A few essential things

[KerikBalm]

3 hours ago, General Apocalypse said:

Not really the technical requirements on the game engine are far bigger than any moder or mod team can handle. You need official support.

You can't , actually FH uses at one point 4 simultaneous instances. You simply can't replicate that. Given that KSP 2 wants to go interstellar they better learn how to walk first.

As for MechJeb it is essential . Rocket missions always use a computer guided preplanned trajectory , there isn't some guy mashing a controls in real time while the rocket engines roar. I was honestly shocked the first time I got into KSP and saw that there was no array of flight planning tools. Even as a kid flying model rockets I had a flight plan , unmoded KSP just gives you your joystick and that's it.
 

Ok, not the falcon heavy, specifically, but you can do a reusable booster is my point. So something more like the starship: 1 booster, 1 orbiter.

And I'd be ok with some level of automation, like I said with maneuver nodes automatically executing, but (having never used mechjeb) I have the impression that mech jeb allows you to just build a craft, and tell mechjeb to take it to destination X and land... so that the entire challenge is just building the craft.

I like having to place and manipulate maneuver nodes... I think that is where KSP really forces you to learn orbital mechanics. If it computes and executes the transfer for you.... that's too much IMO

[fulgur]

MJ allows you to (a) launch to orbit following a perfect gravity turn (b) plot and execute burns, ie rendezvous, trans-[anywhere] intercepts, circularisation (c) landing...

Takes all the fun out of piloting. I do admit that ion manoeuvres are horrible, so I have a 50-line kOS script (including time-to-burn, data printout etc) which took me half an hour to make, copying the tutorial and making a couple of improvements, which does it for me. kOS is about as fun as piloting, because when you become as good as the training manual, it gets boring. Trying to put what you are doing (so I burn retrograde over this continent...) into code (at longitude x, point retrograde and burn for x m/s) is a great challenge.

[shdwlrd]

1 hour ago, fulgur said:

*cue flame war over MJ, again™*

Then why start it again?

Anyway, KSP needs some automation.

MJ is good for newer players to help them figure out what's going on. And for players who don't want to do or don't have the time to do long complicated series of maneuvers. Is MJ perfect, no, but is it helpful, yes.

K-OS is perfectly fine to add too. If you have the drive and time to learn to write, then test, and tweak your scripts. More power to you, go ahead and use it. Some people don't have the programmers mindset, math ability, or patience to use it and will never touch it.

I personally don't have the time, drive, or patience to do learn K-OS or any type of scripting. (I don't enjoy programming.) So I'd rather see MJ or some derivative of it in KSP2.

[Rejected Spawn]

Not fond of just being able to tell a ship to "go win for me" at the click of a button so I'm conflicted here but hey fellas, this would be LEGENDARY:

Have a seat on an orbital shipyard and just enjoy the view as 5 ships come in to dock at the same time. How about also: Several large modules come floating together and turn themselves into a beautiful giant space station while you get to spam the screenshot button from any angle you like... Don't forget: The little colony with a launch platform on top of a mountain, a stream of rovers coming and going to drop off resources from carefully placed extractors around the planet. The Kerbal universe would finally feel alive, you're not the only force in the universe that can make them do things.

Would knock my socks off, like playing multiplayer but without the pains of actually having some other doofus go mess up all my plans with completely different plans. Just realised I'm not wearing socks, hey Star Theory if you put the above in the game I'll make sure to put my socks on while downloading KSP2.

[shdwlrd]

Realism for the sake of realism would destroy the fun of KSP. Adding something that was sorely missing from KSP1 but wouldn't impact gameplay much would be welcomed. (Persistent rotation is one thing that comes to mind. Scansat is another.) 

Weather, I'm on the fence about that. I would love to see clouds and such, but I don't like the not being able to see aspect of it nor it having major effects on your craft. 

Something like tweakscale would be welcomed, just to declutter the parts list some.

