HaS-003: Discovering the process
On finding out my own process for game development
I recently finished another narrative piece for Have a Sip (HaS), published in the game’s companion blog as “The view from here”, a story about stumbling back into children’s vantage point.
It’s good that there’s finally some output, because there has been weeks where the lack of progress has been quite frustrating.
However, when I look back at the last 1+ month, I realized there is a lot of hidden work that went into discovering a structure for how to develop HaS. And it was not what I had in my when I first started.
A lot of the process’s peculiar details came from my own constraints:
Mostly solo development, with no prior background in actual gamedev.
Running this on the side of my full-time work + family duties.
HaS being what it is: a narrative driven game that strive for directed, but also procedural stories.
What I discovered was that: for a particular game & team setup, a good process can really reduce the friction and overhead.
In this case, I stumbled-upon production pipeline was:
→ Write the stories for HaS in blog post form
→ break them apart to create an Interactive Fiction prototype (Twine)
→ imagine what art/music to add
→ translate to engine/code requirements.
This accomplished a few things:
Fleshed out the characters
Now, the characters feels pretty strong. From major ones like PICO and Elodin to the support casts like Eleanor, Martin, Shiori, Diego, etc feels fleshed out. This happened as a consequence to writing and thinking about what they would say or do.
And instead of creating them from a top-down troupes, stitch them back from separate writing pieces allowed them to be nuanced, full of different beliefs and contradictions - the very value that HaS aims to portrait.
On the mechanical side, it also led to the creation of a “framework” for each character: internal thoughts, habits, speech/cognitive pattern, personality types, etc. This keep them coherent across various pieces (a hat tip to Brandon Sanderson’s lecture on building characters).
After some initial investments, HaS’s characters feels like they are alive now. I could then place them into a scene and ask “what would they do/say”, simplifying a lot of the writing process.
Structured the story’s pace
Writing taught me pace-control at various level: the macro of the whole blog post and the micros of paragraphs and sentences. This is emphasized even more as I attempt to turn the writing into the IF prototype.
At many times, it kinda felt like making cut-scenes for the game, the sort that you would see in Stardew Valley when you hit a certain relationship threshold. And if my instinct are right, HaS will be a game with lots of cut-scenes.
Supported the after-hour, solo-dev setup
This turned out to be huge for me. Making HaS required learning about writing, pixel art and (later) music production, besides the usual programming bits. But learning them past a certain level turned out to be seriously hard work, which drained energy quickly.
Thus, a more gradual process that centralized around writing allowed me to make constant progress, which means maintaining momentum and morale. And those are the necessary currency when you are working for passion pay instead of hard cash.
With deliberate practice, it’s also getting easier to slot writing time into the 1-2 hours free time I have every night, without cutting back too much on sleep, family time and other important things.
Set the stage for the visual/audio
I actually didn’t found time to get back into Aseprite or to learn to make music. But because of the writing work, I feel that there are now vivid pictures of what the pixel-art environment and characters should look like.
As for music, it currently comes in the form of a playlist that I kept while writing, as well as a list of artist, composer and style notes whenever something catchy came up. I’m cautiously optimistic that I this will train my music taste somewhat, so I can make some okay tracks later (just not sure about timeline, haha😝).
Kick-started game-design thinking
This mostly happen with the IF prototypes + back-and-forth thoughts. Because HaS’s player interaction is mostly reading, observing and sometimes making decisions, the IF format feels quite appropriate to test them.
Obviously it won’t be enough, but the prototyping helped pinning down better ideas about 2 separate mechanisms: observing characters and story “branching” with the drinks PICO serve.
Sketched the requirements for the game engine
Having explore Bevy more recently, I am of the leaning into the idea that the modular, ECS nature of it will be a good fit for HaS. If each character have a unique system of behavior, it would be feasible to “plug” their systems into the game (we’ll have to see about performance at scale, but I’m hopeful).
Because I work on infra automation at work, I can’t stop comparing this to k8s controller, haha.
Surprisingly here, I found the writing work to be more helpful than the IF prototypes, since the space to describe what a character’s “loop” is more broad.
Obviously, writing alone don’t cut it, so we will have to do a lot more exploration with Bevy in the future. However, being able to see a more concrete shape of the required system is still a great boon.
Built something to show
The last, and somewhat intentional, result of all this work is that I have something to show. Instead of vaguely talking about the game, I can now share blog posts about the story beats to people, showing them what kind of story to expect. This has already earned a few unexpected encouragement (thank you), which really helps with motivation!
After thoughts
All this is to say that I’m somewhat optimistic about the progress so far, even though some weeks things moved at the snail’s pace. The hidden part of the iceberg is being built. And hopefully it will build the foundation to make more things.