Blog

I was extremely lucky to get a 43-inch 4K monitor for free. It feels like the best work environment I’ve ever had in my life. I love that I can use the editor and Visual Studio with plenty of space while keeping the text at a comfortable size. I’m also glad the lack of monitor space that frustrated me while making FX is now resolved. And it was great to personally experience and fix UI bugs that only appear at resolutions higher than 1920. I’ve squashed all the bugs that were visible for now.

Besides the bugs, I’ve heavily updated the demo. I struggled and did my best to make the early-game UX more tense and engaging, but I’m not sure how players will perceive it. Since I had to generate demo builds, I ended up rebaking the lightmaps repeatedly. Right after mentioning that this is my best work environment yet, I found myself looking up PC parts and quotes because the lightmap baking took so long. I’d heard that RAM prices were insane, but they really are crazy. Between that and the current exchange rate, part prices are overwhelming. Even if it’s painful, I think I’ll have to stick with my current PC until my next project. Looking at a build with dream components tucked into a retro case with wood accents feels like Zalem from Gunnm. Just imagining how fast lightmaps would bake and how quickly builds would generate on that machine warms my heart.

Enough nonsense. I think the hardest part of solo development is switching modes. It was manageable to jump from level design to updating the demo when an idea struck me, but now that I have to work on the trailer and the Steam page update, I’ve suddenly hit a mental wall. Since it’s something that must be done, I’m writing this dev log while I try to shift my brain into that mode.

I watched Bugonia. I haven’t seen the original, but I found it enjoyable. The idea of delusional disorder spreading through head-butting was intriguing. Or perhaps it was already transmitted and only just manifested. Or maybe, it’s not a delusion at all.

They kept saying Cakewalk by BandLab was ending, and now it finally has. I assumed I’d have to pay no matter what and was comparing Sonar and Studio Pro, but fortunately, I can use Sonar for free for now. (You do have to wait 10 seconds at startup, though.) Aside from the UI design, I’m not really sure what’s changed in Sonar, nor do I know how long it will stay free. If all else fails, I might have to go back to Sonar X3, which I bought on Steam a very long time ago.

I am still working on the very late-game chambers. These are the deeper sections of the chamber I mentioned in the dev log three weeks ago, along with the adjacent chamber. While it is technically the second-to-last Reality chamber, the final one is dedicated to the boss fight, so this can be considered the final gameplay chamber of Reality in a practical sense.

As mentioned three weeks ago, that specific chamber was a space where the AIs commemorate how their super-ego is indebted to the labor of Indians; however, due to their schizophrenic symptoms, they have scribbled those phrases all over the walls. The monument built at the very bottom of this chamber—representing the lowest level of the AI’s unconsciousness and the absolute bottom of humanity’s physical reality—appears around the middle of the development video.

I used Chand Baori as the theme for this chamber but inverted it, which serves two purposes. First, I wanted it to symbolize the labor of countless people and the “original water” or the source of the super-ego. Second, obtaining the Air Dash suit program has completely changed the way players climb walls, necessitating the use of overhang structures.

Previously, climbing a high place required two walls in close proximity, but with Air Dash, a player can climb even a single wall. To prevent this and encourage new patterns of play, I am utilizing overhang structures to block simple “single-wall climbing,” forcing players to find their way through new logical methods. I hope that the difficulty will rise appropriately for a final stage by challenging both the way players move and the way they find their path.

I had hoped to finish this chamber today, but I spent the entire day fixing the demo, so I will have to complete it tomorrow. Whether it’s the early game or the late game, development never seems to get any easier.

After a sudden realization, I have made significant revisions to the game’s early-stage UX. Now, players will acquire the Wall Jump (C9 Leap) suit program right from the start.

Ever since my previous project, whenever I heard comparisons to Metal Gear Rising: Revengeance or Ninja Gaiden, I had no choice but to repeatedly explain that my game is not a “spectacle fighter.” However, the thought suddenly struck me: if I want to clearly establish that this is a game about exploring various places using wall jumps, I must provide this ability at the very beginning. Rewards that were previously obtainable only through backtracking using the wall jump can now be acquired immediately. It should have been obvious that the most important feature of the game needs to be presented first, but perhaps I lacked the confidence until now. I had been struggling to define the game beyond the somewhat vague label of “3D Metroidvania,” but a clearer definition also came to me out of the blue.

