Discussing the Details before Full Release on March 7th, 2025
I enjoy even the most obscure indie flight games. Where taxiing an F-22A through the crosswalks in Akihabara and flying through a psychedelic landscape of hands with waterfall fingertips are normal. And yet, Exosky by Elevons LLC somewhat raddled my brain. Something about the formula of this game and a few things I learned about it while doing research brought up a lot of questions in my mind.
Fortunately, the solo developer behind the project agreed to an interview that has landed perfectly in the middle of Steam Next Fest: February 2025 (February 24th through March 3rd) where the demo for this game is available for download. The demo also received a considerable update just before game release on March 7th, 2025. What better time than now to discuss it all?
Thanks so much for making time for this interview. We have not spoken much up to now, so I am happy to have the chance to get into a more detailed conversation.
Hey! I’m Jordan Elevons. I am a professional 3D Designer in the Boston and New England areas.
From some of my initial research I see that your background in working with 3D software and related 3D technologies is rather diverse. What are some things you have done with your skills that you are especially proud of?
The things I'm most proud of are projects that can help others. I created a low cost 3D printed prosthetic leg and worked with dentists to make surgical models for training college students. I also designed an airplane cockpit which was really fun and a great learning experience!
Oh, a cockpit? Was that for a training course?
No actually it was for a light aircraft company in the UK called the eGo. They were gracious enough to give me the opportunity and working with them directly really helped me learn a lot about design. It’s a single seat aircraft meant to comply with the UK’s version of ultralight standards. It was also where I first realized that video game modeling tools could be used for real world objects and that any distinction between tools for physical manufacturing and digital manufacturing was completely arbitrary, as I built the designs for the cockpit in Modo, a digital content creation tool.
How did you decide to pursue a career in this field? Was this something you decided to do before you went to university?
3D Design has been what I've done since as long as I can really remember, but I recall that it really swung into gear around 1998-2000 when I was 11ish years old and Freespace 2/Homeworld/IL-2 Sturmovik all came out around ‘99. These games informed and were informed by an obsession with building spaceships and airplanes. So I learned how to make 3D assets for those games and later did some freelancing for it. I also at the same time was learning about CAD through designing spaceships for 1,000 Planets, a startup that was trying to get funding to compete for the Ansari X Prize.
As for your second question, not really no. I went to college originally for New Media Design, because I wanted to make a living and game design even then was fraught with instability and low pay. The program focused on web design, 3d animation, digital photography, etc. I learned so much about so many different digital tools during that program but about halfway through I decided to change majors. New Media was focusing too much on programming websites and advertising, so I joined Industrial Design and learned how to design for manufacturing, hand building models, mass manufacturing techniques and a lot more. Together I think the skills I learned between these two majors constitutes “3D Design”.
The about me page on the Elevons website goes into detail on your education and business ventures, but I wonder, when did you start considering working on video games?
I got my start modding games like Freespace, Unreal Tournament, etc. so I was always kind of doing it. I've always tried to avoid working in the video game industry because it's such a meat grinder so it's ironic I ended up doing game development full time.
In the first Exosky teaser trailer it mentions that the game was designed by a ‘real pilot’. That made my ears perk up a bit considering how fantastical Exosky is. What is your aviation experience?
Ever since I can remember I've been fascinated by aircraft and aerospace. Some of my earliest memories were of this red Waco biplane that offered aerobatics ride on the island where I’m from (St. Croix, USVI). I loved doing that and did it every year on my birthday for a couple years in a row. We did hammerheads, rolls, zero-Gs and more. So that really cemented it as the coolest thing to do ever. When I was 13 I was fortunate to start pilot training, though I didn’t pursue it seriously until I was 16 and obtained my PPL when I was 18. I’ve only got about 150-200 hours, mostly logged in the Diamond DA-20 and the Cessna 152. Other types I’ve flown have been a Cessna 175, a De Havilland Tigermoth and an American Champion Citabria. Fun fact; the .aero domain of the website exosky.aero can only be obtained by people involved in the aviation industry, whether as a pilot or operator in some other capacity.
Thanks for the fun fact, I did not know that. Of the aircraft you’ve flown, do you have a favorite?
I think the Citabria just because of what I got to do in it. I really enjoyed the Tigermoth but I think I would have to pick the DA-20. Not only is it the most familiar but the wraparound canopy frankly gives the best views!
I appreciate that even with your real world experience and professional stance towards aviation you still allow yourself to embrace creativity around it. Before working on Exosky, did you have any other game development projects?
The biggest two would probably be Battlestar Galactica: Beyond the Red Line and Angels Fall First. The former is a space sim dogfighting mod for Freespace 2 where I did the flight dynamics and gameplay balancing. The latter is a sort of Battlefield-in-space where players can fight on the ground, on foot, board spaceships, etc. and for that I made some background props and weapon models. Both of these were back in high school/early college. When I switched to Industrial Design my game work dropped off but I did keep modeling and took contracts for a few other games out there.
