The image on the left is an illustration I generated with MidJourney and refined using Gemini, intended for use as an in-game info kiosk. I’m still debating whether to use this or stick with the current kiosk. I prefer this one, but the current design seems easier to spot from a distance.
The first test is wrapping up. It still stings a bit, but I now have a better idea of what needs improvement, and I’m actively working on fixes. Initially, I planned to gather testers, but I found out that previously released demos can still participate in Steam Next Fest, so I figured I’d put the demo out first and get some early feedback. I aim to finish revisions and translations by next week, then upload the demo to Steam. Right now, I’m thinking of releasing a 2-hour demo first, and then expanding it to a 3-hour demo for Next Fest.
This is my first time running tests so early in the project, and honestly, it’s turning out to be far better than any approach I’ve taken before.
In the last game, I actually wanted to create something with room for exploration, but due to inexperience and late-stage testing, I was forced to quickly reshape it into a linear experience, which left me unsatisfied. Even now, I keep thinking about completely redoing the prologue if circumstances ever allow. So this time, I focused heavily on creating a non-linear experience, but then received a lot of feedback that players felt lost or didn’t know where to go. I ended up clearly showing the destination at all times, but left all paths toward it open for players to choose. I haven’t received much feedback yet on the revised version, but it seems I still need more explicit guidance. I felt I’d already spoon-fed every step to players, practically moving their jaws for them, but apparently, it’s still not enough. In the past, I would’ve agonized over this issue much longer. Now, I’m simply implementing the immediate ideas that come to mind and releasing the demo quickly. I think gaining more feedback is better than agonizing alone. (Although I’m not sure how many people will actually play the demo and give feedback.)
The game currently supports 11 languages. Translations and reviews are handled by AI, but managing text assets across multiple languages is proving quite cumbersome. Previously, I edited text in Excel, exported as CSV, and imported that into the game. Now, I’ve shifted to Google Sheets, managing all 11 languages in one sheet, which I’ll also export as CSV. I made this change considering it might be easier if I ever need human reviewers, but now I wonder if it was unnecessary.