[Vinhero100]

5 minutes ago, Rejected Spawn said:

Not fond of just being able to tell a ship to "go win for me" at the click of a button so I'm conflicted here but hey fellas, this would be LEGENDARY:

Have a seat on an orbital shipyard and just enjoy the view as 5 ships come in to dock at the same time. How about also: Several large modules come floating together and turn themselves into a beautiful giant space station while you get to spam the screenshot button from any angle you like... Don't forget: The little colony with a launch platform on top of a mountain, a stream of rovers coming and going to drop off resources from carefully placed extractors around the planet. The Kerbal universe would finally feel alive, you're not the only force in the universe that can make them do things.

Would knock my socks off, like playing multiplayer but without the pains of actually having some other doofus go mess up all my plans with completely different plans. Just realised I'm not wearing socks, hey Star Theory if you put the above in the game I'll make sure to put my socks on while downloading KSP2.

I'm gonna be honest with you, that will probably NOT be in the base game (for a while, even if they do decide they want to implement it). That's hours upon hours of coding and debugging that they can't really afford as of now. They're just trying to make a game that won't disappoint, but I'd be surprised if it goes any farther than that on launch day. (I'm being really pessimistic because I don't want to feel betrayed on launch day, but it's kinda the truth, this isn't going to be in the base game when KSP 2 launches no matter how much people want it)

[DStaal]

1 hour ago, fulgur said:

MJ allows you to (a) launch to orbit following a perfect gravity turn (b) plot and execute burns, ie rendezvous, trans-[anywhere] intercepts, circularisation (c) landing...

Takes all the fun out of piloting. I do admit that ion manoeuvres are horrible, so I have a 50-line kOS script (including time-to-burn, data printout etc) which took me half an hour to make, copying the tutorial and making a couple of improvements, which does it for me. kOS is about as fun as piloting, because when you become as good as the training manual, it gets boring. Trying to put what you are doing (so I burn retrograde over this continent...) into code (at longitude x, point retrograde and burn for x m/s) is a great challenge.

I think it partly depends on where you find fun in the game.  Is it in piloting ships?  Designing ships?  (Or planes?  Or monstrosities?) Or is it in planning out a space program?

None of those is the 'right' answer - but they each have different needs.  MechJeb is useful for some of them, nearly essential for others, and likely harmful for others.  MechJeb may replace your flying, teach you to fly, or get in the way of your flying, depending on your use.

If they're doing some better tutorials, I think some form of graduated MechJeb where there's some hand-holding on flying that can be disabled as you get better is probably a good idea.

[fulgur]

In the tutorials perhaps, but not in the real game.

MechJeb completely replaces piloting. It also doesn't work very well for many craft. It prevents you from learning essential skills, and also it's just plain boring - imagine something which built a craft to-spec for whatever mission you wanted to fly...

Plus, I believe that MJ is cheating and [censor bars] [more censor bars] [help, we're running out of censor bars]! Anyway, they should be working on Linux/Mac support, not an autopilot. Or I won't buy the game.

[Rover 6428]

I believe there should be some kind of MechJeb, but not a full one. For example, it can get quite tideous to keep manually doing tons of resupply missions to build that one interstellar spaceship. It would be nice to do one mission, kind off "record" it and then just put it on autopilot so it runs in the background while you do more interesting stuff.

[shdwlrd]

2 hours ago, fulgur said:

In the tutorials perhaps, but not in the real game.

MechJeb completely replaces piloting. It also doesn't work very well for many craft. It prevents you from learning essential skills, and also it's just plain boring - imagine something which built a craft to-spec for whatever mission you wanted to fly...

Plus, I believe that MJ is cheating and [censor bars] [more censor bars] [help, we're running out of censor bars]! Anyway, they should be working on Linux/Mac support, not an autopilot. Or I won't buy the game.

cool, you like flying and doing maneuvers manually. Now your opinion is known. But using an autopilot and having automated maneuvers is very common today for standard aviation and spaceflight. So why can't it be used or included in KSP? We might not learn a skill? What if someone can't figure out how to do that skill? What if someone doesn't have time to figure out or perform the skill they learned? What if someone doesn't have the ability to do a skill you think is required to play KSP? Tell them they can't play KSP? 

KSP wouldn't have lasted this long or reached as wide of an audience that it has if the helper tools weren't in the game or available as mods.

