[RichardStorm]

In fact it needs optimization, but it works for now and i'm on other features and fixes to do, I'll improve it later.
Consider that i posted it for zazo and i think it's the "simplest" working version to understand, it's a modified version of which i really use, too.

@ zazo: I forgot to add a comment saying that you need to add "case GOBELIN" among all enemy types i wrote there (and you can remove them from there if you want).

[zazo]

interesting to study...
I thought the canseespr was easier to use...

[zazo]

zazo, on 14 October 2012 - 02:19 PM, said:

interesting to study...
I thought the canseespr was easier to use...

also, I just discovered that there is ally AI in "duke plus"

[Danukem]

zazo, on 14 October 2012 - 02:36 PM, said:

also, I just discovered that there is ally AI in "duke plus"

Also, all of the enemies in DukePlus can become allies if they are flagged that way in a map or if hit with the shrinker in mindlbast mode.

[zazo]

Trooper Dan, on 14 October 2012 - 06:32 PM, said:

Also, all of the enemies in DukePlus can become allies if they are flagged that way in a map or if hit with the shrinker in mindlbast mode.

another little problem: how to creat an animated projectile in 5 sides:
"
define flyingaxe 5500
defineprojectile ...... etc (settings for projectile)

action flyingaxeframe 0 4 5 1 10
useractor notenemy flyingaxe 0 flyingaxeframe enda
"
only the first frame appears but not the entire animation ?

[CruX]

zazo, on 15 October 2012 - 04:10 AM, said:

another little problem: how to creat an animated projectile in 5 sides:
"
define flyingaxe 5500
defineprojectile ...... etc (settings for projectile)

action flyingaxeframe 0 4 5 1 10
useractor notenemy flyingaxe 0 flyingaxeframe enda
"
only the first frame appears but not the entire animation ?

An action has to be declared for each frame of the animation

action flyingaxeframe1 0  1  5 1 10
action flyingaxeframe2 5  1  5 1 10
action flyingaxeframe3 10 1 5 1 10
action flyingaxeframe4 15 1 5 1 10
useractor notenemy flyingaxe 0  flyingaxeframe1 enda //initialized to the first action

Then you rig the animation with a peractor variable in EVENT_GAME

gamevar countvar 0 2
gamevar picnum 0 2

onevent EVENT_GAME
getactor[THISACTOR].picnum picnum
ifvare picnum flyingaxe {
addvar countvar 1 
ifvarg countvar 1 { 
setvar countvar 0 
ifaction flyingaxeframe1 action flyingaxeframe2 else
ifaction flyingaxeframe2 action flyingaxeframe3 else
ifaction flyingaxeframe3 action flyingaxeframe4 else
ifaction flyingaxeframe4 action flyingaxeframe1 
}
}
endevent

[zazo]

EmericaSkater, on 15 October 2012 - 04:31 AM, said:

An action has to be declared for each frame of the animation

action flyingaxeframe1 0  1  5 1 10
action flyingaxeframe2 5  1  5 1 10
action flyingaxeframe3 10 1 5 1 10
action flyingaxeframe4 15 1 5 1 10
useractor notenemy flyingaxe 0  flyingaxeframe1 enda //initialized to the first action

Then you rig the animation with a peractor variable in EVENT_GAME

gamevar countvar 0 2
gamevar picnum 0 2

onevent EVENT_GAME
getactor[THISACTOR].picnum picnum
ifvare picnum flyingaxe {
addvar countvar 1 
ifvarg countvar 1 { 
setvar countvar 0 
ifaction flyingaxeframe1 action flyingaxeframe2 else
ifaction flyingaxeframe2 action flyingaxeframe3 else
ifaction flyingaxeframe3 action flyingaxeframe4 else
ifaction flyingaxeframe4 action flyingaxeframe1 
}
}
endevent

very subtle, but ils work !
I have never found myself! thanks

[Fox]

EmericaSkater, on 15 October 2012 - 04:31 AM, said:

gamevar countvar 0 2
gamevar picnum 0 2

onevent EVENT_GAME
getactor[THISACTOR].picnum picnum
ifvare picnum flyingaxe {
addvar countvar 1 
ifvarg countvar 1 { 
setvar countvar 0 
ifaction flyingaxeframe1 action flyingaxeframe2 else
ifaction flyingaxeframe2 action flyingaxeframe3 else
ifaction flyingaxeframe3 action flyingaxeframe4 else
ifaction flyingaxeframe4 action flyingaxeframe1 
}
}
endevent

I would prefer to just use ifactioncount instead of "countvar".

[zazo]

Is it possible to create a sprite blocking for the actors but not for projectiles: can be traversed by bullets, but not by the bodies of the characters .. ? a special cstat value?

[Fox]

Yes, it is very possible. The cstat flag that control the sprite blocking is 1, and the projectile collision 256.

The problem is that when a map is loaded, by default all sprites with a cstat of 1 are set to 257 automatically. Meaning that you must force it to remain at 1 via CON.

[zazo]

Fox, on 17 October 2012 - 11:05 AM, said:

Yes, it is very possible. The cstat flag that control the sprite blocking is 1, and the projectile collision 256.

