Converted Waypoints for RCbot-AngelScript, Updated & Tested for Git Options

[madmax2]

I've been updating/fixing all four last series maps the last few days. I should be adding them to the pack soon.

I set aside desertcircle update again, cause those darn flag sprites still get stuck . Without bots I can duplicate the problem. I believe flags get stuck when the flag is descending, so if a player or bot lets go when getting shot, or dies, the flag sprite may get stuck. At this point, I don't think this is fixable for the script?

Today I was able to get the first flag fully stuck without bots, really stuck so I was unable to capture it. Most of the time, players can still capture the flag even if the sprite gets stuck. Once the music plays, the flag is captured and the next flag is enabled. (the sprite position doesn't determine if a flag has been captured)

So, for players, it isn't a big problem, as they can cap the stuck flags most of the time. I don't think a fully stuck flag happens often, or this would not be an official map?

For bots, the easy solution is not to use the script, that way they will never be stuck at any one flag because the sprite never reaches the top. Of coarse bots will be back to running past flags again, but players can cap them too, I guess that's better than bots getting stuck at one flag? I could try changing the scripted flags to a non-entity (-1) goto/waypoint? They won't know if they have capped a flag, but maybe they won't run by the flags so much?

For the git current wpt, the 1st & 3rd flag sprites are the most likely flags to get stuck. I've put some hours into these flag wpts, changing the positions, etc, but not sure if it is any better? I'm working on different approach angles now, see if that helps reduce the stuck sprites? Need to protect the bots from being shot or dieing while raising a flag, likely not possible though...A script change will likely be the best or only solution?

Here is a search on 3rd flag with it at the top, I don't think there is anything else to use to indicate a flag capture. Ideally, the music playing indicates a flag capture...

CODE

] as_command rcbot.search
[RCBOT]1 : player frame=181.023438 distance = 0 (x=-3381.720459,y=2840.049316,z=380.03125) visible=1,solid=3,angle.x = -6.026001, angle.y = 20.286255 active = 1
[RCBOT]4 : player frame=0 distance = 198.951141 (x=-3261.890137,y=2794.024902,z=228.03125) visible=1,solid=3,angle.x = 6.597681, angle.y = -3.832489 active = 1
[RCBOT]25 : squadmaker frame=0 distance = 180.629776 (x=-3208,y=2852,z=332.015625) visible=0,solid=0,angle.x = 0, angle.y = 278 active = 0
[RCBOT]71 : env_sprite frame=5.350231 distance = 181.143326 (x=-3384,y=2842,z=561.149719) visible=1,solid=0,angle.x = 180, angle.y = 90 active = 1
[RCBOT]72 : momentary_rot_button frame=0 distance = 32.800926 (x=-3356,y=2844,z=400) visible=1,solid=0,angle.x = 0, angle.y = 49.815845 active = 0
[RCBOT]77 : ambient_generic frame=0 distance = 56.470577 (x=-3360,y=2836,z=432) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]86 : trigger_setorigin frame=0 distance = 77.95314 (x=-3352,y=2864,z=448) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]194 : env_sprite frame=5.350231 distance = 181.143326 (x=-3384,y=2842,z=561.149719) visible=1,solid=0,angle.x = 180, angle.y = 270 active = 1
[RCBOT]264 : trigger_multiple frame=0 distance = 26.332506 (x=-3356,y=2844,z=376) visible=0,solid=1,angle.x = 0, angle.y = 0 active = 1

P.S. Seem so be problems with Bots-United for the last 24 hours. Connection/database errors...

[Cheeseh]

hmm can you check "active" on the rotary button or squad maker to see if they change?

[madmax2]

Sorry for late reply, I get busy at the end of the month...

I checked what you suggested, I don't think it will work. Active doesn't change on buttons or squad maker, and squad maker isn't present on the 1st flag. Angle y on the button does change but varies with the flag. At first I thought it was giving different readings when the flag was at the top, but then I realized if I wait long enough the value returns to zero! Here are the search's on the 1st flag...

