Hacker News Reader

It doesn't seem to do much. It looks like it just generates lots of calls from the byte code. It doesn't do any type inference.

The big win with a smart Python compiler is detecting when you don't need to represent data with the general "CObject" type and can use machine integer or floats. Then arithmetic becomes much faster. PyPy does some of that.

Being able to make an executable with no dependencies is worth something. It makes deployment easier. No need to fuss with containers or installing the right libraries on the target.

Wrong. Look at the list of optimizations it does. Arithmetic optimizations are only done with constants and type hints. But inlining and SSA var tracking is impressive.

That is super cool. Any data on how it compares to pypy's JIT?

https://news.ycombinator.com/item?id=15354613

how different is this from cx_freeze? is it faster? (compile or execution)

Nuitka: A Python Compiler