The author shares practical experience running Python in production environments, covering multiple aspects including improving project quality, ensuring maintainability, and operational deployment.
TL;DR
How to run Python in production, why Browser Use raised $17 million for browser automation, and 15,000 lines of verified cryptography land in CPython.
Articles & Tutorials
× 12Browser Use is a recent viral star project - a pure technical open-source project by two people with 8,000 lines of core code that raised $17 million! What exactly did it do, and why is it worth so much?
Shares some Django Admin themes: Dracula, Daisy, django-jazzmin, django-admin-kubi, django-jet-reboot, django-semantic-admin, Simple UI, Grapelli, django-admin-interface, Unfold.
Introduces the journey of Python adopting the verified HACL cryptography library. The author proposed replacing Python's hash implementation in 2022 due to SHA3 vulnerabilities, and after 2.5 years of effort, Python's default hash and HMAC algorithms are now provided by HACL.
Inspired by last week's article suggesting not to write __init__ methods, this author shares an extreme __init__ case they wrote, explaining why the code had errors.
Python is about to introduce a new ASN.1 API that improves performance with a pure Rust parser, reduces discrepancy risks by sharing parsers with PyCA Cryptography, and enhances usability with a dataclass-style interface.
Shares a case study of using regular expressions for text manipulation, introducing several practical features like enabling verbose mode for better readability, capturing text through named groups, and optimizing matching logic with non-capturing groups.
The author reflects on the wonderful journey of Python's evolution over 30+ years while pondering how Python should develop and face challenges in the future. The author calls for simplifying program distribution and improving mobile platform development experience.
Introduces how to use the Copier tool to generate Python project templates with one click, integrating README, LICENSE, project structure, dynamic configuration, CI workflows, and other basic configurations for automated project scaffolding.
Explores the problem of encountering CPU bottlenecks when handling I/O-intensive tasks in Python, and proposes a solution combining multithreading and multiprocessing: using thread pools in multiple processes to fully utilize CPU resources, solving the problem of high single-thread CPU usage.
How do AI agents work? The article explains in an accessible way, revealing that agents are actually directed graphs composed of nodes, branches, and loops, and demonstrates with code how to define nodes, build, and execute an agent.
A story about Google's journey from the crisis triggered by ChatGPT's release to re-establishing AI leadership, featuring detailed recollections from 50+ employees/executives, showcasing Google's turbulent two-year period, the trade-offs and decisions made.
Projects & Resources
× 12This week's super hot project - a text-to-speech model with only 1.6B parameters that can generate highly realistic dialogue, including laughter, coughing, throat clearing, etc. (11K stars)
Microsoft's open-source agent beginner course, teaching you to develop an AI agent from scratch. Includes Chinese translation. (17.2K stars)
A reproducible Jupyter notebook toolkit powered by uv, supporting PEP-723.
Visually define database models through a UI interface, select services to integrate, and quickly generate complete containerized projects with tests and APIs.
A Xiaozhi voice client implemented in Python, aimed at learning through code and experiencing AI Xiaozhi's voice features without hardware requirements. (1.1K stars)
Presentation materials from a talk at PyCon Taiwan 2024.
A financial management system developed using the Django Web framework, supporting double-entry bookkeeping, hierarchical chart of accounts, financial statements, purchase orders, financial ratio calculations, account management, OFX/QFX file imports, etc. (1.1K stars)
A community of agents where you can create powerful agents with one sentence, freely combine with other agents, and collaborate to complete complex tasks.
Uses AI to analyze GitHub codebases and create beginner-friendly tutorials with clear visual diagrams, explaining in detail how the code works. (5.9K stars)
A project for AI engineers that provides visual workflows to more efficiently develop, test, and deploy AI agents. (4.7K stars)
Uses Microsoft's Florence-2 model to precisely detect watermarks and the LaMA model to erase watermarks and naturally fill in the areas. Provides CLI and PyQt6-based graphical interface.
An agent for web automation that supports multiple large models, intelligently operates web elements, customizes browser settings, and provides real-time session updates with operation recording. (1.8K stars)
Podcasts & Videos
× 2core.py podcast episode 21, diving deep into CPython garbage collection topics. Other topics include: whether to do type checking on the standard library itself, free-threading development progress, http.server HTTPS support, etc.
"Latent Space: The AI Engineer Podcast" episode featuring the creators of MCP (Model Context Protocol), discussing the protocol's origins, challenges, and future development.
Discussions & Issues
× 2What to build for a graduate course project, and how to make it innovative?