Auto Pilot/Auto Maneuver would be nice

[Guest]

12 hours ago, Sesshaku said:

Having automation tech tree not only would allow players more freedom of play, but, if properly implemented, it could also be an essential part of the tutorial. What better way of teaching a new player how to orbit a planet, or how to use gravity to accelare or reduce speed? Just make him see an automated flight that has full instructions.

That’s a really bad way of teaching a skill. It’s essential to get your hands dirty ASAP. Even an annotated video would be more effective. A classic tutorial that pauses when you need to do something, explains why you need to do it, then guides you through it is even better.

[mcwaffles2003]

[snip]

On 3/4/2021 at 12:24 AM, Brikoleur said:

My opinions aren’t about how you play the game. They’re about what the game you play ought to be like.

So you're not telling me how to play the game you are just suggesting limiting the ways in which I can play the game more narrowly so that I have to play the game in the way you intend?

C'mon dude.... logic.

Having the autopilot type system doesn't necessitate someone uses it. Lacking an autopilot disables anyone from using it. Don't force me to play the game to your image of how it should be played, I'm not suggesting that to how you can play.

Edited March 5, 2021 by Snark
Redacted by moderator

[Guest]

44 minutes ago, mcwaffles2003 said:

So you're not telling me how to play the game you are just suggesting limiting the ways in which I can play the game more narrowly so that I have to play the game in the way you intend?

A game is defined by the way it limits your freedom to play it, i.e. its rules and features. There is no way to discuss a game's characteristics without discussing the ways it limits your way to play it. They're two sides of the same coin.

A bishop can only move diagonally in chess. Suppose you propose that it should also be able to leap like a knight. Arguing that not allowing it to leap like a knight is an imposition on your freedom to play chess like you want is silly. If you think it's fun, by all means get together with someone who agrees with you and play chess that way – but don't expect your rule to become part of the standard rules of chess on the strength of the "my freedom to play as I want" argument. 

This discussion is exactly analogous. In my view, introducing a stock MJ-style autopilot* would make KSP2 worse, much like allowing a bishop to also move like a knight would make chess worse. And just like there's absolutely nothing wrong with playing a house-ruled game of chess where bishops also move like knights, there's absolutely nothing wrong with making or installing a mod that introduces, changes, or removes features to your liking.

*unless effectively limited to the mid/late game, which would mean not having sandbox mode anymore

44 minutes ago, mcwaffles2003 said:

Having the autopilot type system doesn't necessitate someone uses it. Lacking an autopilot disables anyone from using it. Don't force me to play the game to your image of how it should be played, I'm not suggesting that to how you can play.

You can make the same argument about literally any feature, and it's just as silly. FTL drives, guns, infinite-fuel drives, wormholes, ... you name it: "Put it in! If you don't like it, just don't use it!" It doesn't add anything meaningful to the discussion. Why not just state "I really want a stock MJ because A, B, C" and go from there?

Edited March 4, 2021 by Guest

[Guest]

10 hours ago, mcwaffles2003 said:

but don't force the rest of us to perform the same monotonous maneuver a million times because it doesn't fit your standard

We already know that the game will have:

• Brachistocrone trajectories running in the background
• Automated supply runs
• Off-planet launchpads

They've already confirmed that reducing the monotony of repetitive mission is one of their design goals.

A MJ equivalent would not serve that purpose, you're proposing a solution to an already solved problem.

 

If the point is making flying optional then someone has to openly say it so we can switch to the next step and debate if flying should be optional in a flight sim.

[mcwaffles2003]

53 minutes ago, Brikoleur said:

This discussion is exactly analogous

No, it's not. Changing how a bishop moves in chess forces everyone playing that game to use the bishop in that way, adding an autopilot feature does not do that.

53 minutes ago, Brikoleur said:

In my view, introducing a stock MJ-style autopilot* would make KSP2 worse, much like allowing a bishop to also move like a knight would make chess worse. 

It actually doesn't make chess worse, it's just called shogi then. Also, you're stating that like it's an objective fact. In chess, you can make a measure for quality of play based on how many endgames end in a tie (less is better), a subjective measure, but at least based in a reasoned argument. And chess as we speak is actually being reformed, so funny you mention that.

Related: https://www.chess.com/news/view/new-alphazero-paper-explores-chess-variants

But anyways, I digress. Back to this specifically:

53 minutes ago, Brikoleur said:

In my view, introducing a stock MJ-style autopilot* would make KSP2 worse

Really? Have you played KSP 2? Have you've built an interstellar civilization with many colonies that demands large quantities of resource transfers to function in a reasonable amount of time (<1,000 hours). Please, go on, how does adding an autopiloting system make the game worse? How does making everyone manually execute every maneuver for the whole game make that game better? Does it all of a sudden make it better if the runs are entirely simulated off screen to avoid milk run simulated: space edition, because if that is your argument, then wow....

