Frequently Asked Questions

The drag-and-drop games were produced by Gemini 3.1 Pro Preview. It provided the best language coverage, especially for endangered languages like Navajo. Claude, by its own admission, is lacking in coverage of lesser-used languages.

The top-level language pages like French were produced by Claude Opus. It was Opus that recommended the solution for combining multiple games into a single app, and the development workflow using Claude was superior to Gemini.

The home page, FAQ, and "Video Guides pages were made with Gemini.

Minimally intrusive targeted ads (not on the games themselves). The site also serves as a portfolio asset and proof of concept.

Installing the Progressive Web App (PWA) allows you to launch the games directly from your home screen without opening a browser. It provides a seamless, full-screen experience, enables offline play so you can practice anywhere without an internet connection, and takes up a fraction of the storage space of a traditional native app downloaded from an app store.

The basic speech synthesis utilizes the Web Speech API, a native feature built directly into most modern web browsers. It relies on the text-to-speech engine and voice packs already installed on your device's operating system. It enjoys broad support across Chrome, Safari, Edge, and Firefox. Because it uses local resources, it supports major global languages like English, Spanish, French, and German without requiring an internet connection. However, the exact voices and languages available will vary depending on your specific device and operating system.

Ukrainian speech was added as a test case for speech synthesis of second-tier languages. Because local device support for Ukrainian is highly inconsistent across different operating systems, we integrated Google Cloud Text-to-Speech. This service generates the audio on Google's servers using advanced neural network models, delivering highly realistic and accurate pronunciation. This ensures that every user gets the exact same high-quality audio experience for Ukrainian, regardless of their device's built-in capabilities.

This Python program can be used to "hydrate" the hydrate-gtts.py. Note that you have to add your own Google Text-to-Speech API key to the script.

To obtain a Google Text-to-Speech API key, you must first create a Google Cloud account. Once logged in, navigate to the Google Cloud Console and create a new project. Go to the "APIs & Services" library and enable the "Cloud Text-to-Speech API" for your project. Finally, access the "Credentials" tab, click "Create Credentials," and choose "API Key." It is highly recommended to restrict your API key to prevent unauthorized usage.

Use the MUTE button at the top of the game.

"Vibe coding" is a term that gained popularity in the AI development community around late 2024 and early 2025. It describes a new paradigm of software development where a human developer interacts with an advanced AI coding assistant using natural language. Instead of manually typing out syntax, the developer guides the AI with high-level concepts, architectural decisions, and intuitive prompts—relying on the "vibe" or flow of the conversation to rapidly generate and iterate on functional code.

Apple has historically restricted the capabilities and ease of installation for Progressive Web Apps (PWAs) on iOS to protect the ecosystem and revenue of their proprietary App Store. Unlike Android, which can proactively prompt users to install a PWA, iOS requires users to manually discover the option by opening the Safari Share menu, scrolling down, and explicitly selecting "Add to Home Screen." Additionally, Apple limits certain web capabilities like background syncing to prioritize native iOS applications.