Answers to some questions we had to skip over during the AMA but I still wanted to get to:
What percentage of the parts in KSP2 was created by you personally?
Depends how you measure it. Effectively zero because I don’t do the asset work, by one definition. In terms of maybe inception/concepting, in the EA release I’d say I had a hand in about 10%?
What is the largest part will be in KSP2?
The largest part I have in my list right now is in the 80m+ size category. It’s a lot harder to measure these colony parts versus vehicle parts though…
Do you participate in the creation of parts for the colonies?
I participate in the concepting and design phase yes. It’s where I’m focusing a lot of my ‘thinking time’ these days. Colony parts are both similar and different from vehicles – in what they look like, how they assemble, etc. As we get to those milestones we refine our designs from player feedback.
How difficult is it to add a new part to KSP2? Is there a big difference? Is it harder than creating a new part for KSP1 for a modder?
Most things in KSP2 end up being more complex than KSP1. As an example at a basic level the PBR shading model that we use requires more texture maps than KSP1. That is mitigated by having access to internal tooling and a faster iteration loop (click Play in Unity rather than load the game).
is there any more concepts for more air-breathing engines like the J-90 smaller or larger
There’s been team interest in larger air breather engines, but as always that’s not so simple – adding an air breather of say, 2.5m size requires us to also look at the supporting parts in that size, like intakes and cockpits, so the player can have a good experience when using those engines. That balloons the required work significantly.
I would want to push out the different technologies rather than footprints first. Nuclear jets, propellers, all unlock interesting new player stories!
is there gear that is going to be angled from the fuselage not straight up and down and finally more tires/wheels in the concept stage, or even remotely thought of...
We definitely have people who want that on the team .
How will the sizes of different stars be scaled with respect to Kerbol? Will they be scaled at 1/3 their real-life analogs like Kerbol and the Sun?
Specific scaling of the actual meshes is less important than defining their specific insolation numbers for input into solar panel math but yeah they’ll be Kerbol-relative.
How do you plan to implement proper motion of other star systems, and how do you expect that to add to the challenges of interstellar travel?
Hah, interstellar travel is going to be hard enough already. Proper motion is something we need to balance carefully there.
Hi, Chris! Im sure you've been deep in colony part design. What are your thoughts on greenhouses and simple life support with snacks for example? How do you see conveying that colonies are both real places where kerbals live and 'working machines' much the way vessels are?
Honestly I don’t like basic life support (by basic I mean something like having Kerbals on a ship consume a resource). I’ve played all* of the KSP1 mods for it, and I haven’t found something that is interesting and holds my interest beyond frustration for more than a few hours – just not my cup of warm beverage.
More seriously though, systems like this need to have a bunch of considerations:
- They need really carefully crafted player stories. Those stories need to support lots of different player archetypes – not just advanced players.
- They often should work on a carrot rather than stick-based approach. KSP has a lot of sticks right now.
- They need scalable solutions that are plannable and toolable. That’s a big thing and that’s where LS gets expensive in dev-hours
We have some things in the works around Colonies that ape some of the ‘results’ of life support, which I hope will get at the idea of colonies being a little more kerbal-involved than just plunking Kerbals in a command part.
* I think all? It has been a while, maybe some new ones have cropped up.
Has the concept of heating changed at any point based on the feedback posted to its thread?
I read every post in the thread, which was nontrivial because it was a long and uh, vibrant thread. The short version is no, the long version is yes but…
A lot of the interesting discussions sat around things that are further down the roadmap, and they provided us with a couple additional things to consider. Interestingly, the player stories we have were well aligned with the comments that I read, but the way the player stories were addressed were not unanimously approved. That’s fine – part of the EA conversation– and in particular with a lot of discussion being on items later in the roadmap, this makes me confident in the iterative model.
