Developer Insights #15 - Writing for Kerbal Space Program

[intelliCom]

13 minutes ago, Geonovast said:

Let's try to stay on topic, please.  We have other places for debating the merits or !merits of the Orion engine.

You're right. I've turned it into a post on science & spaceflight, we can continue there.

[Angelo Kerman]

Nice to see that you've been keeping busy with your writing, @JustJim!

[Rutabaga22]

Very cool! The part descriptions were always one of my favorite parts of ksp!

[The Aziz]

3 hours ago, Just Jim said:

I really can't say any more on this subject than I already have, sorry. 

And I absolutely fully respect that. It's crazy how one answer creates more questions, isn't it?

[StrandedonEarth]

7 hours ago, J.Random said:

Ugh, stupid Orion is still in.

Yes, because it's just so, so.... so kerbal!

Keep up the great work, @Just Jim! Looking forward to seeing the punny humor you and the team come up with!

[Pthigrivi]

4 hours ago, The Aziz said:

Writing aside, this thread brought up an interesting thing that should be noted.

We're still having experiments. Which creates tons of questions that probably can't be answered now. But since the science system is reworked, I hope we won't have to clutter our craft with dozen of different parts only to get some experiment results from a funky rock. Then these results have to be used as something to develop new technology in one way or another, and I hope it's not "you got x amount of sciencies from these experiments which will be used in our lab to develop new tech" because that's still very much KSP1 science system.

This was hinted at way back in the Celestial Architecting vid (right around 6:30), so I don’t think Jim is too far out in confirming some broad ideas. Im personally happy to see experiments and biomes and science come back but Im right there with you on the clicky grind factor. Hopefully they’ve got some tricks up their sleeves to alleviate that. And of course just knowing that leaves a lot room for new and improved mechanics. 

Edited August 31, 2022 by Pthigrivi

[Gargamel]

2 hours ago, intelliCom said:

You're right. I've turned it into a post on science & spaceflight, we can continue there.

We already have a thread on Orion, quite  lengthy one.    Please direct the conversation there. 

[MechBFP]

40 minutes ago, StrandedonEarth said:

Yes, because it's just so, so.... so kerbal!

Exactly. Things like this and the writing is what adds *soul* to the game. Otherwise you end up with something like Simple Rockets which despite some of its technical achievements feels more like a tech demo than a game. Perhaps that is what some people want, and that is all good, but I am very happy that is not the direction they are going here. 

[Kerbart]

Thank you, @Just Jim. A lot of the more experienced forum member surely know who you were and we knew little more after you were beamed aboard the mother ship than that you were "writing," which surely left a lot to our imagination. Would there be convoluted story lines? Passion? Lust? Intrigue? Would career mode turn into a fierce cut-throat competition with Zaltronics, and the loser would be delegated to working in the zalt mines?

Reality, even in Kerbal Space Program, turns out to be a bit less prosaic, regarding the writing job. It's literally* writing. Alas! But your post is appreciated not less of it. Not just because it's great to hear from you (well, that too), but also because it emphasizes what a tremendous project this undertakings is, and why it's taking such a long time. I hope at least that there's a mission where you have to build a ship loaded with 25 lawyers to escape the solar system. I'm sure everyone at Private Division will get a good laugh at that. Well, after reading your post, I'd say almost everyone...

* See what I did there?

[intelliCom]

12 minutes ago, Gargamel said:

We already have a thread on Orion, quite  lengthy one.    Please direct the conversation there. 

Sorry for making more work to sort out. I will remember that thread when I can.

Edited August 31, 2022 by intelliCom

[James M]

I love how heartfelt and compassionate this dev blog was! 

12 hours ago, Intercept Games said:

There are somewhere around 600 parts currently planned for KSP2. I do not have the exact number off-hand, but somewhere around there.

And how funny this was 

Its nice to meet you Jim, and thank you for working on KSP!

[Klapaucius]

@JustJimIt is so great to see you on the forum in your new role. You have an amazing job, and I love the fact we are learning all these details about the process.

