fault
patreon
Vilserio from Dei; Story time with AiD's New Programmer
🕑 Added 2022-06-20 14:05:52 +0000 UTC
Now that the LIGHTKRAVTE dust has settled —
For the longest freaking time, AiD has communicated with the public solely through myself, Munisix, at least when it comes to English. This is because out of the two representatives of the group, I was the only one that can speak English. That is, until now.
Folks, please let me introduce you all to Dei-san, our new (?) programmer. It feels weird to introduce him as “new" since he's been around for 15 or so months now but I feel as though this is actually the most appropriate time for his introduction now that we have successfully released LIGHTKRAVTE.
Before I go on, I want to reiterate that I did not ask him to write this article. Dei-san took the initiative, chose to introduce himself and share this brief history of what was happening behind the scenes at AiD for the past several years so we (AiD and the community) can all be on the same page once and for all. That’s the sort of above-and-beyond tenacity and integrity that gives me confidence that I can start sharing plans and scheduling for the future to the public.
But I digress.
Folks, let me introduce you to Dei, David Teng, the final piece to our AiD puzzle.
—Munisix
Hello, fans, supporters, and patrons of ALICE IN DISSONANCE (AiD).
I’m Dei, the latest addition to the AiD team and the person who handled the programming for our latest game. If you’ve been hanging around our Discord channel, you may have noticed the presence of a new purple user up at the top of the members list. A purple user who, for the past year and a half, has said next to nothing and made no attempt to engage with the community. Some of you might have been wondering what in the world was going on with this Dei fellow. Was the person just not sociable? Unfamiliar with Discord? Did AiD add a bot to their team to make themselves look bigger?
Well, I’m here today to set the record straight. I was not, in fact, trying to appear mysterious and unknowable. I just wasn’t sure what I was doing up there at the top of the list either.
But with the release of our newest title, fault - StP - LIGHTKRAVTE, I figured it would be a good time to introduce myself to everybody, and tell the story of how I came to be a part of AiD. However, my personal involvement is just one of many chapters in the longer arc of AiD’s trials and tribulations over the past few years, and I felt it would be better to tell the whole story, starting from the beginning. Or as close to it as possible within the constraints of feasibility. After a discussion with Munisix, he graciously agreed to be interviewed (by which I mean a meandering conversation over Discord during which I wasn’t even taking notes and just hoped I’d remember enough of the details later) about the situation and circumstances of AiD over the past few years. He divulged some fascinating tidbits about the challenges he and Hare Konatsu faced, and gave me the go-ahead to compile them into a piece of writing, which you now see before you.
This post, therefore, will be an odd one, containing elements of both self-introduction and chronicling. It is lengthy and meandering, as it is both my official greeting to all of you, as well as my personal reflection on the journey I’ve taken with AiD. For those new to AiD and the world of fault, I have included as much context as I could to ease you into this wayward tale I’m about to tell. For longtime fans and patrons, please bear with me as I crunch through the written equivalent of an establishing shot. Many of you likely know far more about the history of AiD than me, but with this unwieldy chimera of a post, I hope to deliver a few morsels of interest that perhaps even you, the wise AiD historian, hadn’t known about.
Let us begin.
- First, let me provide some context. ALICE IN DISSONANCE (AiD) released the first game in the fault series, titled fault - milestone one, in 2013 to significant commercial success. Through a trio of female protagonists, it told a compelling story of adventure and human drama. This was followed in 2015 by the second entry, fault - milestone two side:above, meant to be a two-part story that would be concluded in a subsequent entry, fault - milestone two side:below. The second game was also a smashing success, and currently, lifetime sales of the series has exceeded half a million copies. By all accounts, AiD was on a roll, and both the team and fans were excited for the series’ upcoming developments. Prospects were bright, and morale was high. From the gripping stories and charming characters to the expansive settings and rich lore, there was so much to explore in the world of fault. With a future abound in possibilities, readers eagerly wondered what marvels they would be treated to next. All they had to do was wait a year or two, and further enjoyment would surely be delivered to them. What could possibly go wrong?