During our research about Exosky, there were a few references to a game called “Yesplane!”. Is this the predecessor of Exosky?
Yesplane was the predecessor, yes! The original alpha included the basic flight model, some of the aircraft and a quick deathmatch dogfighting mode. It was named by one of my nephews. I had brought an RC plane to visit and I asked him if he wanted to name it, so he said “Yes!”. I asked him what we should name it and he said “Yes!” and I liked it so much I wanted to use it for my game, Yesplane!
How did the concept between Yesplane and Exosky change?
The biggest thing was that Exosky was supposed to be just one of the game modes of Yesplane, where the primary game mode of that game was aerial laser tag. I also wanted to incorporate a bit of sportsball, like Rocket League. However when I really got into developing the game I realized that all of that was too much and scaled down the scope significantly. Lots of stuff stuck around though; for example, the ability to fly in a custom atmosphere was because the game was originally supposed to simulate this sport in the clouds of Venus and so we needed a custom atmospheric model for that.
This is a favorite question of mine. Why did you choose to make a non-combat flight game?
That’s a good question and it has a couple answers! The first and most important is because it’s the game I wanted to play. I grew up playing IL-2, MS Combat Flight Sim 2 and MSFS. The former taught me how to fly and the latter helped me practice aircraft systems and procedures while I was getting my pilot’s license. However while I consider myself a good pilot I was never very good at gunnery and I found civilian flying to be too boring. So I created Exosky as the in-between game I always wanted; a game that was about flying skill that would be a pure flight challenge.
From a larger standpoint games in general are pretty stale. We have this magic box that can create immersive experiences for literally anything we can imagine. You could be a cat, or a fly, or as big as a planet and experience what all of that is like. However for the most part games boil down to “enemies vs. you”. It’s definitely harder to create non-combat based games because it’s harder to find that challenge, but I really think it’s a worthwhile pursuit to be more creative in games and gaming objectives and really stretch our imaginations and experience more. The universe is infinitely full of possibilities, so when I design games I want to challenge myself to find unique games that aren’t just about shooting something else. For example, I’ve built a small game for HTC where the player is a bored teenager and the objective is to graffiti a space station. Stuff like that - there’s so much potential when you get beyond filling the level with enemies.
The internet loves cats, so I have to ask. Is Norton, the artificial entity, based on anyone's cat?
Yes! Norton was my cat, who passed into Valhalla in 2022. I got him right out of college when I walked into an animal shelter in New Jersey and said “give me whichever cat you want to go to a good home”. They pointed me towards him and after that it’s history. My spouse actually says that Norton was the reason she kept hanging out with me when we first started dating, since they figured anyone who fawned over his cat like I did must not be all that bad.
Aerobatics in an arcade flight game is all well and good when you can just throw the aircraft around the sky with little worry, but Exosky has a rather detailed flight model. Can you tell us some more about it?
Sure, so the flight model was originally built by a contractor and then improved on by me; I added double precision, some extra features, atmospheric transitions, etc. The double precision was necessary for some very fast aircraft, which would generate some large forces for their size. This ended up with force numbers drifting over time or spiking to extreme values at different points, since they would change so much between each physics tic. Tracking them as doubles meant higher precision so we could see smaller fluctuations between each tic and keep the simulation more stable.
The atmospheric transition was me taking the original atmospheric code and enabling it to be applied inside of a volume as well as in the entire level. I have a main “Aircraft” component that gets initialized at the beginning of the level and gets all of the atmospheric data from another system. The transition happens when an aircraft enters a trigger collider, that collider basically jumps in and temporarily overwrites the atmospheric data that the aircraft is reading from. On trigger exit, we replace the new values with the original level values and keep flying.
As for the specifics of how the model works, we take each wing component and sum up the lift forces for it and apply that to a rigid body within Unity and I designed it in such a way as to be easily editable; just slap a plane down in Blender and export it with my plugin and you will have a wing. It’s very fast to concept out an airplane by tugging on vertices. The tool also lets you build the fuselage and then it will calculate the frontal area for you for drag calculations.
That being said it does have quite a few compromises, as I’m sure any sim pilot out there will feel when they play the game. The wing does not take spanwise lift into account and so each wing section force is summed individually. I don’t take the airstream into account either, so blown wind (propellers, jets) will not have an effect on nearby surfaces. Transonic compression is very simply modeled and very large aircraft are not feasible without dropping the physics tic down to a very low amount - something I decided not to do for performance reasons. Post-stall behavior is also something I’m constantly tweaking. I really want to get the sort of violent stall behavior that I saw in IL-2 but at the moment it’s too gentle for my tastes.
