Where to go from here - What went wrong and what does the community actually want from a future KSP sequel?

[Sappire262]

Hello guys, gals, and nonbinary pals! The purpose of this thread is to act as a place where we can brainstorm ideas for how we would want to see an eventual sequel to KSP be developed so that it can avoid the pitfalls the current KSP 2 fell into. That is to say, I want to try identifying what went wrong with KSP 2 development, and much more importantly, figure out a plan of attack for how a hypothetical future attempt to develop a KSP sequel could work. After all, there is an argument to be made that a lot of the issues that occurred with the KSP 2 development were from the people in charge not understanding the community and what it actually wanted, so by getting together as a community and brainstorming all of this now, if someone does try making a KSP sequel in the future they’ll already have this as a good jumping-off point for what they should do to make us happy.

 

 

So, without further ado, to start us off with here are my thoughts:

 

Based on the recent digging by ShadowZone along with trawling acorss the various threads on these fourms, I think it’s fair to say that KSP 2 failed for the following primary reasons, in rough decreasing order of importance;

 

1 - Company upper leadership not understanding how game design works, leading to such problems as a refusal to commit to changing engines or refactoring code despite the scope of the project requiring it

2 - An overcomittance to secrecy causing years of work being wasted on reverse-engineering KSP 1 code rather than just asking the KSP 1 devs, among other associated problems

3 - Improper management resulting in massive efficacy losses (e.g. developers being moved around constantly and not being given time to settle)

4 - An unwillingness to pay developers respectable amounts

5 - A lack of interest in community input resulting in prioritization of the wrong things

 

When designing a KSP sequel, there are basically two options. The first is to use the unity engine, reuse much of the KSP 1 code, and essentially aim to run a short, cheap project that delivers a highly polished version of KSP 1 with a bunch of the best mods being incorporated into the stock game. This is, to put it mildly, a bad idea, given that the KSP 1 engine and code is almost completely incapable of properly handling that, as anyone who has played heavily modded KSP 1 will know. Implementing anything like interstellar, colonies, and especially multiplayer, would require so much work to update the existing code and systems, and work around the limitations of unity, that it would be cheaper and faster to go with option 2.

 

Speaking of which, the second option is to start from scratch in a new engine, and develop a game that is a true successor to KSP 1 rather than a polished version. I think this is the objectively correct course of action to take, however it seems the upper leadership for KSP 2 disagreed. The original plan as I understand it was to do option 1, delivering a better version of KSP one with some mods made stock, using the same engine and code. However, scope creep then very rapidly increased the goals to include interstellar, colonies, and probably the thing that killed it; multiplayer. This would require a new engine and refactored code to work, but the upper leadership was made up of business executives that had never coded a game in their life and who decided to explicitly make that not an option. Thus, the KSP 2 team was forced to develop a game of impossible scale using old, outdated code, and a bad engine, and to make it worse, while they were handed all of the KSP 1 code to work from, they were not allowed to talk to the original KSP 1 devs. That decision alone probably cost them a year or more of development time.

 

 

With all that out of the way, if a development project for KSP 3 was announced tomorrow (or if KSP 2 was restarted from scratch), and I was somehow placed in charge of setting the development goals and pacing, here’s what I’d do.

 

Firstly, start from scratch in a new engine, probably Unreal. As far as I’m aware the only good alternative to this would be building a custom engine, but from how I understand it the development costs and amount of time required would be incredibly prohibitive, so Unreal it is.

 

Secondly, don’t have any secrecy. Have all of the developers talking with the public, be very transparent about what people are working on, and what the roadmap is, and be prepared to listen to community feedback and change accordingly. Release to EA as early as physically possible, and while we would probably have to charge full price for it, make it exceedingly clear (and in legal writing) that if we don’t fulfill our roadmap goals within a certain amount of time we’ll refund all the copies sold regardless of playtime.

 

Also set realistic goals, and prioritize getting code and core system mechanics working first, then optimisation, and only then user experience and graphics. Of course in reality all of these would have to be developed in parallel, but I would want there to be a heavy emphasis on leaving anything that's purely visual/graphical for as late as possible (though obviously you still have to develop your systems so the visual stuff is fully integrated when the time comes and isn’t just duct-taped on top like most KSP 1 graphics mods are).

 

Design the game's code and systems from the start for modability. I’m not sure if we could do something quite as far as what, say, Hades and Hades 2 do, where literally all of the game's code is unencrypted and open for anyone to see and edit, but if that’s possible it would be great (I’m not aware of actual downsides to doing this, it’s just that there’s so much cultural momentum in the industry against it that it would be hard to the leadership to allow it even if it would make the game way better for no extra cost).

 

Pay the developers properly, and don’t put people in management positions who don’t have the experience required to fulfill that role properly. In fact, if possible a worker-led development program, such as one done in a cooperative or other worker-led company, would be ideal. All empirical data collected to date indicates that worker-led companies are multiple times less likely to fail, are more stable, produce higher-quality products, have way better working conditions, and the workers enjoy their jobs orders of magnitudes more. Interestingly, the people in worker-led companies usually vote to reduce wages rather than lay people off in times of hardship, which is usually the right move because losing talent is really bad. The only downside is that the per-unit cost of their products is usually higher because they pay their workers more, so even if they never go out of business, they have trouble climbing up the ranks in the market as it were. Still, for developing a game like this a high amount of workplace democracy would seem to be key.

 

Oh, and don’t implement multiplayer. From what I’ve seen, everyone on here who’s familiar with game design seems to agree that it’s a massive undertaking that would take an insane amount of effort to pull off even when starting from scratch, and it’s not super clear to me if the demand is even remotely high enough to justify that. I would love to hear all of your thoughts on that though.

 

So, with all that being said, here’s a rough roadmap of what I would expect from release onwards. Of course, I’m not a software engineer, so I may be overly optimistic or pessimistic here. Also, keep in mind that while I lay it out in just a couple of major updates, I think in reality it would be better to do this with a much larger number of individually smaller updates, but I don't have the patience to write all of that out, sorry about that. Oh, and I would love to hear all your thoughts on the order I've put all of this in, especially the career mode release, since I'm not certian myself if it's the best.

 

 

Initial full release (1.0)

 

- Recoded from scratch in Unreal engine, utilizing 100% new code, and heavily optimized.

- Graphics on par with KSP1, but using all-new systems and with allowances in place for them to be properly improved in the future without it just being a series of patchworks and band-aid fixes like it is in KSP1.

- All new parts, most redesigned from the ground up to be more internally consistent and fill gaps, and with most stock KSP1 parts represented out of the box.

- Complete UI revamp across the entire game to improve user experience as much as possible, including a redesigned parts window and filters in the VAB so that when thousands of parts are added in the future finding and sorting them doesn't become a pain.

- New resources and resource systems, including new ISRU systems, for a more detailed and realistic experience while still being streamlined and easy to understand.

- Systems in place to allow for easy integration of robotics parts and kerbal-deployable parts in the future, maybe with some limited number of them already implemented.

- Very basic science mode implemented, but no career mode yet.

 

 

First major update (1.1)

 

- Graphics improvements.

- More parts, think stuff from stockalike station expansion, planetside, the near future mods, etc, along with all the KSP1 DLCs and some of the stuff they planned for KSP2 (Orion etc).

- Basic life support system, with 5-6 new resources, as well as simple crew habitable volume requirements, all togglable in options.

- Interiors properly modeled and visible through windows, and going on EVA fully animated with the hatch opening and everything.

- Kerbals now only carry small SAFER-style jetpacks in their inventory, and go on EVA by default using by fully modeled tethers and climb along the outside of vessels using a completely new set of climbing mechanics. MMUs are separate parts like command chairs that they can get in to fly around properly.

 

 

Second major update (1.2)

 

- More graphics improvements (now up to the standard of heavily modded KSP1).

- More parts, mainly focusing on high-tech stuff.

- Other star systems and interstellar gameplay mechanics.

- More structures on Kerbin, including entire cities, along with new launch sites.

 

 

Third major update (1.3)

 

- More graphics improvements.

- More parts.

- Career mode, completely overhauled to use a better progression system that feels more like an actual space agency, with programs instead of contracts, and part unlocking based both on science collection and the programs you’re running.

- Colony system similar to what was planned for KSP2.

- FreeIVA polished and implemented into stock, maybe with VR support.

 

 

Real History DLC

 

- Basically Bluedog Design Bureau, Tantares, SOCK, KNES, and every other stockalike historical parts mod you can think of, but all put on several metric tonnes of steroids.

- Virtually every spacecraft ever conceptualized or designed, let alone actually flown, in the history of the human race, all meticulously modeled to 99% accuracy (more accurate than BDB for example), while maintaining a stockalike style (though they would all be fully in scale to each other, not shifted to conform with KSP's 2.5m, 3.75m, etc diameters).

- Plus all the launch stands and pads to go along with them.

- And a revamped VAB parts list system to allow for you to actually navigate all that.

- Would probably require a significant team of 3d modelers working around the clock for years to complete, hence why it’s a paid DLC.