Well, those of you with a sense for narrative flow can probably smell the impending development in this story already. Indeed, with things having gone swimmingly in Act 1, it was time for the protagonist to be hit with the big obstacle that pushes the story into Act 2. The team decided to sandwich a prequel story, fault - SILENCE THE PEDANT, between the two parts of milestone two to give more context to the world, lore, and characters. SILENCE THE PEDANT would be set five years before the events of the milestone series, and a demo with a considerable amount of content was put out in 2017, further whetting the appetite of fans. The excitement was tangible at Anime Expo 2017 as AiD was invited as guest of honor by our then-publisher, Sekai Project, and a beautiful booth backdrop for SILENCE THE PEDANT was put up, undoubtedly catching the eyes of many a passerby.
And so, the fans waited eagerly for the full release of SILENCE THE PEDANT. And waited. And waited. And...nothing.
In 2018, a spin-off title, Mhakna Gramura and Fairy Bell, was released, telling a quiet but compelling story about two orphans in search of eternal happiness. It was presented in a beautifully hand-drawn style evoking fairy tales and children’s books, and was very well-received by the community. However, it wasn’t a continuation of the fault story, nor was it the long-promised prequel of SILENCE THE PEDANT. The fans continued to wait to be reunited with the trio of protagonists from milestone one.
Fast forward to mid-2020. Still no milestone two side:below. And still no SILENCE THE PEDANT either. Even for the most loyal of fans, desperation had probably turned into despair before melting into melancholic acceptance that the story of Selphine, Ritona, and Rune might have stopped for good. But wait! A teaser suddenly dropped in September containing exciting, albeit entirely pre-rendered, footage of milestone two side:below! With a slated release date of Spring 2021! This was followed by a more gameplay-rich trailer in June of 2021 announcing a delay of milestone two side:below to September 2021.
Considering I am currently writing this in June of 2022, it’s not hard to imagine what happened next. Regardless, the team kept their heads down and their noses to the grindstone, and after much sweat and toil, in May of 2022, finally released...
fault - StP - LIGHTKRAVTE, a story set in the prequel’s timeline about an ordinary, unremarkable boy’s dream of becoming a painter. A boy that, for all intents and purposes, had not existed in the story up to that point. Thankfully, owing to the compelling story writing and absurdly (subjective opinion, but hey, I’m proud of what we made here) expressive animations, LIGHTKRAVTE managed to satisfy many of our longtime fans. Nevertheless, there were sporadic voices of bewilderment and mild outrage. Even the harshest of criticisms, however, frequently seemed to come from a place of passion, representing the frustrations of diehard fans whose expectations had admittedly been betrayed. fault had established itself with a trio of female protagonists. The people wanted more of the girls and their adventures. It’s understandable that a slower, introspective coming-of-age story about a boy they’d never heard of rubbed some of them the wrong way.
AiD certainly bears some of the blame with regard to setting fan expectations and muddy messaging. I’d like to invoke the indie card and argue that for a team of two—yes, at the time, two—people, trying to keep their messaging clear while struggling with what was obviously development hell was simply too much. Still, the responsibility ultimately falls to the team, and we are certainly reflecting on the entirely valid criticism we’ve received from the community.
So why did we do what we did? Why did AiD suddenly release LIGHTKRAVTE instead of milestone two side:below? And what in the name of jeebus took us so long? What was that 5 year (7 year, if you don’t count Mhakna and the SILENCE THE PEDANT demo) drought all about? Well, that’s the tale I’d like to tell today. I apologize for the extremely long preamble. Now, let’s get into the meat of this story.
Remember how I introduced myself as “the person handling the programming for the game”? That’s because I’m not a programmer. Not a professional one, anyway. Keep that in mind, because it’s central to AiD’s struggles over the past seven years.