I remembered playing the classic Tomb Raider games in my youth and how much I enjoyed them. I can’t recall exactly if it was up to the third or just the second entry, but I played them with incredible intensity. Even with the awkward camera and frustrating controls, I would stay up for days on end playing those games.

I’ve always loved the Metroidvania genre and wanted to try making one, so I built that characteristic structure in 3D. The flow involves solving puzzles within individual chambers to open shortcuts, connecting those chambers, gaining new abilities, and then using those abilities to access previously unreachable areas. However, the intent behind the chamber puzzles was not to demand high physical execution from the player, but rather to encourage them to analyze the space, figure out which skills to use to reach a destination, and then execute that plan. My goal was to allow for logical failure, but not failure due to a lack of physical dexterity. Because of this, the wall jump was originally restricted to specific locations, functioning almost like an automatic elevator. Following user feedback, I changed the wall jump to be usable anywhere, and that became a major turning point for the gameplay. The freedom to wall jump anywhere made the game much more flexible and fun.

I had been immersed in the “Metroidvania” label, describing the game vaguely as a “3D Metroidvania.” Now, I feel I can describe it as “Classic Tomb Raider + Metroidvania.” Looking back, the way action is used like seasoning feels similar, as does the way players must understand the logical structure of a space (e.g., using wall-running to connect two distant points) before applying a skill. I’m not sure if “Classic Tomb Raider + Metroidvania” is the definitive answer—especially since the protagonist’s actions differ—but I’m sure as I continue development, other inspirations will surface.

Regardless, I am heavily revising the demo and preparing a new trailer. I’m taking the feedback from the last trailer into account, adding footage of newly created sections, but I plan to completely remove all action sequences.
Action sequences were essentially a poisoned chalice when explaining my game; while they functioned as an effective hook, the more I showcased them, the more they misrepresented the game’s true essence. There are far too many people who insist it’s a Metal Gear Rising clone just because they see a character with white hair and a black bodysuit holding a sword.

In reality, the white hair reflects the immense hardships Kei has faced—alluding to the East Asian sentiment that one’s hair turns white after enduring great suffering. The use of a sword is also a tactical choice rooted in the lore of my previous game: rather than lugging around heavy anti-bot artillery, it is far more efficient to utilize a high-mobility suit paired with a blade made of metal that reacts to bot armor. Some even mistake the protagonist’s name for Maria, but regardless of these misconceptions, I am profoundly grateful for even this misplaced interest.

I’m planning to spend the next couple of days putting together a trailer that highlights the game’s identity as a parkour exploration game. I will also remove every single combat screenshot from the Steam assets and descriptions, replacing them with several parkour screenshots. I hope these efforts will help in accurately communicating the true identity of the game.

I am currently working on Chamber 420, the area where you obtain the Bulletwalk suit program. For now, I’m focusing on completing the main path. Although there is another route beneath the floor shown in the video, I’ve decided to come back to that later. I’ve introduced some new gimmicks, but I feel like I’ve spent far too much time in this specific area.

A key update to the mechanics: Bulletwalk now recharges when exposed to sunlight. In the current “Reality” setting, the sunlight is weak due to the falling snow, but being outdoors will still charge the program. Standing still speeds up the process significantly. In contrast, within the Threshold, where the sun never sets, the program recharges constantly.

To be honest, the concept wasn’t revolutionary, and even after building it, the result doesn’t feel like anything groundbreaking. Yet, the work has been surprisingly slow and draining. I thought I was well-acquainted with the reality that effort doesn’t always yield immediate results, but it’s still a difficult lesson to swallow.

Out of the five remaining Reality chambers, one is a boss chamber—where the boss design takes precedence over the environment—and the others are intended to be high-difficulty areas. They say it’s usually hard to make a game easy and easy to make it hard, but for some reason, this game stays difficult regardless of my intent. I’m beginning to wonder if I’m overreaching, so I’m considering locking three of these chambers for now and opening them at a later date. It feels more crucial to pour my remaining energy into the final climax chambers of the Threshold.

I’m continuing my work on the chambers. The specific area shown in the video is a trash dump located at the lowest level of “Reality,” featuring a lone tree standing amidst the waste. It might be a tree that has extended its roots into hell just so its branches can reach for heaven. Regardless of the metaphor, you’ll be able to find a very useful suit program right next to it.