We’ll get the basics of the system focusing on reentry stories out to everyone. We’ll evaluate how that works with the playerbase. As we move towards the next milestones, we can use the information encoded in the thread, which I’ve collected internally, to make sure we’re making choices (engineering or design-wise) in conjunction with the feedback from reentry to get good solutions. One thing that jumped out for me was that there’s a lot of talk about macro vs micro solutions. I’ll be the first to admit that the current solution is a macro solution. So future design work will probably focus on whether there’s more microscale interaction to look at.
If I know the peak or average specific heat flux a vessel is gonna go through on its final orbit/landing spot, what stops me from just adding enough negative heat flux parts to counteract it?
Nothing. That’s what you should be doing. Of course, it’s not really that simple. If this is atmospheric heat from going fast, adding a big radiator is likely to just increase the amount of next flux, because it has a large surface area. Most heat mitigation tools need something else too – a radiator might need electricity, which means you need to supply that, which will enforce additional constraints.
Considering its possible uses on the automated logistics network, long missions, and just straight up anything that only requires time to pass, how do you balance not timewarping versus just letting things happen in ultra-fast time?
These are the best questions because they’re the hard ones. Often we trend towards supporting a player path that doesn’t reward excessive timewarping, but doesn’t exclude it either. A good case study is resource extraction and deposit concentrations. There’s definitely fun in seeking out and finding the best deposit for mining. Obviously though timewarp makes that kinda moot in timing. You could just start mining a hypothetically low-grade deposit and warp for 50 days. That tells us that time and rate -based mechanics need to have more to work well. A specific example here is that a newly accessible resource should be constrained differently – challenging location, resource transport limitations, etc.
We try to move the real player decisions to things that are interesting with and without time as a mechanic. Mostly hypothetical examples, but here’s a few ways of thinking of these things on top of my head:
Put a locational constraint on something. If you need to do something in orbit over a specific part of a planet, make it take longer than the average orbital cycle. This might encourage a player to put a satellite in GEO orbit over that place. If you do the work to put it in GEO, you get the benefit of being able to timewarp.
Use binaries instead of gradients. Does ore concentration really benefit from a really detailed gradient from 0.0001% to 100%, or can you look at it as a yes/no? Trade that, see if you’re damaging player stories with that simplification.
Use supporting systems. Sure, you could mine that deposit at high timewarp. But the deposit is on a planet with a day length of 200 days, and you need power, and the area has no fissionables. How are you going to power it? If you solve this problem, it is satisfying and you get a cookie. You did the work, enjoy your timewarpable extraction!
These are really big problems we look at for all of the more complex systems because hey, an interstellar transfer could be 100 years. Players will timewarp that and that’s… the whole length of a KSP1 campaign. Fun with and without timewarping like this is essential.
What are your favorite tips and tools for new modders?
My biggest tip is to do what you want to do and not focus on what others want. Lots of the most creative KSP1 mods didn’t hitch themselves to any one concept of the game, and that’s what made KSP1 modding so successful. You want RO? You’ve got RO. You want to launch kerbals in a cardboard box rocket? That’s there too. You want life support? Oh hey there’s about 5 different concepts out there to pick from.
Also don’t try to form a team day 1 . Get some experience, release some stuff, and the team will come to you!
Tools - Blender is an amazing piece of free software, and there are a ton of good coding tools out there for the software-minded as well. It has never been a better time to be an independent purveyor of these kind of things, you don’t need to suffer through e.g. gmax or the trial version of Milkshape3D anymore.
Is there any consideration of 1.875 meter parts
Building out a whole family of 1.875m parts that includes the core stuff (engines and tanks) plus the necessary ancillaries is a lot of work and not something the team is committing to right now.
While we do know it wont be added in the short term, the team has previously been wishy washy if radiation/life support will make it into the game. Are these topics something that the team has decided wont be in the game until maybe after 1.0, something the team has a firm answer on what they want to do with but does not wish to disclose it (though if you do wish to disclose please do), or something that the team is geniunely undecided on
See answer to Pthigviri about LS stuff.
