code to get the illumination at any point on the map

[Immortal_BLG]

35 is SVC_WEAPONANIM.

I can't help you, because my code does not send messages (with the exception of the ShowMagic, but it is not involved), but I also don't know what message is sent during the construction of the message SVC_WEAPONANIM.

P.S. in order that I could help - you should give me more information

[KWo]

I'm not sure if we are talking about the same group of messages. I thouht it would be rather one of those captured i.e. in pfnMessageBegin function - like "BarTime", "ResetHUD", "AmmoPickup" and so on. I mean those I need to know for them MSG_ID (GET_USER_MSG_ID). Actually I cannot reproduce that bug on a windows server (it happened only on a linux server), but I found another one on windows server (not related at all to Your code). It is something with pathfinding, I couldn't believe it could happen.

hlds.exe caused an Access Violation at location 09c261da in module podbot_mm.dll Reading from location cdcdcdf9.

Registers:
eax=09d65c48 ebx=09bd8478 ecx=cdcdcdcd edx=09ccb934 esi=0012ddfc edi=0012f648
eip=09c261da esp=0012ddfc ebp=0012f648 iopl=0         nv up ei pl nz na pe nc
cs=001b  ss=0023  ds=0023  es=0023  fs=003b  gs=0000             efl=00000202

Call stack:
09C261DA  podbot_mm.dll:09C261DA  BotCheckTerrain  bot_navigate.cpp:5517
    ...
                      && (pBot->pWaypointNodes != NULL) && (!g_bRecalcVis) && (pBot->bMoveToGoal))   // KWo - 28.01.2012
                   {
>                      if (pBot->pWaypointNodes->NextNode != NULL)
                      {
                         fDistToCurWP = (pEdict->v.origin - paths[pBot->curr_wpt_index]->origin).Length();
    ...

09BFCF1A  podbot_mm.dll:09BFCF1A  BotThink  bot.cpp:8865
    ...
          && (pBot->f_spawn_time + (((g_fTimeRoundEnd + g_f_cv_FreezeTime > gpGlobals->time) && (g_fTimeRoundEnd < gpGlobals->time)) ? (g_f_cv_FreezeTime + 0.0):(0.2)) < gpGlobals->time)) // KWo - 28.05.2010
       {
>          BotCheckTerrain(pBot);
       }
    
    ...

09C355BC  podbot_mm.dll:09C355BC  StartFrame  dll.cpp:2249
    ...
          {
             g_i_botthink_index = bot_index; // KWo - 02.05.2006
>             BotThink (&bots[bot_index]);
             g_iNum_bots++;
          }
    ...

6B2075F6  metamod.dll:6B2075F6

It looks like is not enought to check if the pointer is null or not to prevent the crash...
That ">" shows lines of code execution in last called functions.

[KWo]

It looks like I have fixed the problem with windows crash (the crash dump above), the windows server is not crashing already more than 2 hours (before I needed to wait about 20 minutes to crash). The problem with linux (FATAL ERROR - msg 35) still exists there, but I had to wait about 2 hours now (after the change of the code concerning to the windows crash - not sure why it didn't crash on linux for the same reason). Yesterday it was a crash about every 10 minutes. Maybe there is some problem with network (linux test server is a virtual machine)? I couldn't reproduce that problem on a windows server. Why the hell does it behave different way, if for bot servers I'm using the same configuration of podbot and AMX X files (CSDM)?

[Immortal_BLG]

I'm sorry that it is so late wrote.
Don't even know how to help you, try to debug bots on LINUX - this is the only thing I can advise to you.... Sorry

[KWo]

It looks like on real linux servers it doesn't crash anymore since latest update.
http://forums.bots-united.com/showth...4409#post64409
It might be some problem with my virtual PC and network access...

[tschumann]

The GoldSource engine is currently getting some love and I raised this as an issue on the GitHub tracker: https://github.com/ValveSoftware/halflife/issues/1774