Learning the Long (Wrong) Way
I spent roughly thirty hours building tools to help me learn Japanese via reading, and I rarely use either of them because at the end of the day tools help but don't replace actual reading. Here's a short story about my adventures in procrastinating by doing work to avoid actually doing what's important.
On my computer, I have a great reading setup but I want to read on my phone, like a Kindle, something I can pick up on the couch or take anywhere.
The catch is I need a one tap dictionary. On desktop, yomitan is genuinely the best: free, fast, and configurable. It doesn't run on iPhone. Android can run it and the experience is great. I even tried it on an Android phone I had sitting around, but I wasn't going to carry two phones.
I already tried the solid free tools on iOS. 10ten paired with ttsu reader covers most of what I need. I just can't shake the feeling I'm being inefficient by not using the best tools.
So I paid for iOS reading apps made for learning Japanese. I even read a book with them. They were fine, but still never quite there.
Then I built my own solution. First, an Apple Shortcuts workflow: capture the word, definition, sentence, and screenshot from whatever I'm reading, translate it, send it to Anki. A few hours in Apple's painful Shortcuts builder, but it worked, and I could even pull sentences from YouTube.
Then I built the app. Not a prototype. A full-featured iOS app, built over evenings until I stopped finding bugs. Feature complete. Roughly thirty hours, across both app and shortcut.
I knew I was procrastinating under the guise of productivity. Building tools was fun. Learning Japanese is hard. That was thirty hours I could have spent reading.
Eventually, I went back to 10ten and ttsu reader. They do everything I need and get out of the way. Since I stopped chasing the perfect setup, I've finished my second book in Japanese and I'm halfway through a third.
I'm sure the perfect setup is out there, but I'm already reading.