53 minutes ago, Brikoleur said:

there's absolutely nothing wrong with making or installing a mod that introduces, changes, or removes features to your liking.

But there is something wrong with making people depend on mods to play the way the game requires in order to progress within it..

53 minutes ago, Brikoleur said:

*unless effectively limited to the mid/late game, which would mean not having sandbox mode anymore

I don't think anyone suggested autopilot comes on the first rocket. And if you wanna play sandbox and don't like autopilot... perhaps then don't use it.

 

Overall, I just don't think you've backed up from how you've experienced KSP 1 and have considered how drastically different of a game KSP will most likely be.

16 minutes ago, Master39 said:

We already know that the game will have:

• Brachistocrone trajectories running in the background
• Automated supply runs
• Off-planet launchpads

They've already confirmed that reducing the monotony of repetitive mission is one of their design goals.

A MJ equivalent would not serve that purpose, you're proposing a solution to an already solved problem.

 

If the point is making flying optional then someone has to openly say it so we can switch to the next step and debate if flying should be optional in a flight sim.

So it's okay to fully simulate trips off screen, but riding along is bad? Go on...

Also, haven't played MS flight sims since MS flight sim 2000, but if I recall correctly, that game did have an autopilot feature and I believe the new one does as well since its mimicking real planes. Because a game that expects you to manually fly a plane for 3 hours straight as the only manner of play probably wouldn't sell well.

Edited March 4, 2021 by mcwaffles2003

[Guest]

45 minutes ago, mcwaffles2003 said:

how does adding an autopiloting system make the game worse?

If it's available from the start, it will nudge new players to rely on it rather than learning to fly the craft themselves. They will never discover whether they actually enjoy flying them. It will also make it much more difficult for them to learn how to design craft, because they won't be able to judge how they handle. This would result in players who are puzzled and frustrated because their craft flip, or fail to make orbit, or otherwise collapse into rubble and they don't understand why. 

If it's better at flying the craft than a skilled human pilot, then it invalidates a central gameplay feature, which is a bad move in my view. If all it does is automate routine manoeuvres and is effectively restricted to the mid/late game, then I'm not opposed to it in principle. If it's good enough to, say, auto-land aircraft or be supremely efficient at auto-landing spacecraft, then it ought to be gimped.

And the final point is opportunity cost: any effort spent on the autopilot is effort not spent on something else, and there are a ton of features I believe would contribute more to the gameplay experience than autopilot.

45 minutes ago, mcwaffles2003 said:

How does making everyone manually execute every maneuver for the whole game make that game better?

That's not what I'm arguing for.

45 minutes ago, mcwaffles2003 said:

Does it all of a sudden make it better if the runs are entirely simulated off screen to avoid milk run simulated: space edition, because if that is your argument, then wow....

Some kind of automation for milk runs is a necessity; otherwise the mid/late game would become unbearably tedious. You do not need a MJ style autopilot for that, however -- a much simpler system that abstracts the milk runs out would address that need just as well.

45 minutes ago, mcwaffles2003 said:

But there is something wrong with making people depend on mods to play the way the game requires in order to progress within it.

I'm not sure I understand you here. Are you implying that a full MJ-style autopilot is a necessity for being able to progress in KSP2? If so, please elaborate.

45 minutes ago, mcwaffles2003 said:

I don't think anyone suggested autopilot comes on the first rocket. And if you wanna play sandbox and don't like autopilot... perhaps then don't use it.

I already pointed out that "just don't use it" is a non-argument.

45 minutes ago, mcwaffles2003 said:

Also, haven't played MS flight sims since MS flight sim 2000, but if I recall correctly, that game did have an autopilot feature and I believe the new one does as well since its mimicking real planes. Because a game that expects you to manually fly a plane for 3 hours straight as the only manner of play probably wouldn't sell well.

It most definitely has autopilot – or, rather, many planes in it do. 

It's also a really detailed, realistic flight sim. You can't just jump into a plane and click the "fly" button. Even programming in waypoints requires knowledge of what you're doing. If your plane supports MLS, then you need to autoland at an airport that provides it. It's simply not possible to use the autopilot in it without knowing what you're doing. Or, in other words, it's limited enough that it doesn't remove the need to actually fly the planes.