Autopilots are a tool, nothing else. If someone prefers to rely a tool for most things in KSP, so be it. Don't say using a tool is cheating because you feel differently. We all play KSP in our own way.

[Rejected Spawn]

Here's another thought... well, not just one actually.

(1) How about having autopilots tied to difficulty settings to some degree? Could look like this maybe; Easy - Autopilots in everything, can get you anywhere. Medium - Autopilots can go where you've boldly gone before. Hard - Autopilot probes off but read point 2 below...

(2) What form should autopilots really have in the game? How about having the kerbals themselves be the pilots? Rather than the rank system in KSP1 it would be neat if they could learn actual skills from experiencing flights firsthand. Let's say we'd like a kerbal to learn how to get to a stable orbit from KSC, to gain this skill the kerbal would need to be in a pilot seat and experience the flight firsthand a small number of times. To train multiple kerbals at once they all have to fit in some manner of "pilot grade seat" (anywhere with a bunch of buttons and displays and stuff) regardless of where it happens to be located in the craft. To learn docking they need to dock a few times, to learn landing they need to land a few times. You get the picture.

(3) Autopilot probes could need supporting "supercomputers" at KSC and colonies, or be a really heavy contraption to bring. You need a very short distance between a probe and what's controlling it, otherwise the lag can easily ruin the day. Going off to explore unknown areas would have to be done by the player if no "supercomputer" is nearby (or built into the ship) so there is still the element of having to do some flying yourself but if you invest enough you can set up "zones" where any ship can be automated. Obviously in real life the actual computer needed to autopilot things isn't all that big or heavy, however this is made by kerbals with their iconic severely skewed skillsets of being able to build some amazingly complex gadgets while being too dumb to even use them properly.

Small note here; I don't expect any of what I've said in this thread to be in the game at launch, I'm only hoping that Star Theory absorbs all the information and plans accordingly so that any core functions in KSP2 won't get in the way of implementing stuff like this at a later date. "Victory loves preparation."

[mcwaffles2003]

I dont get why having an autopilot feature is considered "cheating". It's almost like people think people at NASA manually fly all the space probes they launch. This game is about to get a lot more expansive, and personally, I would rather not be bogged down by being forced to manually run EVERY. SINGLE. MANEUVER. of a program where I hope to have 100's of probes running simultaneously. Forced manual flight is holding the game back from being kerbal space "program" and instead makes it kerbal space "pilot"

If people wanna manually fly craft then fly them, whats stopping anyone from turning an autopilot feature off? But lets be real, theres a difference between enjoying flying a cessna or an F-22 and running Southwest or Delta and if we''re restricted to manually flying everything then no one is going to make it to that level. I really dont want to manually fly every supply transport ship in a network of 10s-100s of colonies strewn amongst possibly 100's of planets over 10's of star systems

Edited February 6, 2020 by mcwaffles2003

[Guest]

None of these are essential in any sense of the word. This is a plain-vanilla "I want KSP2 to be EXACTLY like my favourite mod collection, only with more polish" opinion.

(FWIW I would only love to see one of the items in the list, and am somewhat neutral about one or two others. The other 7-8 I would not want to see in the base KSP 2.0 game.)

[DunaManiac]

3 hours ago, mcwaffles2003 said:

I dont get why having an autopilot feature is considered "cheating". It's almost like people think people at NASA manually fly all the space probes they launch. This game is about to get a lot more expansive, and personally, I would rather not be bogged down by being forced to manually run EVERY. SINGLE. MANEUVER. of a program where I hope to have 100's of probes running simultaneously. Forced manual flight is holding the game back from being kerbal space "program" and instead makes it kerbal space "pilot"

If people wanna manually fly craft then fly them, whats stopping anyone from turning an autopilot feature off? But lets be real, theres a difference between enjoying flying a cessna or an F-22 and running Southwest or Delta and if we''re restricted to manually flying everything then no one is going to make it to that level. I really dont want to manually fly every supply transport ship in a network of 10s-100s of colonies strewn amongst possibly 100's of planets over 10's of star systems

Just because NASA does it in real life, or something like that doesn't mean that is beneficial for game play at all. Even if you do like MJ does not necessarily mean it should be in 2.0. Also, I agree with @fulgur that kOS would be better in KSP and a more real life analogue than MJ.