- Full compatibility with the RO DLC mentioned next if you have it installed.

 

 

Realism Overhaul DLC

 

- Pretty much RSS/RO, but with the polish one would expect if it was properly integrated into the game itself, and even more attention to detail and features, for example:

- Procedural crewed pods and modules, with editable interior layouts and systems.

- Procedural engines on top of a selection of most real ones.

- Procedural tanks, with the internal bulkhead arrangements and all that being customizable.

- Human rather than kerbal astronaut models (togglable if you don’t want that).

- Improved structural simulation system, with internal part stresses properly modeled, so the weight of your tankage and structure now actually depends on what it’s supporting and how many gees it’s expected to endure, etc.

- An aerodynamic simulation system that could make FAR blush, plus a thermodynamic simulation system to match (togglable).

- Principia-level n-body physics (togglable).

- The option to play in either RSS (plus real nearby star systems) or 10x stock.

 

 

From what I can tell KSP 1 got developed to 1.0 in around 4 years, by about 12 people, and KSP 2 went from having all work on it restart to 0.2 in a little over 3 years with a team of 70 people, while also dealing with massive management inefficiencies and trying to work systems into the KSP engine and code that would be faster to just do from scratch.

 

Therefore, I would optimistically estimate that we could probably get from development go-ahead to 1.0 on my roadmap here in about 3 years, assuming a development team of 70 people like KSP 2 and it’s properly managed and well-funded. From how I understand it the fully burdened cost (e.g. the entire cost the company pays, not just their salary alone) of a software development employee for this sort of role is typically about 200k per year, and the KSP 2 team had it capped at 150k by upper leadership that led to severe issues.

 

For some margin of error, let’s say 250k then. That means, for 70 people for 3 years, the total dev cost to 1.0 will be about 52.5 million dollars. To get to 1.3 and then the DLCs is harder to estimate for me, but let’s assume 5 years, so 8 years total. That means the total dev cost from development go-ahead to DLC#2 will be 140 million dollars.

 

KSP 1 has around 100,000 reviews on Steam. On average, games on Steam have 63 sales per review, and though this can be lower, it increases the older a game is so it’s probably actually higher. In any case, going by those figures that means KSP 1 sold 6.3 million copies. The price of it changed a bit over time, but was usually around 40-50 dollars.

 

That means we have an existing playerbase of 6.3 million players to attract, not counting new players. If we price KSP 2 at 50 dollars, and only half of them transition over to KSP 2, that’s 157.5 million dollars. If the two DLCs each sell for 10 dollars and attract a total of 0.5 million players, that’s another 5 million dollars for a total of 162.5 million. That gives us a total profit margin of 22.5 million dollars, which is very tight, but might be just about doable.

 

Keep in mind, that was assuming 70 people and a total dev time of 8 years. Given the DLCs are unlikely to be very profitable relative to the cost of making them, if we delete those and scale down to, say, 3 years to 1.0 and 2 years to 1.3, so 5 in total, and redo all the math, accounting for no sales from the DLCs, we get a total dev cost of 87. 5 million and a profit of 70 million dollars, which is much more workable.

 

 

So, does anyone have any thoughts on these ideas? Am I being completely nonsensical and not understanding how game design works? I would love to hear some input on all that, and also if you have completely different ideas on how to go about redeveloping KSP 2 I would love to hear them as well! There was after all also that recent tweet after all from Jundroo, the Juno New Origins devs, about potentially getting some of the KSP devs onboard and reworking Juno into essentially a KSP sequel, I would love to talk about that but unfortunately I just don’t have the knowledge about Juno to be able to comment, especially given I’ve never played it or followed it’s development, but I'd be happy to hear others discuss it.

 

 

EDIT:

Okay, so, I wrote that post pretty late last night, and I’ve now had the chance to think about some stuff, and I want to expand a bit on my thoughts.

Firstly, timescales. I’ve gone crawling across the forums once again for some more information, and I realized I misremembered some of the stuff relating to KSP 2’s development, specifically they didn’t restart development in 2020 but kept on using the old work they had, but started refactoring it. However, given the way people talk about how this decision cost them time, I think it’s fair to assume they could have restarted from scratch code-wise and still gotten to the point they did, or even further perhaps, in 3 years with their team of 70 people, so my estimates still seem reasonable. Also, apparently the KSP 2 devs were really close to having colonies and interstellar in the game for the initial EA release, and the fact they’re still not out even now was mostly due to them getting tied down fixing things and optimizing and whatnot, along with a lot of mismanagement. So it’s possible we could aim to have those in the base game for the initial release of this hypothetical reboot, though I still think it would be a better idea to build the game’s code ready to accept that, but not actually include it until a future update.

Also, I must have been really tired yesterday because it seems I missed this, but there are actually already a couple of threads on here talking about pretty much exactly this. Sorry about that, if the mods deem that this one is redundant and close it I fully understand.

 

Now, the key question. What do people actually want from a KSP game?

There’s been some talk of a colony-designer game where you start out with something akin to cities skylines or something like that, and then start launching rockets later. I don’t agree with this idea personally, while I do like the idea of something city-builder style for the colonies I think that should be a late-game thing.

However, one thing I am thinking about would be for the KSC upgrade system in career to be replaced with a city-builder style system where you get to actually build the KSC. So you start out with just, I don’t know, a small airfield and tiny pad for sounding rockets, and you get to, using a cities-skylines style system and interface perhaps, redesign and expand it over time until eventually you have an entire space center.

Of course, unless you just want to be placing down upgraded versions of the same 8 buildings over and over, we would have to find new buildings to add and things for them to do. Maybe have it so that placing multiple launch pads could have an actual benefit, such as introducing a system like kerbal construction time where refurbishing pads take time? You could have that be togglable in settings as well -  I know adding togglable stuff like that massively increases development costs but for stuff like construction time etc it doesn't seem like it would be a massive issue compared to some other ideas. And maybe we should make the players put a lot of thought into how they lay out their space center, with the way buildings are connected to each other and the distances involved all being important. For example, make your crawlerways too long, and it takes a while for the rockets to get to the pads. Make them too short, and if a rocket explodes not only does your pad get destroyed but your VAB might as well.

While we’re talking about this, I may as well brainstorm other ideas for career mode as well. Instead of contracts, maybe we could have programs, which each contain several goals and milestones, and give you research benefits. And instead of like it is in KSP 1, we could lay the programs out on a tree. Selecting a program to do could give you research benefits to the parts that would be involved with it, like a lunar landing program reducing the unlock cost of the LM-style lander pod. And to stop the game from becoming linear, we could make it so that you can skip further down the tree and initiate a later program if you want to without doing the ones leading into it, but that would incur penalties like lower rewards or less research benefits. And we could make it so that if you landed on the moon without ever actually selecting the lunar landing program, it would autocomplete it but not give you any rewards for doing so.

I think a good way of doing this would be for each program to have a set end goal, or maybe multiple, and also have several milestones along the way. For example, if you take the lunar landing program, it will give you individual missions for a crewed lunar flyby, orbital mission, and the landing, each with their own rewards and research benefits, on top of the rewards and benefits from the overall program, which could be set so that the more of the individual missions you complete in a program the higher the total bonus reward you get will be from the entire program itself when you complete it. I’d love to hear other people’s thoughts on this though, given I’m pretty much just spitballing here. I’m also not sure if this would completely replace the contract system, or if we should still have some conventional one-off contracts as well to supplement it.

In any case, to round all of this off, I think there should be three main difficulty settings for career mode. In easy mode, the city-builder elements of the KSC are disabled, and it just expands automatically over time or maybe it starts out really big. The core gameplay loop here would be similar to KSP 1. There would also be no life support, no commnet, and no construction time. On regular default difficulty, life support and commnet is enabled, as are the city-builder elements for the KSC, but construction time is still turned off. On hard difficulty, not only is construction time turned on, but life support now takes into account not just the resources you have, but also how much livable space there is, and also radiation becomes a factor. When starting a new game, after picking career mode large buttons would appear for these three difficulties, with description below them fully explaining all of this to avoid confusion, and of course there would still be the option to go into the advanced settings and mix and match all of this.

 

Next, changes outside the space center part of the game. Going over to spacedock and looking at the most popular mods can give us an idea of what the community seems to want most, and so let’s go over that real quick.

Firstly, visual improvements. Given we’re switching to a new engine and rewriting everything from scratch, building these into the base game to a level that even surpasses KSP 1 should be possible. I would prioritize getting the underlying systems needed for the visuals to work done first, and then actually adding all the visual stuff later though.

