These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Upcoming Feature and Change Feedback Center

 
  • Topic is locked indefinitely.
 

[March] Speed changes and Orbit paths

First post
Author
SurrenderMonkey
Space Llama Industries
#61 - 2016-02-10 16:50:36 UTC
Captain Campion wrote:
I'm not quite following what this change means, any chance you could dumb it down for me, perhaps with pictures? :D


Ever notice that if you're just in orbit around something and you turn on your MWD, sometimes you fly off in a significantly different direction/orbital plane than you had previously been flying in?

This fixes that.

"Help, I'm bored with missions!"

http://swiftandbitter.com/eve/wtd/

Mister Ripley
Ministry of War
Amarr Empire
#62 - 2016-02-10 18:40:58 UTC  |  Edited by: Mister Ripley
Captain Campion wrote:
I'm not quite following what this change means, any chance you could dumb it down for me, perhaps with pictures? :D

Now:

  • Click orbit -> ship orbits in a certain inclination
  • Changing your max. speed makes your ship change its orbit inclination and sometimes even the orbit direction (clockwise to counter clockwise or vice versa). Here's another image, just ignore the description and look at the different orbits.
  • This leads to pretty irritating situations where you activate your AB or MWD while orbiting and your ships stops, makes a u-turn and then starst orbiting with a different inclination or even direction
  • Same for webbing. It can make your ship do weird **** if you orbit and get webbed.

New:

  • If your max. speed changes (AB/MWD/Web), your ship will just keep orbiting with the same inclination and direction as before.


As far as I understand it the initial orbit inclination will still be random or whatever the formula is and it will change if you hit orbit again after your max. speed has changed.

SurrenderMonkey wrote:
GetSirrus wrote:
CCP Fozzie wrote:
Stasis webifiers will no longer cause direction changes.

Would this statement include the use of webifiers to reduce time to warp for large or capital?

wat.

I think he confuses orbit direction with align direction and therefor align time.
Mr Floydy
Viziam
Amarr Empire
#63 - 2016-02-10 20:08:08 UTC
Thumbs up here.
Daemun Khanid
Corbeau de sang
#64 - 2016-02-10 20:09:39 UTC
This is a change I can whole heartedly get behind. I can't count the times I felt like I would have lived through a fight had my ship not stopped mid flight to change direction, sapping my velocity and transverse speed in the process. Some people may come to the conclusion that it will potentially make kiters a little harder to kill but I don't think it's in a manner that is unreasonable or beyond expectations of how ships should be behaving in the first place.

Daemun of Khanid

Beta Maoye
#65 - 2016-02-11 09:27:49 UTC
I don't want manual orbiting. I want a dogfight mode like Valkyrie.
Reaver Glitterstim
The Scope
Gallente Federation
#66 - 2016-02-11 17:03:17 UTC  |  Edited by: Reaver Glitterstim
I think anyone click-orbiting at close range in an interceptor is taking valuable time away from doing other, more important--and more strategic--things. It seems more of a veteran move to multi-task your ship, and more of a rookie move to micro-manage your orbit while you ignore everything else. If we want to encourage skill, we should want a simple enough orbit command that it becomes possible to multitask into other processes while still managing the orbit.


Post #36
Annia Aurel wrote:
There is a blatant error in Eve's physics calculations related to orbiting.

Angular velocity is calculated in an approximate fashion that takes into account the relative speed and position of the two ships in question but ignores their rotation.

Easy example, a frigate orbits a stationary target in a circular orbit. In real world physics, the angular velocity of the stationary target, as seen by the orbiting frigate is zero. In other words, the guns of the frigate will always point at the target without the need to turn in any way.

It calculates the target's position across a static background. When orbiting a target, you see it move across the static background. If your camera is positioned relatively close to your ship and you see a target seem to briefly stop moving across the static background, regardless of what causes it to seem to stop, you will have an easier time tracking it.

I successfully used this understanding of tracking on a MWD fit Maelstrom to hit NPC frigates with 1400mm artillery at about 10km without webbing or painting, by matching my velocity briefly with theirs.


Post #38
Dr Cedric wrote:
I'm surprised that with all of the deeper mechanics and code they have been looking at that they haven't looked at gun tracking and some of the other formulas. I'm sure there is room for improvement somewhere in there, whether its for server performance, or realism or something.

One thing comes to mind: the fact that range to target has no direct effect on tracking, which actually makes it easier to hit a moving target the farther they are from you due to the way angular velocity calculations work. By giving turrets a range-to-target accuracy effect, there will mostly no longer be a need to "get under the guns", as it will be safe to orbit at long range also, provided you simply have a high enough perpendicular (transversal) velocity and low enough signature radius.

