This is a brilliant idea! I have so many books and bookcases. I can never find anything, though I do kind of enjoy searching through them all to find the one book I need. Will we ever get an android version? I'd love to know how you went about learning and creating this. What was the effort and lift like?
With this app, I spent around 3 weeks learning Swift and Xcode. I'm not an expert in Swift by any means, but learned enough to be able to put together proof-of-concept. I used the Swift docs for reference, but mainly ChatGPT 4o (the older one) and had it explain concepts to me and debug my code when I was stuck. The best part about that approach, I found, is that it matches the way I like to learn—I like to start with something scrappy and then learn on the go as I work on making the thing better.
Since then, so for the past two weeks, I've been working on getting the proof-of-concept ready for release. I probably put in 10 to 12 hours a day. The target date for release is October 22, and the one thing left right now is making sure enough of the detections are accurate that a first-time user won't go, "Pfft ... this sucks!"
The OS's Vision library does a decent job for the English titles I've been using for testing, but it's not perfect. So my approach right now is supplementing that with AI suggestions from Gemini. That way, your photo is never sent off-device and you can, if you wanted, disable external processing on the detected text all-together.
I don't know anything about Android app development either, which probably means, yes, there will be one at some point. :) I plan on open-sourcing the code post-release, so you're welcome to help out with that.
This is a brilliant idea! I have so many books and bookcases. I can never find anything, though I do kind of enjoy searching through them all to find the one book I need. Will we ever get an android version? I'd love to know how you went about learning and creating this. What was the effort and lift like?
Looking for a book is definitely half the fun!
With this app, I spent around 3 weeks learning Swift and Xcode. I'm not an expert in Swift by any means, but learned enough to be able to put together proof-of-concept. I used the Swift docs for reference, but mainly ChatGPT 4o (the older one) and had it explain concepts to me and debug my code when I was stuck. The best part about that approach, I found, is that it matches the way I like to learn—I like to start with something scrappy and then learn on the go as I work on making the thing better.
Since then, so for the past two weeks, I've been working on getting the proof-of-concept ready for release. I probably put in 10 to 12 hours a day. The target date for release is October 22, and the one thing left right now is making sure enough of the detections are accurate that a first-time user won't go, "Pfft ... this sucks!"
The OS's Vision library does a decent job for the English titles I've been using for testing, but it's not perfect. So my approach right now is supplementing that with AI suggestions from Gemini. That way, your photo is never sent off-device and you can, if you wanted, disable external processing on the detected text all-together.
I don't know anything about Android app development either, which probably means, yes, there will be one at some point. :) I plan on open-sourcing the code post-release, so you're welcome to help out with that.