collied

[hubsonfire]

Engrish is overrated.

[ink]

He screwed up his calculation. The latency using half the speed of light is 0.267 seconds. 267 milliseconds.

Edit: NASA actually has to deal with insane latencies in controlling its satellites. It's 180 or so light-seconds to Mars.

The distance from Dallas to London is about 5000 miles.

The latency at the speed of light would be 0.032 seconds. That is 32 milliseconds.

Given that Euro latencies are (anecdotally from this thread) approximately 300 ms, then Internet data travels at around 1/10 the speed of light at least in the realm of HTC.

is it me or is your calculations wrong?
you say in top quote latency at half the speed of light is 267 ms
 second quote you say latency at the speed of light is 32 ms

I am no mathematics expert by any stretch of the imagination,  but something don't add up:headscratch:

[Lusche]

Given that Euro latencies are (anecdotally from this thread) approximately 300 ms

Mine is ~160ms

[klingan]

Mine is ~160ms

Gone from 130ms in 2006 to 200ms at present time 

[Ghosth]

Want an easy way to see the difference that lag can cause?

Next time your in fairly good formation, ask your buddy, wingman, whoever is with you on vox what distance he see's your icon at from him. Compare it to what you see his at.

9 out of 10 there will be a significant difference in the range between yours and his.
Do this every day for a week and you'll see a significant "flux" or range in days.

More so prime time and on weekends than the relatively quiet daytime hours.
But don't take my word for it. 

Lag happens, Collisions happen, asking for both sides to take damage will never work until we have totally lag free networks.

[Vinkman]

Twitchy, it's been explained before earlier in the thread. As far as exploits - it's possible but not practical or desirable to exploit it as one cannot reliably estimate or guess the netlag of one's opponent. It's often easier to actually gun your opponent down than it is to try to make him collide with you.

As I proposed, you don't have to know the exact net lag for every plane you want to ram. Lag is never Zero and seldom as high as two sec. Assume a normal distribution between them, and you can determine the  average or 50-ith percentile lag. Now based on relative speeds between your plane 200mph-400mph and the bomber 280-220 mph you can calculate the 1 person collision window. This would mean a pilot could dive past the nose of bomber between [example] 0.5 and 1.5 plan lengths and get a single plane collision on the bomber not the fighter 60%, 70% or what ever percent of the time. That is how you do it with out knowing the exact lag of the plane you are

I don't advocate that this is a smart or even effective practice. The engineer in me realized that this was not IMPOSSIBLE or nearly impossible to calculate or execute for that matter and I wanted to share how.

[boomerlu]

is it me or is your calculations wrong?
you say in top quote latency at half the speed of light is 267 ms
 second quote you say latency at the speed of light is 32 ms

I am no mathematics expert by any stretch of the imagination,  but something don't add up:headscratch:

My top quote was calculating around the whole world at half speed.

The second is from Dallas to London (which is about 1/5 of the circumference of the globe and thus 1/5 of the time) at the full speed of light (which is twice as fast, and thus 1/2 as long).

I calculated it as a comparison to see how our real latencies stack up against theoretical "half the speed of light" which someone mentioned earlier. From that, clearly our achieved speeds are around around 1/10-1/20 the speed of light.

Trust me, there's no way you're getting 3.7s out of around the world at half the speed of light. Something you did was wrong. And no, it does not take a mathematics expert. Just punch in (24000/186000 * 1/2) into Google.

[boomerlu]

As I proposed, you don't have to know the exact net lag for every plane you want to ram. Lag is never Zero and seldom as high as two sec. Assume a normal distribution between them, and you can determine the  average or 50-ith percentile lag. Now based on relative speeds between your plane 200mph-400mph and the bomber 280-220 mph you can calculate the 1 person collision window. This would mean a pilot could dive past the nose of bomber between [example] 0.5 and 1.5 plan lengths and get a single plane collision on the bomber not the fighter 60%, 70% or what ever percent of the time. That is how you do it with out knowing the exact lag of the plane you are

Suppose we have a netlag that averages 0.2 seconds. That's pretty high. In 0.2s, our fighter plane dives at 300 mph and travels a distance of 26 meters. That's VERY close to the nose of the bomber, seems to me it's just as easy to RAM it.