This is not at all like KSP, where you can easily tell MJ what you want your rocket to do, click a button, and watch it go – and it'll do it better than a human, a lot of the time. I.e., MJ effectively invalidates a huge gameplay area in KSP, which is why I feel very strongly that any stock autopilot, including late-game ones, must be significantly more limited than MJ.

-- Look: I think this discussion isn't very productive; it's devolving fast into "is not! is too!" In my view, a better way to frame it would be something like this:

• If stock autopilot is implemented in KSP2, what capabilities and limitations should it have? When and how should it be made available to players? How would this affect the gameplay experience? How would it change the way new players learn the game?
• If stock autopilot is not implemented in KSP2, what problems would that cause? What alternative approaches are there to solve those problems?

Edited March 4, 2021 by Guest

[mcwaffles2003]

On 3/4/2021 at 5:00 AM, Brikoleur said:

Some kind of automation for milk runs is a necessity; otherwise the mid/late game would become unbearably tedious. You do not need a MJ style autopilot for that, however -- a much simpler system that abstracts the milk runs out would address that need just as well.

So, once again...

[snip]

On 3/4/2021 at 5:00 AM, Brikoleur said:

it will nudge new players to rely on it rather than learning to fly the craft themselves. They will never discover whether they actually enjoy flying them.

Huh, sounds like that's their problem... or maybe it's not a problem at all.

On 3/4/2021 at 5:00 AM, Brikoleur said:

I'm not sure I understand you here. Are you implying that a full MJ-style autopilot is a necessity for being able to progress in KSP2? If so, please elaborate.

If anyone is going to make reasonable progress in a game that is expanded by orders of magnitude on another game that is already famous for requiring 1,000s of hours out of people it is ridiculous that by someones 50th flight the game not make it available to automate something as mundane as an inclination change based on the argument that the pilot should prove that they could manually perform it, despite having done so already 10's of times. We don't even expect this out of real astronauts flying real space ships....

[snip]

Edited March 5, 2021 by Snark
Redacted by moderator

[Guest]

28 minutes ago, mcwaffles2003 said:

So it's okay to fully simulate trips off screen, but riding along is bad? Go on...

It's ok to have any kind of automation integrated with the progression, as I wrote multiple times already, it's not ok to demand for flying to be entirely optional from the start, it's (at least partially) a flight sim we're talking about.

Since that kind of automation has already been confirmed, and nobody argued to remove it, and most of the people talking here were the ones highlighting the "milk run problem" back when we didn't know about the supply routes, I don't think that the "i don't want to run the same mission a million times" has any place in this discussion. It's an already solved problem.

 

I mentally split the game (KSP2) in 3 phases:

• Exploration: You're exploring a new body or system, gathering science, finding spots for bases or stations.
  • I figure this portion of the game running more or less like a KSP1+, mostly crafts custom tailored for the mission at hand, the most repetitive part is getting away from the starting point, the solution to that repetition is that you'll be able to build launchpads on other bodies, changing the conditions dramatically.
• Colonization: You put your first bases on said body, primary function is to become autonomous and generate science
  • Colonization is a broad term, it includes everything ranging from your Laythe City all the way down to your first LKO space station, it's were you see the first permanent or semi-permanent outposts appear and it's were you'll start to see the Milk Run Problem appear, here the supply route system makes its appearance, you fly manually your first Demo-1 mission for your newly built Crew Phoenix but then let the system run the crew rotation missions for you.
  • Unrelated note: here is also were I hope most of the science is conducted.
• Infrastructure: You set up the facilities, relays, mining operations, launchpads, and supply routes to allow for further exploration to start from your new colony
  • Here's where the supply route system will shine, you do a bunch of manual missions to prove various designs for various routes and then everything just works, if you don't like optimizing every route you just design a bunch of over-engineered jack-of-all-trades heavy lifting landers and/or interplanetary tugs and call it a day, but if you do enjoy designing and flying a lot of ships you can make custom-tailored designs for every route and costantly keeping them updated with the latest tech.

At phase 3 you'll possibly have tens if not hundreds of flights going on at any given time, building things and moving amounts of resources that are simply not possible in KSP1 and begin to approach the point in which, if you stop to see every single mission fly (that means no simultaneous ones) you'll slow down the progress of everything almost to a halt.

All of this is what I call the "off flight scene" automation, and it takes care of most, if not all, the tedium and the repetitiveness of trying to run a Jool mission while your KSS in LKO requires crew rotations every 6 months and 3 supply runs a year.

 

What's left for the "in flight scene" automation is automating orbiting, maneuver nodes and rendezvous and docking, which can range from making some routine maneuvers automated after you've done them all the way up to making flying entirely optional. On that scale, I'm more on the "put those things in a tech node (even better if a separate one for each if those functions)  at mid to late game".