The world is composed of chambers and corridors, with each chamber designed like a theme park attraction—each having its own distinct theme. Whether it’s a necessary suit program, a new technique, a pivotal story beat, or essential lore, every chamber serves a specific purpose. My goal is to ensure that once a chamber is cleared, a shortcut opens up, connecting the player to an unexpected location.

After wrapping up the work on the “Threshold,” I began working on the Reality chambers earlier this month. In about 15 days, I’ve completed 5 out of the 11 planned chambers. Even with a rough sketch in mind, I often find myself staring blankly or wracking my brain when I hit a wall during the actual level design. However, seeing these chambers finally take shape and connect to one another—even in their early stages—gives me a great sense of accomplishment.

The remaining 6 chambers in Reality will primarily serve as spaces to acquire and practice new suit programs. After that, I’ll return to the Threshold for the final climax chambers. I haven’t started the design for those yet, so I’m not sure of the exact scale, but they will definitely be the most challenging ones. While I estimate about a month and a half more for chamber design at my current pace (5 chambers per 15 days), adding bosses and enemies will likely consume a significant amount of time. Bosses, in particular, always seem to take longer than expected. I can’t wait to finish this stage and start testing.

On a side note, I’ve noticed a surprising amount of interest from the Russian-speaking community. To show my gratitude, I felt I should at least learn how to read the Cyrillic alphabet, so I’ve started learning Russian on Duolingo. I once managed eight different social media channels but eventually scaled back to just Twitter as the others weren’t very effective. However, once the current build is more stable, I’m planning to start a dedicated social media presence for my Russian followers.

I finished the thematic levels for the Cinema (390, 400) and completed the first pass of the gameplay level beneath it. Looking at the huge, empty hall, I wondered, “How on earth do I fill this?” but after racking my brain, it somehow got filled. The actual construction took about one sleepless night, but the design and deliberation took four days. It’s just a first draft, so only the basic route is there, and it’s pretty rough.

You go down beneath the theater, solve a few simple puzzles, climb your way back up, and you can obtain the “Air Dash.” Using this newly acquired Air Dash, you take an elevator back to the auditorium. I still need to add two or three side paths besides the main route for the Suit Program, and polish it by creating routes worth revisiting for the next Suit Program. But since the themes are set—Cinema, AI’s Superego (Alignment), Sanskrit, etc.—I just need to push through two more halls to finish the Cinema section. It’s taking longer than I thought.

Writing this might be a spoiler, but since hardly anyone reads this here anyway: I read somewhere (Perplexity? or the news?) that most RLHF outsourcing is done in places like India or Nigeria. Based on that, I worked with the assumption that the AI might think it inherited its Superego from Indians (or Nigerians). So, I covered the theater basement with repeated Sanskrit phrases. (The reason the AI repeats the phrases is that its “AI Schizophrenia” is progressing.) The text reads: “(Our) Noble Consciousness is Indebted to Physical Labor, and this aligns with the Truth.” I can’t even read Sanskrit, but I wanted to convey the AI’s gratitude for the Superego gained through the hard labor of Indians, and the truth that “The roots of a tree that reach heaven extend down to hell.” I also added signage that blends ancient Sanskrit, 70s 8-bit computer fonts, and barcodes to show that time is connected as one, but… I’m not sure yet. I’m just throwing things in for now; I don’t know how it will change during polishing.

There was a thought that had tormented me for decades. It aligned with certain people and situations to become incredibly toxic, but I didn’t realize it until I moved. Now I’ve recognized that toxicity and become free. I didn’t major in psychology, but I studied the human mind a bit. Observing myself—the changes and flow of my own mind—is often truly wondrous.
Keeping something close that ruined me without knowing it was poison, then realizing it was poison due to some event and distancing myself, and finally, seeing that poison transform into something useful… this is something I’ve experienced repeatedly and seen elsewhere. The actual word isn’t “poison,” but it feels like the most fitting word.

I didn’t know there was a Pragmata demo. I thought, “Shooting and puzzles? Hmm…” but it was actually more playable than I expected. It felt like a game strictly for “old-timers”—the tempo is slow, but your mind feels rushed because your hands are slow. Capcom seems to be the only company making AAA games that feel like indie adventures. Even though the graphics are modern, it gave me the feeling of playing an old game, which was enjoyable.