The problem is that when a map is loaded, by default all sprites with a cstat of 1 are set to 257 automatically. Meaning that you must force it to remain at 1 via CON.

getplayer[THISACTOR]. .....
member structur to get the player health as a variable ?

[RichardStorm]

getplayer[THISACTOR].extra

Why don't you get a look to this ? http://wiki.eduke32....ructure_members

// EDIT
Sorry i read wrong. If you need to use it inside the aplayer actor you could simply write getactor[THISACTOR].extra. In other cases use getactor[player[THISACTOR].i].extra.

[zazo]

RichardStorm, on 17 October 2012 - 01:33 PM, said:

getplayer[THISACTOR].extra

Why don't you get a look to this ? http://wiki.eduke32....ructure_members

// EDIT
Sorry i read wrong. If you need to use it inside the aplayer actor you could simply write getactor[THISACTOR].extra. In other cases use getactor[player[THISACTOR].i].extra.

thanks, with the cstat values i search a property for actor to cross any other: this actor can not be stopped by actors "solid" in cstat 1.257 ...: it could be a single projectile (or actor) that pierce a line of ennemies and finally stops in the walls ...
Is it possible ?? also would try to force the coordinates with the x,y variables...

[zazo]

zazo, on 18 October 2012 - 01:43 AM, said:

thanks, with the cstat values i search a property for actor to cross any other: this actor can not be stopped by actors "solid" in cstat 1.257 ...: it could be a single projectile (or actor) that pierce a line of ennemies and finally stops in the walls ...
Is it possible ?? also would try to force the coordinates with the x,y variables...

possible to change the spritepal of the weapons in hands in first person view (tiles 2524) in game ?
in the player code, spritepal change only the color in third person view (F7).

[Danukem]

zazo, on 18 October 2012 - 04:50 AM, said:

possible to change the spritepal of the weapons in hands in first person view (tiles 2524) in game ?
in the player code, spritepal change only the color in third person view (F7).

You can change the spritepal of hud items, but it requires that they be drawn using CON scripting commands. By default all the hud stuff is hardcoded. EVENT_DISPLAYWEAPON can be used to override the hardcoded weapon display and make your own. Also, Hendricks266 wrote a mod that perfectly replicates all of the hud display using CON script, meaning you can use that mod and easily change the pal values as you desire.

[zazo]

Trooper Dan, on 18 October 2012 - 07:01 AM, said:

You can change the spritepal of hud items, but it requires that they be drawn using CON scripting commands. By default all the hud stuff is hardcoded. EVENT_DISPLAYWEAPON can be used to override the hardcoded weapon display and make your own. Also, Hendricks266 wrote a mod that perfectly replicates all of the hud display using CON script, meaning you can use that mod and easily change the pal values as you desire.

another annoying problem: the shrinker weapon ( shoot SHRINKER ) not working when it is shot by a new enemy code ( useractor or actor)...
??

[Hendricks266]

shoot SHRINKSPARK

[Fox]

How can I mimic a "per player per player" variable? For example, I want to display a scoreboard of RPG-only kills. I need to store in a per-player variable how many times player 1 was killed by player 2, player 3, etc. I can more-or-less do it having 8 per-player variables, but I wouldn't like that solution.

[Hendricks266]

You would have an array of however many players EDuke32 supports for each player, either containing how many kills achieved or how many times killed by.

[Fox]

I remember back then TerminX (?) posted a picture of an experimental multiplayer with more than 8 player. So I had like to be safe just in case.

Overall it would be very usefull to have some some sort of a gamevar with ranging ID, similar to ammo_amount #. Instead of the player scoring system, what if I wanted to have something like that for actors? It would require 16384 gamevars!

[zazo]

shoot SHRINKSPARK nothing happens !
the shrinkspark is shot with shoot SHRINKER
i changed the max player health. Is that the reason?

[zazo]

zazo, on 20 October 2012 - 12:34 PM, said:

shoot SHRINKSPARK nothing happens !
the shrinkspark is shot with shoot SHRINKER
i changed the max player health. Is that the reason?

all right, the error comes from my side: I moved and badly replaced some } in the code.
now ils work

[Darkus]

Hi everyone! It's been a while since I had posted anything on this forum...
I experimenting stuff with monsters and I have a little problem;
- I noticed that the enemies move differently with the chosen Difficulty: For example, in easy mode, their movements seem more 'smooth', but in normal mode or above, they move seems more 'realistic' (when they walking).

The problem is when I create an enemy that moves very fast. In easy mode it's okay, but in higher difficulties modes, he seems to "skip" or "teleport" in his movements.
So I asking if it's possible make him move 'smoothly', like in easy mode on other difficulties.

[Mblackwell]

Are you modifying a vanilla Duke3D/EDuke32 installation?

[Fox]

I did some tests and he is right.

Even if you are not using a custom CON file, you can still notice this behavior with the Newbeast.

[High Treason]

I seem to remember this enemy jitter being discussed before, I remember long words like "interpolation" being used but very little beyond that.