 

I have to say, however, that I am mighty disappointed that K.R.A.S.S.H. Industries (Klapaucius Reliable Air, Space, Sea and Hardware) lost the bid to manufacture engines.  As our motto clearly states: Practicality is for the unenlightened.  Besides, I would have thought the frog would have been a selling point.

[Vl3d]

Can you tell us anything about lore and game story? @Just Jim

[prestja]

14 hours ago, Just Jim said:

The barometer is a great example. In low orbit it would read something along the lines of "The readings are zero, almost as if it was in a vacuum." Great joke, but it repeats the same thing in orbit of any celestial body.

I'm sure all of us would understand if you left the vacuum barometer text as-is, no pressure!

[Anth]

15 hours ago, Just Jim said:

"The readings are zero, almost as if it was in a vacuum."

First time I ever laughed in KSP. It was absurd because it was so obvious

[Just Jim]

7 hours ago, Vl3d said:

Can you tell us anything about lore and game story? @Just Jim

I'm afraid not... sorry.  

[JoeSchmuckatelli]

19 hours ago, Just Jim said:

in KSP2 each and every experiment "result" will be new and different and unique to the location and situation and type of experiment. And Kerbal sounding

Whelp - sounds like you are earning your pay, then! 

Won't lie - I'm particularly looking forward to seeing what you and the team have done w/r/t 'doing science' in the game.  For me at least, that was one of the immersive elements: having a reason and purpose for sending my ship and probes to a place.  Getting to 'do a thing' in addition to the 'fly to a rock' parts of the game was fun. 

Don't know why - but the Breaking Ground DLC is what made the game for me... It brought the Kerbals back to KSP, giving them something to do beyond planting flags and clinging to the outside of a ship for an EVA report.  Hoping you guys have built on that! 

 

[Pthigrivi]

2 hours ago, JoeSchmuckatelli said:

Whelp - sounds like you are earning your pay, then! 

Yeah and it makes me curious about how experiments and biomes have been refined because in KSP1 you're at like 10 biomes per CB, landed, high and low above, etc. times 4-9 different experiments depending on whether there's an atmosphere or not, and potentially tripling or even quadrupling the number of planets + moons. With all that we're taking about thousands of individual reports. I feel like from a grind perspective and also Jim's mental health I hope there's been a bit of an effort to emphasize quality over quantity. 

Edited August 31, 2022 by Pthigrivi

[JadeOfMaar]

20 hours ago, Strawberry said:

This definitely isnt intended but the name "atomic fusion" makes me wonder if there will be a drive that uses thermonuclear bombs to have a good mix of volume density and low weight, a fusion of fission and well, fusion (perhaps as a precursor to full on fusion drives?)

Mini-Mag Orion comes to mind. Its pulse units are very low mass and offer very high impulse. The engine is a magnetic nozzle and uses Z-Pinch to implode the pulse unit, sending the fissile fuel to critical mass. The pulse unit is just a sphere of Curium-245 with a magnetic coil and ultra-light sheets of conductive material connecting the sphere and coil.

I would argue that hybrid fission-fusion isn't automatically a precursor to full fusion. Some, if not all fission-fusion drives exist to easily trigger one process by using the other as a jump-starter. In the case of bombs: you're mainly a fission process but you use a fusion bomb to kickstart it. On the other hand: Lithium saltwater is mainly a fusion process but uses neutron bombardment to kick off a fission process which leads into fusion.

A stock Mini-Mag produced by "Shake-proof" would be quite appropriate. The agency "earns its name" even more by developing it.

[Strawberry]

1 hour ago, JadeOfMaar said:

In the case of bombs: you're mainly a fission process but you use a fusion bomb to kickstart it.