Next, stockalike parts mods, especially the near future series along with some others. Adding a much larger selection of parts in a stockalike style should be possible, hell it’s probably one of the easiest parts of the entire development process, though if we’re redoing everything from scratch I think we should take the opportunity to completely redo the stock parts as well rather than just copy them. However, after doing some thinking I do believe the stock 1.25m, 2.5m, etc scales should stay, but with additions like 0.9375m, 1.5m, 2.1875m, 3.125m, 4.25m, and 6m, along with maybe even larger scales than that. We would need a new parts sorting system to deal with all that without it getting confusing, but I don’t think it’s anywhere near impossible to come up with one that would still be easy to use and beginner-friendly. I considered upscaling everything by 1.5 to 2 to be more or less actual scale (in terms of the spacecraft), but I think part of the kerbal charm as it were is that they’re slighter smaller than humans are. This decision may come back to bite us later with RO/RSS DLCs and mods though. Oh, and more spaceplane parts sizes would be nice, right now we just really have three. If we’re really clever with how we choose them and model the adapter pieces for them, we could even allow people to make unusually shaped vehicles like an X-33 by stacking various adapter pieces in front of each other.

After that, we have interstellar extended, which again should be easy to implement compared to everything else. There are also a lot of quality-of-life things, better burn time and docking port alignment indicator for example, which again should be relatively easy. Kerbal attachment system, kerbal inventory system, and ground construction as well, we’re already close to that with the braking ground stuff anyway for the former and the latter would come with colonies.

You have to keep going a while before finding any planet mods, even kerbalism seems to come before that, but nonetheless I think an overhaul and expansion to the stock kerbal system, along with more star systems for interstellar, would be nice. Actual asteroid belts, way more asteroids and comets, rings that have actual particles in them, all sound like amazing ideas if they could be properly implemented. Another gas giant as a Saturn analoge, and an ice giant or two, would also be really nice, and completely revamping all of their moons to be more realistic would be awesome. Jupiter has almost a hundred moons and thousands of smaller objects orbiting it, and Saturn has even more. Most of them would be very small, but even then, it would add a lot of interest to those systems I think, and I can’t imagine it would be hard to implement compared to all the other stuff we’re talking about. It would also be nice to see some more interesting features on the planetary bodies themselves, especially ones that require specially designed equipment, and effort and skill, to get to, like deep ravines that require kerbals to bring climbing equipment or winches to get to the bottom of.

Also, stuff like FreeIVA and through the eyes of a kerbal are really cool, and it would be neat if we could get those fully implemented at some point, maybe even with VR support, though for all I know that might be as difficult as multiplayer would be. But in any case, fully modeling the interiors of spacecraft and having them be visible through windows also seems like an awesome idea, and since KSP 2 did it it probably is possible, though we might want it to be togglable for low-end computers. Also, I really like the idea of having kerbals go on EVA properly, by getting in an airlock, depressurizing it, and opening the hatch to step outside, all fully animated. We could have it so that kerbals can’t just exit from any random hatch, thus making having actual airlocks important on larger ships. Some smaller pods, Apollo-style ones for example, obviously could just have the entire thing depressurize as they did IRL. And once on EVA, it would be cool if we could have fully modeled tethers, and redo the climbing system from scratch so the kerbals have to clamber across the sides of the spacecraft until you unlock EVA jetpacks. I’m not sure whether the EVA packs should be modeled like they are in stock, or if they should be large, clunky things that have to be stored externally and entered like an external command seat, like the real ones are, I’d love to hear your thoughts on that. Of course, we could still give the kerbals smaller and less powerful SAFER-like packs that they can carry on them.

 

Alright, that’s about all I can think of for now. I’m sure I’ll come up with more ideas in the future, and I might keep on posting them, but for now I’d love to hear some feedback on all this. I’m trying to figure out what the community wants most from a KSP sequel, not just what I do personally, so it’s important I don’t just ramble on about my own thoughts unchecked.

Edited May 28 by Sappire262
changed text coloration

[Dakitess]

I'm really sorry but I'll go the TLDR route for that one.

What I want from a KSP2 (a real one) ? Only ONE thing : to be technically up to date. I don't care about new content, new parts, new planets, new gameplay. I want my KSP1 to be recreated from scratch PROPERLY, with all the new techs, the new dev method, the new tool, with a pro-team, with a real budget to get the things done right.

It would be KSP1 but beautiful (and I mean it, not what we got with KSP2, uh ? It's so friggin outdated, and aesthetically questionable !...) and running well and that's it, as a base. Then only, it would evolve, being Early Access if they wish to go that way, the old KSP1 fashion; but at the very least it would be KSP1 as it exist nowadays, but developed on a real up to date new fresh basis, ready to embed a whole new decade of contents, mods.

I don't see how it cannot be the priority. Really, I don't. But that's something personal and I might be wrong, it also happens to me xD

Edited May 28 by Dakitess

[HebaruSan]

@Sappire262, could you please clear the text coloring from your post? It's unreadable on a dark theme.

[Fizzlebop Smith]

I would like any an all endeavors that utilize early access.. to enter and engage with that system in the manner and spirit in which it was intended.

Primarily.. communication. 

Monthly 

Technical Reports / blog entries & Dev Diaries

Use of a widely accepted form of official communication (forum)

Should small development studios hold themselves under the shelter of a much larger umbrella should have a degree of autonomy

Quit putting creative directors in charge of development projects. Allow creative director to head the creative department and share those visions with *qualified* engineers.

I understand budget restrictions and salary caps. Why such an imbalance of art producers?

[Scarecrow71]

I've got plenty to say on this topic, so please bear with me as I multi-quote a bunch of stuff.

On 5/26/2024 at 5:30 PM, Sappire262 said:

When designing a KSP sequel, there are basically two options. The first is to use the unity engine, reuse much of the KSP 1 code, and essentially aim to run a short, cheap project that delivers a highly polished version of KSP 1 with a bunch of the best mods being incorporated into the stock game. This is, to put it mildly, a bad idea, given that the KSP 1 engine and code is almost completely incapable of properly handling that, as anyone who has played heavily modded KSP 1 will know. Implementing anything like interstellar, colonies, and especially multiplayer, would require so much work to update the existing code and systems, and work around the limitations of unity, that it would be cheaper and faster to go with option 2.

 

Speaking of which, the second option is to start from scratch in a new engine, and develop a game that is a true successor to KSP 1 rather than a polished version. I think this is the objectively correct course of action to take, however it seems the upper leadership for KSP 2 disagreed. The original plan as I understand it was to do option 1, delivering a better version of KSP one with some mods made stock, using the same engine and code. However, scope creep then very rapidly increased the goals to include interstellar, colonies, and probably the thing that killed it; multiplayer. This would require a new engine and refactored code to work, but the upper leadership was made up of business executives that had never coded a game in their life and who decided to explicitly make that not an option. Thus, the KSP 2 team was forced to develop a game of impossible scale using old, outdated code, and a bad engine, and to make it worse, while they were handed all of the KSP 1 code to work from, they were not allowed to talk to the original KSP 1 devs. That decision alone probably cost them a year or more of development time.

There is a 3rd option here that you are discounting:  Use a new engine, but reuse whatever code from KSP1 that you can.  I highly lean towards this option, knowing that I know nothing about the existing codebase for either game, as well as having very limited/not-even-amateur-level knowledge of either Unity or Unreal (which seems to be the popular pick here in the community to re-do the game with).  Even just having access to the code from the first game can provide a lot of insight, especially as it relates to bugs that the first game solved, as well as providing a roadmap of sorts of what not to do.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Firstly, start from scratch in a new engine, probably Unreal. As far as I’m aware the only good alternative to this would be building a custom engine, but from how I understand it the development costs and amount of time required would be incredibly prohibitive, so Unreal it is.

As I said above, this seems to be the engine of choice for the community.  I have zero idea how much effort it would take to build a custom engine, but with my very limited knowledge I was able to follow a tutorial from someone and build a rotating planet in Unreal 5.4.  And the tutorial was written for Unreal 4.x, so a lot of the concepts, ideas, and techniques from older versions may still be very applicable.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Secondly, don’t have any secrecy. Have all of the developers talking with the public, be very transparent about what people are working on, and what the roadmap is, and be prepared to listen to community feedback and change accordingly. Release to EA as early as physically possible, and while we would probably have to charge full price for it, make it exceedingly clear (and in legal writing) that if we don’t fulfill our roadmap goals within a certain amount of time we’ll refund all the copies sold regardless of playtime.

This is actually supposed to be part of the agreement between the developers and the customers during EA.  Something which Take Two, Private Division, and Intercept Games seemed like they wanted to do at first...but then failed miserably at.  But even with that "agreement" in place, I still would rather have the developers working on the game itself, with Community Managers engaged with the community and acting as a go-between wherever possible.  I think we need more communication, no matter who it comes from, but the developers should only be involved in community engagement when the concepts they are talking about are way over the heads of the CM's, or when we are discussing bugs that need to be fixed.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Design the game's code and systems from the start for modability. I’m not sure if we could do something quite as far as what, say, Hades and Hades 2 do, where literally all of the game's code is unencrypted and open for anyone to see and edit, but if that’s possible it would be great (I’m not aware of actual downsides to doing this, it’s just that there’s so much cultural momentum in the industry against it that it would be hard to the leadership to allow it even if it would make the game way better for no extra cost).

