Oh sorry no— the boilerplate is the value! The issue is when I rely on an LLM for important code and it breaks — now I need to debug somebody else’s code, and that somebody else has usually made things far more complicated than they needed to be.

For me it’s really about the boilerplate/helpers. I similarly dislike when the actual logic or utility isn’t in my style or I don’t completely understand it (when it inevitably doesn’t work, it’s much harder to debug)

That said, paying $100/year for family or personal apps is kind of annoying.

Recently did exactly this and it is a fun experience. I’ve found that Claude is by far the best for SwiftUI, but all of the big LLMs tend to make up functions or provide deprecated code (for some reason, they always mess up the ‘onChange’). Easy enough to fix even with virtually no Swift experience.

Very cool, I’ll have to check it out! Thanks for the link

Super cool work and I love the incredible detail of the appendix! It looks like a ton of effort went into making this a reality, I admire the vision/collaboration of the team. Would be great to get access to that super simulator someday 😁