Edited February 6, 2020 by DunaManiac

[mcwaffles2003]

1 hour ago, DunaManiac said:

Just because NASA does it in real life, or something like that doesn't mean that is beneficial for game play at all. Even if you do like MJ does not necessarily mean it should be in 2.0.

Im not requesting this feature out of a desire for realism necessarily, but instead out of seeming inevitable necessity. Currently, a campaign is limited by how many ships a person can manage simultaneously. With the addition of interstellar travel, colonies, and orbital & extraterrestrial launch platforms it sounds like it will be a herculean effort just to keep up with progress over time. So, IMO, having some type of mission planner/autopilot seems almost necessary, why do I need to personally complete every single maneuver when I have like 20 different ships distributing materials among my colonies, like 4 or 5 probe craft exploring new planets (each in different solar systems), and a manned mission setting up a colony simultaneously. Having 100% focus on flying craft takes away from ones ability to truly develop a space program with constant incessant  micromanaging. So is it really that outlandish to ask for SOME level of automation?

In no way can I see this subtracting from gameplay. Everyone can still pilot ships as much as they want, it doesn't interfere with building ships or anything else in the game, meanwhile not having it does interfere with anyone's desire to actually make a sizable expansion to their overall space program. We're all stuck playing the sims while playing sim city and it'd be nice if we could have at least a little bit of help focusing on the city aspect.

Edited February 6, 2020 by mcwaffles2003

[dave1904]

On 2/3/2020 at 10:34 PM, General Apocalypse said:

If possible mods like TweakScale should be baked in the base design of the game. This would lower the number of parts and make things more interesting.

I disagree with this so much. I use tweak scale with great caution. It can be abused. You cannot simply scale rocket engines up and down in real life. Actually you cannot simply scale anything in real life. It cannot be more realistic and have tweakscale like the mod works. Maybe structural parts only. Procedural parts would be a better idea. Tanks, structural etc.... Would reduce parts by alot. The mod as we have it now is not as detailed as other tanks so I only use it when desperate too. 

 

[mcwaffles2003]

9 minutes ago, dave1904 said:

I disagree with this so much. I use tweak scale with great caution. It can be abused. You cannot simply scale rocket engines up and down in real life. Actually you cannot simply scale anything in real life. It cannot be more realistic and have tweakscale like the mod works. Maybe structural parts only. Procedural parts would be a better idea. Tanks, structural etc.... Would reduce parts by alot. The mod as we have it now is not as detailed as other tanks so I only use it when desperate too. 

 

I'm in a similar boat, there's something dirty feeling about shrinking a NERVA to fit a 0.625 m probe. I wouldn't mind procedural parts, just feel like they might hurt early careers lego format, so maybe if it was a midgame tech that would be awesome. Also, procedural parts gives better control over length vs width type stuff where tweakscale doesn't, meaning if you want to make a tank longer you have to make it wider also....

Overall though, structural parts really need an overhaul, this I flat out agree on

[shdwlrd]

48 minutes ago, dave1904 said:

I disagree with this so much. I use tweak scale with great caution. It can be abused. You cannot simply scale rocket engines up and down in real life. Actually you cannot simply scale anything in real life. It cannot be more realistic and have tweakscale like the mod works. Maybe structural parts only. Procedural parts would be a better idea. Tanks, structural etc.... Would reduce parts by alot. The mod as we have it now is not as detailed as other tanks so I only use it when desperate too. 

 

You're right, tweakscale can be abused. But if you only limit the types and sizes parts you can use use it with, I don't see any real downside. Plus it would clean up the parts list. For example, instead of one tank for each size in the parts list, you can have one part, but change the size of it. Helpful.

Now it shouldn't be used for command pods or engines. (Unless you're making a replica of something.)

Also tweakscale would be easier to implement than a procedural part system. Tweakscale is just changing the absolute scale of the model used plus some live tweaks to the part configuration. A procedural part has to be set up to be pushed or pulled in any dimension to change its shape. Then a program written to allow you to do that and calculate the specs for it.

