PlayerStart query

[EvilGrins]

Doing an Xvehicles edit of an existing CTF4 map, where the tanks are placed far from where most of the other PlayerStarts are.

So, I placed 4 extra (1 each) next to the teams' tanks, and configured them for the team they're on.

Doesn't appear that any players are spawning out near the tanks.

In the past when I've done similar I've duplicated existing PlayerStarts and moved the duplicates near to XV, but given this time XV is MUCH FATHER away, I went with placing new ones.

Is there something extra I need to do after placing PlayerStarts on the map in order to "activate" the new ones?

[OjitroC]

Is there something extra I need to do after placing PlayerStarts on the map in order to "activate" the new ones?

Have you rebuilt the path network?

How many starts do you have altogether?

How many bots are you using?

Other than TeamNumber, you haven't altered any default properties?

Does the log say anything relevant?

[Buggie]

Use existing PlayerStart. Just Duplicate it and drag to new location.

[EvilGrins]

Use existing PlayerStart. Just Duplicate it and drag to new location.

As explained above, that's what I usually do.

Have you rebuilt the path network?

I did not.

How many starts do you have altogether?

28.

How many bots are you using?

Testing with 12, occasionally 11 when I'm playing too.

Other than TeamNumber, you haven't altered any default properties?

Nope.

Does the log say anything relevant?

Always forget to check that thing... but nothing unusual so far as I can tell.

May be a moot issue, now. One of my successive tests shows a bot did manage to get to the green tank... so they do work, but I guess they don't get spawned to those specific locations very often.

[Buggie]

16 player starts per team. Paths must be rebuilt. Or bots wander forever, not able find way to flag.
Player start must be in general grid or bots not able find flag.
Also bot wanna carry some weapon for try take enemy flag. Vehicle too count as weapon, but after exit, they can start chase weapons, instead of grab, since feel self naked.
On respawn usually picked spot which not near with enemy if possible.

MapChecker report if there found mismatched count of player starts. Can be used for ensure if this even count of Player starts.

If all Player starts set to red, such map too will work for blue team, but not as expected. So need ensure of teams of PlayerStarts.

[OjitroC]

Have you rebuilt the path network?

I did not.

May be a moot issue, now. One of my successive tests shows a bot did manage to get to the green tank... so they do work, but I guess they don't get spawned to those specific locations very often.

If the PlayerStarts aren't 'integrated' into the path network by rebuilding it, then it will only be by chance that the bots go anywhere?

I've no idea how the multi-CTF mutators handle PlayerStarts and whether they need to be in the 'navigation list' for those mutators to use them.

[EvilGrins]

16 player starts per team. Paths must be rebuilt. Or bots wander forever, not able find way to flag.

If all Player starts set to red, such map too will work for blue team, but not as expected. So need ensure of teams of PlayerStarts.

Bots are finding flag fine, and I did configure at starts to proper team assignment.

I've no idea how the multi-CTF mutators handle PlayerStarts and whether they need to be in the 'navigation list' for those mutators to use them.

From my understanding it's no different than standard CTF.

[sektor2111]

If you stick with 436, ALL PlayerStarts should be into Navigation Chain. If no Start is found by scanning said linked list, 436 is using a foreach Iterator and will "forget" Team - I think even discarding if are active or not. A "new comer" should join into chain or else is useless.
Next - IT'S NOT mandatory to have Paths in/out them if they are Higher over ground - UnRreachable but there are several Nodes in spot. First Node closer to Pawn in EyeHeight visibility is taken if has all needed links (ReachSpecs). And yes, a CTF-Coret Modified due to it's original Flaws it's working normally without outgoing connections from starts but these are linked in navigation chain. Stage happens as follows: Pawn Spawns in spot and starts falling, After fall it takes a little pause, then it moves to a valid connected PathNode/InventorySpot - there is no any issue at this point. PlayerStart higher over ground embedded with Paths has no deal with them anyway. 436 is NOT ALLOWING any access at Pawn with physics falling - if Pawn Falls, it's just falling and nothing else. If you are only duplicating a PlayerStart as stated above this is a 469 practice - 436 doesn't work like this, for anything added/removed Paths-Net must be rebuild. It's not about Specs as much as it's about NAVIGATION Chain, at rebuilding paths, in custom pathing tweaks you can earn bugged specs or lost specs. This is why I asked X times embedding in Editor some control over paths like for brushes and forgetting dumbness with constants that are variables not constants. If you need help with these Starts use MapGarbage, selected New Nodes (PlayerStarts in this case) are added into chain inserted nearby other navigation nodes that are already in navigation chain. If you suspect some borks just check Navigation chain with the same builder...
You have tools, do use them.