This was supposed to be a thing with KSP2.  We were told by Nate and a lot of other people involved that the modding community was a big part of the reason for KSP1's success, and they wanted to make sure the modders were involved with KSP2.  The big problem is that the game has to be built and be stable before you can create API's that are effective enough to mod the game.  Although, as we have seen - and as I myself even did - the modding community didn't care about having API's and just went about their business.  There are a lot of great mods for KSP2, and I have no doubt that if the company had stuck to their word about the game being mod-friendly we would have even more than we do now.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Pay the developers properly, and don’t put people in management positions who don’t have the experience required to fulfill that role properly. In fact, if possible a worker-led development program, such as one done in a cooperative or other worker-led company, would be ideal. All empirical data collected to date indicates that worker-led companies are multiple times less likely to fail, are more stable, produce higher-quality products, have way better working conditions, and the workers enjoy their jobs orders of magnitudes more. Interestingly, the people in worker-led companies usually vote to reduce wages rather than lay people off in times of hardship, which is usually the right move because losing talent is really bad. The only downside is that the per-unit cost of their products is usually higher because they pay their workers more, so even if they never go out of business, they have trouble climbing up the ranks in the market as it were. Still, for developing a game like this a high amount of workplace democracy would seem to be key.

This goes along with having a proper budget, which I didn't see you mention anything about.  We heard about the average salary for developers in SZ's video, and so the budget becomes a very real, very needed thing.  If you pay the developers what they are worth and/or what the market commands, and give them the right tools for the job, which would include testing rigs, we are talking about $250k per developer.  You want a team of 4?  That's $1 million per year, and we know that the team should be at least 5-8 people.  We estimate high (8 people), so we get $2 million per year for the developers alone.  And that doesn't include office space, CM's, other overhead we aren't thinking of.  That $10 million for 2 years that TT gave Intercept was probably in the right ballpark, but still well short of what they needed.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Oh, and don’t implement multiplayer. From what I’ve seen, everyone on here who’s familiar with game design seems to agree that it’s a massive undertaking that would take an insane amount of effort to pull off even when starting from scratch, and it’s not super clear to me if the demand is even remotely high enough to justify that. I would love to hear all of your thoughts on that though.

Couldn't agree more.  Too many issues with multiplayer for this genre of game.  It works best in FPS games where the object is to kill your opponent (and yes, KSP could very well have that element with air-to-air and air-to-ground combat situations).  But for a space game where the vast majority of things are building and getting to orbit?  Just doesn't work in my humble opinion.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Initial full release (1.0)

 

- Recoded from scratch in Unreal engine, utilizing 100% new code, and heavily optimized.

- Graphics on par with KSP1, but using all-new systems and with allowances in place for them to be properly improved in the future without it just being a series of patchworks and band-aid fixes like it is in KSP1.

- All new parts, most redesigned from the ground up to be more internally consistent and fill gaps, and with most stock KSP1 parts represented out of the box.

- Complete UI revamp across the entire game to improve user experience as much as possible, including a redesigned parts window and filters in the VAB so that when thousands of parts are added in the future finding and sorting them doesn't become a pain.

- New resources and resource systems, including new ISRU systems, for a more detailed and realistic experience while still being streamlined and easy to understand.

- Systems in place to allow for easy integration of robotics parts and kerbal-deployable parts in the future, maybe with some limited number of them already implemented.

- Very basic science mode implemented, but no career mode yet.

Again, I disagree with 100% new code.  If you have access to the original game's codebase, it would be highly detrimental to not utilize what you can.

If we are talking about this being a 1.0 release, I think we need to back up a bit and do this in pieces.  For example, Resources/ISRU could be left out of the initial release into Early Access, with it then being added in at a later point.  Same thing with Science mode - which needs a massive overhaul to begin with as the science points system just doesn't seem to work/fit any longer - and Career mode.

On 5/26/2024 at 5:30 PM, Sappire262 said:

First major update (1.1)

 

- Graphics improvements.

- More parts, think stuff from stockalike station expansion, planetside, the near future mods, etc, along with all the KSP1 DLCs and some of the stuff they planned for KSP2 (Orion etc).

- Basic life support system, with 5-6 new resources, as well as simple crew habitable volume requirements, all togglable in options.

- Interiors properly modeled and visible through windows, and going on EVA fully animated with the hatch opening and everything.

- Kerbals now only carry small SAFER-style jetpacks in their inventory, and go on EVA by default using by fully modeled tethers and climb along the outside of vessels using a completely new set of climbing mechanics. MMUs are separate parts like command chairs that they can get in to fly around properly.

Graphics improvements should be coming out with every patch or bug-fix if possible.  I can see having improvements in a 1.1 release, but let's not forget they should be coming all the time (again, if possible).

I love the idea of more parts, especially if we can get some of the Near Future, Stock-alike, Space Station, and RL US/Russia/Japanese parts that we all want.

I wholly disagree with Life Support.  This is a topic that is a heated one on the forums, with 2 very distinct factions - one for, one against.  I'm not against having it so long as it's an option to turn on or off...but if the vast majority of the community (in my estimation) will keep it off, then why have it?

I love the idea of tethers, with the option to remove them and do EVA maneuvers.  I also would like to see being able to move the tethers from one part to the next instead of having just one long one.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Second major update (1.2)

 

- More graphics improvements (now up to the standard of heavily modded KSP1).

- More parts, mainly focusing on high-tech stuff.

- Other star systems and interstellar gameplay mechanics.

- More structures on Kerbin, including entire cities, along with new launch sites.

Interstellar is a hot topic, and brings up a lot of questions.  Such as "How do we deal with Time Warp under acceleration" and "How close/far should they be" and "Do Kerbals live forever".  I would love to sit in on discussions on this, even if I'm muted and can't ask questions.

I think other launch sites should be included before this update, though.  We have them in KSP1 thanks to DLC; why wait this long in KSP2?

On 5/26/2024 at 5:30 PM, Sappire262 said:

Third major update (1.3)

 

- More graphics improvements.

- More parts.

- Career mode, completely overhauled to use a better progression system that feels more like an actual space agency, with programs instead of contracts, and part unlocking based both on science collection and the programs you’re running.

- Colony system similar to what was planned for KSP2.

- FreeIVA polished and implemented into stock, maybe with VR support.

Why is career mode so far after initial release?  You can see here I'm starting to ask questions on decision-making, which is probably more me wanting to understand where you are coming from.  Career mode could be implemented before this, and should be implemented before Colonies.

I have no comment on IVA as I don't use it.

On 5/26/2024 at 5:30 PM, Sappire262 said:

Real History DLC

 

- Basically Bluedog Design Bureau, Tantares, SOCK, KNES, and every other stockalike historical parts mod you can think of, but all put on several metric tonnes of steroids.

- Virtually every spacecraft ever conceptualized or designed, let alone actually flown, in the history of the human race, all meticulously modeled to 99% accuracy (more accurate than BDB for example), while maintaining a stockalike style (though they would all be fully in scale to each other, not shifted to conform with KSP's 2.5m, 3.75m, etc diameters).

- Plus all the launch stands and pads to go along with them.

- And a revamped VAB parts list system to allow for you to actually navigate all that.

- Would probably require a significant team of 3d modelers working around the clock for years to complete, hence why it’s a paid DLC.

- Full compatibility with the RO DLC mentioned next if you have it installed.

I guess a lot of this depends upon how compatible those models - not the mods themselves, but the actual models in those mods - are with KSP2.  I honestly don't know if they are or not.  I do know that KSP2 kind of uses the same lego-building system that KSP1 used, so I'd really be curious as to how much effort this would take.  Beyond that, my question is "Why is near future part of 1.1 but all the other real-life stuff part of a DLC?  I get Near Future is probably needed for Interstellar, but wouldn't that fall under the same premise you are using here with needing all that effort to recreate the parts for KSP2?

And RO is another one of those things that I'd support if it had the option to be turned off.  There's a reason why RO is a mod and not part of stock - or even part of DLC.  I know people use it, but are there enough people who use it to make it part of a purchasable DLC?  Or are you only throwing it in here because it's part of DLC with a lot of parts?

On 5/26/2024 at 5:30 PM, Sappire262 said:

Realism Overhaul DLC

 

- Pretty much RSS/RO, but with the polish one would expect if it was properly integrated into the game itself, and even more attention to detail and features, for example:

- Procedural crewed pods and modules, with editable interior layouts and systems.

- Procedural engines on top of a selection of most real ones.

- Procedural tanks, with the internal bulkhead arrangements and all that being customizable.

- Human rather than kerbal astronaut models (togglable if you don’t want that).

- Improved structural simulation system, with internal part stresses properly modeled, so the weight of your tankage and structure now actually depends on what it’s supporting and how many gees it’s expected to endure, etc.

- An aerodynamic simulation system that could make FAR blush, plus a thermodynamic simulation system to match (togglable).

- Principia-level n-body physics (togglable).

- The option to play in either RSS (plus real nearby star systems) or 10x stock.

Why have RO in the first DLC package and not this one?

What about JNSQ?  Or KSRSS?  Un Kerballed Start?  Alternate tech trees (depending upon how Science mode is set up)?

Why are you getting rid of Kerbals in favor of Humans?  If I want a human look-alike, I'd go play Juno, or Rocket Science, or one of the other space clones.  Leave Humans out of it here, please.