Edited February 6, 2020 by shdwlrd
Clarification

[dave1904]

2 hours ago, shdwlrd said:

You're right, tweakscale can be abused. But if you only limit the types and sizes parts you can use use it with, I don't see any real downside. Plus it would clean up the parts list. For example, instead of one tank for each size in the parts list, you can have one part, but change the size of it. Helpful.

Now it shouldn't be used for command pods or engines. (Unless you're making a replica of something.)

Also tweakscale would be easier to implement than a procedural part system. Tweakscale is just changing the absolute scale of the model used plus some live tweaks to the part configuration. A procedural part has to be set up to be pushed or pulled in any dimension to change its shape. Then a program written to allow you to do that and calculate the specs for it.

Actually I have another idea. Its already in the game. There are some parts that can have different model variants. Apply that to tanks. Just choose the 1 tank and switch width, length and colour. Tweakscale does absolute size like you said and that can be unpractical. Procedural parts stretches the model and that looks bad and can limit the external features. It would also get rid of the stupid way KSP deals with tanks. I hate the tank stacking so much. Its unreal and unpractical especially if you have LH2 mods.

 

 

Edited February 6, 2020 by dave1904

[shdwlrd]

2 hours ago, dave1904 said:

Actually I have another idea. Its already in the game. There are some parts that can have different model variants. Apply that to tanks. Just choose the 1 tank and switch width, length and colour. Tweakscale does absolute size like you said and that can be unpractical. Procedural parts stretches the model and that looks bad and can limit the external features. It would also get rid of the stupid way KSP deals with tanks. I hate the tank stacking so much. Its unreal and unpractical especially if you have LH2 mods.

 

 

That's a good idea. From what has been shown so far, KSP2 would need someway to switch the resources in the tanks. Maybe star theory already has something working that is a kin to a part, resource switching. We already know that texture switching is going to be stock.

That is definitely something that star theory can leverage to reduce the visible part counts, but give the players some more flexibility with the size and style of parts to use without going full procedural parts. But the only downside I can see is having to create or modify multiples of models. I can also see just changing only the necessary axes to correspond with the size or height you want.

On the other hand, they can have some really clever solution that would solve the whole tank stacking weirdness when it comes to resources. 

[General Apocalypse]

In regards to MJ and TweakScale -  they are wonderful mods that give you the option to play the game in certain ways. 
Just like the impact of G-forces they can be made into an OPTION. 

I personally use MJ because after a few launches fuel resupply flights become a snooze and because it has a built in planner for interplanetary transfers. Manual corrections are often needed . As for TweakScale it's the best thing since sliced bread because the base game lacks far to many options. The fight over "IRL you can't make things bigger" is quite embarrassing TBH . Imagine the progress in aerospace engineering and materials science we would have if we were only able to do things that in KSP are basic like landing people on another planet.

[Guest]

6 hours ago, General Apocalypse said:

In regards to MJ and TweakScale -  they are wonderful mods that give you the option to play the game in certain ways. 
Just like the impact of G-forces they can be made into an OPTION. 

Software developer here.

Optional features S U C K. They're just as expensive to implement as mandatory features, and they double the testing load: you have to test the system with the optional features enabled, and with them disabled. Because they're optional, they also tend to get deprioritised, which means they tend to break a lot.

With a game like KSP which has a relatively small development team but massive moddability and a vibrant modder scene, it makes much more sense to keep the core game as lean as possible, with modding interfaces that are as stable and as well-documented as possible. 

I would rather see a KSP 2 with no optional features at all, but the features that are present working really well, with high efficiency, high scalability, and stable interfaces. If something is important enough to implement, it should work the same way for everybody. If someone doesn't like it, let them mod it out, or in, or changed.

[Rejected Spawn]

1 hour ago, Brikoleur said:

Software developer here.

Optional features S U C K. They're just as expensive to implement as mandatory features, and they double the testing load: you have to test the system with the optional features enabled, and with them disabled. Because they're optional, they also tend to get deprioritised, which means they tend to break a lot.