Radiation is a bit more interesting to me because I have a fair bit of history in mods with it, and I’ve eagerly assimilated the early concept work the team has done for KSP2. There are two tradespaces in terms of vessel design, point sources and ambient radiation that we at least nominally want to think about including.
Ambient radiation is basically a time trade. How long can you spend in a radioactive environment? You can throw things like radiation shielding, storm shelters, etc but ultimately it all comes down to time to Bad Things. It’s harder to help a player to plan. You have to give them tools to determine how much radiation there is around somewhere and how to figure out how long they can spend there, etc.
Point radiation is nuclear engines and reactors. This is harder to implement but is definitely relevant in terms of craft design, because it is a big part of why fictional interstellar ships look the way they do. Interestingly it is easier to model and communicate to the player because you know lots of the variables at vessel build time. One of the messy things here though is that as soon as you throw in radiation, you railroad players into building ships with nuclear engines in a very specific way. We have to craft a solution that hits a nice middle ground. See this comment.
I’m candidly going to say that we don’t have the ideal solution in the bag right now – but that’s what EA is all about. I’m sure I’ll write some kinda discourse on radiation eventually for a dev blog and everyone can weigh in on why I’m wrong :P.
Are there any features you modded into KSP 1 that you are bringing into KSP2? What is your favorite?
I wouldn’t want to port anything specific without a good justification, but I really want to bring in more planning tools. The only ones I built were around heat and power management, but yeah. Something like that.
One of the cool things about this job is that I get to start again, so to speak, with the support of people who have been in the industry for a while. So if I want to bring in nuclear reactors, I can take my concepts from Near Future Electrical, talk to some Real Designers ™ and get their opinions on what works and what didn’t work, and make something cleaner for KSP2.
What are next parts that are comming into the game?
Science parts! But also those gridfins we teased a while ago should appear.
Are there any kinds of parts you're going to be adding to KSP2, that as far as you know, weren't even really available as mods in KSP1? Some unexpected bits and bobs, maybe
The entire colony loop is more or less stuff that was never really available in KSP1 mods from a system perspective. Modding KSP1 was really wide though – hard for me to say.
Are all parts from Your mods to KSP1 will be implemented in KSP2? Especially large solar panels, station parts & MK4 spaceplane?
Hah, no not at all. I like to re-use concepts, but this is a great opportunity to start afresh and to fix some stupid things I did in development of those in my mods.
Gotta somehow get more Thunderbirds in the game though.
Will we get all size variants for all parts? Example, hydrogen tank with the smallest radius, only has long option. Why "semi procedural" parts weren't considered, where you can select a tank and set its lenght/radius to some of the predsfined available values?
You definitely have me to blame for no smaller hydrogen tanks – just don’t think they’re useful with the low density.
Why wobblyness still exists? What are design choices and reasons to keep it, if there is a way to remove it? If there is indeed a way to remove it...
I have a post on my thoughts about this as a player. Generally though – it’s not where we want it to be and we’re trying to figure out how to get it there. That’s extremely non-trivial, there are various posts in the forum that do a good job of explaining some of the whys.
How advanced will the Kerbal's technology be, will there be very advanced parts such as anti-particle devices?
We’ll definitely get way up there in the tech tree. I do want to keep those under wraps for now tho.
Are you able to tell us 'something' about science and career modes, there's been an alarming lack of any real information regarding the two.
Well! Science mode is cool. It is designed to be a progression-based mode that takes the aspects of KSP1’s Science mode that we like and build upon them to create a solid progression experience that has higher level of agency and approachability. You can expect the return of the experiment loop, with changes, and the inclusion of a very different mission paradigm from Career.
One of the fiddlier aspects of the last few months has been taking our full set of concepts from KSP2 1.0 and figuring out how they break down into the early access structure.