On 5/26/2024 at 5:30 PM, Sappire262 said:

From what I can tell KSP 1 got developed to 1.0 in around 4 years, by about 12 people, and KSP 2 went from having all work on it restart to 0.2 in a little over 3 years with a team of 70 people, while also dealing with massive management inefficiencies and trying to work systems into the KSP engine and code that would be faster to just do from scratch.

 

Therefore, I would optimistically estimate that we could probably get from development go-ahead to 1.0 on my roadmap here in about 3 years, assuming a development team of 70 people like KSP 2 and it’s properly managed and well-funded. From how I understand it the fully burdened cost (e.g. the entire cost the company pays, not just their salary alone) of a software development employee for this sort of role is typically about 200k per year, and the KSP 2 team had it capped at 150k by upper leadership that led to severe issues.

 

For some margin of error, let’s say 250k then. That means, for 70 people for 3 years, the total dev cost to 1.0 will be about 52.5 million dollars.

3 years of development up to 1.0?  Are you accounting for setbacks, changes to scope, issues with staff coming/going?  What about things like studio buyouts, shifting priorities, changes in office locations?  You brought up 12 people for 4 years to get to 1.0 with KSP, but what you are talking about here is a far larger project than 1.0 in KSP1.  We have to assume that all currently working features in KSP1 are part of KSP2 at 1.0, plus you want to redo all the parts, all the code, have ISRU and resources, and a whole lot more.  Just in 1.0.  Are you sure 3 years is enough time to do all that and have contingencies for things that fail or go wrong or change?

On 5/26/2024 at 5:30 PM, Sappire262 said:

What do people actually want from a KSP game?

What I want, in no particular order (primarily because I'm just writing this off-the-cuff):

• Any sequel must have the same working core functionality as the original.  Every core system must make the transition and be stable and working.  All of these issues with orbital lines missing, or decaying orbits, or chutes not opening, or spontaneous combustion on the launchpad simply because the game says so?  Gone.  None of that can happen in the sequel, even in early access.
• You simply cannot have the game only function on NASA-level computers.  To have mid-range equipment and get 10 FPS at best just because I have 100 parts is ridiculous.  There is ZERO reason to have any game do this with the technology we have today.  20 years ago, sure.  Today?  Simply cannot happen.
• Several popular mods for KSP1 should be included in stock.  MJ, KER, Transfer Window Planner, and Alarm Clock should all be part of stock.  Now, some of their functions could be unlockable based on what you've gotten done in a career game (MJ already does this; you don't start unlocking functions until you've upgraded the buildings AND gotten Advanced Flight Control).  This is a space program; I refuse to believe that any competent species that has spaceflight capability doesn't have the ability to program their own computers to display critical pieces of information OR have the computer take over for launching or landing.  We see it in TV and movies all the time - the computer controls launch and landing unless something breaks.
• Procedural Tanks.  Let me say that again because it's that important:  PROCEDURAL TANKS.  There is absolutely no reason other than "we don't want to code that" to keep this out of the game.  Juno does it.  I think Rocket Science does it.  Why does the sequel to the genre-leading game not have it?
• A mission/contract system that is not repetitive, but allows for replayability through different careers.  The hard rails with limited actual missions in KSP2 is sad and boring.  I get that every game will have some variation of "Go to the Mun" and "Go to Dres" (assuming you believe it exists).  But there is no reason to stop there.  Make the system generate random contracts to go do stuff or go explore other places.  No, I don't have the answers, so don't ask.  Yes, I happen to like KSP1's contract system, especially when you start adding the contract packs.
• Change from Science Points to Tech Points. 
  • You can still run science experiments.  Heck, I'm a fan of weather balloons and weather tracking.  Still need biomes, primarily because weather patterns are different over deserts than they are over the ocean (as an example).  But we should still be able to perform experiments so we can account for different things when designing and flying.
  • We should progress based on what we did IN ADDITION TO what we know.  I mean, how does knowing the barometric pressure on Kerbin get us the ability to fly to Jool?  We should be going through a series of different activities to increase our knowledge and gain access to new things.  And make this in addition to experiments.  We cannot just throw those out; they contain valuable information (such as how different materials react to different temperatures, as an example).  So make this a combination of the two.
• Bring.  Back.  Kerbal.  Specialties.  I liked having to account for "I need to run experiments in these three different areas, but one of them needs to be reset each time, so I need to bring a scientist...but I also gotta have a pilot because Bill isn't skilled enough to fly or have SAS...and I better bring along an Engineer in case something breaks".  This brings design considerations into the game that simply don't exist in KSP2.  Right now, it's just "Well, who has the coolest-sounding name" instead of "I need a pilot, an engineer, a medic...".
  • And on this topic, I liked having to have them "level up" to get better...but I didn't like how that was implemented.  I don't have an answer on this one myself.

I could probably keep typing away, but you should get the point here.  There are so many opportunities to do things in KSP2 that I don't think the developers took into account.  Or that they did take into account but TT told them "Um, no".

[kerbiloid]

https://www.google.com/search?q=wishlist+site%3Aforum.kerbalspaceprogram.com

[chefsbrian]

At this point, all I want from a future sequel is a finished game. There's no realistic scenario in which I consider giving them a cent otherwise now. I've backed a lot of early access, and I will continue to do so in the future, but this franchise has now burnt that trust and it ain't coming back.

[Sappire262]

5 hours ago, Dakitess said:

I'm really sorry but I'll go the TLDR route for that one.

What I want from a KSP2 (a real one) ? Only ONE thing : to be technically up to date. I don't care about new content, new parts, new planets, new gameplay. I want my KSP1 to be recreated from scratch PROPERLY, with all the new techs, the new dev method, the new tool, with a pro-team, with a real budget to get the things done right.

It would be KSP1 but beautiful (and I mean it, not what we got with KSP2, uh ? It's so friggin outdated, and aesthetically questionable !...) and running well and that's it, as a base. Then only, it would evolve, being Early Access if they wish to go that way, the old KSP1 fashion; but at the very least it would be KSP1 as it exist nowadays, but developed on a real up to date new fresh basis, ready to embed a whole new decade of contents, mods.

I don't see how it cannot be the priority. Really, I don't. But that's something personal and I might be wrong, it also happens to me xD

Honestly, I can see this point of view making sense. Rewriting all of KSP from scratch in a new engine, and focusing on just getting the same core mechanics in place, and expanding it later. The problem, of course, is that an enourmous amount of the new content you're going to be adding after the inital release would be way quicker and cheeper to impliment if it was developed in parallel rather than haphazardly duct-taped on top after the fact.

 

5 hours ago, HebaruSan said:

@Sappire262, could you please clear the text coloring from your post? It's unreadable on a dark theme.

Sorry! I've editied it now, so that should be fixed.

 

2 hours ago, Scarecrow71 said:

There is a 3rd option here that you are discounting:  Use a new engine, but reuse whatever code from KSP1 that you can.  I highly lean towards this option, knowing that I know nothing about the existing codebase for either game, as well as having very limited/not-even-amateur-level knowledge of either Unity or Unreal (which seems to be the popular pick here in the community to re-do the game with).  Even just having access to the code from the first game can provide a lot of insight, especially as it relates to bugs that the first game solved, as well as providing a roadmap of sorts of what not to do.

That's not possible though? You can't reuse Unity scrips and settings in Unreal. You can resuse 3d meshes and some animations, but those are often some of the easiest parts of a development project like this to make anyway from how I understand it. Anything else has to be recoded from scratch. Of course, that's not to say you can't look at KSP 1/2 code to get insperation and for use as a guidebook, especially if you're able to talk to the actual KSP 1/2 devs, but a lot of the systems KSP 1 and 2 used aren't optimal anyway and we'd want to rework them when starting fresh in any case.

 

2 hours ago, Scarecrow71 said:

This goes along with having a proper budget, which I didn't see you mention anything about.  We heard about the average salary for developers in SZ's video, and so the budget becomes a very real, very needed thing.  If you pay the developers what they are worth and/or what the market commands, and give them the right tools for the job, which would include testing rigs, we are talking about $250k per developer.  You want a team of 4?  That's $1 million per year, and we know that the team should be at least 5-8 people.  We estimate high (8 people), so we get $2 million per year for the developers alone.  And that doesn't include office space, CM's, other overhead we aren't thinking of.  That $10 million for 2 years that TT gave Intercept was probably in the right ballpark, but still well short of what they needed.

I mention all of the budget related stuff later. But yeah, I basically agree with this assessment.

 

2 hours ago, Scarecrow71 said:

Again, I disagree with 100% new code.  If you have access to the original game's codebase, it would be highly detrimental to not utilize what you can.

If we are talking about this being a 1.0 release, I think we need to back up a bit and do this in pieces.  For example, Resources/ISRU could be left out of the initial release into Early Access, with it then being added in at a later point.  Same thing with Science mode - which needs a massive overhaul to begin with as the science points system just doesn't seem to work/fit any longer - and Career mode.