The changes of flight characteristics when entering some of the out of this world atmospheres in the different levels is a nice touch. How did you come up with these unusual levels made up of different atmospheres and… computer graphics cards?
Growing up my dad owned a computer repair store and I used to spend ages tracing circuit lines in miniature cities. My father really hated my interest in computers, ironically, so he would only begrudgingly get me any sort of equipment. This meant that I grew up finding old parts laying around the shop and trying to piece together a gaming PC from them. I got very comfortable and familiar with the literal nuts-and-bolts of computer hardware and my dream some day is to host public computer classes where I just toss an old motherboard in front of people to demonstrate just how sturdy computer parts are and how you don’t have to treat them with kid gloves.
When I really got into defining the theme of Exosky, I decided that instead of making a game set in the future, I would make a game that someone in the future might play. That’s when I decided to do all these wacky levels and the idea of flying through those miniature cities really appealed to something deep inside me so I just went for it. Some of the level concepts came from my own experience as a pilot; the Motherboard level takes place in and around a scaled down map of Fajardo airport, Puerto Rico. I had a pretty intense experience there during my training where I struggled to land due to a strong updraft from a nearby land feature. My spouse also gave me a ton of great ideas, like the Norton easter eggs in every level and the sea turtle in the Graphics Card level.
After the January 16th demo update I found myself having an amazing experience playing with a Hands-On Throttle and Stick controller more than I did with a gamepad previously. What were some of the improvements made to the flight model?
There were so many small changes but I can break out the three main ones. The first was flow-aligned drag/lift. In the previous model the wing would make the same amount of lift/drag regardless of it’s alignment with the forward velocity. So if a wing was parallel with the ground and falling straight down it would generate both lift and drag, making the aircraft very “floaty” and unable to really drop. Conversely, if the aircraft was flying in a normal configuration, it would generate the same amount of drag as it was falling down. This was obviously unrealistic so a big change was changing the lift calculation to factor in vector alignment.
The second change was changing how effective control surfaces were - you can see some of it there, but basically the biggest change was artificially boosting control surface effectiveness at low speed. This is a bit of an artistic tweak to make the planes “feel” more correct, especially when they neared the slow-speed stall flight regimes. You can see the code below;
The final big change was just optimization. Awhile back I converted a lot of float values to doubles for better precision at high speed and increased the frequency of physics ticks to make larger aircraft fly correctly. This came with the unfortunate side effect of really destroying performance on lower end hardware. So I went in and was smarter about what was converted to doubles and improved performance overall to where it was back to the pre-conversion state.
Here’s is the actual code if you’re interested:
Flow-Aligned Drag/Lift Code
Control Surface Adjustments
The documentation for the flight model is a part of helping players mod aircraft into the game. It is very detailed, which is quite helpful. Could the flight model documentation also be used by other developers to create or edit their own projects?
Sure - hey maybe they could help me? Haha…yea I think that the documentation could definitely be helpful for others and I’d be happy if someone wanted to use it as a basis for their own projects. Like I explained previously there are compromises in the model and areas where I had to fudge things to make it “feel” right (looking at you low-speed control forces), but I think the documentation does a good job of taking a lot of complex concepts and centralizing them into a spot where someone can build on top of it.
In Exosky players fly along strings of procedurally generated waypoints to complete a level. Pre-January update it was definitely harder to fulfill level requirements for unlocking new levels. It seemed like it was difficult to stray off the path and focus on aerobatics. While it does feel more attainable, I cannot say it is exactly easy. What changes were made?
The biggest change came after some tester feedback on how difficult it was to meet the score requirements of each level. This was because the requirements were a large monolithic value without a lot of direction on how to attain it.
The newer version of the scoring system breaks them out into smaller specific requirements. Not only does this guide players towards scoring opportunities but it also allows me to design score requirements around specific levels; if one level has a lot of interesting objects then the score requirements coild be heavily weighted towards Points-Of-Interest. If another level has lots of tunnels then it could be weighted towards Proximity scores. The actual scoring values didn’t change much, it was the breaking down of larger score values into discrete chunks.
The amount of Extras currently in the demo are quite substantial. There are many radical aircraft designs from throughout the years. Are any of these designs going to be included in the game in the future?
To be honest I think the answer depends on how bored I get with the existing aircraft. I’ve been toying with the idea of releasing DLC packs of new aircraft in the future, but since I don’t know how successful the game will be I don’t know if that would be worth it from a cost/benefit analysis. This means it comes down to whether or not I want different planes to fly. If I do then I’ll probably start with implementing those aircraft - a few of them are already modeled and one of them is even mostly textured. Frankly I’m really hoping some people latch onto the mod tools and make some creative new planes for everyone to try out!