Delving deeper, what can we expect from science mode, is it the same ‘click and reward’ setup as KSP1 or are you going for a ‘science over time’ sorta approach more akin to Kerbalism?
The system as designed is independent from things like Kerbalism, but you could say there’s some concepts that aren’t dissimilar in there. It has been a while since I have played with that mod tough. We definitely want to get to more player agency in science. Instead of it effectively being mandatory to hide 4 tiny science experiments on every craft you send anywhere, we want you to make a more informed decision about what you take with you, and make the actions you take a bit more specific too.
I should write a little dev blog on this.
What sort of part numbers are we looking at, is there going to be the same sorta number of experiments as KSP1, or significantly more? What does that entail, are the experiments something more dynamic this time, looking at things like NASA’s GRACE mission for example?
I should definitely write a little dev blog on this. Similar number, more impactful.
In terms of career mode, is there a more dynamic contract system in place rather than the rather ‘rinse and repeat’ system of KSP1? There’s still going to be funding, reputation?
I believe we are on record about not using the same framework there. Funding and reputation weren’t our favorite systems and didn’t have the gameplay impact we wanted.
As a side question, stations and bases. Are these going to have something of a real use this time around, given that stations were limited to more or less just fuel depots in KSP1. I'm thinking more along the lines of long term research projects, with big pay-off for significant durations of time. Is there some sort of requirement to resupply the stations, perhaps required crew rotation, stuff like that?
The progression we want to deliver for bases and stations mirrors IRL conceptions about how these things should work. You will start out with outposts that have limited utility – let’s call that KSP1-like. Fuel depots, maybe comms relays, etc. As you progress through the tech tree, you’ll get access to stuff that provides them with greater utility. That’s shipyards and docks, fuel factories, launch pads, etc. Eventually you’ll get the biggest parts, which are mostly focused on giving you the full capabilities of the KSC at a colony.
A core piece of the utility in my mind comes with resource gathering (which is a ways off in the roadmap,) when the specific positioning and configuration of a colony becomes really important. Placing a colony with good access to progression-related resources and having easy access to heat management/power sources will allow you to build specific functions and cool vibes into each colony.
Crew rotations and resupply are not currently something we would want to enforce. I hope that when we get resources and delivery routes fully operational though, that this is something modders will hit really hard because the framework of stuff like delivery routes will be there.
Pineapple on pizza or not?
I don’t like it, but recently I was made aware that liking baked potato pizza was weird so I can’t really judge.
Having done both, what do you think are the main differences between adding a part (or set of parts) to the game as a modder, vs as a paid member of the team?
Accountability and justification are big. It’s easy enough to incept a new part as a generalist modder. I just say that I want it and make the time to model/integrate/QA it myself.
In a professional context, that involves the use of studio resources and we have to balance that versus other work we want the staff that would be executing that work to do. A new part needs a concept, it needs artist time, it needs designer time, and it needs QA time. We have to really be sure we want a part before we do it.
Will the salt water nuclear engine make a return?
I’d like instead introduce the artisanal nuclear fresh water engine, using only the purest Vall-ian glacial meltwater and hand-centrifuged Pol-ian uranium. But yes.
You've made several mods for KSP1 in the past. Will parts from any of those mods, like Restock, Far Future Tech, Near Future Tech, and Stockalike Station Parts make a comeback in KSP2?
Never exactly, though there’s similar roles. I have a 3.75m command pod in Near Future Spacecraft that is pretty similar in role and profile to one in KSP2, for example.
What was the transition like going from being a modder (or, more honestly, a pillar of the modding community) to working on the development team?
It was really weird to come into the project and find pictures of my work as references in the team wiki.
But it has been great. We have a really solid team working to replicate what amounts to 10 years of hard KSP1 development work. Ways to go though.
Is the same approach to design & diameter consistency going to be applied to KSP2, similar to what you did with ReStock?