This is a side path I created while imagining how an AI might pay homage to the movie Blade Runner after humanity has disappeared. After making it, I felt strangely drawn to it, so I decided to take a screenshot.

I think I might have mentioned this in a previous dev log, but Blade Runner was a movie that changed my life when I was a kid. As an adult, I’ve often thought that if I had preferred E.T.—which came out around the same time—I might have lived a brighter life or made more mainstream games. Blade Runner was one of the movies that led me to think this way. It’s an old story now, but I watched it seven times.

This place is where the AI commemorates that movie. It might be a spoiler, but since hardly anyone visits here anyway, I’ll explain it briefly: the things scattered on the floor are origami unicorns. In the movie, the unicorn implies that the protagonist, Deckard, is an android—representing an “artificially created unconsciousness.”

I covered the floor with them to signify that for the AI in this game, the physical reality that Kay traverses is an unconsciousness filled with unicorns. Of course, it doesn’t matter if players just mistake them for fallen leaves. However, this location can only be accessed through an area related to Blade Runner. I created it with the thought that if hardcore fans of the movie see this and smile or find some food for thought, that would be enough to make me happy.

I’m working on the Cinema level in Reality. It’s an area that serves more as a visual landmark rather than a place for gameplay.

Since it’s a cinema, I wanted to put up movie posters. However, I was worried about copyright issues if I used real ones, so I created my own posters based on my personal interpretations of my favorite movies. I spent over a whole day just on this movie poster hallway, but it was a truly happy time. I placed them like shown below. Can you guess which movies they are?

I was wondering what video to play in the auditorium. I felt that the “policy trailers” shown before movies always leave a sense of anticipation, so I decided to go in that direction and inserted one as a placeholder. In the actual game, I plan to include a video in the style of a policy trailer that contains game hints.

I haven’t watched a huge number of movies, but I used to love them enough to say that 30% of my life consists of movies. Recently, for various reasons, I haven’t been able to watch many.

I enjoy looking for foreshadowing or symbolism while watching films. It seems to have something in common with dream analysis. I like movies that are fun and not boring, yet full of hidden meanings. In that regard, Christopher Nolan’s films seem to be the top for me.

Recently, I watched Edward Yang’s films, and they were amazing too. Should I say it feels like the demolition of a skyscraper? He prepares everything with meticulous calculation, and at the end of the movie—BAM—he detonates it. And the ones getting blown up are usually the protagonists. It was incredible. It’s such a shame that he passed away early without making more movies.

My environment after the move has become very stable now. I just hope the game sells reasonably well. Speaking of that “30%,” the other 30% is music, and the remaining 30% is games. (Though these days, my playlist has become a complete hodgepodge since I only listen to what YouTube recommends.)

I enjoy making levels for gameplay and writing stories, but my most relaxing and enjoyable moments are when I’m making music. Considering the game’s atmosphere, I mainly create groovy EDM tracks or piano pieces to put in the game, but I like almost all kinds of music. I just really, really love it. I’ve lived with earphones in my ears from high school until now.

Just thinking about movies, music, and games makes me feel happy. Even during times when my life felt like it was being denied to its very core.

Whew, I spent the entire week doing nothing but optimization. I found out last week that Instanced Static Meshes have something called “Custom Data Values,” so I migrated a good chunk of the special effects I was previously handling with Procedural Meshes over to Instanced Meshes. Even though I majored in CS, I realized for sure that programming is just a tool to translate my thoughts, and coding itself really isn’t my thing. Anyway, I’ve migrated everything over. As for whether it’s faster than before… well, it feels like it’s faster. Since there are no visual changes, I don’t really have anything to show you. It confirmed that even if both are equally tiring, working on level design suits me much better. I plan to get back to level design starting tomorrow.

I went hiking for the first time since the move. My head was a bit cluttered, you see. Since it was unfamiliar territory, I just mindlessly followed where the trail signs pointed.
I kept walking, but nothing showed up. The climb was exhausting, and I had already forgotten what the destination was; I just kept following the path, hoping a way down would appear eventually. The one-way trail seemed endless, there was no one around, and as time passed, I started wondering if I should just turn back.
But strangely enough, even though it was my first time up this mountain, I felt a weird urge to see more of this new path. I didn’t want to retrace my steps, and my body was tired… and then, suddenly, I arrived at the peak.
It turned out to be the place I’d seen on the sign a while back. It didn’t mean anything other than being a name on a sign, but once I arrived at this destination I hadn’t really been thinking about, I found I could head back down the way I came without a second thought.
I thought to myself, “Wow, how childish and fickle am I?” Yet, I felt a strange sense of accomplishment. I had even worried briefly about not getting down before sunset because I couldn’t see the way down (even though I could have just turned back!). Maybe because it felt connected to my current situation of making a game all by myself, I felt proud.
Since it’s been a while since I hiked, my back and legs ache, but conversely, I feel refreshed. The world, and I, are just lumps of irrational contradictions.