The concept art, test videos and other development related content show that the development arc that led to Exosky was quite elaborate. Can you summarize what Exosky or the project before it was planned to be?
Exosky is…big. I’ve been working on it for a long time. When the metaverse craze took off I wrote an article that defined a true metaverse as “...a self-referential entity; something that is created to create itself.” In another article I wrote:“The metaverse offers us an amazing opportunity to not only imagine but also to consciously design and test different futures for life on Earth. We need to use that; great leaps in humanity were not built to sell products, they were made because we wanted to push the boundaries of what we could do.”
I’m not bound to the word “metaverse”, in fact I would prefer to not use it, but what Exosky is really meant to be is a visionary future for our species. It’s a grand social experiment; can we build a simulation of the future so compelling that it actively itself becomes that future?
Exosky/YesPlane were made up of two distinct components. The first is the flight sim, which Exosky (as released) is a cut-down version of. The original vision was to have a celebration of flight, with several different gamemodes ranging from a sort of Aerial Rocket League to a multi-level aerial laser tag simulation, where large airships would be commanded by captains using virtual reality to steer the ships and pick out targets amongst smaller aircraft which were piloted by both players and bot pilots. There’s other stuff there and I even have an unlaunched Kickstarter with the list of game modes and such.
The second component is the overworld - the simulation of a future society, one that exists in a universe where the Robotosphere (as mentioned in the game’s intro) has been made manifest. The initial release would have players navigating Venus from a top-down grid view, plotting the course of their airship across the planet as they explore, trade, mine, etc. “Combat” would tie into the laser tag experience, where players could either fly matches against each other or even bet on other player’s teams. Players could let others board their airships to hang out, trade, etc.
The social simulation aspect would be the next layer on top of that. In the original draft it was a sort of “economic system deathmatch”, where there would be different types of economies existing in parallel, including my own. Through player actions these economic factions would compete and strive for dominance in synergy with new government strategies as well.
This would test to see which really was the most effective system for this new world. We would also test if the Robotosphere itself could work as designed and more than that, what AI’s role in society could be through AI run NPCs.
Eventually, I could see this simulation becoming more and more real. If players started trading real money using these new economic systems, started using the in-game governments and groups to organize and affect change in the real world, started trading virtual goods for real ones and vica-versa then this system in time could become the actual system people live by. It would be a universe that created the conditions for itself to exist.
Years later, as Exosky nears release, how do you personally feel?
At the moment, pretty tired haha. I have a couple other projects also all due around the same time that the game is releasing. If I take a step back from that though I guess I’d say happy and proud of what I’ve accomplished. It’s kind of weird, when you work on a project this long each task sort of blurs together. Release is just another task, made up of a list of smaller tasks. It’s a big moment because the game will be out there but it’s just another spot on a timeline in the grand scheme of the game. It’s releasing on my birthday though so I am definitely going to be having a party, lol!
Pre-happy birthday to you! It is a bit early to ask this, but after early access release do you have any ideas about what else needs to be added towards making this a 1.0 release?
So it’s funny that you ask that, because the March 7th date is the 1.0 release. I remember back in like, 1999-2001 when DSL was new and many people were still transitioning from 56k modems. Steam wasn't even around until 2004. Back then games had to be ready to go out of the box because you didn't have a way to reliably patch it or update it once it shipped. Ever since then we've gotten more comfortable with unfinished games and I decided I wanted to do it old school. That's why I'm going straight to full release with no EA or open beta period. That means all the game content is in there, everything has been tested multiple times, etc. New features may come later but those are in addition to the completed spec that is being shipped.
Oh my goodness. I am so used to early access releases I made the assumption it was not a full release. Haha! Well, thank you so much for this interview. I appreciate getting into the fine details about Exosky.
I just want to say that I really appreciate this interview opportunity. I’m really excited to share my game with the world on March 7th. Regardless of how well it does it’s been a labor of love and a great learning experience and hey, maybe in the future those bigger ideas will take flight too. In the meantime, I hope players enjoy the unique flying challenges I've created and perhaps even build their own unique aircraft with the modding tools I’ve included in the game.
I also wanted to say that it means a lot to be able to talk about the project in such great detail and I am happy that I got to do it on your platform. Skyward Flight Media reminds me a lot of the old Internet, where individual websites dedicated to niche interests ruled the web.
See you in the skies! :)
Connect with 'Exosky'
Discord: https://discord.gg/nnM2cUPUW4
Website: https://exosky.aero/
About the Writer
Aaron "Ribbon-Blue" Mendoza
Co-founder of Skyward Flight Media. After founding Electrosphere.info, the first English Ace Combat database, he has been involved in creating flight game-related websites, communities, and events since 2005. He explores past and present flight games and simulators with his extensive collection of game consoles and computers. Read Staff Profile.
.png)