This is already ongoing – we sneak in consistency work where we can depending on the team’s bandwidth. We’ve sorted at least a dozen parts since EA release. The part-ists are probably sick of my hOw’S tHe SiDe CoUnT questions.
Hello Chris Adderley. How detailed will the reentry VFX be, on the vessel's parts? Will we be able to see the heat propagate relative to what part of the ship is hitting atmosphere the most? (As in, will there be a glow on the entire vessel that spreads as atmosphere becomes more dense, in a reentry? Or will the heating visuals display in every single parts of the vessel individually?)
I will leave this one completely to allow future dev communication to represent it. It’s really cool and I think the path to get to what we think is our final solution would be a fun thing to tell people about.
What steps is the development team taking to make KSP2 accessible and appealing to new players who may not have played the previous game or are new to the genre?
Obviously, the tutorialization we worked into EA will continue as we add new systems. Eventually though we want to enable players to do more with the same skill level. There’s some really big difficulty jumps in the game, and while we are more confident in the ‘get into orbit’ jump, we still need tools and strategies to tackle the next one, which I’d peg as going to another planet. After that, go to another solar system.
I saw a really cool concept from the UX team about this last week which made me squeal in happiness. I hope we get to it.
Can you give some more detail on the supply route system? Can you automate the construction of supply vessels, or does a vessel have to be built to assign an automated route to it? In other words, when the route is finished, does the vessel have to be intact?
That system is a ways off and while I think our concepts are pretty solid, they have to survive another round of detailed design, and the EA feedback we get through that time period. So let’s save that for a dev diary later.
Intactness is an interesting thing that the system does need to consider. On the one hand, we obviously want you to not crash your ship to create a delivery route. However, we also don’t want to disallow multi-stage approaches to routes. You should be able to create a delivery route with a two-stage rocket. It won’t be as resource effective as a single stage one, but particularly for routes that launch from high G or atmospheric planets, we need to have a design that eventually supports this. It is possible that this could be delivered in phases for effective development – consider a V1 of routes that focuses on single-stage-to-place deliveries and a V2 that is more comprehensive.
Also, will metal to build basic rockets and methalox fuel be limited in the early game, or will there be infinite fuel on Kerbin? If so, how is this balanced against the ability to send an arbitrary number of refueling ships to a colony, as opposed to what I think you probably want to encourage, which is ISRU?
If you want to create an interstellar empire based on shipping methalox light years from Kerbin, I don’t want to discourage that. That’s kinda cool and would be a big investment in player time and resources, so we would reward that by not constraining it. You’re also probably not going interstellar on methalox… so you are going to be incentivized to not do that in a particular way.
In KSP1 some realism enhancements can be achieved with a relatively simple MM patch, because those mechanics are already in the game, but not used in stock (i.e. engine spool up time, throttle depth limits). Are there any realism mechanics that you wanted to put into KSP2, but couldn't because of the gameplay balance? Any of those that you or somebody else sneaked in for config tinkerers to find? What are the limits of stock realism options and will there be something extra under the hood, in a space between stock and full blown mods?
Yeah some of those do exist in the game. Part of that comes in the engine module that supports most of the ‘fancy’ stuff from KSP1 like spool up. As for new things, yeah I’m pretty sure there are some things we’ve asked for but not ended up using. I can’t really think of them off the top of my head.
What's your favorite part of the game to work on?
I really enjoy the small part of my job that’s artistic – making sketches, concept models and stuff to pass over to the team is quite fun. I also like to make the project plan go brr, ticking off things on milestones makes me happy.
Will there be hydrolox fuel type given how we already have hydrogen as a fuel type for nuclear engines?
If we get the NERV-US in that will be a need for Hydrolox there.
Are there going to be more design challenges implemented with more fuel tanks and such? E.g. will there be fuel tanks that don't have a centered COM?
Fuel tanks are a basic component of ships that we don’t want to have players need to manage too much. There are some interesting trades about that for far future fuel types though. As we get there we’ll examine if they’re interesting to support or whether to leave it to the modding community.