An autopilot for planes with things like altitude and direction hold and tools like ILS it's also another completely different but related feature, a welcomed one for anyone that wants to fly anything more advances than a WWII aircraft.

 

Making the supply route system runs completely physical and "followable" outside seeing them arrive and dock/land at the destination colony/station?  It would be a fun addition, I have doubts on if it would be worth it in terms of spent resources, but it would change nothing in terms of gameplay.

[mcwaffles2003]

3 minutes ago, Master39 said:

It's ok to have any kind of automation integrated with the progression, as I wrote multiple times already, it's not ok to demand for flying to be entirely optional from the start

Please point out where I said or even eluded to "from the start"

I have been arguing the feature should exist in the game. At no point have I suggested, eluded, or implied that autopilot from the start is a good idea... And if I recall correctly, mechjeb itself isn't even available at the start of a career.

Edited March 4, 2021 by mcwaffles2003

[Starhelperdude]

4 hours ago, Brikoleur said:

That’s a really bad way of teaching a skill. It’s essential to get your hands dirty ASAP. Even an annotated video would be more effective. A classic tutorial that pauses when you need to do something, explains why you need to do it, then guides you through it is even better.

I've learned to make an orbital rendervous through mechjeb, I've watched what mechjeb did and later did it myself, the tutorials leaned me the basic stuff but not what to do in the specific thing

I would say there will be probably the colony supply sytem and that stuff, and there will probably be a mod like mechjeb for well, mechjebbing, also I find it ok if others fly with mechjeb, for example sometimes I'm doing someting else during playing ksp and then I just use the mechjeb ascend guidance, or I use it if I fly for example Vertical SSTOs (no spaceplanes) where I need to make the gravity turn very carefully

though I have nothing against a sort of stock mechjeb

[Guest]

6 minutes ago, mcwaffles2003 said:

Please point out where I said or even eluded to "from the start"

I have been arguing the feature should exist in the game. At no point have I suggested, eluded, or implied that autopilot from the start is a good idea... And if I recall correctly, mechjeb itself isn't even available at the start of a career.

When everybody agrees that automation shouldn't be available from the start and that flying shouldn't be optional and you come in asking what automation would remove from their game if they don't use it and why they want to force everyone to play the game their way, unless you're searching the debate for the sake of it, you're saying that.

The repetitiveness of milk runs is a solved problem, and (almost) no one here has problems with automation features being unlocked through gameplay, more than that there's only having those features from start and thus making flying optional.

[mcwaffles2003]

12 minutes ago, Master39 said:

When everybody agrees that automation shouldn't be available from the start and that flying shouldn't be optional and you come in asking what automation would remove from their game if they don't use it and why they want to force everyone to play the game their way, unless you're searching the debate for the sake of it, you're saying that.

The repetitiveness of milk runs is a solved problem, and (almost) no one here has problems with automation features being unlocked through gameplay, more than that there's only having those features from start and thus making flying optional.

I don't see where I've implied "from the start" and that is not my stance. So if you'd like to, please quote it. @Brikoleur has numerous times, on the other hand argued against developing an autopilot feature as it's a waste of dev time and goes against what they believe is the way the game should be played:

19 hours ago, Brikoleur said:

It is a bad idea to spend time and effort to implement optional features, especially if they go directly against one of the main pillars of the game's design intent.

19 hours ago, Brikoleur said:

I am saying that KSP is a game about building and flying spacecraft, and if you're not interested in building and flying spacecraft, then you will most likely find some other game more to your liking. It doesn't mean you're a terrible person or bad or wrong, it just means your interests don't align with the kind of game KSP is. No game can please everybody.

20 hours ago, Brikoleur said:

The argument is, “if it exists, lots of people aren’t going to learn to fly manually and will thereby miss out on one of the most important and enjoyable aspects of the game, so it shouldn’t be stock.”

As other have said no doubt MJ will be one of the first major mods for KSP2, and if someone wants to install it without ever having played it before, I’m not going to stop them. I do think they will likely be missing out on a lot of fun, and therefore I think it would be a bad idea to nudge players to do it by making it a standard feature.

On 3/2/2021 at 11:36 PM, Brikoleur said:

In my view autopilot goes against the grain of KSP where flying your craft is such a central gameplay element.

etc...

