Is KSP's timewarp system flawed?

[abowl]

Hello,

Is the timewarp system flawed? What if we could jump time instead? How many times have you overshot something because of the timewarp system? What if you could select a point in orbit and jump time instead. The game will fade to black, and fade in immidately and you will have arrived at the destination. Of course time and the the game will have progressed, planets will have moved etc. But you are free from having to speed up time and slow it down at the correct time.

With a mechanic like this, it opens up for much larger solar systems. It opens up for the posibility of adding galaxies. Bodies that are decades away. And it wont change the difficulty of the game. You still have to plan your mission, bring enough fuel, equipment etc. and execute the proper manouvers at the correct time. But getting to the manourvre points and end destinations will be smoother, and the size of the game could expand where you have multiple star systems because you would be free of the time warp mechanic which can only go so fast. One of the reasons multiple star systems havent been added yet is because it would be hundreds of years away, and doing that kind of time even at maximum timewarp is going to take alot of time. I guess what im asking is, what are the benefit of a time warp system, rather than a jump to time system?

[HoloYolo]

I liek this idea.

[pincushionman]

How is this different from the warp-to feature? Other than you need to wait through it?

[Endersmens]

How is this different from the warp-to feature? Other than you need to wait through it?

The warp to feature doesn't work correctly sometimes. Many landings have been botched because "warp to" warped past.

[pincushionman]

Oh, I agree with that. But "this system has bugs" is different from "this system (concept) is flawed." On-rails warp(-to) simulates the evolution of the system; as mentioned in the OP, the simulation would still have to be carried out, it just wouldn't be animated. It's identical under the hood.

The OP may wish to try one of the mods that increases maximum time warp, as well as Kerbal Alarm Clock, which I believe is more robust than stock warp-to right now (set a placeholder maneuver node and warp to that)

[samstarman5]

Using the Warp To function takes care of preventing overshoots pretty well. That idea sounds like a lot of extra calculating to be done, since every body in the system has to be accounted for in that time frame. Planets, moons, tracked asteroids, debris, bases, landers, etc. Would probably take longer to do that than it would to just use the timewarp system.

[windows_x_seven]

Would be nice to have, although the current system works for me.

[uglyduckling81]

Kerbal Alarm Clock works perfectly. It never over shoots. I don't see any need to over complicate the issue.

If your not using KAC already then your not playing KSP correctly. It should honestly be brought into the KSP fold as default.

[nohelmet]

I'd like to see "warp to" bug-free. Other than that, I really like it.

[Alshain]

The warp to feature doesn't work correctly sometimes. Many landings have been botched because "warp to" warped past.

Then they need to fix the warp-to feature.

[Thomas988]

I really like this idea, but above all I'm still worried about timewarping through SOIs. Is it safe now?

[windows_x_seven]

I really like this idea, but above all I'm still worried about timewarping through SOIs. Is it safe now?

I think it got fixed a while ago.

[selfish_meme]

What would be a nice little improvement, especially for multiplayer eventually is an instant space warp drive. Currently multiplayer is really difficult because of time warp, a space warp drive could fix that. Unfortunately I think Squad have explicitly ruled out FTL though.

[PB666]

Is the timewarp system flawed?

The highest time warp ignores atmosphere, you could set your Pe at 10k alt and the highest time warp would fly right through it.

The physics warp often causes terrain collisions, and can kill large ships. (Dont use 4x for wheel vessels on kerbin or landing a space craft)

Breaking warp takes too long, the game is not as responsive to reducing warp as it should be.

For long journeys (like to Eeloo, its too slow).

Individually these problems can be repaired by

1. Anticipating ellipses that intercept atmosphere and dewarping before arriving (Its easy enough to determine the mimimum amount of time required to dewarp, and then find the point on the ellipse that is below 70k alt its time in the future and when the dewarp needs to begin (Sweep rule, keplars laws, orbital bodies sweep the same area per unit time at every point an orbit).

2. Travel too point and click.

3. More responsive down warping.

4. Another time warp level available for hohmann transfers outside of planetary SOI.

[rkman]

Oh, I agree with that. But "this system has bugs" is different from "this system (concept) is flawed." On-rails warp(-to) simulates the evolution of the system; as mentioned in the OP, the simulation would still have to be carried out, it just wouldn't be animated. It's identical under the hood.

It would be different in that when 'jumping' instead of warping there would be no chance to overshoot.

The OP may wish to try one of the mods that increases maximum time warp

Increasing time warp increases the chance of overshooting, so it would make the problem worse instead of solving it.

That idea sounds like a lot of extra calculating to be done

The calculations are fairly trivial because everything is on rails, so for each body a specific moment in time directly corresponds to a specific point on their orbit, no actual physics simulation needs to be done.

Edited July 14, 2015 by rkman

[Commander Jebidiah]

First off I like this idea.

Next, have some rep

Third, I'm not in favour of replacing the current timewarp system as it's useful within planetary sub-systems like the kerbin sub-system or the Jool sub-system. This idea would be really good for interplanetary and interstellar space. Currently we have the 'warp to here' function which is great, but if you told it to warp to your Eeloo encounter sometimes it can take 10 minutes at full warp to get there.

Fourth, this is in the wrong sub-forum, I think.

[Superfluous J]

The calculations are fairly trivial because everything is on rails, so for each body a specific moment in time directly corresponds to a specific point on their orbit, no actual physics simulation needs to be done.

