Warzone 2100 Forums

NoQ wrote:I mean, if we can load maps just-in-time, why can't we load mods without restarting the app?

We can, and we do (you can drag a mod to autoload while Warzone is open, and it will be loaded upon game start - and will kick you for cheating if you try that in an MP game).

We just don't have a framework to transfer mods like we transfer maps yet.

NoQ wrote:

Warzone_Fan wrote:theres a problem with download
i tried to download it using mega manager
and when i tired to install it they say that the files been corrupted

Did you try downloading it without the mega manager?

o dowonload it directly from firefox and installed normally

cybersphinx wrote: The problem is that there is a line-of-sight calculation units use to see if they can fire, which is different from the path the projectiles will then take.

Why can't the line of sight calculation be based directly on the trajectory the projectile will take?

Couldn't the AI fire a sped up, simulated test projectile, and if it collides with the target, then the AI knows whether it has a clear shot over the terrain or not?

I notice, there is no readme.txt file (where describes wz commands & different info), only license stuff txt files.

My bad havn't seen doc folder.

Deus Siddis wrote: Couldn't the AI fire a sped up, simulated test projectile, and if it collides with the target, then the AI knows whether it has a clear shot over the terrain or not?

No, not like you propose. Balistique calculation I a bit tricky and the collision detection is worst. Especialy if you consider you must express the world in 1/128 of square to be precise. If you think about the hundreds of simultanous units that fire to each others, making the pure true calculation for everything would be crazy.

So the game need to make approximations and shortcuts all around. This one is just particulary bad.

while i'm here i just want to say : I dont like this change.
That's interesting in bumpy natural map, but that's awful in artificial/high cliff environement. We need to do something with that line of sight that give that much false positives.

Probably only cliff tiles should block bullets?

NoQ wrote:Probably only cliff tiles should block bullets?

That sounds like a good compromise to me.

cybersphinx wrote:The problem is that there is a line-of-sight calculation units use to see if they can fire, which is different from the path the

How can this be a problem? If I'm a warzone unit and see something, I can shoot at it, since the projectile will always be at least as high as the line of sight.
There can only be a problem, if the unit aims at some point other than the one it it sees. And that's the impression I got in the game: I get the impression that units are always pointing their guns at the lowest point of the target. Thus when they see e.g. a tank turret they will aim at the tracks, even if the tracks are hidden behind a hill. The projectile might then hit the hill.

If that is the case (I don't really know, wrt. to WZ I'm just a player not a developer), the correct fix would be to aim higher.

Philipp

P.S.: I just wanted to add: IMO Change: "Allow projectiles to go through terrain again" is bad news, even though it was broken, I'd have strongly prefered it to be fixed instead of removed.

My pennyworth. Is this not a matter of calculating by "centres" of objects, rather than, say, the actual dims? This might explain why the actual terrain boundaries - cliff edges etc - are ignored. That is, radar can see the centre of a droid because it is not obscured by the centre of the intervening terrain, even tho' the droid is actually out of the radar's sight.

PkK wrote:

cybersphinx wrote:The problem is that there is a line-of-sight calculation units use to see if they can fire, which is different from the path the

How can this be a problem?

You all do realize that my post didn't end there?

Why not use rectangular or circular collision math?

if x < tx+widht and > tx then x collide
if y < ty +length and > ty then y collide
if z < tz + height and > tz then z collide
if all three then you are in the hit prism

circular is more complex in 3d
try using a 3d version of tangent to get the distance and if it is less than the circle radius it is a hit.

if it is a straight projectile, then find the highest point of the terrain in the way and see if it interrupts the line of fire.
if it curves, then plot the curve on a 2d graph and place the terrain on it, and if they intecept then the projectile won't hit.

effigy wrote:

NoQ wrote:Probably only cliff tiles should block bullets?

That sounds like a good compromise to me.

Bad idea, by the way. Indirect weapon shells should always hit the ground at some point, and it's pretty hard to tell if they're going through it or into it.

NoQ wrote:Bad idea, by the way. Indirect weapon shells should always hit the ground at some point, and it's pretty hard to tell if they're going through it or into it.

That would not be a problem at all. You know where it is supposed to shoot at, so you can easily ignore all other surfaces. As to whether it would be a good idea, I'm a bit unsure.

@rockachu2: That is an attempt at a joke, right?

Hmm, really. Well, i guess, indirect-fire weapons should never hit any intermediate terrain; how exactly the trajectory is drawn is a different story, but they simply shouldn't, because there always exists a projectile trajectory to any point within range.