AiD’s brand is based on pushing the limits of cinematography and cinematic presentation in the medium of visual novels. Being fundamentally two-dimensional, there are obvious challenges with trying to implement pillars of modern film-making, such as narratively expressive camera work and scene composition. Nevertheless, AiD took up the challenge and were rewarded for their efforts with milestone one and milestone two side:above. For SILENCE THE PEDANT, however, they started to run up against technical limits of Ren’Py, the visual novel engine they were using. The 2017 demo was, to put it nicely, a whole lot of smoke and mirrors. From a technical perspective (according to Munisix, since I wasn’t there during its development), it was a house of cards built on sand. Time and again, the application would crash out of the blue, apparently due to resource constraints. Despite the best efforts of Munisix and Hare Konatsu (the two-person team that comprised AiD at its conception and during their seven-year-long struggle), they couldn’t get the game into a sufficiently stable state, making further development a concerning prospect.
Neither Munisix nor Hare Konatsu possess any programming expertise. The programmer they’d been working with for the first two milestone titles had moved on to other projects, so they brought on a new person to help with the scripting for SILENCE THE PEDANT. Let’s call this person Alex. Alex had some Python experience, and they were able to help out during the development of the demo. However, due to the technical constraints, AiD decided to shift development to Unity, which would not only afford them more creative freedom but also pave the way for porting to the Nintendo Switch, which was all the rage at the time. However, compared to Ren’Py, Unity was an entirely different beast, and its complexity required a level of expertise that Alex did not have. So, AiD reached out to their initial publisher for help. The publisher generously agreed to provide a programmer, let’s call them Blair, who agreed to build a foundation in Unity that would provide AiD with a scripting environment they could use to make SILENCE THE PEDANT. In the meantime, Alex could observe Blair’s work and learn the ins and outs of Unity and the C# programming language in preparation for a baton pass down the road. The development of the demo was scheduled to proceed in parallel with the Unity work, so by the time the 2017 demo was released, a good deal of progress had already been made on the Unity side. That’s what AiD assumed, at least. Since neither of their core members were programmers, they could only take Blair at their word. There were periodic reassurances that progress was being made.
A year passed.
When AiD pressed for some signs of progress, to be shown something, anything, of the promised framework, it became apparent that for all intents and purposes, nothing had been accomplished. Whatever complications there might have been behind the scenes, the end result was that Blair had dropped the ball, and AiD had lost a full year of development time. Without a usable framework, Alex’s hands were also tied, and the project faced the imminent possibility of stalling.
AiD immediately asked their publisher for a replacement programmer, which they graciously provided again. Let’s call programmer number three Chris. With Chris, AiD restarted the waiting game for a framework in Unity but, feeling a little nervous from the previous experience, asked to see some proper progress at a little after the six months mark. At the same time, AiD went back to Ren’Py and started making a smaller spin-off game that wouldn’t stress the engine too much, figuring they might as well make the best use of their time while the Unity framework was being built. This was how Mhakna came into existence. Employing a children’s story book art style, it rarely reused sprites and told its story almost entirely through unique hand-drawn illustrations. Reception was highly positive, giving AiD some breathing room from a release schedule perspective. They could tell their fans, “SILENCE THE PEDANT is having some development trouble, but it’s still in works, and here’s a nice little story for everyone to enjoy in the meantime.”
Many months passed. The time had come for Chris to deliver on their promises. Again, AiD was presented with nothing. Nothing functional.
On the Unity front, almost two years had been spent in vain. Without Unity, AiD couldn’t proceed with development for their main games. Scrambling for a programmer, AiD found a fourth, let’s say Dylan.
Dylan was nice. Dylan got some work done. But Dylan was terribly careless. Or perhaps, Dylan simply didn’t care. Upon reviewing work from Dylan, AiD would identify, for example, problems A, B, and C. Dylan would implement some fixes, and on a second review, problems A and B would be fixed, but C was still present. After further fixes and a third review, C would be fixed, but now B was a problem again. It quickly became clear that Dylan would not be up to the task either, especially considering AiD’s focus on artistic quality and polish.
At this point, AiD had talked to all sorts of people, searching for ways to push the story forward while they waited for progress in Unity. One of these talks resulted in the potential of working with a new publisher. Hearing about AiD’s woes, they offered to provide a programmer of their own. The fellow, whom we’ll call Ezra, was apparently a young but brilliant coder with accolades to their name. Ezra had won awards in programming competitions. Ezra was the real deal, an up-and-coming prodigy who could whip up a visual novel engine in no time.
