fails, meaning that an out of bounds value exist in ideal_yaw when pfnRunPlayerMove() is called. I am completely puzzled why it fails. If somebody has an explanation I'm all ears.

The code is

   // save the previous speed (for checking if stuck)
   pBot->prev_speed = fabs (pBot->f_move_speed);
 
   // Reset Damage
   pBot->iLastDamageType = -1;
 
   UTIL_ClampVector (&pEdict->v.angles);
   UTIL_ClampVector (&pEdict->v.v_angle);
   UTIL_ClampAngle (&pEdict->v.idealpitch);
   UTIL_ClampAngle (&pEdict->v.ideal_yaw);
 
   assert ((pEdict->v.v_angle.x > -181.0) && (pEdict->v.v_angle.x < 181.0));
   assert ((pEdict->v.v_angle.y > -181.0) && (pEdict->v.v_angle.y < 181.0));
   assert ((pEdict->v.angles.x > -181.0) && (pEdict->v.angles.x < 181.0));
   assert ((pEdict->v.angles.y > -181.0) && (pEdict->v.angles.y < 181.0));
   assert ((pEdict->v.ideal_yaw > -181.0) && (pEdict->v.ideal_yaw < 181.0)); // ASSERTION FAILURE
   assert ((pEdict->v.idealpitch > -181.0) && (pEdict->v.idealpitch < 181.0));
 
   g_engfuncs.pfnRunPlayerMove (pEdict,
					 vecMoveAngles,
					 pBot->f_move_speed,
					 pBot->f_sidemove_speed,
					 0,
					 pEdict->v.button,
					 0,
					 msecval);

   return;
}

and UTIL_ClampAngle() which fixes ideal_yaw before the assert is

void UTIL_ClampAngle (float *fAngle)
{
   if (*fAngle >= 180)
	  *fAngle -= 360 * ((int) (*fAngle / 360) + 1);
   if (*fAngle < -180)
	  *fAngle += 360 * ((int) (-*fAngle / 360) + 1); // don't forget the "-"
 
   if ((*fAngle >= 180) || (*fAngle < -180))
	  *fAngle = 0; // heck, if we're still above the limit then something's REALLY fuckedup!
}

[Terran]

@walten:
Who do you invoke those commands? Do you type them in the servers console (or HLSW) or in your clients console? For the later one you have to "rcon" them.

PMB,

With the new .so the POD bot plugin fails to load with map changes

L 03/12/2004 - 07:47:20: [META] ERROR: dll: Not unloading plugin 'POD-Bot'; already unloaded (status=opened)
L 03/12/2004 - 07:47:20: [META] ERROR: dll: Failed to unload plugin 'POD-Bot' for reloading


Thanks,
Cooley