With a game like KSP which has a relatively small development team but massive moddability and a vibrant modder scene, it makes much more sense to keep the core game as lean as possible, with modding interfaces that are as stable and as well-documented as possible. 

I would rather see a KSP 2 with no optional features at all, but the features that are present working really well, with high efficiency, high scalability, and stable interfaces. If something is important enough to implement, it should work the same way for everybody. If someone doesn't like it, let them mod it out, or in, or changed.

Self proclaimed generally knowledgeable and logically adept guy here.

There is little wrong with your argument if seen from a Generic Game Development perspective but you're applying the wrong logic to the wrong game and in the wrong way. "There's a mod for that" is essentially your closing argument to excuse not putting options in a game but that's not how it works. Steam Workshop functionality and other integrated systems that let players browse and use mods with no knowledge of how installation works have in recent years made the mod scene more appealing for the general audience that only cares about getting to play a game in a way they like, however if you reach above a certain percentage of the playerbase using the same mod then you can conclude the base game itself has a very real problem with its design. If we assume a mod to have a substantial number of players deeply against it I dare say it's safe to conclude that putting it in the game as a mandatory feature is controversial and unwise since riled up players tend to voice their opinions at least as much as overjoyed players - both groups greatly influence those who are still unsure if they want to buy a game. The Number One Long Term Sales Rule: try to upset as few players as you possibly can at every single moment. Rather than try to make a game full of dream content one should first and foremost strive to make a game that "doesn't have anything that sucks" and after that has been taken care of go for the bonus round of wonderful extra stuff - this part seems to align with your viewpoint as well. Our take on what's essential to have on launch day seems to differ greatly though.

KSP2 won't be about looking in the rearview mirror at underfunded NASA mission replays, it will be about the near future and embody wild ambition. Taking such spirit into account it's very difficult to justify not having some degree of craft automation already in the game on launch day, however there will be those who can't restrain themselves from abusing it and think it makes the game too easy - these are the prime candidates for playing on a higher difficulty that will let them challenge themselves fully under the restrictions of the official game content and never need to go download a mod, it sets a standard they can use to measure their success and compare with others if they choose to. One should also mention that some players just Don't want to mod their game and don't need to justify their choice, there is no shortage of players who will never mod a game unless it is a complete disappointment and can't be enjoyed at all without being altered.

At the end of the day there are only two factors deciding what goes in a game and what doesn't. #1: What the devs want to include, people produce better results when they feel motivated and essentially "not miserable". #2: What's profitable. Even an optional feature that only a subset of players will enjoy can cover its own development cost many times over, once again the key is that people on both far ends of the spectrum of love and hate for something are far more likely to talk about it and be the ones that potential new players will see. Doing nearly anything that can be done to pull a few people out of the hate end and put some others in the love end is essential in the long run. Even if many optional bits are not in KSP2 on launch day it would be absolutely foolish to not have the game built to be painlessly upgraded with as many of those as possible in mind. If nobody hates a feature then make it standard as soon as possible - if some love it and some hate it then try to make it optional as long as it fits the vision the devs have for the game, it's as simple as that.

[Guest]

1 hour ago, Rejected Spawn said:

if you reach above a certain percentage of the playerbase using the same mod

Which percentage, which mod?

[DStaal]

I expect some form of partial autopilot will likely be included, as it would help one of the stated goals of making the game more accessible to newcomers.  How extensive and what modes of the game it would be in are of course open questions - it could just be in some limited tutorial mode.  If I were designing it I'd have the ability to execute the next node, and maybe a helper for gravity turns.  Those two would touch most of the 'frustrating to learn to pilot' issues, and leave plenty of skill for people to learn about transfer planning, etc.

As for tweakscale/procedural parts: I think having some limited form of part scaling - that can apply to specific parts only, within limitations that would depend on the part - would likely reduce the overall load on the developers.  (Though it would move the load from the modeling team to the programming team.)  I also think something like it built-in would be something that could enhance modding for the game.  Personally, I'd limit it to structural parts and tanks, at least in stock.  This would mean fewer parts in stock, and fewer total parts in many ships, easing some of the physics issues.