DM Server: Spawn Farthest?

[console]

It would be pretty easy to keep the random spawning, but add in enough smarts in the code to try to avoid telefrags, like +sf does...

Could be done, if there's enough interest.

[[BTF]Gator]

i think it is funny, and consider that situation everytime i jump over a spawn pad.  In the end, its just a silly game, and i personally dont get too hung up over the score.  Unless someone is winning, then that is f***ing bullshit man!

[DaHanG]

 

[Whirlingdervish]

Wow. that sounds like quite the solution.
We get the best of both worlds, out of that plan.

 

[UUD-40]

i think it is funny, and consider that situation everytime i jump over a spawn pad.  In the end, its just a silly game, and i personally dont get too hung up over the score.  Unless someone is winning, then that is f***ing bullshit man!

 

[DaHanG]

i think it is funny, and consider that situation everytime i jump over a spawn pad.

[console]

Alrighty...

I rewrote SelectRandomDeathmatchSpawnPoint to still be random, but try to avoid telefrags.

The new code is now running on vanilla, mutant, and dm servers.

Seems to be working nicely so far.

Here's the code:

Code: [Select]

edict_t *SelectRandomDeathmatchSpawnPoint (void)
{
        // avoid spawn pads where the client is closer than LIKELY_TELEFRAG_DIST
        const int LIKELY_TELEFRAG_DIST = 128;
        const int MAX_SPAWN_POINTS = 256;
        edict_t *preferred_spots[MAX_SPAWN_POINTS];
        edict_t *all_spots[MAX_SPAWN_POINTS];
        int preferred_count = 0;
        int all_count = 0;
        edict_t *spot = NULL;
        float range;

        while ((spot = G_Find (spot, FOFS(classname), "info_player_deathmatch")) != NULL)
        {
                range = PlayersRangeFromSpot(spot);
                if (range > LIKELY_TELEFRAG_DIST)
                {
                        if (preferred_count < MAX_SPAWN_POINTS)
                        {
                                preferred_spots[preferred_count] = spot;
                                ++preferred_count;
                        }
                }

                if (all_count < MAX_SPAWN_POINTS)
                {
                        all_spots[all_count] = spot;
                        ++all_count;
                }
        }

        spot = NULL;  // we'll return NULL if no spawn points at ALL could be found

        {
                edict_t **spots;
                int spots_count;

                if (preferred_count > 0)
                {
                        spots = preferred_spots;
                        spots_count = preferred_count;
                }
                else
                {
                        spots = all_spots;
                        spots_count = all_count;
                }

                if (spots_count > 0)
                {
                        // Having read an article about how the lower bits of rand() on
                        // many systems' stlib aren't very random, we'll try to mix it
                        // up a bit here.  (hopefully in some meaningful way)

                        int selection;
                        unsigned short rndval;
                        rndval = rand();
                        rndval = (rndval >> 8) | ((rndval & 0xff) << 8);
                        rndval ^= rand();  // hopefully this isn't stupid :)

                        selection = rndval % spots_count;
                        spot = spots[selection];
                }
        }

        return spot;
}

If anyone notices anything I've done that looks stupid (i.e. buggy) please don't hesitate to let me know.


Let me know how it plays...


Regards,

[Whirlingdervish]

very cool. I'll have to try it out, after my netgraph settles down a bit...
(I seem to be getting timeouts somewhere in dallas just before the hop to The Planet)