AiD was ecstatic. At last, they’d found a talented programmer who could get the series back on track. So, they waited for Ezra to knock it out of the park.
Ezra ended up bunting. The work that AiD received was in no way acceptable for their standard of quality. Even a coding whiz, it seemed, wasn’t capable of delivering.
By this time, AiD’s momentum had ground to a halt. With no way to push their main storyline forward, they decided to refresh the visuals of their previous games as a way of renewing interest in the series and hopefully bringing in some new readers. Hare Konatsu’s artwork had improved significantly in quality since their early days, and it would be nice to present milestone one and milestone two side:above to readers with a new coat of paint. Most importantly, it’d give AiD something to do instead of twiddling their thumbs and wallowing in frustration. The Nintendo Switch was selling like hotcakes, and porting the previous games over with fresh visuals seemed like a good idea.
It was then that AiD found another programmer, let’s say Frances, who had experience in the industry making games, and promised to build a visual novel engine in Unity from scratch. By Frances’s estimate, it’d take about six months. Not too long. It was only a visual novel, after all.
It did not take six months. It took quite a bit longer. But Frances did deliver, and AiD managed to remake milestone one and milestone two side:above for Nintendo Switch in Unity using Frances’s framework. At long last, AiD had put out some titles, albeit rehashed from previous ones. The whole process of porting the two games to Switch took almost two years and was in no way a trivial effort. By the end, Frances seemed to have run out of gas. Perhaps it was the constant push for polish and refinement from AiD that exceeded what Frances had been expecting from a visual novel gig. Or perhaps it was the whirlwind of bug reports and game-breaking issues that pop up shortly before release during quality checking (and to some degree, after release, to the dismay of everyone involved). Whatever it was, Frances declined to continue developing the engine after the two Switch ports.
AiD was now back to exactly where they’d started—two artists sans programmer. I can’t even imagine how it must have felt for Munisix and Hare Konatsu then. In fact, I’m pretty sure they were this close to throwing in the towel.
Because they were willing to take a chance on me.
This, dear readers with the patience of three Buddhas, is where I enter the story.
I spoke to Munisix in August of 2020, in the midst of the lockdown pandemonium that was early COVID. I was winding down my career as a pharmacist in preparation for marriage and a move to Japan. Obviously, the pandemic had thrown a wrench into the “move to Japan” part. Figuring that I’d have some extra time (as I so naively assumed at the time) on my hands while waiting for COVID to pass, I decided to tap Munisix over Discord and see what was cooking. I’d met him once in 2018 and exchanged some pleasantries, but no communication had occurred since. I hadn’t really kept up with AiD’s situation, but I did enjoy their titles a great deal, so I figured I’d see if I could help with anything while I waited for the global situation to calm down.
Being a freelance JP-EN translator on the side, I offered to translate some stuff for AiD. After a quick talk, Munisix sent me the manga-novel-script fusion artbook that had been made for Limited Run Games’s fault - milestone one Collector’s Edition for Switch. I sent back the translation, we gave each other a thumbs up emoji, and went our separate ways.
...
...
Just kidding.
I then asked if there was anything else I could help with, and Munisix, in what I can only assume was a tone of profound weariness, admitted that what AiD really needed was a programmer. I am not a programmer, so that was not an issue I could reasonably solve. The last time I did any programming was in high school, and I have zero industry experience. I was completely honest in our discussion and said that, frankly, it would be insane to have me do the heavy lifting on the programming side. After all, no matter how simple it might seem to make a visual novel, it’s still a video game. Everything I’ve ever learned in life told me never to underestimate the complexity of a video game, and especially not one where the developers have high standards for quality and are trying to push some envelopes.
But you know what Munisix did? He gave me the proverbial pat on the back and told me to give it a try. So, I signed on to develop their next game. Which would not be SILENCE THE PEDANT (due to an internal pivot to appease fans who’d been waiting so very long for a continuation of the main story) but fault - milestone two side:below. I, a complete non-programmer, would be handling the programming for what was arguably AiD’s most anticipated title.