Most irl thermonuclear bombs due to the relative ease of fission bombs have a mostly fissile (with some fusionable materials to act as a booster) primary core compressed by explosives to start the reaction, then uses the neutrons and heat from that reaction to trigger lithium deuteride to fise into deuterium and tritium which then fuse to produce energy. The reason for this is that while fusion is very efficient, currently the only way we know how to make more energy from fusion then we put in is through this type of thermonuclear bomb, thus a fusion primary wouldnt really make sense. Some nukes have a third fusion stage triggered by the second (mostly experimental us nukes and maybe tsar bombas), and these further increase the efficiency greatly yet add a lot of weight (if you want a horrific thing to look into, shoutout to project sundial which would be a 10000 megaton nuke that would be triggered by a 1000 megaton nuke (project Gnomon)). General rule of thumb is the larger the explosion the more reliant it is on fusion to do the legwork, tsar bomba's energy came from 97% fusion. 

[GigFiz]

22 hours ago, Pthigrivi said:

Yeah and it makes me curious about how experiments and biomes have been refined because in KSP1 you're at like 10 biomes per CB, landed, high and low above, etc. times 4-9 different experiments depending on whether there's an atmosphere or not, and potentially tripling or even quadrupling the number of planets + moons. With all that we're taking about thousands of individual reports. I feel like from a grind perspective and also Jim's mental health I hope there's been a bit of an effort to emphasize quality over quantity. 

Yeah, it'll be interesting to see what direction they go. On the one hand, lots of experiments is very true to life in some ways; part of the goal of space exploration is the furthering of scientific knowledge; I know the ISS always has a ton of experiments running at any time (I recall a video with Chris Hadfield where he said at least 50-100 at any given time, I think), though that something moreso reflected by the research labs. On the other hand, as you said, doing the same handful over and over, just in different places, gets repetitive, not to mention that it doesn't always make sense to be running the same stuff over and over; something like taking soil samples in lots different biomes makes good and true to life sense but, similar to what was mentioned, taking barometer readings over and over in different orbits over different bodies wouldn't exactly add a whole lot to scientific knowledge. Then there is stuff like the nebulous 'mystery goo' and 'materials study', whatever those may be (really just an abstraction of a huge number of different things). Plus, on the purely gameplay side, doing stuff like building a biome hopper early on to cheese a crapton of science out of the moons (especially minimus) is, perhaps, not the most compelling gameplay.

As a bit of contrasting opinion though, I thought there were too few, not too many, experiments in KSP (though I just added more with mods. Of course that would be overkill with how easy it is to unlock everything in stock, but CTT and other accompanying mods give me a reason to need all that sweet, sweet extra science), though perhaps for similar reasons, ie: gathering science feeling the same most of the time. On the other hand, I think you are right about it getting repetitive just cramming the same handful of parts into every vessel; though I do conceded that just going the route of adding a greater variety of experiments does have a cost/benefit there, because if the underlying framework of science gathering remains the same, then adding a greater variety of experiments is just adding dev time to create, model, and write all the text for them, for what amounts to a purely cosmetic benefit, really. So it's still back to needing changes to the underlying science system and gameplay treadmill for it to mean anything, really.

Of course, if was established that the existing science regime needs to be reworked or at least spruced up, then we must ask: what do you replace it with? Even if experiments are passé, we need to gather science somehow. The research lab may be more realistic in some ways, but just clicking 'start research' and waiting is even less compelling gameplay. Not to mention that fact that the much greater scope of KSP2 and the accompanying parts and tech involved should mean a much bigger tech tree (which I am totally in favor of, by the by; pump that science grind crack right into my veins, baby), which should translate into needing way more science over the course of unlocking it.