CODE

1st flag - at bottom

] as_command rcbot.search
[RCBOT]1 : player frame=201.965088 distance = 0 (x=2817.24707,y=-2527.96875,z=132.009308) visible=1,solid=3,angle.x = -4.72229, angle.y = -69.906006 active = 1
[RCBOT]12 : momentary_rot_button frame=0 distance = 36.06044 (x=2828,y=-2548,z=160) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 0
[RCBOT]54 : env_sprite frame=13.409779 distance = 29.029301 (x=2800,y=-2548,z=120.009811) visible=1,solid=0,angle.x = 180, angle.y = 90 active = 1
[RCBOT]55 : trigger_setorigin frame=0 distance = 116.187103 (x=2824,y=-2528,z=248) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]79 : ambient_generic frame=0 distance = 98.542564 (x=2836,y=-2540,z=228) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]192 : env_sprite frame=13.409779 distance = 29.029301 (x=2800,y=-2548,z=120.009811) visible=1,solid=0,angle.x = 180, angle.y = 270 active = 1
[RCBOT]262 : trigger_multiple frame=0 distance = 23.085733 (x=2828,y=-2548,z=128) visible=0,solid=1,angle.x = 0, angle.y = 0 active = 1

at top

] as_command rcbot.search
[RCBOT]1 : player frame=161.63208 distance = 0 (x=2817.24707,y=-2527.96875,z=132.009308) visible=1,solid=3,angle.x = -2.854614, angle.y = -68.131714 active = 1
[RCBOT]12 : momentary_rot_button frame=0 distance = 36.06044 (x=2828,y=-2548,z=160) visible=1,solid=0,angle.x = 0, angle.y = 64.806274 active = 0
[RCBOT]54 : env_sprite frame=12.355457 distance = 183.427017 (x=2800,y=-2548,z=313.521729) visible=1,solid=0,angle.x = 180, angle.y = 90 active = 1
[RCBOT]55 : trigger_setorigin frame=0 distance = 116.187103 (x=2824,y=-2528,z=248) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]79 : ambient_generic frame=0 distance = 98.542564 (x=2836,y=-2540,z=228) visible=1,solid=0,angle.x = 0, angle.y = 0 active = 1
[RCBOT]192 : env_sprite frame=12.355457 distance = 183.427017 (x=2800,y=-2548,z=313.521729) visible=1,solid=0,angle.x = 180, angle.y = 270 active = 1
[RCBOT]262 : trigger_multiple frame=0 distance = 23.085733 (x=2828,y=-2548,z=128) visible=0,solid=1,angle.x = 0, angle.y = 0 active = 1

I don't think there is much we can do, but change the flags to non-entity important points. I can test and adjust the wpt for that, see how they do... It doesn't totally ruin it for the players, when a flag sprite gets stuck, but it can mess the bots up, to where they aren't as much help.

Last series is going well, the 2nd & 4th maps are probably ready, still need to retest the other 2.

Last2 is the 1st map I tried the grapple with, took me a while to get some near vertical pulls working right, but it's good now... In order to get one working good, I had to noclip the landing wpt into the ceiling directly over the landing point, above & beyond the grapple wpt. Then they could land on the platform there. Raising the grapple wpt above the slime didn't help. The original landing wpt was on the platform so they would let go once they pass that height. So I thought Tarzan swing to a higher point is what worked... It feels like they need to pull a bit more to the grapple wpt? Other than that, grapple seems pretty good, I'll know more when I get to that royals2 map or the sectore maps...

[Cheeseh]

meant to say perhaps you can make the 'z' much lower for desert circle, because the flags will only move up if they are available, so even if it is like 1 unit above the resting position, bots will think it is complete, but the bots that are currently using the flag won't stop until the invisible button stops moving. So even if it is slightly high bots will ignore it, but when it reaches the bottom again bots will go to it.

[madmax2]

meant to say perhaps you can make the 'z' much lower for desert circle, because the flags will only move up if they are available, so even if it is like 1 unit above the resting position, bots will think it is complete, but the bots that are currently using the flag won't stop until the invisible button stops moving. So even if it is slightly high bots will ignore it, but when it reaches the bottom again bots will go to it.

Ok, good idea, i'll try to get to that tonight while working on last series...

[Edit] Yeah, this looks promising, set all flags to one unit above the bottom position. If flag gets stuck and they are killed, they won't cap it, but when a player caps it, bots will be able to cap the next incomplete goal. (Instead of getting stuck at one flag goal for the rest of the game.)

Request:
Would it be possible to add a waypoint_move command, one that retains the flags & connections? I think that would be very useful fixing waypoints, a big time saver. What do you think?

[Cheeseh]

Ok, good idea, i'll try to get to that tonight while working on last series...

Request:
Would it be possible to add a waypoint_move command, one that retains the flags & connections? I think that would be very useful fixing waypoints, a big time saver. What do you think?

I can add "waypoint cut" and "waypoint paste", and it will try to keep the same waypoint ID as not to break scripts. If can't keep the same ID then it can issue a warning message about the new ID

[madmax2]

I can add "waypoint cut" and "waypoint paste", and it will try to keep the same waypoint ID as not to break scripts. If can't keep the same ID then it can issue a warning message about the new ID

Excellent, sounds good...