Insane, right? Well, I’m here to tell you that... Yes, it was absolutely insane, and if a single piece of this ludicrous puzzle hadn’t by miraculous coincidence fallen exactly into place, this whole thing would have been a complete disaster.
First, AiD arranged for a quick hand-off between me and Frances, where Frances would give me a quick explanation of the code base. Anyone with any programming or game development experience can probably picture how this went. Imagine explaining the code base of a one-man job for a visual novel to someone who not only hasn’t programmed in over a decade, but doesn’t even know how Unity works. Oh, did I mention that yet? At the time, I’d never even touched Unity. In other words, Frances was like, “This script controls the transform for the camera’s game object” and I was like, “What’s a game object?”
And then Frances left.
I was now on my own, staring at a code base I didn’t even know how to begin comprehending. It was here that an old friend of mine came to my aid. His advice was not only timely, but desperately needed and absolutely crucial. He stared me in the eyes over Discord and said, “No, man. That’s not how you game dev. Go get a visual novel engine. Look, I’ll find one for you on the Unity Asset Store. What’s this? Naninovel? This looks fine. Use this.” Needless to say, the conversation was a little more involved than that, but his point was clear—don’t fiddle with someone else’s code, don’t start from scratch, use an existing visual novel engine.
I followed his advice, bought the engine asset, and began to poke around as I tried to simultaneously learn Naninovel, Unity, and the C# programming language. Now, full disclosure, I have some C++ experience from high school, but my coding was entirely algorithmic. My grasp of object-oriented programming was shaky at best, and my understanding of industry best practices was nonexistent. C# expressions like () => {DoAThing();} bewildered me, get set declarations made no sense, and I hadn’t the slightest clue what events and delegates were. That was where I started in August of 2020, all the while juggling translation, leaving pharmacy, and an uncertain visa situation.
I’ll spare you the details of the next year and a half, but needless to say, progress was unsteady and difficult. On many, many occasions (more than I can count), it took literal miracles to solve the problems I struggled with. That is, completely coincidental flashes of inspiration caused me to try things that seemed completely irrelevant, only for them to lead me to solutions. Had any of those miracles not happened, I probably wouldn’t be sitting here writing about AiD’s newest release.
Furthermore the team had decided to use Spine from Esoteric Software to animate our sprites. It quickly became clear, however, that the combined usage of Naninovel and Spine presented its fair share of unique challenges, and we needed a lot more practice. Real practice. The kind that involved shipping a complete game with the new set of tools we had. This led to a series of emergency team meetings in late 2021, which resulted in a pivot. Using what we’d managed to build up, we’d split SILENCE THE PEDANT into a collection of smaller stories and put one of them out first. It’d allow us to familiarize ourselves with the new technology and also function as a sort of rehab for making a full game from start to finish. The goal was to just release a game.
And in the end, AiD (now a total of four members, having also on-boarded a very hard-working individual from Russia for the much-needed extra manpower) managed to release a game. It is titled fault - StP - LIGHTKRAVTE, and by god, it is real and sitting on Steam, and people are playing it and enjoying it. Every time I look at it, it still feels surreal.
And it technically only took seven months from pivot to end. Freaking record pace, that.
All’s well that ends well, I suppose.
But not really. Because that doesn’t explain AiD’s seven-year-long marathon of frustration, nor does it satisfy the many fans who are still waiting for milestone two side:below. As such, we are left with a question. An open question, which we pose to anyone reading this.
What happened?
What were those seven years of disappointment and discouragement? None of us in AiD are professional programmers, and our understanding of the coding industry is limited. Is this just how indie development is? Are programmers who can see a project from start to finish simply in short supply in the visual novel scene? It certainly seems possible, given the terrible incentives for strong programmers to work in the indie scene, never mind the visual novel scene. Or was AiD simply unlucky and suffered a terrible streak of incompatible coders?