 

I disagree

Edited March 4, 2021 by mcwaffles2003

[Guest]

[snip]

Just to sum up, here's my position on stock autopilot – and it has evolved somewhat since the start of the thread, as such things do:

• It needs to be limited enough that it doesn't make flying entirely optional at any point in the game.
• It must be effectively gated to the mid/late game. If it's through the tech tree, that leaves open the question of sandbox mode. Perhaps sandbox mode would only be unlocked after you've unlocked, say, the first tier of the tech tree in adventure mode?
• Autopilot of any kind is not strictly necessary, as there are alternative ways to automate away mid/late game grind. Therefore it needs to be prioritised against other not strictly necessary features.

And that's pretty much it, really. I'm happy to discuss any of these positions [snip]

On 3/4/2021 at 5:33 AM, Starhelperdude said:

I've learned to make an orbital rendervous through mechjeb, I've watched what mechjeb did and later did it myself, the tutorials leaned me the basic stuff but not what to do in the specific thing

I think that's more of a reflection on just how bad KSP1 is at teaching players the basics of orbital mechanics than the utility of an autopilot at doing it. FWIW I had immense trouble learning to RV and dock too.

Edited March 5, 2021 by Guest
Redacted by moderator

[mcwaffles2003]

[snip]

On 3/4/2021 at 6:05 AM, Brikoleur said:

• It needs to be limited enough that it doesn't make flying entirely optional at any point in the game.

Define flying. I'm assuming you mean every flight must have some manual control during its course, if thats the case, I disagree and will explain further if you confirm that is what you mean.

On 3/4/2021 at 6:05 AM, Brikoleur said:

• It must be effectively gated to the mid/late game. If it's through the tech tree, that leaves open the question of sandbox mode. Perhaps sandbox mode would only be unlocked after you've unlocked, say, the first tier of the tech tree in adventure mode?

I don't want autopilot in the start of a career. (I am making this as clear as I can)

As far as sandbox mode goes, I don't get the point of delaying it and forcing people to play adventure mode. I resent games that force you to play tutorials before you can play the game, it is just aggravating to deal with (just my opinion).

I, personally, have pretty much only played career and have barely touched sandbox outside of testing ship designs. That said, I don't get the point of the delay other than ensuring the game excludes people who only want to play using autopilot. Its called sandbox because everything is just set in front of you and you can just go play. I don't get the hang-up on making sure there is absolutely no way someone can experience the game without manually flying. This is a very controlling attitude and I don't know why you insist. Just let people play in the sandbox.

Would you force a kid to make a sand castle by hand before allowing them to fill a pale with sand and flipping it? I say leave the kid alone and let them do their thing.

On 3/4/2021 at 6:05 AM, Brikoleur said:

• Autopilot of any kind is not strictly necessary, as there are alternative ways to automate away mid/late game grind. Therefore it needs to be prioritized against other not strictly necessary features.

I agree that a mechjeb like system should not be the only form of automation, I hope not every flight has to be loaded and fully physically simulated if the route has been proven to be possible.

That said I think its worth prioritizing and is practically necessary in the same way a steering wheel is necessary in a car. Could you take the steering wheel off and just twist the steering column directly with your hand? Yeah, but it'd be extremely inconvenient and I'd be riding a bike more often instead to avoid the hassle.

On 3/4/2021 at 6:05 AM, Brikoleur said:

I think that's more of a reflection on just how bad KSP1 is at teaching players the basics of orbital mechanics than the utility of an autopilot at doing it. FWIW I had immense trouble learning to RV and dock too.

Is it that hard to admit that an autopilot feature can have any intrinsic value?

Edited March 5, 2021 by Snark
Redacted by moderator

[Guest]

On 3/4/2021 at 6:34 AM, mcwaffles2003 said:

Define flying. I'm assuming you mean every flight must have some manual control during its course, if thats the case, I disagree and will explain further if you confirm that is what you mean.

Yes, that's correct, except the milk runs fully abstracted out with the supply line automation mechanic. 

On 3/4/2021 at 6:34 AM, mcwaffles2003 said:

I don't want autopilot in the start of a career. (I am making this as clear as I can)

Good, in that case we're in agreement about this point, so no further discussion is necessary.

On 3/4/2021 at 6:34 AM, mcwaffles2003 said:

That said I think its worth prioritizing and is practically necessary in the same way a steering wheel is necessary in a car. Could you take the steering wheel off and just twist the steering column directly with your hand? Yeah, but it'd be extremely inconvenient and I'd be riding a bike more often instead to avoid the hassle.

 

Why do you think it's as necessary as a steering wheel in a car? From where I'm standing it's about as necessary as a Level 2 autonomous driving system is in a car.

