Author Topic: Collision modeling  (Read 6112 times)

Offline Norad

  • Zinc Member
  • *
  • Posts: 98
Collision modeling
« on: October 31, 2005, 01:38:09 PM »
Okay i've held my tongue long enough about this subject..... i counted everytime I got "collided" with today,  8 times in 3 hours.... and everytime... i was the one losing a wing and tumbling down to the ground , i get the... "so and so has collided with you" and so and so flys away undamaged.   If there is a "collision".  BOTH planes should tumble out of the sky, not just the guy with the faster internet connection.  This is getting ridiculous, I know i'm not the only one fed up with the collision model.   All is takes is some dweeb to HO and collided with him and get the kill credit and fly away undamaged.  having a plane fly away undamaged in a mid-air collision is totally unrealistic!!

Offline Schatzi

  • Platinum Member
  • ******
  • Posts: 5729
      • http://www.slowcat.de
Collision modeling
« Reply #1 on: October 31, 2005, 02:36:48 PM »
Do some search on that topic. It has nothing to do with who has the faster connection. It has something to do with who actually *sees* the collision.

Now i can understand your frustration and venting. But would you rather have someone fly by 100 yds off and *you* loose a wing cause on the other guys FE there happened a collision?
21 is only half the truth.

Offline Kev367th

  • Platinum Member
  • ******
  • Posts: 5290
Collision modeling
« Reply #2 on: October 31, 2005, 06:03:24 PM »
Schatzi - partly WRONG.

It's not the speed but the latency (ping rate good indication).
Funnily enough spent some time today talking to Skuzzy about just this subject, heres the outcome -

1) A guy playing from Texas with a low (say 64ms) ping rate collides with a guy playing in Texas with a high (say 100ms) ping rate.
Guy in with 64ms will suffer collision almost 100% of the time.

2) Start applying that to Europe and rest of the World, where latencies will 'generally' be higher.

3) Asked him if one FE detects a collision why it can't be applied to both.
He said there'd be just as many complaints because people taking damage from a collision they 'didn't see'.

Basically if you have a low latency your hosed, just about any collision will result in you taking damage.
Reason - you see collision 1st, by the time the other guy has 'caught up',
the collision no longer exists.

I was speaking to him because I was in buffs today in the tail gun and a LA7 (surprise) comes up my 6, opens fire then flies right through me. I lose a buff, he flies on undamaged.
« Last Edit: October 31, 2005, 06:05:44 PM by Kev367th »
AMD Phenom II X6 1100T
Asus M3N-HT mobo
2 x 2Gb Corsair 1066 DDR2 memory

Offline Tilt

  • Platinum Member
  • ******
  • Posts: 7358
      • FullTilt
Collision modeling
« Reply #3 on: November 01, 2005, 04:35:04 AM »
Quote
Originally posted by Kev367th

1) A guy playing from Texas with a low (say 64ms) ping rate collides with a guy playing in Texas with a high (say 100ms) ping rate.
Guy in with 64ms will suffer collision almost 100% of the time.

2) Start applying that to Europe and rest of the World, where latencies will 'generally' be higher.

 


I always thought that for both players to be certain of a collision then both pings would have to be low. ie if one pinged at 64 (round trip) then that is 32 + 32 if another pinged at 150 then that is 75 + 75. So lag from one FE to another is 107 plus the AH server lag.

Hence its FE to FE lag which determines how large a lag error is(with respect to collision). Not the FE to server.

Given this both players need to be in Grapevine to achieve a greater chance of  mutually assured collision.

Once the FE to FE lag is high enough to provide "collision separation" then either player (short or long delay) could uniquely see the collision.

Is stuff "solid" during warping?
« Last Edit: November 01, 2005, 04:39:03 AM by Tilt »
Ludere Vincere

Offline Schatzi

  • Platinum Member
  • ******
  • Posts: 5729
      • http://www.slowcat.de
Collision modeling
« Reply #4 on: November 01, 2005, 06:58:50 AM »
Im aware that ping rate has something to do with what you see/happens in a 'collision'. Im also aware that colision model is in a sense inaccurate and unfair. Actually, thats what i wanted to say with my post. Im sorry if it came across wrong (i guess i was 'venting' as well for seing the same beaten dead horse brought up again - i apologize :)).

But i also believe that any change in colision model wouldnt make it really more 'fair', just place the unfairness (and whines) elsewhere. Beat on dead horse with different colour fur.


I myself die from collisions every now and then. More often than someone else dies from colliding with me it seems - but then that might just be perception psychology. I usually am §$"%"§$%$§%&$§/ for a sec or two, then take a deep breath and reup.


Lord, grant me the serenity to accept the things i cannot change, the courage to change the things i cannot accept, and the wisdom to hide the bodies of those I had to kill today cause they got on my nerves....



PS: Tilt, im not sure, but id say things are not 'solid' during warp. I dont even think they *move*. Id say they just reappear at their corrected position. its just our eye/the film viwer, that shows it as 'warp'. If things were solid during warp, how could planes warp through hills and ground?
21 is only half the truth.

