A curated roundup of Python news and tutorials covering Polars vs SQL differences, Python's built-in constants, using deque, a Pyrefly type checker sponsor spot, PEP drafts, Mojo going open source, debugging Python with AI agents including Claude Code, a security issue with str.lower(), reproducible builds on PyPI, a React/FastAPI 'Mullet Stack' approach, a Django counting trick, and several new open-source projects and libraries shared by the community.
Questions this post answers
Is the Mojo programming language open source now?
Yes, Mojo has been made open source by Modular. This follows Mojo's development as a Python-family language designed for high-performance AI and systems programming, positioning it as an alternative for developers who want Python-like syntax with lower-level performance control. Developers evaluating new AI-focused languages can follow Mojo's evolution alongside Python news on daily.dev.
Why can str.lower() be a security vulnerability in Python?
Because some internet standards only support ASCII, Python must translate Unicode strings to ASCII-compatible forms in certain contexts, and using str.lower() for this kind of normalization can introduce mismatches or bypass checks due to Unicode case-folding quirks, creating a security gap between the intended and actual string comparison. Developers hardening string handling logic can track Python security gotchas like this one via daily.dev.
What is Python's Mullet Stack architecture?
The Mullet Stack refers to pairing a React and TypeScript frontend with a FastAPI and Pydantic backend, combining a JavaScript-based UI layer with a Python-based API layer. The term, coined by Sean Helvey, plays on the mullet haircut idea of 'business in front, party in back' to describe this common full-stack pairing. Developers weighing frontend and backend framework pairings can compare stacks like this one on daily.dev.