[OjitroC]

Bots are finding flag fine, and I did configure at starts to proper team assignment.

If the additional PlayerStarts aren't part of the pathnetwork (that is, the network hasn't been rebuilt) then whatever happens from those PlayerStarts happens by chance. Run one of the tools (Buggie's MapChecker or sektor's PathsChecker or MapGarbage) to check this.

From my understanding it's no different than standard CTF.

CTF4 is subclassed from Botpack.CTFgame but alters the code to deal with the rules of that game. MultiCTF is subclassed from TournamentTeamGame and includes much more additional code, again to deal with the rules of that game.

[sektor2111]

CTF4 aka MultiCTF aka whatever might get called has extra-options compared to Normal CTF, such as at recovering Flag would be needed to be transported back to its Base not just touching it - it's an option that can be toggled On/Off. But in this MultiCTF PlayerStarts must have TEAM declared and if you want to see Players/Bots using them I'm very convinced that they need to be thrown in Navigation chain otherwise are used only original ones.
Any Navigation Node added (which might be connected later or not) can be added into chain using pointed option:

AddNodes.PNG

Extra: Next updated version of this helper (not used - and then maybe it won't be needed any future development) will have a bit of extra-work when builder is triggering Editor to do Paths - Original Paths. It won't only clean nodes from debris, but it will also reorder navigation chain, goals first, then PathNodes and others, PlayerStarts as last. Main references are FlagBases for CTF (+AlternatePaths) and ControlPoints for Domination.

On/Around Topic:
MapGarbage also can show which PlayerStart belongs to which "Team" by toggling Textures. If memory doesn't cheat me I think the last public version has this option extended. Old versions were aiming only two teams - textures are toggled with colors, big ones, and then mapper/user can figure what PlayerStart for which "Team" where is located. This way it's easier than checking them One by One.

[EvilGrins]

MultiCTF PlayerStarts must have TEAM declared and if you want to see Players/Bots using them

That's standard to normal CTF too, as it prevents someone from the other team spawning in your own base and too close to your flag.

Think I've got this ironed out though, though as is still not spawning from the new PlayerStarts as much as I'd like.

Suppose I could go back and remove them all, and then add not as many which might cause greater frequency... but it's okay.

[OjitroC]

Think I've got this ironed out though, though as is still not spawning from the new PlayerStarts as much as I'd like.

So the new PlayerStarts are now linked in to the navigation chain? And into the path network?

Presumably if the tanks are far from most of the PlayerStarts, there are paths from those Starts going to the tanks and from the tanks to the various flags?

[sektor2111]

I don't think he added them into chain - at least I don't see explanations about how he did that and that's why Bots/Players won't use them. If he don't do what we explained nothing like a sudden magic will operate here for making them useful. 469 operates something in run-time but map is just SCREWED and these practices are not helpful. When engine solves mapper's problems it's not like the mapper will learn to do things correctly.
It would be better to show us this edit for inspecting it - with PUBLIC assets which anyone can use after all...

Recommended Process for learning/testing.
Creating a copy of map CTF4-MapTitle_ Tmp.unr - or something like that. Doing X changes there and probing - builders checkers are helping here. If everything is normal rename it to something relevant.

To the point
Add PlayerStart - Select it - access properties, set Team, keep it selected - Open MapGarbage scroll at pointed option, turn it True (it will stay True after first use so you can even close it for the moment) - Hit Build Button - open log and see if task was successful.
After adding "Starts" and doing fore-mentioned process - open MapGarbage - deselect option for linking nodes and use "bCheckNavChain" for checking chain integrity. No, they won't have paths unless you rebuild them... or connect them semi-automatic with XC_PathsWorker... or just pull them higher from ground.
Hint and Hit: Before placing additional Starts class can have adjusted CollisionHeight at 90 - no, we won't save Engine file but only the map after adding these adjusted classes.
Estimated time: 10 - 30 minutes + a coffee.