Anyway, main reason I am in here, second time in 24 hours I've been stumped, first ran into a problem with this and scrapped the code (a good thing, it was inefficient anyway).

Basically, I have a waypoint for some AI, I use eventloadactor to get it's lotag into a peractor variable, say, "WAY_LT" but I need this info in another actor, as it is used to tell that actor special things about that point of the path (and running more findnearactor code is likely not good for CPU usage). Anyway, the actor that uses the waypoints does several things, one of them is that once following a path, it will use findnearactor to get the ID of a waypoint in range to a var "WAY_ID" - as soon as WAY_ID is not -1 it stops looking and asks a bit more about the waypoint, a lot like this;

ifvarn WAY_ID -1
{
	getactorvar[WAY_ID].WAY_LT WAY_MD // The WAY_MD var is global, WAY_LT is peractor
	getactor[WAY_ID].ang STALKER_TANG // This seems to work every time
// Some other stuff is in here, which all seems to do what it is supposed to...
}

The actor successfully gets the angle of the waypoint, which is useful as this is what it tries to change it's own angle to by default, but WAY_LT is always 0. It doesn't stop there, in my previous effort, I set one waypoints xvel (I was using that in place of the lotag at the time, that was already in use) and from then on, it always ended up setting to that one waypoint's xvel, even at other waypoints (dndebug revealed that each waypoint's xvel was 0, as set in eventloadactor, it would only happen with that particular one.

This is all being tested in a new map so I don't think it can be my map being broken, this is the latest EDuke, but I was using one from June when I first ran into the bug (hence updating).

It is more than likely I have misunderstood the purpose of the getactorvar command? I still don't pretend to be any good at CON coding, I can usually get things to work now though without asking for help, so I've let myself down. I can post more code if required.

[Hendricks266]

It looks like you used getactorvar correctly. I would try replacing WAY_MD with WAY_LT, considering that WAY_LT is peractor anyway, unless that actor also uses WAY_LT for something else.

The only other thing I can think of is that findnearactor will always return the same actor within its range if the same range is used from the same coordinates, but I don't remember if that actor is necessarily the closest one.

[Mblackwell]

Why not store an array of all of your waypoint nodes (in order, by actor id) and then do a dist to find the nearest one. In my testing doing so is relatively efficient. For example use HITAG to number them, and use PAL to label which path (and use that as an offset). Then also store LOTAG as a set of flags which the actor can look at when it cares about the waypoint in question. Then rather than necessarily caring about what ang the waypoint has instead have your actor face the waypoint. With intelligently designed paths you should have no trouble.

[Reaper_Man]

High Treason, on 22 October 2012 - 03:44 PM, said:

Basically, I have a waypoint for some AI, I use eventloadactor to get it's lotag into a peractor variable, say, "WAY_LT" but I need this info in another actor, as it is used to tell that actor special things about that point of the path (and running more findnearactor code is likely not good for CPU usage). Anyway, the actor that uses the waypoints does several things, one of them is that once following a path, it will use findnearactor to get the ID of a waypoint in range to a var "WAY_ID" - as soon as WAY_ID is not -1 it stops looking and asks a bit more about the waypoint, a lot like this;

You may also want to look at the older version of my A* pathfinding code here. The associated ART tiles are missing but the code for waypoint building, path finding, and a basic AI for an actor that moves along a path of waypoints from A to B is all there. And it sounds like it did exactly what you're trying to do, at least in terms of creating a global array of waypoints as Mblackwell suggested and building a navigation path from one to the next. Feel free to use the code in part or in whole if it's what you need.

Though depending on what you're trying to do exactly, I may suggest my newer A* since it's better in every conceivable way.

[TerminX]

Darkus, on 22 October 2012 - 03:54 AM, said:

Hi everyone! It's been a while since I had posted anything on this forum...
I experimenting stuff with monsters and I have a little problem;
- I noticed that the enemies move differently with the chosen Difficulty: For example, in easy mode, their movements seem more 'smooth', but in normal mode or above, they move seems more 'realistic' (when they walking).

The problem is when I create an enemy that moves very fast. In easy mode it's okay, but in higher difficulties modes, he seems to "skip" or "teleport" in his movements.
So I asking if it's possible make him move 'smoothly', like in easy mode on other difficulties.

This is actually a very good post. Your request will be granted, though since it effectively directly changes per-tic movement rates for enemies and how often these movements occur, it will have to be tested pretty thoroughly to make sure the overall rate of speed still more or less matches the original game.

I am also taking the opportunity to add a new sprite flag that enables smooth movement. It turns out that in the original code, skills 0 and 1 allow the enemy to move once every other game tic whereas skills 2 and 3 allow movement at twice the normal speed for 3 tics in a row with every 4th tic having the movement cancelled. I don't think this is optimal at all, so I'm changing it so that all skills allow movement once every other tic. I had tested it with movement being allowed every tic, but it was TOO smooth and looked really, really strange, which is where the new sprite flag comes in... since some people have requested it in the past, the new flag will allow movement every tic without anything involving the skill coming into play.