Reusing code is literally impossible (you could maybe put some of it though converters? But you'd still end up spending more time doing that and patching the result into a workable state than it would take to redo it from scratch).

Yeah, I think releasing stuff in smaller batches would be a better idea. I mostly just didn't want to type out a couple dozen update lists to be honest.

 

2 hours ago, Scarecrow71 said:

Graphics improvements should be coming out with every patch or bug-fix if possible.  I can see having improvements in a 1.1 release, but let's not forget they should be coming all the time (again, if possible).

I love the idea of more parts, especially if we can get some of the Near Future, Stock-alike, Space Station, and RL US/Russia/Japanese parts that we all want.

I wholly disagree with Life Support.  This is a topic that is a heated one on the forums, with 2 very distinct factions - one for, one against.  I'm not against having it so long as it's an option to turn on or off...but if the vast majority of the community (in my estimation) will keep it off, then why have it?

I love the idea of tethers, with the option to remove them and do EVA maneuvers.  I also would like to see being able to move the tethers from one part to the next instead of having just one long one.

Again, I agree, a higher number of smaller updates is better. Life support, yeah, okay I can see the arguments against it. I guess the modding community would have that covered anyway. I'm glad you like the idea with the tethers though!

 

2 hours ago, Scarecrow71 said:

Interstellar is a hot topic, and brings up a lot of questions.  Such as "How do we deal with Time Warp under acceleration" and "How close/far should they be" and "Do Kerbals live forever".  I would love to sit in on discussions on this, even if I'm muted and can't ask questions.

I think other launch sites should be included before this update, though.  We have them in KSP1 thanks to DLC; why wait this long in KSP2?

Other launch sites would be included earlier, I was talking about cites and things. I'm not sure if the demand for that is high enough though - it probably isn't, on second thought. Interstellar is basically non-negotiable in some form or another though, given how popular it is even in the buggy state that KSP 1 mods have it in.

 

2 hours ago, Scarecrow71 said:

Why is career mode so far after initial release?  You can see here I'm starting to ask questions on decision-making, which is probably more me wanting to understand where you are coming from.  Career mode could be implemented before this, and should be implemented before Colonies.

I have no comment on IVA as I don't use it.

Yeah, I thought I might have made a mistake with the placement of that. I guess putting career mode closer to or on the inital release might not be that hard upon second thought.

 

2 hours ago, Scarecrow71 said:

I guess a lot of this depends upon how compatible those models - not the mods themselves, but the actual models in those mods - are with KSP2.  I honestly don't know if they are or not.  I do know that KSP2 kind of uses the same lego-building system that KSP1 used, so I'd really be curious as to how much effort this would take.  Beyond that, my question is "Why is near future part of 1.1 but all the other real-life stuff part of a DLC?  I get Near Future is probably needed for Interstellar, but wouldn't that fall under the same premise you are using here with needing all that effort to recreate the parts for KSP2?

And RO is another one of those things that I'd support if it had the option to be turned off.  There's a reason why RO is a mod and not part of stock - or even part of DLC.  I know people use it, but are there enough people who use it to make it part of a purchasable DLC?  Or are you only throwing it in here because it's part of DLC with a lot of parts?

I'm not talking about mods here, so I don't really understand what you're asking? The real-life stuff is part of a DLC because it's a too much content for mods to make, but too niche for it to be in the base game. It would completly overwhelm like 70-80% of players and make them stop playing. When I say every real spacecraft, either flown or even just designed, would be in it, I wasn't kidding. Even just the command pods section of the parts list would have well over a thousand parts in it. I'm also super skeptical if this would ever actually happen, I mostly just included it as a whish-list item. It would be a lot of effort to make a DLC that not many people would buy.

 

2 hours ago, Scarecrow71 said:

Why have RO in the first DLC package and not this one?

What about JNSQ?  Or KSRSS?  Un Kerballed Start?  Alternate tech trees (depending upon how Science mode is set up)?

Why are you getting rid of Kerbals in favor of Humans?  If I want a human look-alike, I'd go play Juno, or Rocket Science, or one of the other space clones.  Leave Humans out of it here, please.

What? The second DLC package I proposed was the one with RO in it, the first didn't. I think it would be pretty cool if we had not just RSS, but also real-scale kerbal systems and various other real and fictional star systems as well. And I didn't get rid of kerbals in favor of humans, I explicity said it was a toggleable option, probably one that would be off by default. Again though, this would be a lot of development work for a niche product, so it's mostly a wish-list item. I guess given the modding support here will be much better than in KSP 1, we could just rely on community mods to do all of this for us anyway.

 

2 hours ago, Scarecrow71 said:

3 years of development up to 1.0?  Are you accounting for setbacks, changes to scope, issues with staff coming/going?  What about things like studio buyouts, shifting priorities, changes in office locations?  You brought up 12 people for 4 years to get to 1.0 with KSP, but what you are talking about here is a far larger project than 1.0 in KSP1.  We have to assume that all currently working features in KSP1 are part of KSP2 at 1.0, plus you want to redo all the parts, all the code, have ISRU and resources, and a whole lot more.  Just in 1.0.  Are you sure 3 years is enough time to do all that and have contingencies for things that fail or go wrong or change?

Yeah, okay, that might be optemistic. I figured that the 4 year figure for KSP 1 could be improved with a larger team size and a clearly defined roadmap to start off with, and the 3-4 year timeframe for KSP 2 was while dealing with massive problems and inefficencies that we wouldn't have. But, yeah, it is possible if things go suboptimally that we could be looking at 4 years to 1.0 rather than 3.

Keep in mind though, you say we would have to redo all the code from scratch as if that would make it harder this time around... but the KSP 1 devs had to do that as well. Obviously. Because they were making the game for the first time. So we know coding KSP from scratch takes 3 years with a team of 12, and the features I'm talking about adding to it here aren't substantial enugh that it would massively increase that time, especially given we'd have a team of like 70 people. And doing it in Unreal shouldn't make it automatically take longer than Unity. So it should be fine, but hey, I don't have much actual software engineering experiance (or any really) so I'm probably wrong.

 

2 hours ago, Scarecrow71 said:

What I want, in no particular order (primarily because I'm just writing this off-the-cuff):

• Any sequel must have the same working core functionality as the original.  Every core system must make the transition and be stable and working.  All of these issues with orbital lines missing, or decaying orbits, or chutes not opening, or spontaneous combustion on the launchpad simply because the game says so?  Gone.  None of that can happen in the sequel, even in early access.
• You simply cannot have the game only function on NASA-level computers.  To have mid-range equipment and get 10 FPS at best just because I have 100 parts is ridiculous.  There is ZERO reason to have any game do this with the technology we have today.  20 years ago, sure.  Today?  Simply cannot happen.
• Several popular mods for KSP1 should be included in stock.  MJ, KER, Transfer Window Planner, and Alarm Clock should all be part of stock.  Now, some of their functions could be unlockable based on what you've gotten done in a career game (MJ already does this; you don't start unlocking functions until you've upgraded the buildings AND gotten Advanced Flight Control).  This is a space program; I refuse to believe that any competent species that has spaceflight capability doesn't have the ability to program their own computers to display critical pieces of information OR have the computer take over for launching or landing.  We see it in TV and movies all the time - the computer controls launch and landing unless something breaks.
• Procedural Tanks.  Let me say that again because it's that important:  PROCEDURAL TANKS.  There is absolutely no reason other than "we don't want to code that" to keep this out of the game.  Juno does it.  I think Rocket Science does it.  Why does the sequel to the genre-leading game not have it?
• A mission/contract system that is not repetitive, but allows for replayability through different careers.  The hard rails with limited actual missions in KSP2 is sad and boring.  I get that every game will have some variation of "Go to the Mun" and "Go to Dres" (assuming you believe it exists).  But there is no reason to stop there.  Make the system generate random contracts to go do stuff or go explore other places.  No, I don't have the answers, so don't ask.  Yes, I happen to like KSP1's contract system, especially when you start adding the contract packs.
• Change from Science Points to Tech Points. 
  • You can still run science experiments.  Heck, I'm a fan of weather balloons and weather tracking.  Still need biomes, primarily because weather patterns are different over deserts than they are over the ocean (as an example).  But we should still be able to perform experiments so we can account for different things when designing and flying.
  • We should progress based on what we did IN ADDITION TO what we know.  I mean, how does knowing the barometric pressure on Kerbin get us the ability to fly to Jool?  We should be going through a series of different activities to increase our knowledge and gain access to new things.  And make this in addition to experiments.  We cannot just throw those out; they contain valuable information (such as how different materials react to different temperatures, as an example).  So make this a combination of the two.
• Bring.  Back.  Kerbal.  Specialties.  I liked having to account for "I need to run experiments in these three different areas, but one of them needs to be reset each time, so I need to bring a scientist...but I also gotta have a pilot because Bill isn't skilled enough to fly or have SAS...and I better bring along an Engineer in case something breaks".  This brings design considerations into the game that simply don't exist in KSP2.  Right now, it's just "Well, who has the coolest-sounding name" instead of "I need a pilot, an engineer, a medic...".
  • And on this topic, I liked having to have them "level up" to get better...but I didn't like how that was implemented.  I don't have an answer on this one myself.

