POD-bot back into shape.

[Austin]

Originally Posted by Pierre-Marie Baty

Austin: I'd say... according to the expertise my nose gives me, a few hours of work (to do something clean, I mean. It's the entire aiming code that needs to be changed, I don't want to "patch" it). But have you tried the bot yet ?

No I havn't, but I can't wait to try it. This week has been terrible work wise but I get some relief starting tomorrow and promise I will try the bot on friday and this weekend and give some feedback. I will grab and complie the latest version from your site as of late tonight around 10pm my time. I like to run new plugins under th debugger!

I will send some feedback soon and thanks agian!

[Dethpod]

Hey all,

I'm really glad to see that podbot for Linux has not died.
However, I'm trying to run the metamod version on my 1.5 server and when I try to add a bot, here's the crash I get:

status
hostname: Test Server
version : 46/3.1.1.1 2615 insecure
tcp/ip : 172.21.10.175:27015
map : de_dust at: 0 x, 0 y, 0 z
players : 0 active (32 max)

# name userid uniqueid frag time ping loss adr
0 users
pb addbot
Adding new bot...
hlds_i686: bot.cpp:430: void BotCreate(edict_t *, const char *, const char *, const char *, const char *): Assertion `pBotNames != __null' failed.
./hlds_run: line 339: 17523 Aborted $HL_CMD

I'm using metamod ver 1.17 from the metamod.org site
Anything I can do to fix this?

-Dethpod
cstrike.rusin-central.com:27015
cstrike.rusin-central.com:27016

[Pierre-Marie Baty]

Originally Posted by stefanhendriks

it seems RB does not like other bots yet... gonna hunt tha bugs for that.

Like I tell you, RealBot will work fine with all the other bots which use the FL_THIRDPARTYBOT flag and NOT the FL_FAKECLIENT one, for the reason I told you one page or two before. But since TurtleRocker told us to put the FL_FAKECLIENT flag back (for engine consistency reasons) your bot is confused about which edict is a RealBot and which other is a third-party one. If I was to put the FL_THIRDPARTYBOT flag in POD-bot these two bots would get along together quite nicely, but since it is RECOMMENDED to put the FL_FAKECLIENT flag back, you have to check your code and fix it the way I told you.

If you don't know what I'm talking about, email me your latest RealBot source code and I'll do the changes for you. It's a matter of 5 minutes.

[Pierre-Marie Baty]

@Dethpod: the bot is probably not installed correctly, and it couldn't find the files called "BotNames.txt". Ensure you have the "podbot" folder in your "cstrike" folder and ensure that all your config files are present and look okay.

[DreamLord34]

That other podbot dll file that was fixed by someone here is able to work with realbot.

[Dethpod]

Pierre-Marie Baty,

Thanks for the help. It was a case issue. My file was botnames.txt.
I changed it to BotNames.txt and that did the trick.
And it runs with RB WIP#9!!!!!!!!!!!!!!!!!!!!!!

Thank you very much,

-Dethpod
cstrike.rusin-central.com:27015
cstrike.rusin-central.com:27016

[Austin]

PM, the bot seems to run fine.
The main problem was that all bots always took the exact same route. I was playing dust and started with no pxp files.

All 4 bots on my team would run around the perimeter of the t's spawn and then head for the long hallway overpass / tunnel and then go up and around and then through the short tunnel to the bomb site.

I was on a LAN with 1.5 and I did several rounds.
Is there something I am missing with their navigation?

Nice job.
They seem to run great!

[Pierre-Marie Baty]

If they always took the same route that's precisely because they have no experience yet, so the pathfinder can't "weighten" the returned paths. As a consequence, as long as the bots didn't notice that this path was dangerous for them, they will keep using the shortest path. Just wait for them with a grenade or two at hand to temperate their pretentions in this area and they should start to think again

Or perhaps what could be done is to lower the heuristic of the A* algorithm according to the age of the .pxp files...

[stefanhendriks]

i see, i have to eliminate the checks based upon FL_Fakeclient but run then om UTIL_BotPointer() so i am sure it is a realbot and not an other bot

[Dethpod]

Pierre-Marie Baty,

podbot works pretty good but it's crashing after a map change.

Especially after a map vote.

Here's what my log says:

L 02/27/2004 - 17:22:20: Server cvar "sv_voiceenable" = "0"
L 02/27/2004 - 17:22:20: Server cvar "sv_wateraccelerate" = "10"
L 02/27/2004 - 17:22:20: Server cvar "sv_waterfriction" = "1"
L 02/27/2004 - 17:22:20: Server cvars end
No Experience Data File or old one - starting new !

And then the server reloads.

Has anyone else experienced this?


-Dethpod
cstrike.rusin-central.com:27015
cstrike.rusin-central.com:27016