I can see two ways to implement this feature:

Method 1
Change tracking to be based on transversal velocity instead of angular velocity. Seems most of the playerbase believes it is based on transversal anyway. I'll explain:
Consider a Wolf is orbiting you at 5km and a Jaguar is orbiting you at 20km. Both have the same sig radius and are moving at the same velocity.
Tracking is based on angular velocity, meaning that just as that Jaguar appears to be moving 1/4th as fast as the Wolf (when your camera is zoomed in on your ship), your turrets will in fact track that Jaguar 4x as well as they track the Wolf.

If tracking were based on transversal velocity, there would be no difference in your ability to track either target as they are both moving at the same intrinsic velocity.

Method 2
Keep tracking based on angular velocity but implement apparent signature radius into the calculation. This way the signature radius of target used in the calculation of tracking is determined by dividing the target's actual signature radius by a distance factor. I would use the base optimal+falloff of the turrets as 1, such that at a distance equal to the base optimal+falloff (before skills and effects), you track the target as if its signature radius were 100% of what it is. If the target is closer, its apparent signature radius is increased.

This alteration would make it far easier to hit webbed frigates at close range, but far more difficult to hit smaller ships towards the end of your range.

An adjustment to this idea could be to give each turret a fixed value range modifier (much like signature resolution) for their signature / range units = apparent signature. This way you avoid strange effects such as pulse laser Abaddons tracking frigates at 10km better than blaster Megathrons can.

FT Diomedes: "Reaver, sometimes I wonder what you are thinking when you sit down to post."

Frostys Virpio: "We have to give it to him that he does put more effort than the vast majority in his idea but damn does it sometime come out of nowhere."

Arla Sarain
#67 - 2016-02-11 17:57:17 UTC  |  Edited by: Arla Sarain
Annia Aurel wrote:
There is a blatant error in Eve's physics calculations related to orbiting.

Angular velocity is calculated in an approximate fashion that takes into account the relative speed and position of the two ships in question but ignores their rotation.

Easy example, a frigate orbits a stationary target in a circular orbit. In real world physics, the angular velocity of the stationary target, as seen by the orbiting frigate is zero. In other words, the guns of the frigate will always point at the target without the need to turn in any way. Build a paper model if you don't believe me.

In Eve physics, on the other hand, the same 'angular velocity' is calculated for both ships. Which is only correct if none of the ships are changing direction and is thus plain wrong in case of orbiting.

dicated physics loving space nerd

Thoughts?

Huh, sounds right.

Guess it is just looking for the change in X and Y in the local space rather than the direction relative to the other observer.

Odd, I always had the impression that the angular speeds would be cumulative.

Still, EVE ships don't have a heading/direction. They are just particles/dots, are they not?
Reaver Glitterstim
The Scope
Gallente Federation
#68 - 2016-02-11 18:10:45 UTC
There was actually a post a while back in a thread about this very topic, explaining in detail how exactly the turret mechanics in EVE at current actually are accurate with physics, and how they should not, in fact, rotate with the rotating ship. Put simply: a turret locked in place will follow the rotation of its ship, but a turret freed to turn easily and with extremely low fiction will tend to stay where it's facing rather than follow the turn of the ship.

FT Diomedes: "Reaver, sometimes I wonder what you are thinking when you sit down to post."

Frostys Virpio: "We have to give it to him that he does put more effort than the vast majority in his idea but damn does it sometime come out of nowhere."

Arla Sarain
#69 - 2016-02-11 18:18:19 UTC  |  Edited by: Arla Sarain
Reaver Glitterstim wrote:
There was actually a post a while back in a thread about this very topic, explaining in detail how exactly the turret mechanics in EVE at current actually are accurate with physics, and how they should not, in fact, rotate with the rotating ship. Put simply: a turret locked in place will follow the rotation of its ship, but a turret freed to turn easily and with extremely low fiction will tend to stay where it's facing rather than follow the turn of the ship.

It doesn't seem like it. The turret tracking speed is as of now a function of it's change of XYZ of in local space.

If you hold hands with another person, and just rotate around a point in between each other whilst constantly facing each other - the locations of your X1, X2 and Y1, Y2 are rotating around each other. But the observers really aren't. They don't change angular locations relative to each other, and end up constantly facing each other. If the turret was modelled after the neck/head, then it wouldn't change angular positions over time either.