[EDIT] I don't need it now, for updating wpts, but could you also add the autowaypoint function back in? I autowaypointed Judgement with the old rcbot then converted it and edited it with rcbot-AS. It worked ok, but had some glitches using the old bot in SC5

Gotta sign off, past 1am here, and a travel day... Thanks Cheeseh...

[Cheeseh]

Excellent, sounds good...

[EDIT] I don't need it now, for updating wpts, but could you also add the autowaypoint function back in? I autowaypointed Judgement with the old rcbot then converted it and edited it with rcbot-AS. It worked ok, but had some glitches using the old bot in SC5

Gotta sign off, past 1am here, and a travel day... Thanks Cheeseh...

will try, will take a lot more effort than waypoint cut paste tho

[madmax2]

will try, will take a lot more effort than waypoint cut paste tho

Guess I should of been more clear, the autowaypoint would be a nice addition, but it's not a high priority for me. I thought you could take the code from rcbot 1 or 2, but perhaps you have to re-write it in AS? The cut & paste wpt is more important to me, as I've seen a lot of situations that will be useful. So, just put autowayoint on the list of possible additions, for when & if you want to add it?

The last series waypoints are taking a bit longer than I thought, ran into some problems that slowed my progress, but most of it is resolved now. Once I get last3 into shape, I should be able to add them to the pack. Last3 has an awful lot of wpts & paths, I've removed some of that... (btw, I may of come across a bug with barney wpts, I'll try to post the details soon, with some debug info)

[madmax2]

Added Updated Last Stand waypoints, original waypoints by Mard... (4 maps)

I will post some details tomorrow ...

[Cheeseh]

Added Updated Last Stand waypoints, original waypoints by Mard... (4 maps)

I will post some details tomorrow ...

cool

I added some new commands

CODE

waypoint_cut
waypoint_copy
waypoint_paste
waypoint_move1
waypoint_move2

cut/copy/paste just remembers the waypoints flags
waypoint move1 / move2 will move the waypoint to a new location. Use move1 to select a waypoint first, and move2 to place it to the new location

[madmax2]

cool

I added some new commands

CODE

waypoint_cut
waypoint_copy
waypoint_paste
waypoint_move1
waypoint_move2

cut/copy/paste just remembers the waypoints flags
waypoint move1 / move2 will move the waypoint to a new location. Use move1 to select a waypoint first, and move2 to place it to the new location

Excellent, I will use this a lot...

After playing through the whole Last series, I may need to do a bit more tweaking on some waypoints. Bots do good for most areas, just some things could be better. Last4 is good, just looking at 1, 2, & 3 now...

Generally, for most all waypoints, I did a lot of path edits and removed some waypoints, there is more that could be done. And the new move commands will be great for fixing/improving waypoints like this. Also, removed a number of important givetypes, mostly on door buttons. Added some Unreachable wpts. Since the maps are survival now, most of the ammo has been removed or moved, some health/batteries changed too. While I fixed some of these locations, I left some unused ammo locations to keep bots going there. Bots do fine this way, but I may remove more & see how the do?

Outline Details of the major fixes/additions... Got a bit long winded...

Last4: Fixed all ladders so bots go up them well, get the RPG, bots will occasionally call a airstrike. Waypointed the new tower, so bots can go to the top & snipe/RPG the enemies below, then return tho the main area.

Last3: Has the most wpt/path/ammo/health/armor/weapon edits, more could be done. There are 2 shortcut doors in this map that don't work by the use key, player must push against the panel to open the doors. I fixed this by putting a path through the panel to an important wpt, bots will open these doors at random, they can use the shortcuts once open.

I don't have a checkground on the bridge atm, I may add that back on, still testing that area. I tried to improve the wpts at the rotating machine with platform wpts, but couldn't get it working better than what Mard had there, I did add an extra wpt there that helps a little. Bots do fairly well there as is, at least 50% success...

The next area with the impact machine & HT into the vent was completely reworked due to a map change there. Bots can go in the vent & turn the machine on then come back out the vent, but they are a little slow at it, and they have a little trouble at the HT/vent ladder, but no permanent jam. I originally was going to leave this vent for players only, I'm still testing it and may change it again?

Fixed it so bots can activate the laser & get the crystal, but they may not get the crystal the first time, which is ok..

Reworked the last ladder so bots can jump up to ladder and into vent, it may take a couple attempts...