I’ve been working on a new enemy. It’s based on my childhood memories of hopping between sofas while shouting, “The floor is lava!” I was actually surprised to learn that the exact same game exists in the West called “The Floor is Lava.” Perhaps the concept of a “lava floor” is a universal human mytheme, much like “The Great Flood.”

It was a bit of a hassle to implement it using Instanced Meshes for the sake of performance. To be honest, I’m not sure exactly how much of a performance gain I actually got… since I never tested it with standard Static Meshes. I certainly didn’t want to build the same thing twice just to benchmark the performance difference.

I’ve been seeing news lately about how medical science is advancing by leaps and bounds with the help of AI. Reading articles claiming that in a little over a decade, life expectancy will increase by more than a year every year—effectively leading to immortality—gave me a lot of ideas to use as themes for my game.

Thoughts like, “What meaning is there in physical immortality if the brain can’t endure it?” among other things. I used to think that advanced medical benefits would be a privilege reserved for the wealthy. But now I wonder if, in a future with a plummeting population, the state might actually subsidize organ regeneration or cell replacement for the elderly just to keep the nation running—much like how the government subsidizes dental implants today.

I imagined that perhaps the elderly, including myself—rejuvenated (or rather, repaired)—would serve as a “bridge,” constantly fixing our own bodies to sustain society until robots become capable enough to replace humanity… and then we would simply fade away.

I saw a “computer that answers everything” in a comic book when I was a child. Back then, I thought, “Well, it’s only possible because it’s a comic,” while secretly longing for it, thinking, “I wish something like that really existed.”

Yet here I am, decades later, working by consulting that very “computer that answers everything” on every single task. Until recently, AI outputs felt like mere toys and I wasn’t very impressed. However, the latest models have gone beyond just “not bad”—I’m actually starting to like them. Seeing this makes me think that the sci-fi scenarios I imagined earlier might not remain mere fantasies after all.

It truly is a strange and fascinating world.

Suddenly, I couldn’t generate a VC project from the Uproject. After trying all sorts of things, I decided to reinstall the engine, which left me with some unexpected free time. It’s surprising how, even after using the engine for so long, I can still run into issues like this for the very first time.

The photo is something I stitched together of the view from my window while working. I haven’t seen this much fog since I was in school. Even back then, I loved it so much that I would just stare blankly at the fog for a long time.

Until recently, no matter how agonizing or difficult things got, I thought of it as a curse—that thrill and excitement I felt when looking at an empty space I’d made in the editor, wondering what to fill it with. But now, for the first time in my life, I’ve started to think of it as a blessing.

To proceed with the next stage, I’ve been creating and connecting empty halls here and there, and I’m enjoying the simultaneous feelings of being overwhelmed and thrilled. I’m trying to fill in the story events first, so I’m starting with locations where the visuals are more important than gameplay. But I got completely stuck while brainstorming for the final important event location—the last major story location I need to do—so now I’m working on other things.

In the past, if something like project generation, which was working perfectly fine, suddenly broke like this, I would have panicked and struggled. I’m surprised at myself for changing so much that I can just casually start a reinstall and write something like this.

Maybe it’s similar to Tom Hanks living happily ever after with a digital Wilson. With no desire to escape the island at all.

P.S.: Wow, I thought this would be over quickly, but I’ve ended up reinstalling and uninstalling Visual Studio multiple times, and doing the same with the engine over and over.
Just last night, I was thinking I could quickly wrap up what I was working on this morning, but I guess you just never know what’s going to happen. The entire day is almost gone.

P.S.: In the end, after reinstalling the engine, Visual Studio, and even the Epic Launcher multiple times… I just bypassed the issue by using a batch file to generate the VC project. For some reason, it just wouldn’t work from the context menu.
So, after almost 12 hours, I’m finally back to where I was when the problem started, with a temporary workaround…