What about a ship orbiting Kerbin with an orbit that has about a 1/1000 chance of intercepting Mun's SOI on each pass?

[r4pt0r]

What about a ship orbiting Kerbin with an orbit that has about a 1/1000 chance of intercepting Mun's SOI on each pass?

As someone who has lost an important Duna mission because that jerk IKE must have grabbed my probe while I was time warping a different craft, I can attest this is a thing.

[Red Iron Crown]

What about a ship orbiting Kerbin with an orbit that has about a 1/1000 chance of intercepting Mun's SOI on each pass?

Chance shouldn't be a factor. You run the numbers to get the first encounter.

[Sharpy]

Using the Warp To function takes care of preventing overshoots pretty well.

except when it throws you out of warp at 100x time speed.

Luckily the worst that happened to me with these was missing a periapsis by a couple minutes.

[KrazyKrl]

Lots of the underlying calculations in KSP are deterministic. These items cannot reliably be calculated at once when you want to jump; you need to simulate them at some point anyhow. And since the universe has lots of interactive parts to it, especially with ships up there you can dock to, and interact with.

You still need to take physics slices of your trajectory; and perform collision detection on it. SOI changes, atmospheric entries, and things like gravity assists; all must be accounted for, and all affect your trajectory after them.

At which timer resolution do you say is a reasonable factor? What happens if you brush Jool's dense atmosphere for a few seconds, but it may be enough to miss your gravity slingshot to Eeloo or something. The game must then simulate every encounter either when warping, or when changing your trajectory. There is no way around these calculations... Also, you need to perform all these calculations every tick for all craft in your universe.

[abowl]

Kerbal Alarm Clock works perfectly. It never over shoots. I don't see any need to over complicate the issue.

I

Are you kidding? With kerbal alarm clock and the current time warp system that IS the overcomplicated way. If you could select a point in orbit and have the game load up at that point, seems way simpler. 2 clicks, brief load and you are there. Imagine loading up next to jool, and have the game slowly fade it, that would be amazing.

Lots of the underlying calculations in KSP are deterministic..

I dont think deterministic means what you think it means.

- - - Updated - - -

What about a ship orbiting Kerbin with an orbit that has about a 1/1000 chance of intercepting Mun's SOI on each pass?

Is this really cause for concern? It doesent seem neccesary to take into consideration because it doesent seem like its the goal of the player to intercept the muns SOI. I have seen the pictures where people try to get into as many SOI as possible, for a nice picture, but this can still be done. And i dont see why you wouldnt be able to jump to any point on that trajectory. When you think about it, this makes gravity assists more rewarding as you can skip the time warp once you have finally gotten the correct manourvre up and executed it.

- - - Updated - - -

That idea sounds like a lot of extra calculating to be done, since every body in the system has to be accounted for in that time frame. Planets, moons, tracked asteroids, debris, bases, landers, etc. Would probably take longer to do that than it would to just use the timewarp system.

You do realise this is taking place when you timewarp, except it has the graphical overhead as well, which arguably makes the simulation slower.

Edited July 14, 2015 by abowl

[Red Iron Crown]

Jumps are definitely possible if everything is in a stable orbit without SoI changes, you can edit the time in a persistence file to try it. Seems like this is an implementable change.

[pincushionman]

It would be different in that when 'jumping' instead of warping there would be no chance to overshoot.

"Warp-to" is supposed to do this. The fact that it doesn't work correctly on hyperbolic orbits is a bug, not a feature. I'm not talking about manual time warp with the ./, keys. The only thing a jump-to system would add is that it would force the developers' hand and force them to fix it before they implement it. The underlying accelerated simulation is the same for both systems.

Increasing time warp increases the chance of overshooting, so it would make the problem worse instead of solving it.

which is why I continued the sentence with "…as well as Kerbal Alarm Clock." I've already noted that stock "warp-to" is buggy, and manual (./,) time warp shouldn't even be part if the discussion anymore. KAC is known for being reliable, especially since the stock "warp-during-SOI-transition" behavior was corrected. I suggested such mods so you can see if behavior they provide is similar enough to what you're/the OP is looking for, before more talk on changing the user experience.

The calculations are fairly trivial because everything is on rails, so for each body a specific moment in time directly corresponds to a specific point on their orbit, no actual physics simulation needs to be done.

The calculations are exactly the same (note, this is not an argument against jump-to). Time warp already needs to simulate orbits forward, so would jump-to. There's no sense reinventing the wheel. The only thing that changes is the user experience.

[KSK]

I don't like this idea. Fair play to the OP for suggesting it and if it gets implemented as a replacement/update to the current warp-to system then fine. I wouldn't use it myself but clearly some people would. If it replaced the current time-warp system altogether I think KSP would be a lot poorer for that change.

This is spaceflight. Things take time. Much as I dislike the phrase, pressing a button and skipping over most the actual flying-through-space part of spaceflight would be immersion breaking for me. Missing a maneuver node through over-zealous time-warping is annoying but no more so than crashing into the Mun, forgetting some small but essential piece of equipment, running out of fuel or power, or all the many things that can go wrong in KSP. Learn, move on, try again.

Sorry, this is something of a pet peeve and not directed at the OP specifically. It's just something I see far too much of on game forums in general. Some player does something stupid and the knee-jerk response is very rarely 'that was stupid, I'll remember not to do that in future' but mostly 'OMG this broken game needs to be fixed to stop me doing stupid things.'

It gets very tedious.

Short answer - I don't think the current system is flawed, it just requires some care to use.