I'd mock up an animation, but this seems closer to truth. Turrets do spin around the grid as a consequence of the ship changing local position, but the turret don't actually spin relative to the target, so there is no change in angular position, so no angular speed to observe.
Mister Ripley
Ministry of War
Amarr Empire
#70 - 2016-02-11 20:39:28 UTC  |  Edited by: Mister Ripley
Reaver Glitterstim wrote:

Method 1
Change tracking to be based on transversal velocity instead of angular velocity. Seems most of the playerbase believes it is based on transversal anyway. I'll explain:
Consider a Wolf is orbiting you at 5km and a Jaguar is orbiting you at 20km. Both have the same sig radius and are moving at the same velocity.
Tracking is based on angular velocity, meaning that just as that Jaguar appears to be moving 1/4th as fast as the Wolf (when your camera is zoomed in on your ship), your turrets will in fact track that Jaguar 4x as well as they track the Wolf.

If tracking were based on transversal velocity, there would be no difference in your ability to track either target as they are both moving at the same intrinsic velocity.

This doesn't make sense, since a target farther away is actualy moving slower if we talk about tracking. Tracking beeing the degree that your guns have to turn to keep track of a target.

Annia Aurel wrote:
There is a blatant error in Eve's physics calculations related to orbiting.

Angular velocity is calculated in an approximate fashion that takes into account the relative speed and position of the two ships in question but ignores their rotation.

Easy example, a frigate orbits a stationary target in a circular orbit. In real world physics, the angular velocity of the stationary target, as seen by the orbiting frigate is zero. In other words, the guns of the frigate will always point at the target without the need to turn in any way. Build a paper model if you don't believe me.

In Eve physics, on the other hand, the same 'angular velocity' is calculated for both ships. Which is only correct if none of the ships are changing direction and is thus plain wrong in case of orbiting.

Now, the error would be very simple to fix by calculating velocities based on position changes between server ticks and factoring in the corresponding rotation (just subtract this rotation from the present wrong formula, and voila, everything is correct).

Now, it can be argued that a) this will complicate calculations and b) it is desireable from a gameplay point of view.

a) Can only be answered by the devs

b) Is up for debate. I never get why orbiting in a fast frigate should give me any problems with tracking. A change of the formula would benefit people who actively pilot their ships, would benefit people in smaller ships (new players anyone?) and would make much more sense to the dedicated physics loving space nerd

Thoughts?

I think it's "b".

Changing the calculations to match real world physics like you describe it is a nice idea from the "this calculations are stupid" standpoint, but it would break the whole gameplay around tracking. You could just rotate your ship to match the angular velocity to the other ship.

Example: Set your speed to minimum speed and approach the other ship. Done. The speed of the other ship will almost not matter anymore since your ship will rotate faster than the other ship is able to orbit you, no matter if you are webbed or not. (To a certain degree of course. Like frig vs. Battleship) This way it would be possible to get near perfect tracking in almost every ship a few classes bigger than the opponent. It would be "align time" vs. "angular velocity"
Primary This Rifter
Mutual Fund of the Something
#71 - 2016-02-12 03:09:52 UTC
GetSirrus wrote:
CCP Fozzie wrote:
Stasis webifiers will no longer cause direction changes.


Would this statement include the use of webifiers to reduce time to warp for large or capital?

No it wouldn't, as when you're aligning your velocity vector is already along the warp path. This has no effect on the use of stasis webs for that.
Reaver Glitterstim
The Scope
Gallente Federation
#72 - 2016-02-12 03:39:19 UTC  |  Edited by: Reaver Glitterstim
Mister Ripley wrote:
This doesn't make sense, since a target farther away is actualy moving slower if we talk about tracking. Tracking beeing the degree that your guns have to turn to keep track of a target.

Only angular velocity uses apparent/relative motion. Transversal velocity uses intrinsic motion and only functions in Euclidean physics which all games run on, but which real life does not. EVE simulates angular velocity tracking using extraneous calculations in a Euclidean engine, and switching to transversal tracking would inded have the effect I explained, as well as reducing server load as far as tracking calculations.

FT Diomedes: "Reaver, sometimes I wonder what you are thinking when you sit down to post."

Frostys Virpio: "We have to give it to him that he does put more effort than the vast majority in his idea but damn does it sometime come out of nowhere."

Reaver Glitterstim
The Scope
Gallente Federation
#73 - 2016-02-12 03:44:03 UTC  |  Edited by: Reaver Glitterstim
Arla Sarain wrote:
If you hold hands with another person, and just rotate around a point in between each other whilst constantly facing each other - the locations of your X1, X2 and Y1, Y2 are rotating around each other. But the observers really aren't. They don't change angular locations relative to each other, and end up constantly facing each other. If the turret was modelled after the neck/head, then it wouldn't change angular positions over time either.