Last2: A number of changes/additions on this one too. The biggest changes are in the center & end of the map. I removed the armor givetype at the HEV charger in the 2nd half of the map, as they seem to multi-path & use some shortcuts much better without it. Bots can grapple into the elevator shaft & climb the cables to the top. Bots can grapple up to the generator controls, turn the generator off then jump back down, or use the shortcut door there when open. Turning off the generator usually is not fast. Bots would open the generator shortcut door, but they may of quit doing that when I got them using the other shortcut to the boss room more.

Bots can jump across electrified floor and open the shortcut door, but they may not open the shortcut door until all the enemies are killed. There is a grapple wpt there too, but they only use the jump path atm. I have some ideas to improve this whole center area, so may update this wpt again. Fixed the last ladder in vent so bots don't get stuck there, and bots will go both ways in vent now. Some tweaks in the boss room and to end of map. Bots can end the map, but there is plenty of time for a player to end it too...

Last1: Reworked some pathing in the 1st area, around the "tent/radio", and added a path over the tent (just for fun)... some tweaking here & there, and at the teleporter controls. At the next area, changed the HT at the wood plank to a jump wpt, single bots can now jump up, go over the top and open the door there, unassisted. Reworked the paths/wpts in this room so a crate doesn't block that door and bots won't spawn into the crate, haven't seen either for a while now, so hopefully it's fixed.

The barney wpts there is currently not used by otis, I encountered a possible bug here and will post on that soon... For now, I added some wait & staynear givetypes there so bots protect otis from the 1st pit drone, that was killing him a lot, it seems to work most of the time. Right now, otis likely won't open the weapons room without player help, but may open the 1st 2 security doors if he survives. Fixed the ladder room to get the uzi, but bots won't use the uzi much? Bots can end this map, but there should be time for players too...

For The first 3 maps, bots probably would work better if scripted, but I do like some random behavior in the bots too, as it gives the players time to figure out the map and do things. So I would need to think about that?

[madmax2]

Cheeseh, the new move1/move2 commands work great, I've used it several times now on Last Series maps. I've used it at a HT, at doors, at ladders, in vents, at ammo boxes, and to adjust for better navigation in difficult places, etc. I think this is one of the most useful commands, thank-you...

For desertcircle, your idea to set the z low is working good, so bots won't get stuck at a flag now. I put the z settings at one unit from the bottom.

=============================
Updated Last 1 & 3 rcwa's

Last - Improved the navigation around the tent/radio and other areas of map with the move commands. Bots run around & under tent without getting caught on the poles. Bots shouldn't get stuck behind those double doors for long, if ever. Bots protect otis from 1st pit drone better, I think?

Last3 - Improved Human Tower & vent area. Bots will open & use shortcuts faster now. Bots likely won't get grenades, removed the weapon givetype there, this was one of the reasons bots were not using some shortcuts. Some adjustments to ammo wpts and for navigation...

=============================
Updated Desertcircle rcwa + script

desertcircle.rcwa - Ladder improvements at 2nd flag. 1st & 3rd flags have been reworked for different approach, so hopefully those flags won't stick to often. Two more important wpts to encourage bots to use the middle & reverse routes to the fort after the last flag is captured (bots also run the outer perimeter route).

desertcircle.ini - new & improved script

Bots won't get stuck at a flag now, trying to raise a stuck flag that has already been captured.

If a bot is killed while raising a flag, bots will bypass this flag and continue to next goal, and will return to this flag when it reaches the bottom of the pole (they cannot capture the next flag, they will die trying, so will respawn before this flag).

If the flag gets stuck anywhere above the bottom, then bots will continue to bypass this flag, until a player truly captures the flag, indicated by the music playing. At that point bots will be able to capture the next flag, and so on...

2 non-entity scripted wpts added so bots will use middle & reverse routes after all flags are captured

=============================
I'm still testing these wpts, and may update them again? I think they are better than the previous wpts & script... Trying a few things for Last2, but after that i'll work on some other wpts. I think these are good enough.

[Cheeseh]

thanks madmax , I will try to update the bot with these and have a look for that barney waypoint bug -- still haven't got around to doing it but on my list of things to do.

[madmax2]

thanks madmax , I will try to update the bot with these and have a look for that barney waypoint bug -- still haven't got around to doing it but on my list of things to do.

No problem, I've been bogged down in these LAST wpts. Need to get them done & move on ... That barney bug is weird, and may only happen when barney wpts are after a teleport wpt or button. It may be specific to this map only, not sure?