So what then? Do you lower the science cost and instead have monetary and time costs, also? Do you make gameplay elements part of unlocking? For example: getting research from specific experiments, on specific bodies/biomes, to give our scientists enough to make greater discoveries in a related field, or maybe needing a certain number of launches, with certain engines, in order to give our engineers enough data to create bigger/better ones. For colonies, it would make sense that you would need to get to a certain level of size/complexity in order to reach greater tier of new fancy hotness. Etc, etc. Then, if you do something like that, does it then accompany the existing framework, or do you ditch the 'science' currency altogether? That does have of realism advantage, inasmuch as 'science' and 'reputation' as currencies are quantified numbers drawn from nebulous, and totally unquantifiable/subjective concepts, (now, that's not necessarily a bad thing; functional gameplay always requires some level of simplification and quantification, but it is a little incongruous when you think about it, and certainly invite more interesting replacements, if ones could be found and implemented). Although, does totally doing away with it then run the risk of ending up with similar issues as what we have now, just wrapped in a different flavor of purple?

Really, though, this isn't more than just rambling and speculation, cuz they didn't consult me about it (quite rightly because of course not and who are you anyway and how did you get into my office?), so all it boils down to is: yes I hope the science system is different, and that whatever we get is super more fun/compelling/all around awesomerer/blah blah blah, etc etc.

So yeah..........

Anyway, I'm Ted Kermin (or someone who stole his identity at least) and this has been me talking a bunch

[minerbat]

On 8/30/2022 at 6:24 PM, Intercept Games said:

Another very important part of the writing process is we are localizing KSP2 into several different languages

Do you happen to need a Dutch translator

[Pthigrivi]

12 hours ago, GigFiz said:

 On the other hand, as you said, doing the same handful over and over, just in different places, gets repetitive, not to mention that it doesn't always make sense to be running the same stuff over and over; ... doing stuff like building a biome hopper early on to cheese a crapton of science out of the moons (especially minimus) is, perhaps, not the most compelling gameplay.

We've spent oodles of time talking about this over the years, and my ideas have certainly evolved and gotten looser over time. There are a lot of interesting possible solutions. I do personally think the basic structure (find new places, conduct experiments, earn science, spend science on new parts) is not really the problem and Im happy to see them try to fix whats broke rather throw the baby out with the bathwater and come up with, as you say, a completely new system with all the same problems in a different shade of purple. At this point so near release Im sure these decisions are pretty well baked in, but there are some pretty important takeaways (I think) from the many, long discussions we've had on this board on the subject.

1) Get the Grind Out: Oh, the grind. Shadowzone's most recent video is all about treating player's time as THE most important commodity in the game, and the failure to treat it as such in KSP1's science system is one of the more glaring offenders for me. Im psyched to see all the clever and witty things Jim and others have come up with as experiment results readouts, but it's vitally important to consider when to present these readouts so players can take a minute to enjoy them. Ascents and descents for instance are for most players really important, continuous experiences. You really want to have your hand on the wheel and your eye on the ball, and if the game is creating a critical incentive to take readings "near above" and "high above" you definitely don't want scroll through half a dozen or more experiment results and manually store or discard them, let alone hunt around and right click on individual parts to run things while in flight. Does that mean hotkey everything and store automatically and review only when you decide to transmit? Or is even transmission a more-or-less automatic decision? One way or another you only want to present those experiment results when the vessel is at rest in stable orbit or stationary on the surface.

More than that though is eliminating full-on redundant missions. Grind = repetition. No more biome hopping to a dozen different visually indistinguishable places, certainly no partial results that require going back to the same place twice. Treat players time as valuable and only call a biome distinct if its really worth taking the time to go experience. I think some of the things Shana Markham mentioned about the iterative back-and-forth with the art department making unique geographies and tying those into the science system are really encouraging. Above all I'd recommend quality over quantity. If Minmus had 3 main biomes and 3 anomaly specific mini-biomes that required some clever landing or rovers to reach I'd call that about right. 

2) Useful information: This is much more obvious with things like resource scanning than it is with most of the experiments we're used to, but some mods like SCANsat have managed to tap into some experiments for biome mapping, altimetry, and anomaly detection that are way, way more compelling than anything stock. You could imagine a scenario where barometer scans unlock the ability to predict reentry factoring drag, and thermometers let you see heat bars and manage radiator usage, and EM booms let you know about dangerous radiation environments. As much as possible players should see these experiments as legit engineering and navigation tools even if they had completed the tech tree. 

