Python Bytes: Postmodern Errors, PyPy, and AsyncIO Gains
·19m 15s
Shared point
–
Episode Highlights
Postmodern Error Handling
This episode kicks off by exploring the concept of Postmodern Error Handling in Python 3.6. The discussion highlights how developers can leverage modern language features to write more robust, error-resistant code rather than relying solely on traditional exception handling.
- Key tools discussed: Enum classes, named tuples, and Type Hints via MyPy.
- Better code structure significantly aids in preventing logical errors.
Mozilla Open Source Grants
Mozilla is actively supporting the open source ecosystem by awarding significant grants to impactful projects. A major focus for current funding is the Foundational Technology track, which includes:
"PyPy received $200,000 in donations to make that happen. That'll make PyPy a lot more compelling if we can get that updated to 3.0."
Performance Optimization
The hosts dove deep into performance improvements across the Python ecosystem:
- Intel Distribution for Python: Updates now offer massive performance gains, specifically targeting NumPy and SciPy optimizations for multi-core architectures.
- PyPy & AsyncIO: New benchmarks show that PyPy 3.5 is significantly faster than CPython when handling AsyncIO workloads, often reaching 5-10x performance improvements.
- Client-side Async: Using AIOHTTP as an AsyncIO alternative to the standard requests library allowed a performance jump from 12 minutes to just 22 seconds for network-intensive tasks.
Community Milestones
- PyPI Growth: The Python Package Index (PyPI) has officially surpassed the 100,000 packages milestone, reflecting the incredible growth and contribution level of the global Python community.