Found a spot on Malden where the sea depth is >100m and aircraft appear to take off ok.

awesome, thanks!

Ok slight update, due to team availability, we'd need it by this Fri/Sat to get it into our release for next week...

Still really looking forward to this :) I know its going to be a very busy week getting ready for launch, respect the fact we are probably not at the top of your list of things to do :D

Looking forward to it :)

Cheers!

https://cdn.meme.am/instances/500x/68980257.jpg

It's time consuming and a pain to do in game - if possible at all.

Any word?

If Tanoa map is a PBO at release, we can just do the indexing then. If not, it would be good to get an object list output once Tanoa is in final data lock.

Ok, so tested wildcard i.e. (and pleasantly surprised!)

class CfgCommands {

allowedHTMLLoadURIs[] += {
    "http://alivemod.com/alive_news.php*";
};

};

and this appears to work! Thanks!

Close ticket

Added a mission for test purposes. Slight stuttering can be seen in vanilla when spawning 64 AI at once (8 groups of 8). No stuttering is seen when spawning 1 vehicle, 1 AI or 8 AI to players group.

Antony, can you test this mission on dedi?

Once in game:

press 0 for action menu
press 0 for radio menu

Then choose:
press 1 to spawn vehicle
press 2 to spawn unit
press 3 to spawn group
press 4 to spawn 8 groups of 8 AI

SQF compiled function is called from an FSM checking to see if units should be spawned.

for vehicles...

_vehicle = createVehicle [_vehicleClass, _position, [], 0, _special];
_vehicle setDir _direction;
_vehicle setFuel _fuel;
_vehicle engineOn _engineOn;

for custom groups (we keep an array of units that make up any group)...
_group = createGroup _sideObject;
{
_unit = _group createUnit [_x, _unitPosition, [], 0 , "NONE"];
_unit setPos formationPosition _unit;
_unit setDamage _damage;
_unit setRank _rank;
} foreach _units;

Testing this on vanilla dedicated, not seeing any stuttering on client or large FPS drop on server when spawning a vehicle, a unit or a group (to the players group). Will try several groups.

Antony, not sure its up to BIS atm. Its up to us to provide a vanilla repro.

Hopefully I didn't waste 10 minutes of my life...

Stratis:
"2e306b00# 76387: tenthangar_v1_f.p3d, [1698.93,5053.42,1.90735e-006]"
"2e307900# 75767: tenthangar_v1_f.p3d, [1732.45,5193.83,1.90735e-006]"
"3cd55600# 66924: tenthangar_v1_f.p3d, [1858.3,5515.63,9.05991e-005]"
"2df35600# 66577: tenthangar_v1_f.p3d, [1906.17,5639.7,0.000470161]"
"2df36400# 66585: tenthangar_v1_f.p3d, [1954,5629.5,2.67029e-005]"
"2df1eb00# 64308: tenthangar_v1_f.p3d, [1913.16,5820.46,1.90735e-006]"

Altis:
"6e0d9600# 526049: tenthangar_v1_f.p3d, [14401.4,16225.4,-0.000520706]"

Of course we can do that, but the point is that it's a bug with the placement of objects on BIS maps.

Like I said, it's a 5 minute job in Visitor.

If BIS is not interested in fixing bugs, then fine :)

Are you joking?

Lots of MP game modes and missions dynamically spawn aircraft. Rather than have to statically accommodate and work around a BIS error/bug, it would be great if the environment team could take the 5 minutes in Visitor to correct the issue.

Confirmed, fixed. Tested in 1.44

Exhaust dir can be animated, it appears the exhaust pos cannot be.

need to work out what is causing this before submitting bug

I think your issue is because you have set autocenter=0 in the geo lod. Remove this.

Also I started a wiki entry here based on community knowledge, still WIP

https://community.bistudio.com/wiki/Arma_3_Ships_Config_Guidelines

BIS_fnc_exportCfgWeapons actually calls a command diag_screenshot. However, this appears to be disabled in the Dev and Stable versions of Arma 3.

no mods. I'm writing my own persistence mission, so storing player state on server and restoring the state to the client on reconnect. Also written a JSON extension to push state out to CouchDB, but that's another story! Its all working other than vest/uniform state and being able to restore globally ammo count on mags.

Obviously somewhere I'm doing something wrong with vest/uniform as backpack restore works fine.

I still need addMagazineGlobal [_mag,_ammocount] though :)

ok, I see you can't reproduce my addItemToUniform bug either, clearly I've got some issue with my script on reconnect. Will double check. Sorry to waste your time. Its not working for me, so must be something other than these commands..