[snip]

Edited March 5, 2021 by Guest
Redacted by moderator

[Guest]

9 minutes ago, mcwaffles2003 said:

Define flying. I'm assuming you mean every flight must have some manual control during its course, if thats the case, I disagree and will explain further if you confirm that is what you mean.

You should have to learn how to orbit, rendezvous and dock before having access to automation of such maneuvers (this does not include Sandbox, which is an entirely different can of worms).

That's were I draw my personal line.

 

13 minutes ago, mcwaffles2003 said:

42 minutes ago, Brikoleur said:

I think that's more of a reflection on just how bad KSP1 is at teaching players the basics of orbital mechanics than the utility of an autopilot at doing it. FWIW I had immense trouble learning to RV and dock too.

Is it that hard to admit that an autopilot feature can have any intrinsic value?

I wouldn't consider "it's a better tutorial than the existing KSP ones" an intrinsic value. I want tutorials as tutorials, not being forced to automate things the game doesn't explain me how to do only to learn them by mimicking the automation.

The whole point of KSP is that trial and error is a better learning experience than watching someone else doing something in your place.

[mcwaffles2003]

3 minutes ago, Brikoleur said:

Why do you think it's as necessary as a steering wheel in a car? From where I'm standing it's about as necessary as a Level 2 autonomous driving system is in a car.

Fair enough, but with 1.35 million people dying in auto accidents every year that's looking just as necessary as well.

5 minutes ago, Brikoleur said:

Yes, that's correct, except the milk runs fully automated away with the supply line automation mechanic. 

Yeah, thats how I envision the milk runs being automated as well.

I don't see why late game it cant be possible to autopilot whole flights in regions that have already been piloted numerous times. I have used mechjeb plenty, I know it has an automated landing feature and I'm pretty sure it has an automated launch feature as well. I have used neither, but them being in the game seems 100% fine to me and I see no reason not to include them. In the late game I believe mission planning should become a more central role than piloting, though piloting manually should always be available to those who care to do so. If that is not the case I cant see how game progress isn't an even worse crawl than in KSP 1.

1 minute ago, Master39 said:

I wouldn't consider "it's a better tutorial than the existing KSP ones" an intrinsic value. I want tutorials as tutorials, not being forced to automate things the game doesn't explain me how to do only to learn them by mimicking the automation.

The whole point of KSP is that trial and error is a better learning experience than watching someone else doing something in your place.

It's still trial and error in setting up maneuver nodes and if anything that's where most of the thinking happens, excluding hovering/landing manual flight is just following the nodes you already made.

[Guest]

7 minutes ago, mcwaffles2003 said:

Fair enough, but with 1.35 million people dying in auto accidents every year that's looking just as necessary as well.

Somebody nicked the steering wheel from my car a couple of years ago. I had to call a tow truck. My current car has L1 automation which is very nice but I wouldn't consider it even close to equally necessary.

7 minutes ago, mcwaffles2003 said:

In the late game I believe mission planning should become a more central role than piloting, though piloting manually should always be available to those who care to do so. If that is not the case I cant see how game progress isn't an even worse crawl than in KSP 1.

If you're able to automate away routine supply missions, exactly what are the missions that would slow game progress to a crawl if you're not able to MJ them to the point that no manual control is needed?

Edited March 4, 2021 by Guest

[mcwaffles2003]

1 minute ago, Brikoleur said:

Somebody nicked the steering wheel from my car a couple of years ago. I had to call a tow truck. My current car has L1 automation which is very nice but I wouldn't consider it even close to as necessary.

You say that now, in 80 years people will be gawking asking "you DROVE those? How did people not get killed?"

To which the honest reply will be "Oh, lots of people died, but it was normal and we had to work with what was available.

Just like kids today astounded that glove boxes used to be filled with paper maps and them asking "how didn't you get lost" and replying... "we did get lost."

5 minutes ago, Brikoleur said:

If you're able to automate away routine supply missions, exactly what are the missions that would slow game progress to a crawl if you're not able to MJ them to the point that no manual control is needed?

All of the inclination changes, Hohmann transfers, apo/peri gee changes... With mechjeb you don't have to worry about time warping past your maneuver node and nodes get executed in very quick succession. Makes missions so much quicker and way less frustrating. But, you still have to plan and plot out a course, design the ship (thats at least 50% of the game time), and correct for any oversights in your plan on the fly. And if you make a planning mistake you don't always have time to plot nodes or run maneuver planner so you have to fly manually anyway.

Mechjeb has made my experience with KSP much more enjoyable, far less frustrating, and no less rewarding. Not to mention.... I've still learned orbital mechanics.