Now, netlag can be as low as roundtrip of 0.05s (Dallas to Dallas) and as high say 0.5s. That gives you a range of 6.7-67 meters. Considering the average plane has about a 50 ft wingspan and therefore about a 50 ft length, one plane length is about 15 meters. So in our range of variance, we could fit about 4 plane lengths in. Ok, so your argument is that it's probably normally distributed about 0.1s or so with a StDev of about 0.05s which translates to 13 meters +/- 7 meters. So yes in most of the cases you could pass about 13 meters in front of the bomber (that's one plane length) and reliably reproduce this.

Variables wise it looks like you're right. Supposing you could actually fly that well, you could reproduce this reliably for a majority of the cases. But given our imperfect flying and lead adjustment skills, it's just as likely that the pilot rams the bomber for real and dies. If I could aim that well, it's far easier to just shoot the bomber.

[RTHolmes]

If I could aim that well, it's far easier to just shoot the bomber.

 

[CAP1]

Want an easy way to see the difference that lag can cause?

Next time your in fairly good formation, ask your buddy, wingman, whoever is with you on vox what distance he see's your icon at from him. Compare it to what you see his at.

9 out of 10 there will be a significant difference in the range between yours and his.
Do this every day for a week and you'll see a significant "flux" or range in days.

More so prime time and on weekends than the relatively quiet daytime hours.
But don't take my word for it. 

Lag happens, Collisions happen, asking for both sides to take damage will never work until we have totally lag free networks.

i've noticed that when i'm talking to someone, when i release my ptt button, if the other person responds quickly enough, i hear me on his speakers. this seems like almost a full second after i've said what i'm now hearing through his vox.

[boomerlu]

i've noticed that when i'm talking to someone, when i release my ptt button, if the other person responds quickly enough, i hear me on his speakers. this seems like almost a full second after i've said what i'm now hearing through his vox.

I think we have a built in VOX delay. Tune to yourself, for example, and try it.

[dedalos]

As I proposed, you don't have to know the exact net lag for every plane you want to ram. Lag is never Zero and seldom as high as two sec. Assume a normal distribution between them, and you can determine the  average or 50-ith percentile lag. Now based on relative speeds between your plane 200mph-400mph and the bomber 280-220 mph you can calculate the 1 person collision window. This would mean a pilot could dive past the nose of bomber between [example] 0.5 and 1.5 plan lengths and get a single plane collision on the bomber not the fighter 60%, 70% or what ever percent of the time. That is how you do it with out knowing the exact lag of the plane you are

I don't advocate that this is a smart or even effective practice. The engineer in me realized that this was not IMPOSSIBLE or nearly impossible to calculate or execute for that matter and I wanted to share how.

Why do you even try?  Numbers and facts mean nothing to these guys.  Funny how two guys run a couple of numbers and come up with the same conclusions.

[boomerlu]

Funny how two guys run a couple of numbers and come up with the same conclusions.

That's the point of running numbers. To see who's right. If both people are right.. well so be it.

[Shuffler]

i've noticed that when i'm talking to someone, when i release my ptt button, if the other person responds quickly enough, i hear me on his speakers. this seems like almost a full second after i've said what i'm now hearing through his vox.

Same here..... I also noted how great I sound 

[Karnak]

Suppose we have a netlag that averages 0.2 seconds. That's pretty high. In 0.2s, our fighter plane dives at 300 mph and travels a distance of 26 meters. That's VERY close to the nose of the bomber, seems to me it's just as easy to RAM it.

Now, netlag can be as low as roundtrip of 0.05s (Dallas to Dallas) and as high say 0.5s. That gives you a range of 6.7-67 meters. Considering the average plane has about a 50 ft wingspan and therefore about a 50 ft length, one plane length is about 15 meters. So in our range of variance, we could fit about 4 plane lengths in. Ok, so your argument is that it's probably normally distributed about 0.1s or so with a StDev of about 0.05s which translates to 13 meters +/- 7 meters. So yes in most of the cases you could pass about 13 meters in front of the bomber (that's one plane length) and reliably reproduce this.

Variables wise it looks like you're right. Supposing you could actually fly that well, you could reproduce this reliably for a majority of the cases. But given our imperfect flying and lead adjustment skills, it's just as likely that the pilot rams the bomber for real and dies. If I could aim that well, it's far easier to just shoot the bomber.

You are only displacing the aircraft on one axis though. It is in a three dimensional space and could easily be to one side of where you think it is.

Can it be done?  Yes, I've done it in a C-47 when I had no other option.  Is it smart?  No, odds are you'll just get yourself killed while passing back and forth in front of all of the bomber's, or fighter's, guns.