I could probably keep typing away, but you should get the point here.  There are so many opportunities to do things in KSP2 that I don't think the developers took into account.  Or that they did take into account but TT told them "Um, no".

All of this makes a lot of sense, and I more or less agree. The one thing I really disagree on is procedural tanks - I think there should be an enormous amounts of varients avalable on every part, but I don't like true procedural tanks in the stock game because when you have too much free choice, it ironically feels very limiting. If we limit it so you can chose between the standard stock dimaters and a couple dozen length varients, then maybe I guess I could get behind it, but in general, while I like procedural wings (or more accurately, I can tolorate them in stockalike games), and stuff like radiators would probably be fine, I think the tanks should still be lego-like. That's just my opinion though, if the community disagrees as a whole then obviously their thoughts are more important than mine.

 

1 hour ago, chefsbrian said:

At this point, all I want from a future sequel is a finished game. There's no realistic scenario in which I consider giving them a cent otherwise now. I've backed a lot of early access, and I will continue to do so in the future, but this franchise has now burnt that trust and it ain't coming back.

That's fair! But given that any future KSP sequel that doesn't go to EA is guaranteed to fail, I guess that means you're just going to have to wait a few years longer than some others before getting it. Which is fine, EA isn't for everyone by definiton, it's for a small number of people who can act as playtesters while the rest of us wait to find out if it's any good.

Edited May 28 by Sappire262
reworded some stuff

[dprostock]

 

[Sappire262]

20 hours ago, dprostock said:

 

I’m not sure what this means? It’s a good video, but without any actual writing I don’t know how to respond to it. I guess if you’re trying to say you want more soviet-styled parts, I agree, and I even have some basic ideas for how to implement that, but that’s all I can think of. But, well, I guess I might be missing something.

 

 

Anyway, on another note, I’ve been doing more thinking about all this, and something I want to emphasize again is that even though there’s a very good argument to be made that the initial goal at 1.0 release of any KSP sequel should probably be to just have all the core mechanics from KSP 1 down and improved upon to some degree, we still need to develop roadmaps for the content that will be released after the fact, so that we can take that into account when making the game.

 

For example, let’s say we want the game to have a new mechanic in it where certain parts can split apart mid-flight into multiple parts. For example, a pod with an ejector seat in it. In KSP 1, the only way to implement that would be to make the pod, seat, and hatch three separate parts (you might, maybe, be able to combine the pod and hatch with fairings). What if we want one part, like say a Vostok-style capsule or a jet aircraft cockpit of some sort, to have an ejector seat in it? Not only would it have to be able to split apart in-flight into two separate, individual and separately modeled pieces (KSP 1 decouplers can kind of just about do this, but one of the two leftover parts is just a physicsless mesh IIRC), but we might also end up needing to have that one part use multiple stages. Imagine an apollo-style command pod, made of just one part, with built-in parachutes. You would need one stage to decouple the topshell cover, one to activate the drogue chutes, and another to cut the drogues and activate the mains (or you could do something fancy and animate the drogue chutes pulling the main chutes out, etc). In KSP 1 to my knowledge, each part can only have one staging action, but here we need three that need to be individually movable anywhere in the staging tree.

 

Implementing all of that in KSP 1 would be, well, I have no idea for sure but given I know a lot of modders have tried and failed, it’s probably very hard. So if we just try to match KSP 1’s features to start off with and add this later, it will also be very hard. But, I would bet quite a lot that if we knew from the start we would want to do this someday, and designed the game to be able to allow it, then when we do go to implement it later it would be way easier. So, yeah, the point is the game has to be designed from the beginning to be, well, scalable I suppose, or extendable or modifiable, or whatever other adjective you want to use. And the only way to really know how to do that is to have some sort of idea for what you’re going to want to include in the future. Hence, roadmap.

 

 

Right, so, anyway, here’s some more random brainstorming of stuff to add, in no particular order.

 

Some more things to make exploring around celestial bodies interesting would be cool. Right now, you basically just land somewhere, plant a flag, set up some deployable science experiments, and that’s it. Some particularly intrepid players may bring a rover to drive to various biomes, but even that gets old after a while. Some much more diverse and varied terrain, with actual concave features like caves and lava tubes, would probably help (and require an all new terrain generation system but that was already a given), and giving our kerbals more things to set up and do on the surface would also be cool. I’m imagining something like a portable winch system and some climbing picks so that a kerbal can repel down the side of a steep crater or ravine, or into a sinkhole that leads to a massive underground lava tube network.

 

Maybe it would also be a good idea that in career mode, before you get landing programs you get orbital survey ones, and after doing those they give you maps of the surface of the body, annotated with points of interest like overlapping biomes or unique surface features, and helpfully mark out specific potential landing sites. You could then, as part of the lunar landing program for example, get individual missions to go and explore all of those sites, and if you do more orbital surveys in the meantime they might start pointing out more and more interesting locations, culminating in entrances to some massive lava tube networks or perhaps sources of water ice at the south pole, that later down the road once you get ISRU tech you can exploit to make a refueling outpost.

 

I do like what KSP 2 was doing with the story progression and the discoverables and all that, I do really love the idea of there being actual lore in KSP and a story to discover, that’s probably one of the pieces of cut content from KSP 1 that I most regret (for those who don’t know, yes, the KSP 1 devs did plan to have that!). I’m torn though between doing what KSP 2 did and having one or two discoverables on each celestial body, and rapidly progressing past them, and having dozens on each and way more missions and programs to go explore them. Maybe having two different, well, not difficulty settings, but I guess progression settings, for career mode would be good? So you select the button for easy, medium, or hard, and then beneath that there are options for rapid or slow progression modes, which define how many missions you have to do to find all of the various discoverables and advance the story. That would probably increase development time a bit though, but I’m not sure by how much.

 

On other, completely unrelated thoughts, I think it might be a neat idea to have the player interact more in depth with the space center and the people working there. Taking some cues from Buzz Aldrin’s Space Program Manager, for example, and allowing the player to select specific kerbals to lead development efforts, or appoint individual flight controllers, stuff like that. I’m still having trouble figuring out exactly how that would work though, KSP is very different from BASPM after all and we wouldn’t want to distract too much from the build-fly-learn loop intrinsic to KSP. The ideas I had earlier in the latter half of my first post about adding a city-builder like aspect to the KSC where you actually managed the upgrade process yourself in a similar manner (or perhaps even using the same basic game systems, just on a larger scale) to colonies I think is alright, but I would be curious to hear everyone’s thoughts on that and if expanding much beyond just that in scope would detract from other aspects of the game. We do, afterall, want to keep a fairly even balance between emphasizing mission planning, spacecraft design, and actual flying. Adding a fourth thing, space center management, would be accepted very well by some players but I’m not sure if absolutely everyone would like it, and making it togglable may be possible but it might also turn out to be a lot of work.

 

Okay, so, again on an unrelated note, I’ve been looking through those wishlists someone linked earlier, and here are some things that stand out, in no particular order. I’m mostly going to mention new stuff here, obviously there would be better aerodynamics and better graphics and all of that but I’ve already mentioned that in earlier posts. I may end up having some stuff here I’ve already talked about though, I’ll try to avoid it if I can.

 

Credit to Vl3d, Z.O.M.G, and SirDeadPuppy, for most of these, though I’ve tried to expand on them where I could.

 

 

A built-in mod loader with dynamic mod loading in-game if possible.

 

Editable and scalable UI elements, with a lot of customisation.

 

More diverse kerbal appearances, with them maybe even growing beards and their hair longer on long trips.

 

Something like Final Frontier, with ribbons or medals given out to kerbals to keep track of their achievements, and with a half of fame and maybe a memorial as minor buildings in the KSC somewhere.

 

 

Closed part loops and multi-point attachments - basically the ability to make, for example, circles and have them not have a seam somewhere where they don’t actually connect, without having to use docking ports or struts. Also the ability to have two parts join together at multiple points.

 

More complex center-of-thrust indicators, that take into account which stage is firing (so they don’t always show literally every engine on your ship), and perhaps something like RCS Build Aid as well.

 

A part comparison tool to compare, say, engines to each other.

 

Construction and pad refurbishment times. I think this would be really cool, but perhaps only on the hard difficulty settings.

 

Better and perhaps even customizable part interiors, and maybe something like connected living space to merge connected part modules together.

 

 

A mission control room for probes, and maybe even for crewed flights as well.

 

More stuff on the navball, specifically mach numbers and time to impact with the ground or to start suicide burns, would be nice.

 

Atmospheric trajectory predictions in the mapview, taking into account the aerodynamics of your craft, and with options like setting what entry attitude to assume when making the calculations.

 

Automatic station-keeping so your satellite constellations don’t slowly drift out of alignment when time warping at max speed for a while. Maybe you could set this up so it checks if a satellite has any thrusters on it, and automatically removes some of its propellant over time, until you need to replace or refuel it. The draining of propellant during automatic stationkeeping could be togglable in the difficulty settings as well.