Repro mission attached.

Run mission on dedicated server.

The mission has an init entry for the player.

During player init the script (both client and server) removes and adds a uniform, removes and adds a vest. On client, Adds handgrenade to uniform and vest using addItemToXXX, adds a smokeshellgreen using addmagazineglobal. Logs uniform/vest state to rpt before and after additions on both client and server.

Check the server and client logs to see uniform/vest state during init.

Press Radio Alpha to get uniform/vest state post init.
Press Radio Bravo to add items.

Result

On Client - handgrenade added to vest and uniform, no smokeshell added
On Server - only smokeshell added (post init)

Expected

On Client - handgrenade added to vest and uniform, smokeshell added to uniform
On Server - handgrenade added to vest and uniform, smokeshell added to uniform

Apologies, I should clarify. This command fails for a JIP player.

1. Start a dedicated server session and ensure the session is set to persistent = 1 and ensure DisableAI is set.
2. Disconnect from the server
3. Reconnect to the server
4. Run the addMagazineGlobal command on the client and no grenade is added to the inventory.

addMagazineGlobal "HandGrenade" does not work, it adds a handgrenade to the player object on the server, but not on the client.

addMagazine "HandGrenade" does work, adds a handgrenade on the client, but not on the server

updated. It does not add a Handgrenade to the player when run on the client. It does add the handgrenade to the player object on the server. The expected behaviour is to add a grenade to both client and server player objects.

Please run player addMagazineGlobal "Handgrenade" on client and tell me if it adds a handgrenade to your inventory.

We need addMagazineGlobal [_magazine, _ammocount]

Also getting

Error in expression <c) lbsetcursel _cursel;

deletevehicle BIS_fnc_animViewer_unit;
_logic = BIS_fn>

Error position: <BIS_fnc_animViewer_unit;

_logic = BIS_fn>

Error Undefined variable in expression: bis_fnc_animviewer_unit

File A3\functions_f\Debug\fn_animViewer.sqf, line 375

with latest Beta Dev build

There are 3 key things here.

1. Make OPC and OPD stackable and also if possible the other OnXXX commands (such as OnSingleMapClick)

2. Fix the race condition that occurs on OPD where player unit can sometimes get deleted before the event handler starts/finishes

3. Return a "copy" of the player object on OPD to allow easier access to player data on disconnect.

Thanks!

From Columdrum:

Seems that lately there is a lot of people working on persistence scripts, and this event its especially useful to save player data, server status, correctly handle player trying to exploit reconnect... So any improvement on this command could be really usefull. Here a list of changes that can improve it and make it really useful and reliable.

1- Make stackable, refs above all commented there so no need to add anything

2- Return the player object itself as param:
As already stated main use of this command right now its save player data, so most times you want to search for the player object, so that would make that easier and should be that hard to implement.

3- Some way to assure the eventhandler its executed before the unit its deleted:
With the current implementation of this event, to get the player data you should first search for the player object( by name or by UID), get all the data(position,weapons,...), and under heavy load (low server fps) that can cause that the object of the disconnecting player is deleted before getting all this needed data, and therefore its lost forever. Posible solutions to this would be:

-Only really delete the player object after all the onplayerdisconnected events are executed or any of them have returned a value( something similar to the keyeventhanderls where you can return true or false to indicate if the event have handled it completely or not). Main problem of this implementation is that if the script creator makes a mistake that generates an infinite loop inside the event could leave ghost players, could be avoided by introduce a adjustable constant (via config, description.ext, scripting command....) that indicates a timeout to handle this event.

• The player object that its returned by the event its a exact copy of the real player object, but its not the real one, so the real one can be deleted on all the clients or even the server, but you can still access all the data of the player object just like it was before disconnecting. This way you wont have posible ghost players problems, but may be really a mess to implement

This is clearly seen in the combined arms showcase, where the AI controlled transport Heli, "rocks" back and forth during level flight.

This no longer occurs on the dev version

Fireball, it occurs regardless of params, that paste was a typo.

Not sure what is causing this as only some users experience it.

I no longer see this with the latest Steam Client (beta) and the Beta Dev dedi server. Anyone else still see this issue?

This command does work from the debug console... Or any script called from the debug console.

Main issue here is when units are killed by say a grenade or UGL, the killer's primary weapon returned is a rifle or other weapon, but not Throw or the UGL.

And here http://forums.bistudio.com/showthread.php?160755-Stealth-Takedown-Weapon-Resting-and-more