[Guest]

2 minutes ago, mcwaffles2003 said:

All of the inclination changes, Hohmann transfers, apo/peri gee changes... With mechjeb you don't have to worry about time warping past your maneuver node and nodes get executed in very quick succession. Makes missions so much quicker and way less frustrating. But, you still have to plan and plot out a course, design the ship (thats at least 50% of the game time), and correct for any oversights in your plan on the fly. And if you make a planning mistake you don't always have time to plot nodes or run maneuver planner so you have to fly manually anyway.

That sounds perfectly fine by me. I would also like an atmospheric autopilot that holds straight and level or flies to waypoints.

My concern is that MJ as it currently exists is so good that it does make flying completely optional, and I think that's a bad idea, because IME games that effectively play themselves are boring, just like games that force you to do repeated busywork are boring. An autopilot that just executes manoeuvre nodes is fine by me. However an autopilot that you can tell to fly to 100 km orbit, or land at KSC, or transfer to Duna without human intervention is a bridge too far. 

I think the furthest I would be willing to go with autopilot is a system that simulates ILS, with some of its limitations. You would need to first build the infrastructure for it and the craft would need to support it as well. You wouldn't be able to auto-land or auto-dock just anywhere, but if both your base/station and your craft were suitably equipped, they could handle it when you initiated the landing sequence.

[Guest]

17 minutes ago, mcwaffles2003 said:

Mechjeb has made my experience with KSP much more enjoyable,

I'm starting to think that MJ is too big of a mod to be a meaningful example.

Automating running maneuvers? I don't think pressing 3 buttons (warp to maneuver, full throttle, cut throttle) in the right order is a skill or some magic engaging gameplay, and since we'll have background maneuvering it's probably a given, at least for anything interplanetary.

But MJ is much more than that, automating landings, gravity turns, rendezvous, docking and plotting it's an entirely different kind of automation.

And plane autopilot features are another one altogether.

[mcwaffles2003]

1 minute ago, Brikoleur said:

That sounds perfectly fine by me. I would also like an atmospheric autopilot that holds straight and level or flies to waypoints.

 This is why I use the aircraft autopilot mod, without it playing RSS/R0 would not be as fun.

3 minutes ago, Brikoleur said:

My concern is that MJ as it currently exists is so good that it does make flying completely optional, and I think that's a bad idea, because IME games that effectively play themselves are boring, just like games that force you to do repeated busywork are boring. An autopilot that just executes manoeuvre nodes is fine by me.

Cool, how do you feel about including the mission planner feature  it has? Or is that the "bridge to far" you are referring to?:

5 minutes ago, Brikoleur said:

However an autopilot that you can tell to fly to 100 km orbit, or land at KSC, or transfer to Duna without human intervention is a bridge too far. 

 

7 minutes ago, Brikoleur said:

I think the furthest I would be willing to go with autopilot is a system that simulates ILS, with some of its limitations. You would need to first build the infrastructure for it and the craft would need to support it as well. You wouldn't be able to auto-land or auto-dock just anywhere, but if both your base/station and your craft were suitably equipped, they could handle it when you initiated the landing sequence.

Is autolanding on flat new terrain too much? Personally I would be ok with that as an advanced feature that came way down the line. Perhaps to do such a thing you would have to include a laser that measures your ground distance or something... and if youve accidentally picked a hazardous landing spot because of poor oversight then the autoland feature should continue as id the terrain were clear and you wreck your nice ship....  unless you take over and land it yourself at that point.

[Guest]

13 minutes ago, mcwaffles2003 said:

Cool, how do you feel about including the mission planner feature  it has? Or is that the "bridge to far" you are referring to?:

Definitely a bridge too far.

More generally: I'm all for features that provide you with better and more accurate information (although some of that could be locked behind tech tree nodes), but I'm much more skeptical about features that "do the work for you." So for example I'm all for a stock suicide burn countdown, stock transfer window planner, more accurate burn planner that takes into account burn duration, better aerodynamics planning tools, and so on and so forth. 

So: better tools with more data and information for planning missions, I'm all for it. Something that actually spits out a flight plan for you or flies you to or from orbit, thumbs down. 

13 minutes ago, mcwaffles2003 said:

Is autolanding on flat new terrain too much?

Yeah, it is a bridge too far for me. If autolanding is in at all, it ought to be restricted to bases that you've built and equipped for the purpose, and you ought to be able to do that only for relatively mature bases. Getting to that capability would be a nice carrot for you to chase when developing bases.

Edited March 4, 2021 by Guest

[mcwaffles2003]

3 minutes ago, Master39 said:

I'm starting to think that MJ is too big of a mod to be a meaningful example.