 

Communications antennas caring which direction you point them in. Personally I’m not sure about this one, I think it’s a bit over the top. Having them rotate to point at the nearest comm relay if it’s in its gimbal range, and still working even if it isn’t, though, would be cool.

 

 

Uncrewed probes before crewed stuff in the tech tree, maybe with you starting out with both suborbital sounding rockets and some pretty good subsonic aircraft, and able to progress from there down either path as you want (rather than starting out with crewed suborbital rockets and then unlocking planes and uncrewed rockets later).

 

Some sort of system for handling launch gantries, I think doing something akin to Modular Launch Pads would be overkill and not for the average player, but it would be nice if the launch gantries were similar in size to your rocket every time and spawned at a reasonable distance from it. I’m sure some sort of solution is possible, and you could have something like Modular Launch Pads as a toggleable option or DLC content perhaps, with the launchpad parts still being assembled in the VAB but following their own separate systems and mechanics. Maybe you could build a launchpad, and then until you next spend a bunch of funds to replace it, all your future rockets have to fit on that same pad, with the fuel lines to the tower and all that being slightly flexible/movable but only up to a point. That might be a bit over-the-top though to be fair, even as extra content. Maybe we could code it so that systems like that are possible, but leave it to the mods to actually implement them.

 

The colony system could work using standard cities-skylines style building mechanics, though it would be nice if the buildings were still modeled using similar systems to parts and had full physics and whatnot. I’ve seen some people propose space elevators as colony structures, I think those are a bit too large-scale for a game like this, but if a decent system could be figured out to implement them it would be cool. Honestly, the more I think about it I do kind of like the idea that the colonies could start out with a couple prefab modules each less than a dozen meters across, and over several decades of in-game time could grow to entire cities with proper buildings made of in-situ resources and with tens of thousands of inhabitants, and with the colony builder at that point basically just being actual cities skylines but in space, so maybe with that sort of scale it would make sense.

 

Also, terraforming as part of the colony system would be amazing. I’m talking about small-scale, excavating dirt and flattening terrain here, though I suppose planetary-scale terraforming could also work (but that would definitely be too large in scope). It would require an all-new terrain system, but again, we were going to do that anyway.

 

Massive ground or even orbital laser arrays to accelerate solar sails would be awesome.

 

 

More stuff to do on celestial bodies, as I talked about earlier, would be cool. And to go with that repelling and climbing equipment, scuba gear for underwater exploring, flares to light up caves, deployable emergency shelters for very small rovers, all sound great. Perhaps even dedicated astroarchaeology equipment for examining discoverables. I’m not sure if life support in general should be in the base game, but having to manage oxygen supplies on EVA specifically sounds like a great idea even if that’s on its own, or disabled on lower difficulty settings.

 

Maybe having some celestial bodies be discoverable, perhaps with space telescopes, like really small dwarf planets, or a small planet at Kerbin’s L5 lagrange point with Kerbol, or the moons of the outer planets, etc.

 

More science parts would be neat, perhaps with a science management system akin to Kerbalism. I like the idea of getting rid of the sorts of science parts that you slap on to every vessel, and making more custom-tailored ones that are designed for specific missions to an extent. Lots of passive experiments and really specific ones would be nice, like positioning a bunch of satellites to make a gravitational wave detection array, or launching a bunch of lunar missions to build an array of several dozen optical telescopes on the farside, or even setting up a large radio telescope in a crater.

 

Through The Eyes Of A Kerbal and FreeIVA, with fully detailed interiors that you can navigate around and also see into though windows from outside, and with going on EVA fully animated with the hatch opening and everything. Getting this to work in a way that’s bug-free, does not kill your FPS, and works even if you clip twenty pods into the same space, sounds like a nightmare but I’m sure some sort of solution is possible.

 

 

Balloons, airbags, and other inflatable things would be nice, especially if some of them actually take resources like hydrogen or air to inflate. A tiny probe hanging from a balloon, packed into a 1.25 or 1.875 m aeroshell, would make for a super cool and unique early-game eve mission, especially if we make eve’s environment super corrosive and dangerous, and requiring late-game parts to explore at the surface.

 

More complex procedural tube-like parts would be cool. Right now we have struts and fuel lines, but larger things like crew tubes would be neat, especially if we had some way to manipulate the shape of them so they can curve around stuff for example. Imagine being able to have a lander descent stage carry an ascent stage cabin and a surface habitat side-by-side on top of it, like some of the irl NASA concepts, and the two were linked by a crew tube that could be disconnected before the liftoff and ascent into orbit.

 

More engines, especially unique ones like ion drives and nuclear thermal rockets, to say nothing of Orion drives and the near and far future stuff, would be great. And having them not gimbal when set at zero thrust would be really cool.

 

More wheels and caterpillar tracks and landing legs (procedural ones maybe?) and all that good stuff as well, with lots of variants too.

 

Just in general, a large selection of parts, all in a consistent stock-alike style, that can be used in the right combination to build pretty much any real world spacecraft you can think of, or at least kerbalised versions of them. Including soviet stuff, and spaceplanes like dynasoar and dreamchaser, etc. Plus near-future parts, and ones to make wholly unique crafts with no real-world analogue. And more part sizes, I think 0.625, 0.9375, 1.25, 1.5, 1.875, 2.1875, 2.5, 3.125, 3.75, 4.25, 5, 6.25, and even more larger diameter options would be great. Plus more custom-cross section parts for spaceplanes, in stock we only have Mk 1, 2, and 3, and Mk 1 is just the regular 1.25m circular parts but with a different name.

 

Much more editability on fairings would be nice. Having a similar default interface as KSP 2 perhaps, so it’s easy to use in a pinch, but with advanced options if you want to spend time tweaking them. Maybe with the ability to make interesting cross-sections and have certain parts flare further out, for example the Voskhod 1 spacecraft had a bulge partway up on one side of the fairing so that the airlock could fit under it which is impossible to replicate in stock KSP 1. And being able to position hatches on fairings so that you can enter and exit crew modules underneath them (like on Vostok/Voskhod/Soyuz, or the boost protective cover for Apollo/Orion, etc) would be amazing, and stuff like that.

 

More launch escape system parts, or perhaps just some small high-thrust low-burn time SRBs so we can make our own.

 

Parts designed for use underwater would be cool, and maybe you could even build probes designed to borrow through the ice on a place like Val and reach a subsurface ocean, before deploying a small uncrewed submarine. Once you start building colonies, having the option to make them underwater and build larger submarines to explore around from them would be incredible.

 

Maybe you could even use nuclear explosives to tunnel a large hole down through Val’s crust to the ocean below, build an elevator down the length of it and a colony at the top and bottom, and lower down large crewed submersibles to explore around with. You could even have some small or perhaps large (maybe kraken like?) sea creatures to find and interact with.

 

Speaking of which, some floating algae in Jool’s atmosphere to explain its green color would be cool, and maybe some larger flying creatures as well.

 

Weathering on parts would be interesting, we already kind of have this for heatshields in KSP 1 but having them on other parts, and stuff like dents, scratches, and melted or frozen sections would be cool as well.

 

 

More planets in the Kerbol system, and way more moons around the gas and ice giants, including maybe a Titan analog with an atmosphere kerbals can fly in using special wingsuits. And asteroid belts as well, that would also be great.

 

And more unique environments as well, like having lava on some planets, Moho for example, and extreme thermal environments that require specially designed crafts and EVA suits to deal with. 

 

Also destructible or movable terrain, like quicksand, mud, landslides, moving glaciers, icebergs that actually float and move around, lava tubes that can collapse in on you and force you to improvise a plan to dig your kerbals out with explosives or something, maybe even craters left behind from where your craft impacted particularly fast. And footprints, wheel tracks, etc. Plus colidable terrain scatters, and cool particle effects like snow, sand, rain, dust, etc. Maybe weather phenomena, like tornadoes and such? That all sounds very cool, and if we’re recoding the game from scratch I’m sure we can find ways to implement it.

 

A black hole in one of the other star systems added for the interstellar stuff maybe? Not sure how time dilation would be handled though. Also rouge planets, double planet systems with unique n-body physics systems unique to them, interstellar asteroids and comets, oort clouds, tons of dwarf planets (maybe even procedural ones!), etc.

 

Maybe there could be an event where a large asteroid impacts another planet or moon, and you get to send a probe to watch the pretty lightshow as it happens. Or an asteroid is going to impact Kerbin, and you have to deflect it off course.

 

And I mean proper, real-scale asteroids, not the tiny ones in KSP 1, though you could still keep the small ones around as well I suppose so that you can redirect them with smaller spacecraft if you want to capture and mine one.

 

Make it so that during reentry, parts tend to detach themselves from each other a bit before actually burning up, so they leave their own independent reentry trail behind them and it looks like your spacecraft is actually breaking apart rather than just getting smaller.
 

 

Alright, that’s all I can come up with for now, though I’m sure I’ll find and think of more. In the meantime though, does anyone have any other thoughts?

Edited May 29 by Sappire262
Fixed the text formatting a bit