Offline Tilt

  • Platinum Member
  • ******
  • Posts: 7358
      • FullTilt
Collision modeling
« Reply #5 on: November 01, 2005, 07:15:32 AM »
Quote
Originally posted by Schatzi

PS: Tilt, im not sure, but id say things are not 'solid' during warp. I dont even think they *move*. Id say they just reappear at their corrected position. its just our eye/the film viwer, that shows it as 'warp'. If things were solid during warp, how could planes warp through hills and ground?


Instinctively I would agree with this............... but then our FE is constantly "filling in" object movement between packets. This is normally on a very short time duration.............however we have seen folk with net issues just carry on their last trajectory prior to a warp..........and during this period our FE (or the server) is providing this data. Further stuff is solid during this period cos we can shoot it............ just before stuff is updated and it warps.

So when does the FE decide that the data loss is a warp?
Ludere Vincere

Offline Schatzi

  • Platinum Member
  • ******
  • Posts: 5729
      • http://www.slowcat.de
Collision modeling
« Reply #6 on: November 01, 2005, 09:52:40 AM »
Good question. Now if you got someone flying towards the ground and pull up at last moment. If our FE isnt fed the 'pulling up' info fast enough, we see him hitting the ground, right? Then we get the update and hes still flying.... Now... do/did we see him blowup flying into the ground or did he just warp through it? Does our FE wait for confirmation from his end till we see an actual crash? :confused:


Whoa, the more i think about it, the more confused i get...
21 is only half the truth.

Offline Waffle

  • HTC Staff Member
  • Administrator
  • *****
  • Posts: 4849
      • HiTech Creations Inc. Aces High
Collision modeling
« Reply #7 on: November 01, 2005, 10:11:32 AM »
I thought the white text "PLAYER ID has collided with you"...or something like that....was when the other player hit you...

"you have collided" is when you hit them...


is that correct?

Cause I'm still loosing crap with both white and orange text :D

Offline hitech

  • Administrator
  • Administrator
  • *****
  • Posts: 12418
      • http://www.hitechcreations.com
Collision modeling
« Reply #8 on: November 01, 2005, 12:16:34 PM »
Waffle is correct.

Kev either you missunderstood skuzzy.  Or he stated it unclearly.

Connection speed has nothing to do with who collides.


HiTech

Offline killnu

  • Gold Member
  • *****
  • Posts: 3056
Collision modeling
« Reply #9 on: November 01, 2005, 12:35:10 PM »
if i get message, "xxxx has collided with you" and that message only, why am i still taking the damage and he flies away cleanly?  

"you have collided" message means im losing something...correct?

if you get both messages, we should both be losing something...correct?
Karma, it follows you every where you go...

++The Blue Knights++

Offline hitech

  • Administrator
  • Administrator
  • *****
  • Posts: 12418
      • http://www.hitechcreations.com
Collision modeling
« Reply #10 on: November 01, 2005, 12:53:37 PM »
Quote
if i get message, "xxxx has collided with you" and that message only, why am i still taking the damage and he flies away cleanly?


If that message only , you shouldn't be taking damage from a collision. He could have shot you. And you might not see the damage to his plane.


HiTech

Offline Tilt

  • Platinum Member
  • ******
  • Posts: 7358
      • FullTilt
Collision modeling
« Reply #11 on: November 01, 2005, 12:55:07 PM »
Quote
Originally posted by Schatzi
Whoa, the more i think about it, the more confused i get...


Just imagine a system where everything happens at FE's and all the HTC server does is pass on location/status data and manage a terrain, and its objects whilst relaying the odd "system" report and radio traffic.
Ludere Vincere

Offline Schatzi

  • Platinum Member
  • ******
  • Posts: 5729
      • http://www.slowcat.de
Collision modeling
« Reply #12 on: November 01, 2005, 01:47:05 PM »
Quote
Originally posted by Tilt
Just imagine a system where everything happens at FE's and all the HTC server does is pass on location/status data and manage a terrain, and its objects whilst relaying the odd "system" report and radio traffic.


*bluescreen*

Fatal System Error.

Memory Overflow.
21 is only half the truth.

Offline Kev367th

  • Platinum Member
  • ******
  • Posts: 5290
Collision modeling
« Reply #13 on: November 01, 2005, 02:08:57 PM »
Quote
Originally posted by hitech
Waffle is correct.

Kev either you missunderstood skuzzy.  Or he stated it unclearly.

Connection speed has nothing to do with who collides.


HiTech



Not speed, latency, thought I mentioned that?

Well in which case something is up.
I've had "xxx has collided with you" and I've taken damage while he flies off unscathed.
Also had "you have collided with xxx", same result as above.
Basically as it stands any collision whether I collide with him or vice versa, I take damage, the other party doesn't.
AMD Phenom II X6 1100T
Asus M3N-HT mobo
2 x 2Gb Corsair 1066 DDR2 memory

Offline killnu

  • Gold Member
  • *****
  • Posts: 3056
Collision modeling
« Reply #14 on: November 01, 2005, 02:41:37 PM »
same here Kev...as have other in my squad.  thats why i asked for clarification, just to make sure what i was telling them was correct about the messages.  Maybe one of us will catch it on film, maybe one of the others have, ill have to ask.
Karma, it follows you every where you go...

++The Blue Knights++