Was it us? Did AiD do something wrong with regard to how we looked for programmers? Given Munisix and Hare Konatsu’s position at the time, what should they have done? With no understanding of programming, they could only rely on the words of people they knew, such as publishers and friends. They tried candidate after candidate, and they were never hesitant to pay handsomely for good work. Over half a decade passed to no avail. When they finally put out a game again, the person who coded their game...wasn’t a programmer.
What gives?
Is this just indie life? We’re honestly curious. If anyone in the industry has some wisdom for us, we’d love to hear your thoughts.
Such questions are, of course, largely meaningless. The vicissitudes of game development are well-documented, and no perfect solution seems to exist. For now, we are content to count our blessings and simply appreciate the fact that after crawling through a long, long tunnel, we’ve finally made it out into the light. We have a stable team committed to pushing the fault series forward, and we’ve gained invaluable experience through developing LIGHTKRAVTE. We needed this. All of us. It was team building, boot camp, and trench warfare all rolled into one. But we survived—some a little more scarred than others, perhaps, but all in good spirits.
After five years of near-silence on the new title front, the mere fact that AiD still exists is surely a minor miracle in and of itself. That is in no small part thanks to you, dear reader, because you cared enough to keep paying attention. And when we crawled back onto the hill to blow our time-weathered horn, you answered the call. You came back.
Thank you.
Fortune is a fickle thing, and one can never tell what the future holds. I don’t know how long this arrangement we have will last, nor how far I’ll be able to help this team go. However, until fate sees fit to sunder our fellowship or, indeed, death itself does us apart, I intend to keep pushing forward as a member of AiD. Please wish us luck.
That’s all I have for now. Thank you very much for your time. If you haven’t played fault - StP - LIGHTKRAVTE, please consider giving it a try. Your support will do wonders for our motivation. And milestone two side:below which, though already back in development and poised to receive a small mountain’s worth of system and design improvements, will certainly still benefit from a beefier budget. Bear with us while we step up to the plate.
We’re going to swing for our next title. See you in the stands.
—David Teng
Comments
Ah, interesting. I guess that's why the code of the Switch ports (not mine, theirs) of Fault MS1/2sa is majorly different from StP LK is because they switched to Naninovel. Also why some features like live language switching and rollback are missing. I also wonder why A-B loop points (like in the StP demo) are missing, even through Unity supports them natively. Too bad that unfortunately with this switch, the Unity based codebase of Fault MS1/2sa on Switch won't be coming to PC… I was so looking forward to running ILSpy on it and seeing how it works. Now for the rest of the blog post… There's only so much that one person can do. Most of the large visual novel engine projects like Ren'Py and Kirikiri are open source, documented very well, and have many community contributors so they can fix bugs they encounter. Having NIH syndrome could be interesting, but once you get compared to the other engines that have features the client wants, it becomes a pain really fast. Also, as in the post, most NIH stuff isn't very documented because it assumes only one person will work on it and know how to deal with it. I can also relate to the "1 bug fixed, 2 bugs reappear" issue. Engines like Ren'Py has regression tests so you can catch that stuff easily. If you don't have unit tests or regression tests, it will be a painful route to take. Nowdays, there are ports of Ren'Py to consoles (my own port, and Ratalaika Games' port). The good thing about Ren'Py compared to Unity is that it doesn't have broken Linux video support, can compress the assets more or use lossless assets, and if there is an issue in Ren'Py, you can fix it. (Unity core is closed source and their feedback/bug report system is horrible) Personally, I'm interested in what specific issues they had with Ren'Py and why they had to switch off of it. But yeah, I'm glad at least the ball is rolling again and can't wait for the release of Fault Milestone 2 Side Below :-)
I don't usually post either, but I want to say that I'm glad to hear how things have gotten back on track. I've been a supporter for ages now, and I will admit I was pretty curious what was going on behind the scenes that had been causing the delays, so this post is pretty enlightening. One could say the project was hostage to a 3rd party's incompetence, which is unfortunate. Needless to say, from the posts that I've seen over the years from Munisix and Hare they have always felt sincere in their efforts to work on fault, and you seem to share that spark as well, so I hope to continue supporting all of you and seeing more of fault.
