POD-bot back into shape.

[Hot-Doc]

Thank you Pierre for your unstoppable engagement!

Just wanted to mention that unfortunatly there is still an extrem high server load and the slow motion effect present with the latest version (although pvi files were included) - but only after the first mapchange (didn' try a second because it was nearly unplayable).

I tried disabling statsme but that didn't help. Will try disabling mean next time and report.

Nevertheless the first map (italy) was great fun - hostages being rescued, t's changing their "tactics" after having lost in a row, a ct switching to pistol when out of ammo, bots taking enemy weapons from the ground - really impressive. Thank's again!

EDIT:

Without mean the first mapchange worked well, no slomo effect and no high server load - sorry for any false suspicion! :o

Never had problems with mean before, but will go checking out the mean forum now...

[SoUlFaThEr]

i personally dont have mEan......i only have podbot and amx........but before i had amx, it was doing this anyway.

even for me taking it out didnt change anything.....still slo on second map and after that too

[sPlOrYgOn]

maybe its not freeing somethings from the memory...
[edit]
i always have that slow-mo effect since some update way back..
and back then i was only running podbot alone with metamod and the slow mo was happening.
but it got passed off as me having a slow comp.
and i do have a slow comp
[/edit]
[edit2]
if you goto page 23
you'll see that i said it worked better because the pvi files were saved but even if the pvi files load it still is slow.
I'ma look through the code.
[/edit2]

[SoUlFaThEr]

if the bot doesnt go slo mo in the first round.......there should be a way to make it completely reload......after the first map.....as if it were the first map.

might it have to do with the bots taking over the same names for the second map as were in the first one?

[sPlOrYgOn]

[edit]
i'm an idiot ignore this post..
[/edit]

[SoUlFaThEr]

lol spiorygon

PMB we would like to test a new option if you are willing to throw it in:

about bots handling use button switches for doors

we need a "use" flag...

or is this what the button flag does(force a bot to use the USE key?)? they apparently dont do USE on certain doors even tho a button flag is placed on the waypoint.

if its at all possible they should have a WAIT set on them after hitting the button..like 1 second at least.....for the door to open......door sometimes dont move if a player is pushing against them. this is what the bots do presently.....hit the button...and slam into the door.....so it cant open.

i hope i explained that all right

heres how Austin said it:

""For a start I would suggest a very simple thing that could make a big difference right now. Add a USE flag to the waypoints. When the bot reaches this waypoint it hits the use key. Simple, easy, backwards compatible, and this would fix the problem with all doors that need to be "used." (at least for the doors that always open on use and not the toggle open/closed type of doors.)""

EDIIIIIIIIIIIIIIIIIT
Stephanhendricks posted this thread about how he got the bots to use the doors right(de_rats)
here this link:
http://forums.bots-united.com/showth...newpost&t=1344

Hmm, is anybody else having problems on maps like aim_ak-colt and other waypoint-less maps where it crashes a few seconds after it errors? Also, my server crashes like 2 seconds after I add a bot now, but it wasn't like this earlier, I was playing for 10 minutes... any ideas? Any help would be appreciated.

[SoUlFaThEr]

download the last relsease of this dll, its fixed that problem
on page 45 theres the new link

Thank you for the quick response! I'm assuming it's been fixed in the Linux .so port too? Does this fix both problems or? Will install either way, just curious. Thanks again.

Awesome, it stopped the crashing, haven't tried a custo map yet though. One last question.. admin_rcon pb jasonmode on (and I tried 1 too) doesn't work. Does admin_fun have be on? Thanks for your help!