As far as the deeper mechanics go with labs and samples and the structure of the tech-tree itself we'll just have to see. I think if there's a genuine effort to make the process as straight-forward and intuitive as possible we'll be in good shape. 

Edited September 2, 2022 by Pthigrivi

[GigFiz]

4 hours ago, Pthigrivi said:

As far as the deeper mechanics go with labs and samples and the structure of the tech-tree itself we'll just have to see. I think if there's a genuine effort to make the process as straight-forward and intuitive as possible we'll be in good shape. 

Yeah it makes sense that this would be a topic that has been heavily discussed, in much detail, for a long time around here.

Oh god, I didn't even think about the whole 'visiting biomes more than once to get all the science' thing. Probably because I never did it, cuz, like you said, it's awful and shouldn't be part of the game. And yeah, I always did feel a little weird sending someone out to get a EVA report in the upper atmosphere; yes, it's true that you can have an orbit where you just dip down and skim through the top, but it's still...disconcerting.

One thought would be to have biome categories as a scientific meta level above the specific ones, ie: could still have lesser/greater/great/etc flats on Minimus, which could be nice for locating things and for resource searching, but the meta-category would be 'flats', which would all be the same for science porpoises. It would allow them to actually increase the number of biomes, for flavor and geolocation, but at the same time also decrease the number of places you need to gather science. And of course, little specialty locations is a nice idea. Makes exploring fun and lets it be rewarding without just being a long-ass checklist.

I do think it could be cool for there to be same variation in what different conditions matter for different experiment, like the example of only being able to get science from a barometer reading in space one time ever, full stop. Though I do see an obvious danger there: even if it was cool and added realism, if there is too much variety, the inconsistency could easily turn into a time wasting hassle as far as keeping track of what works where, and then we are back to what you said about not respecting the player's time.

RE: useful information. YES! There is a lot of good ground to be had there. There are some things in KSP that involve either a little TOO much trial and error, or just a straight up web search, that could be integrated. For example: first time sending a craft to land on Eve and wondering about parachutes? Right now: close your eye and guess or google it. But it would be so much better if instead you could send a satellite with equipment to do gravity and atmosphere analysis, maybe also have instruments that you could put on a suicide probe into the atmosphere to get useful telemetry with. Once you do that, now you can get information in the VAB to help you. Perhaps stuff like safe deployment speeds and then like; at minimum have it show how low your final speed will be with parachutes fully deployed. Just one possibility, but there is some really interesting space there to take useful outside information people might look for, and make it diegetic.

I was just spitballing, but I did like the thought I had about having some scientific/engineering advancement from actual gameplay choices. Now, there certainly are potential problems if you tied it to full on node unlocking: the last thing you would want is for a player to lag behind on/be locked out of techs they want because unlocking them requires types of gameplay that they don't enjoy, but if it was just little bonuses even, that could be fun. They couldn't be unbalancing, and it would be best to have the requirements be secret and at least somewhat procedural (so you can't just look up a guidelist on the internet), but it could add neat little unexpected, occasional rewards. Say...maybe you like doing insanely fast re-entries, so after some of those, your team might figure out a way to make your heat shields a little stronger or have more ablative capacity. Or you really like to use one specific engine, it might get some performance enhancements, or get an alternate version that can use a different fuel type. Stuff like alternative looks/clustered versions, etc, would be cool in brainstorm land, but not really not a good use of finite dev time in reality, but the other stuff doesn't seem like it would be too terribly difficult/time consuming to do.

Like you said, though, at this stage in development, the science system/treadmill is probably fully ironed out and implemented, and just being worked on at the level of tuning and tweaking. I am definitely glad to hear that the signs you have been seeing are promising. Hopefully we get something that fixes the issues with the current system, and is fun and compelling (I'd enjoy a few surprises, too)

[zer0Kerbal]

If you ever want to moonlight, have so much lorem ipsum flavor text!