Get on a merry-go-round, get it spinning, and then stick your arm out and try to point at any part of the merry-go-round. While those parts appear stationary from the vantage point of your eyes, which are anchored to your rotating head, your arm is much more loosely anchored and will need to push against inertia and continually adjust its alignment in order to continue pointing at the object.

Now point at an object in the distance. It appears to be moving to your eyes, but you can much more easily remain pointing at it--at least briefly before your body turns too far to allow your arm to point at it.

FT Diomedes: "Reaver, sometimes I wonder what you are thinking when you sit down to post."

Frostys Virpio: "We have to give it to him that he does put more effort than the vast majority in his idea but damn does it sometime come out of nowhere."

Rain6637
NulzSec
#74 - 2016-02-12 08:44:28 UTC
Reaver you're confused.

Transversal is a speed. It doesn't account for distance. But Angular V does, and it's Transversal divided by distance.

Angular V is in radians. So is turret tracking. They're both the transversal compared to distance.

Angular V is also the same for both bodies at the same time. You don't want to mess with that.
Rain6637
NulzSec
#75 - 2016-02-12 08:52:20 UTC  |  Edited by: Rain6637
Transversal = arc, in meters per second
Distance = radius, in meters
Angular V = arc / radius = radians per second
Rivr Luzade
Coreli Corporation
Pandemic Legion
#76 - 2016-02-12 10:03:28 UTC
A good change to finally make EVE make more sense. You can still do it. Smile

UI Improvement Collective

My ridicule, heavy criticism and general pale outlook about your or CCP's ideas is nothing but an encouragement to prove me wrong. Give it a try.

Nicola Arman
Deep Maw Salvage
#77 - 2016-02-12 11:01:27 UTC
Great change! +1 Fozzie
Freelancer117
So you want to be a Hero
#78 - 2016-02-12 13:22:17 UTC
CCP Fozzie wrote:

We still highly advise players to learn how to use manual piloting for best performance, as the orbit command does not include any advanced logic for managing transversal or avoiding slingshot maneuvers. However now the orbit command will no longer freak out whenever your speed changes, which should bring it in line with reasonable expectations for how it should behave.
Well said Cool

Eve online is :

A) mining simulator B) glorified chatroom C) spreadsheets online

D) CCP Games Pay to Win at skill leveling, with instant gratification

http://eve-radio.com//images/photos/3419/223/34afa0d7998f0a9a86f737d6.jpg

http://bit.ly/1egr4mF

Camios
Sebiestor Tribe
Minmatar Republic
#79 - 2016-02-12 22:22:40 UTC
That's great, I've been saying this for ages!

Yours sincerely
A Happy Customer.
Faruzen en Divalone
Newbie Friendly Industries
#80 - 2016-02-13 23:40:47 UTC  |  Edited by: Faruzen en Divalone
Annia Aurel wrote:
There is a blatant error in Eve's physics calculations related to orbiting.

Angular velocity is calculated in an approximate fashion that takes into account the relative speed and position of the two ships in question but ignores their rotation.

Easy example, a frigate orbits a stationary target in a circular orbit. In real world physics, the angular velocity of the stationary target, as seen by the orbiting frigate is zero. In other words, the guns of the frigate will always point at the target without the need to turn in any way. Build a paper model if you don't believe me.

In Eve physics, on the other hand, the same 'angular velocity' is calculated for both ships. Which is only correct if none of the ships are changing direction and is thus plain wrong in case of orbiting.

Now, the error would be very simple to fix by calculating velocities based on position changes between server ticks and factoring in the corresponding rotation (just subtract this rotation from the present wrong formula, and voila, everything is correct).

Now, it can be argued that a) this will complicate calculations and b) it is desireable from a gameplay point of view.

a) Can only be answered by the devs

b) Is up for debate. I never get why orbiting in a fast frigate should give me any problems with tracking. A change of the formula would benefit people who actively pilot their ships, would benefit people in smaller ships (new players anyone?) and would make much more sense to the dedicated physics loving space nerd

Thoughts?


Now this would be a problem. Imagine server asking for relative positions of both ships between ticks (1 sec minimum + 2x ping time), and then calculating the damage and hit chance (second tick - we are over 2 sec now). Firing a gun would thus take several seconds to finish. Several seconds latency is not an enjoyable experience in a fight.

However I sure agree with you on the physics part, it would be great but I am afraid its not possible now. But I would like to hear an answer from dev on this too.