Automating running maneuvers? I don't think pressing 3 buttons (warp to maneuver, full throttle, cut throttle) in the right order is a skill or some magic engaging gameplay, and since we'll have background maneuvering it's probably a given, at least for anything interplanetary.

But MJ is much more than that, automating landings, gravity turns, rendezvous, docking and plotting it's an entirely different kind of automation.

And plane autopilot features are another one altogether.

Perhaps then we should discuss what features would be reasonable to take from mechjeb as stock? 

My MAIN concern is automated node execution,  the transfer window planner, and maneuver node planner. Without those 3 features I feel missions would be incredibly tiresome and setting up things like perfect geosynchronous orbits to be just agitating. I personally don't mind mechjeb having all the other features, I just simply choose not to use them and am happy they're there for those who would like to.

Honestly though, I'd like to see what some people could pull off with high powered automation systems... Just in career make it so you have to earn your way up to things like that.

 

Maybe this all comes down to individual philosophies but what I'm hoping to see in a career is a shift in priorities as a game progresses from learning to fly > mastering flight > learning mission planning > mastering mission planning > setting up infrastructure > expanding that infrastructure. By the end of the career I believe there should be little focus on flight and much more focus on management and exploration. Then when you're tired of doing all that you can hop into a special ship you've made with all the bells and whistles and cruise wherever you want

4 minutes ago, Brikoleur said:

Definitely a bridge too far.

More generally: I'm all for features that provide you with better and more accurate information (although some of that could be locked behind tech tree nodes), but I'm much more skeptical about features that "do the work for you." So for example I'm all for a stock suicide burn countdown, stock transfer window planner, more accurate burn planner that takes into account burn duration, better aerodynamics planning tools, and so on and so forth. 

See, I feel like I need that feature just so I can specify what I'm trying to do. I want the ability to type in:

• Set inclination to 0 deg
• Set new apoapsis to 2863.33km
• Set periapsis to 1654.50km
• At apoapsis stage
• Set new craft (the staged one) to burn to circularization
• Repeat 4 times at apoapsis

Now could I do all that manually with the stock maneuver node editor? Mostly. Would it be infuriating? Yes... Would my satellites be slightly out of sync and my geostationary sat network turn into a mess within the next 4 months? Definitely. 

So what, in your view, have I lost in doing this? Why is playing the game this way wrong?

[Guest]

21 minutes ago, mcwaffles2003 said:

Perhaps then we should discuss what features would be reasonable to take from mechjeb as stock? 

I also think that would be a much more productive discussion to have.

21 minutes ago, mcwaffles2003 said:

My MAIN concern is automated node execution,  the transfer window planner, and maneuver node planner.

I have no objections to automated node execution and the transfer window planner. The manoeuvre node planner I would leave out. It gets too close to the game playing itself for my comfort. 

21 minutes ago, mcwaffles2003 said:

Maybe this all comes down to individual philosophies but what I'm hoping to see in a career is a shift in priorities as a game progresses from learning to fly > mastering flight > learning mission planning > mastering mission planning > setting up infrastructure > expanding that infrastructure. By the end of the career I believe there should be little focus on flight and much more focus on management and exploration. Then when you're tired of doing all that you can hop into a special ship you've made with all the bells and whistles and cruise wherever you want

I'm sure individual preferences and philosophies are pretty central here, yes, although I'm pretty much in agreement with you about the progression of priorities. I think where we may differ is that I think that it would be a mistake to add features that make the first two parts entirely optional, even in the late game. Specifically, I think it would be a mistake to let you automate away the flying part of exploration missions.

I also think that if you mostly play RO/RSS then your views will be coloured by that. I would not want to play with those constraints and no MJ, the margins are just so thin that manually piloting craft to or from orbit is too fraught. However, with stock KSP there's plenty of room for error, which makes manual flight enjoyable and sufficiently forgiving. 

Why?

One, flying spacecraft really is central to KSP, and as I've said several times above, I object to game systems that invalidate central areas of gameplay on general principles. 

Two, the devs have said that the four pillars of KSP2 are building cool rockets, exploring new planets, realistic spaceflight, and setting your own goals. Base and infrastructure development and the other systems serve these four pillars. I don't think KSP2 will be – or ought to be – as deep in its infrastructure development stuff as a full-fledged city builder, 4x, or other game where that's a central gameplay pillar. I think that if late-game automation made flying optional, what's left would be thin gruel – there wouldn't be sufficient depth and complexity there to make for a particularly engaging game.

Edited March 4, 2021 by Guest