sql.js is 2 MB big (https://github.com/kripken/sql.js/blob/master/js/sql.js https://github.com/kripken/sql.js/blob/master/js/sql.js ) And it's an in-memory dat

sql.js is 2 MB big (https://github.com/kripken/sql.js/blob/master/js/sql.js https://github.com/kripken/sql.js/blob/master/js/sql.js ) And it's an in-memory database, not like WebSQL. So one has to store the data (array) to localStorage or rather IndexedDB (HTML 5 NoSQL storage). https://github.com/kripken/sql.js/wiki/Persisting-a-Modified-Database https://github.com/kripken/sql.js/wiki/Persisting-a-Modified... So imagine a Firefox user: He has to download an extra 2MB JS file, the huge file has to be run through asm.js JIT and the data is (for example) stored (offline mode) using IndexedDB. Firefox implements IndexedDB on top of SQLite. A SQLite instance runs on top of an NoSQL engine on top of another SQLite instance - wtf!?