It looks like i'll be updating Last3 again! That darn Human Tower was messing up again . So removed the crouch on the top wpt after the HT, and both down paths from it, its a basic HT config now with no extra givetypes, and the lead-up wpt squeezed in tight to the HT. This seems to work the best. Tweaking a few other things, hopefully will upload it tonite?

[Cheeseh]

No problem, I've been bogged down in these LAST wpts. Need to get them done & move on ... That barney bug is weird, and may only happen when barney wpts are after a teleport wpt or button. It may be specific to this map only, not sure?

It looks like i'll be updating Last3 again! That darn Human Tower was messing up again . So removed the crouch on the top wpt after the HT, and both down paths from it, its a basic HT config now with no extra givetypes, and the lead-up wpt squeezed in tight to the HT. This seems to work the best. Tweaking a few other things, hopefully will upload it tonite?

I will upload when you're done, no worries -- I'll need to improve HT tbh

[madmax2]

I will upload when you're done, no worries -- I'll need to improve HT tbh

Updated Last3 again, seems to be good now . Didn't get to Last2 last night... I'll post a few details on Last3 later, after I unscramble my brain LOL...

[Edit 2-25-29]
Updated Last2, will post more info later...

[madmax2]

Finally, think I'm done with these for now...

Last2 was updated a couple days ago... most notable changes are:

Bots prefer to jump across the electric floor but will grapple there as well, and they will be more reliable at opening that shortcut door. Beyond that, bots will also open the laser room shortcut door better.

In the last vent past the ladder, bots would sometimes get stuck looking down thru the side of vent (2 places) in the direction of the action below? Not sure why, as the crouch wpts looked fine and spaced close enough? Tried moving them and adding jumps but nothing worked to resolve the problem until I noclipped 2 normal wpts between 2 crouch wpts, kinda high and off to the side with auto-paths I left in (no cleanup to the paths). It looks messy but fixed the problem! They seem to pick up on a normal wpt about the point they would get stuck before.

More improvements in the lift shaft to the barnacle gun. Added an important flag to a wpt near the gun, to stop some confusion the bots were having, due to no important goal available to them until they picked up the barnacle gun.

A few changes in the boss room, to help bots move around the room better and they will use the teleports to the boss room, but not all the time.

Last3 Updates:

There were a number of fixes/improvements on this wpt, but there is one door bots will sometimes stop at and wait for it to close then open again before they go through, half the time they go right though it. It's not a major issue, and i've tried many things to fix it, but I think the door trigger is to close to the door, so i'm done with that.

Today, added a minor fix for jump into vent at the boxes, with 2 fallback wpts added there.

Added the checkground back onto the bridge but with some new jump down paths/wpts to the area below, so they may attack the voltagores or shoot down from above. Other wpt/path changes too...

More work on the HT, it seems good & reliable now. I added a new approach wpt, straightened the approach, jump flagged the top wpt, and removed some extra paths from the top wpt so they are forced into the vent. It's definitely much better now.

Bots will open that next shortcut door before the crystal room reliably (the shortcut to bypass the giant gear machine), if they go past it to the crystal room, they can go back to it and open it, and often do so.

Added openslater back on crystal, bots won't pile up on glass anymore & will activate laser 1st time...

=====================================================================
On both wpts, I used some fake ammo, health, or armor givetypes to get bots moving around rooms or going places or pressing buttons more reliably/better. It works well, but was a lot of trial & error. It may of been easier & faster to of scripted the main goals? So I may try that the next time, these wpts are working good w/o scripts now...

I used the move commands to make some things work better, it solved some problems just realigning wpts. Like the "big" door on Last3, by moving the door wpts closer to the door buttons, bots no longer fail to open that door and it Only took a few seconds to make the change. What's nice about it is, I can now quickly make a change while the bots are running around and see if the change helps before saving it, and it doesn't disrupt the bots current paths.

Anyways, enjoy, these wpts are final for now...

[Cheeseh]

great -- added to git now

[madmax2]

great -- added to git now

Cheers man...

I've been trying to convert my sc_psyko wpt, but bots seem to be having trouble with the openslater wpts. It's a hub map with 10 doors in the central room and I currently have 13 openslater wpts in it. Bots don't want to go through most doors unless I remove the openslater givetype. I'm trying to re-position them to see if that helps? Might have to make a basic test wpt & see if that changes anything. Think I may need to check the doors & bots with debug, as they will act confused in the central room until I remove an openslater.

Got them making the lava & slime jumps/areas, also the ice area, and partly the cloud & sun areas. Still more work to do though.

I'll do a few more tests on this, but if it continues to be problematic I'll just set it aside for now, and move on to some walkthrough maps I want converted...