I’ve been stuck on Nucleate for about a month now, mostly waiting on feedback from my brother on the beta build after I signed and notarized it for Mac. The problem is I don’t have a proper beta group to get solid input from, so I’m starting to feel like I can’t wait around forever. I’m planning to push for feedback again soon, but if I don’t hear anything by the weekend, I’ll start reaching out to people on Reddit or similar platforms to get some beta users lined up.

Meanwhile, I’m getting ready to package the Windows version of Nucleate. I expect the certification and licensing process will be similar to what I went through with Mac, but I’m still trying to figure out the exact costs involved. That’s something I need to nail down soon because it’ll affect the release timeline and budget.

I’ve decided not to wait for a large beta group before releasing. I’m going to push the release soon even if the feedback is minimal. It feels like waiting longer isn’t helping, and I need to move forward. Hopefully, once it’s out there, I can get more real-world input.

On the development side, I’ve been experimenting with IDE-assisted coding tools like Cursor. The free version seems pretty useful, but I’m not sure yet how well it actually integrates with Godot’s scripting language. That’s still an open question. The whole transition from Godot 3 to 4 has complicated things more than I expected, so I’m feeling a bit cautious about relying too heavily on these tools without fully understanding their limits.

At work, I’ve been using agentic AI and IDE-assisted coding, including GitHub Copilot and some premium AI models, to build a new app from scratch in about a week—a process that would have taken me a year before. That’s been eye-opening and makes me want to push AI tools harder in my own projects. But I’m still figuring out how to balance speed and quality with these tools. I don’t want to end up with a half-baked product just because the AI rushed things.

Progress on Jumpsuit (working title: Orbital Jump) and learning Godot has been slow overall, mostly because of my day job taking up a lot of time. Still, I’m trying to keep moving forward bit by bit.

Right now, the big open questions for me are how well Cursor will play with Godot’s scripting, what the Windows certification and licensing costs will actually look like, whether I can trust AI-assisted tools enough to keep quality up while moving fast, and if soliciting beta users externally will actually get me useful feedback or just noise. I’ll keep testing and adjusting as I go.

Two-Line Takeaway

Key step forward

The most important forward progress step is preparing to release the Windows version of Nucleate soon, even with minimal beta feedback, to move forward and gather real-world input.

Significant barrier or risk

The most important barrier or risk is the uncertainty around beta user feedback quality and the costs and certification process for Windows, which could impact the release timeline and budget.

Generated with Nucleate by Watchlight Studio — distilled from full transcription.