Python Weekly (Issue 671 October 10 2024)

Python Weekly - Issue 671

October 10, 2024

Python Weekly

Welcome to issue 671 of Python Weekly. Let's get straight to the links this week.

Articles, Tutorials and Talks

The video discusses the recent release of UV 0.3.0, a Python packaging tool that aims to streamline the development workflow by integrating features that allow it to serve as a comprehensive solution for managing Python projects. The presenter highlights its speed, ease of use, and potential to replace existing tools, while also addressing current limitations and areas for improvement in the tool's functionality.

Or how I made CRC-32 calculations 18 times faster than Python, and 3 times slower than Python.

The article explains how to use Python's ctypes to wrap SystemV shared memory functions (like shmat, shmget) for interprocess communication on systems restricted to Python 3.7. The author demonstrates creating, reading, writing, and destroying shared memory segments through Python, noting that while this approach isn't needed in Python 3.8+ due to built-in abstractions, it's useful in restricted environments.

The article provides a comprehensive guide to error handling in Python, covering various techniques and best practices. It explores different types of errors, exception handling mechanisms, and strategies for creating robust and maintainable code, including the use of try-except blocks, custom exceptions, and logging.

Learn about the exciting new features in Python 3.13. Get insider insights into the latest updates and learn about the plans for Python 3.14.

Using django-pgactivity for application-level monitoring of database queries.

The article demonstrates how to scale AI-based data processing using Hugging Face and Dask, progressing from processing 100 rows locally with pandas to handling 211 million rows across multiple GPUs in the cloud. It showcases the use of Dask for distributed computing, enabling efficient data loading, preprocessing of large datasets, and parallel model inference, with a practical example using the FineWeb dataset and the FineWeb-Edu classifier.

How Python's recent performance improvements work under the hood.

This post explains how Python’s TypedDict can enhance code clarity and maintainability by enabling more precise type annotations in dictionaries. It discusses how TypedDict ensures type safety and helps with early error detection in dynamic programming environments.

Interesting Projects, Tools and Libraries

Your agent in your terminal, equipped with local tools: writes code, uses the terminal, browses the web, vision.

A toolkit to create optimal Production-ready RAG setup for your data.

Lightning fast OLAP-style point queries on Pandas DataFrames.

A foundation model for zero-shot metric monocular depth estimation, capable of producing high-resolution depth maps with exceptional sharpness and detail in less than a second.

Build real-time multimodal AI applications.

Monitor DNS queries by host processes using eBPF!

A Customized Cloth Texture Generation Pipeline for 3D Virtual Try-On.

A tool for embedding Python into .NET projects.

Fast, accurate and scalable probabilistic data linkage with support for multiple SQL backends.

Niquests is a simple, yet elegant, HTTP library. It is a drop-in replacement for Requests, which is under feature freeze.

New Releases

The newest major release of Python introduces several new features including an improved interactive interpreter, an experimental free-threaded build mode, and a preliminary JIT, along with various optimizations and changes to the standard library.

This release includes the following announcements:

  • Run Python tests with coverage

  • Default Python problem matcher

  • Python language server mode

Upcoming Events and Webinars

There will be following talks

  • Concurrency in Python: What No-GIL Means for Developers

  • Building AI SaaS on the Cloud: A Hands-on Guide

  • Does Python Need Tool Qualification for Safety?

There will be following talks

  • Bringing a RAG From Prototyping to Industrialisation: Things I Wish I Knew Earlier

  • From Data to Insights With the SHAP Explainable AI Library

  • Beyond Pandas: Polars and DuckDB for Data Processing at Scale

There will be following talks

  • The Data Architecture behind John Lewis Partnership's Return to Profit 

  • Revamping our A/B testing methodology - everything is a histogram if you squint!

Join us for a meander through the maze of Python development-tools, from project set-up to packaging. Four of us will each present on one of the tools he uses—whether it’s pyenv, venv, poetry, or uv.

There will be following talks

  • Visual Place Recognition

  • Faster Models, Faster Answers: Discover Emulation for Your Workflow

There will be a talk, F-strings.

Our Other Newsletters

 - A free weekly newsletter for programmers.

- A free weekly newsletter for entrepreneurs featuring best curated content, must read articles, how to guides, tips and tricks, resources, events and more.