Are there plans for adding nostalgia/legacy parts? aka, adding some revamps of the KSP1 parts? I mean, some old users would be delighted with these.
I’d argue that anytime we have a part that comes from KSP1 it is already a revamp, so I’d be interested to understand what that actually means to you.
In the upcoming Science update - does conducting experiments give you science points? Will there be a tech tree?
There will certainly be a tech tree, and science points!
For colonies, do we know if/how lifesupport will work? Simple colony expansion or more complicated management of individual resource routes? Will users be surveyed for whether or not we want lifesupport?
See answer about life support from Pthigviri.
For interstellar, will there be astronomy aspects required to detect/map the other system(s)?
Fun things for the future! I can’t be more specific at this time.
Why Quenya and not Sindarin, Telerin or Noldorin? Do you have something against Elves that went to Middle Earth? By the Ninth, I must know the answer.
The real answer is that the corpus of Quenya is a lot more complete than say, Sindarin, so when I went to try to learn it, that’s where I went.
What real life concept / scientific work gave you the most headache? Is there something you are really proud of, that your creations will introduce to players?
Heat and radiation are the hardest concepts to map to gameplay, so I’ll say those.
Every time we get a system that is showing a new scientific or engineering reality I get excited. Example - with 0.1.3’s new extensible engines, we’re showing the community that doesn’t follow aerospace precisely than extending engines exist and are useful in some ways.
Are we going to get some engines like the Orbital Maneuvering System from the Space Shuttle because the current (and only) monopropellant engine is not very liked among the community.
The Puff is pretty OMS-like. I’d turn that around and say that something more conventional in terms of attachment modality is probably more useful than something that tries to ape the OMS a lot.
When will we see more interiors for the command parts?
We want to fully define the IVA system and experience before we commit to more interiors so we limit possible rework.
Will the team add RCS to the space shuttle front cockpit section eventually?
This is not planned.
How do you go about balancing new engines with twr/isp/cost/size/etc?
Check out the Engine Archetypes dev blog for the framework – but the overall concepts we use are related to…
· Spreadsheeting versus comparables,
· Looking sneakily at how mods have done things when possible,
Follow up question, with the full 1.0 tech tree, aside from cost/resources, will there be a reason to still use the basic methalox atmo/vac engines we have now, over newer engines/fuel types?
Resources accessible to a colony will drive this. Say you’re mining a frozen ice ball of a planet with water ice – that’ll be something that would drive you to hydrogen engines. However, maybe you’ve got a colony on a world with trace atmosphere of CO2 – that might make methalox attractive.
I routinely exceed 150 parts for spacecraft in KSP 1, would the team consider a higher baseline for the “typical” vehicle? Do you have stats on how many parts players use for their EA KSP 2 craft?
We are building our analytics pipeline to give us that data. We have lots of legacy data from KSP1 to help us in the meantime.
So someone in the KSP2_general channel have pointed out that the "brass line" vacuum engines in KSP2 have some resemblance to your previous modded content as Nertea.
How is the process like with implementing these similar designs into KSP2? Do you do it entirely by yourself, texturing and all? Do you do 3D models, coding, or maybe nothing? You just manage the team to do it?
I do very little of those things. Effectively I…
1. Try to incept the concept and discuss its utility with the rest of the team,
2. Make sure we can support it with the engineering that has been done,
a. There’s a whole side thread about when we need to ask for new gameplay functions.
3. Make concept models,
4. Hand it off to the art team,
5. Coordinate other things we might need for the model – VFX, SFX, animations,
6. Come back once we’ve got all that sorted and do the final integration into the game, and some tuning later on.
If you as a team manager delegate others to recreate your parts, how does it feel to let others rummage with your own engines?
To be clear, we’re not really recreating parts – when things are similar, there’s often just convergent evolution. But our art team is equal to the task!