Post Reply 
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
There is no 'snap to enemy construct' effect with explosions, including flak.
2017-09-14, 08:05 AM
Post: #11
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-13 01:17 PM)MizarLuke Wrote:  While the flak has a large radius, the large radius only affects missiles and players. There is a smaller radius which DOES damage enemy constructs (or any constructs, including yours), but it is smaller.

Precisely.
The explosion radius is capped at 11 for the purpose of interacting with constructs but has no limits for those few entities which are not constructs but can still be damaged. (I'm sure NPCs would fall into that category as well if they were still in game, but they are not.)
Find all posts by this user
Quote this message in a reply
2017-09-17, 09:40 AM
Post: #12
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-12 02:06 PM)Gladyon Wrote:  There's a bit of code that probably did that, but it must be broken.
I've seen the problem while playing with the explosions, but I haven't dug into that bit of code to see why it doesn't work.

One way to 'fix' it is to use a larger maximum radius for the explosions, which is why I didn't bother... I'm basically waiting for Nick to decide to extend the max radius as about 99% players are asking for.

Only if we get a fix for turret pop and sub-constructs in the same patch and it doesn't kill the FPS.

-Do not bring forth an argument as fact that can be disproven with a 10 minute Google search.
Find all posts by this user
Quote this message in a reply
2017-09-17, 03:27 PM (This post was last modified: 2017-09-17 03:28 PM by Richard Dastardly.)
Post: #13
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
I was under the impression that HE missiles snapped their explosion to the nose now without a fuse - but testing with a ~100m long torpedo showed the explosion appearing in the centre of the warheads still *even with a prox fuse* for *nearly* all torps.

I suspect this thread would be more useful in bug reports.

Poke my boat! mostly pre-2.0 learning & catalogue thread - Heavy & light tanks 18/11/17. 6 ships made 2.0 aware. If it's not AotE I can't do it regularily - sea -> post processing -> eyestrain.
Find all posts by this user
Quote this message in a reply
2017-09-19, 01:48 PM
Post: #14
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-17 03:27 PM)Richard Dastardly Wrote:  I was under the impression that HE missiles snapped their explosion to the nose now without a fuse - but testing with a ~100m long torpedo showed the explosion appearing in the centre of the warheads still *even with a prox fuse* for *nearly* all torps.

I suspect this thread would be more useful in bug reports.

Now do a test with glass blocks where it hits, try to see if it is just a visual glitch or an actual error. see how much glass at the end is destroyed compared to around the middle.

2000mm HE dakka enthusiast.
Find all posts by this user
Quote this message in a reply
2017-09-20, 03:46 AM
Post: #15
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-19 01:48 PM)MizarLuke Wrote:  
(2017-09-17 03:27 PM)Richard Dastardly Wrote:  I was under the impression that HE missiles snapped their explosion to the nose now without a fuse - but testing with a ~100m long torpedo showed the explosion appearing in the centre of the warheads still *even with a prox fuse* for *nearly* all torps.

I suspect this thread would be more useful in bug reports.

Now do a test with glass blocks where it hits, try to see if it is just a visual glitch or an actual error. see how much glass at the end is destroyed compared to around the middle.

For a test with 100m long torpedos with I don't know how many HE warheads, the difference is "makes a huge hole in the ship" vs "does no damage at all" - and it's quite obvious Tongue I will fiddle with it in a more controlled manner at some point when I have time. It might also be a problem with overly long missiles, of course.

Poke my boat! mostly pre-2.0 learning & catalogue thread - Heavy & light tanks 18/11/17. 6 ships made 2.0 aware. If it's not AotE I can't do it regularily - sea -> post processing -> eyestrain.
Find all posts by this user
Quote this message in a reply
2017-09-21, 07:26 PM
Post: #16
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
After looking into it, I know why the 'snap' algorithm I saw do not work.

Some technical background:
Code:
There are 2 bounding boxes around each construct.
The first one is an AABB, it is aligned with the world.
There's a check that is perfectly correct to find all constructs whose AABB counding box intersect with the explosion.

Then, there's another bounding box, used to store the main blocks array. This one is aligned with the construct and is the one responsible for the slow-downs when you have a huge vehicle and you build a stick on top of it (the whole bounding box has to be extended and re-copied).
There's an algorithm to take the center of the explosion, and 'drag' it inside this bounding box.
From there, the explosion algorithm begins.
And that's the problem. The 'dragged' position is very often farther than 11m from the closest block.
So, the explosion will not hit the construct, because of the 11m limitation.

There's a way to avoid that, it's just a matter of dragging the center of the explosion closer to the construct.


I just did that and now it snaps (without SpinBlocksMadness).
I do not know why Nick and Khaz haven't done it, maybe they saw a balance issue that we're not aware of.

Also, the test example of the OP is very dangerous, because the explosion will also destroy the gun and detonate all the shells/ammo...
Now that it snaps, it also snap to the vehicle that fired it if it is in the range of the shell... maybe it was that the issue Nick and Khaz saw...
Find all posts by this user
Quote this message in a reply
2017-09-22, 07:57 PM
Post: #17
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-21 07:26 PM)Gladyon Wrote:  After looking into it, I know why the 'snap' algorithm I saw do not work.

(...)

There's an algorithm to take the center of the explosion, and 'drag' it inside this bounding box.
From there, the explosion algorithm begins.
And that's the problem. The 'dragged' position is very often farther than 11m from the closest block.
So, the explosion will not hit the construct, because of the 11m limitation.

There's a way to avoid that, it's just a matter of dragging the center of the explosion closer to the construct.


I just did that and now it snaps (without SpinBlocksMadness).
I do not know why Nick and Khaz haven't done it, maybe they saw a balance issue that we're not aware of.
Huh, very interesting. Is the AABB box that much bigger that it doesn't bring the large flak explosion on the example within range of either of the test constructs, then? How come it's so big?

(2017-09-21 07:26 PM)Gladyon Wrote:  Also, the test example of the OP is very dangerous, because the explosion will also destroy the gun and detonate all the shells/ammo...
Now that it snaps, it also snap to the vehicle that fired it if it is in the range of the shell... maybe it was that the issue Nick and Khaz saw...
That was entirely deliberate. Using something which would detonate at the slightest provocation makes it clearer that there is no hit whatsoever in the game's current build.
Find all posts by this user
Quote this message in a reply
2017-09-22, 08:05 PM
Post: #18
RE: There is no 'snap to enemy construct' effect with explosions, including flak.
(2017-09-22 07:57 PM)BioPhoenix Wrote:  Huh, very interesting. Is the AABB box that much bigger that it doesn't bring the large flak explosion on the example within range of either of the test constructs, then? How come it's so big?

I don't know waht the size of the AABB is.
But the test is made to know if the sphere of the explosion intersects with it, so that test is the good one.
The problem is the size of the main array box, which is a lot larger than the vehicle, so bringing the center of the explosion on its border means placing the center of the explosion more than 11m of the vehicle.


(2017-09-22 07:57 PM)BioPhoenix Wrote:  That was entirely deliberate. Using something which would detonate at the slightest provocation makes it clearer that there is no hit whatsoever in the game's current build.

Then it's a good test.
When I tried it with the fix, the construct exploded and the Avatar with it.
I'd say: mission successful.
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)