asgi-gzip 0.3

2026-04-09 · Source: Simon Willison's Weblog · Field: Technology & Digital — Software Development & Engineering, Cloud Computing & IT Infrastructure · Depth: Intermediate, quick

Summary

asgi-gzip 0.3, a gzip middleware for ASGI applications, has been released with a critical fix for Server-Sent Events (SSE) responses. The update addresses an issue where previous versions incorrectly compressed `text/event-stream` responses, which was identified during a new feature deployment to a production Datasette instance. This problem stemmed from a lapsed GitHub Actions workflow designed to port updates from Starlette, from which asgi-gzip was originally extracted. The workflow has been reactivated, and the necessary fix, originally implemented in Starlette, has now been integrated into asgi-gzip and datasette-gzip, ensuring correct handling of SSE responses.

Key takeaway

For DevOps engineers deploying ASGI applications with streaming features like SSE, you should verify that your gzip middleware correctly handles `text/event-stream` content. Regularly audit automated dependency update workflows, especially for extracted libraries, to ensure critical fixes are not missed, preventing unexpected compression issues in production environments.

Key insights

Regularly verify automated dependency update workflows to prevent critical production issues.

Principles

• Middleware can inadvertently break streaming protocols.
• Automated update checks require active maintenance.

In practice

• Implement scheduled workflow monitoring.
• Test middleware with streaming protocols like SSE.

Topics

• asgi-gzip
• Starlette
• Datasette
• Server-Sent Events
• Gzip Middleware

Code references

• simonw/asgi-gzip
• simonw/datasette-gzip
• Kludex/starlette

Best for: Software Engineer, DevOps Engineer

Related on AIssential

• Running Python ASGI apps in the browser via Pyodide + a service worker — Python ASGI web apps can run serverless in browsers via Pyodide and service workers.
• I Thought Data Engineering Was Just Writing Scripts. I Was Wrong. — True data engineering extends beyond scripting to build reliable systems through idempotency, persistence, and scheduling.
• I Gave My AI Server Access. It Became A Great DevOps Partner. — AI tools can serve as effective DevOps partners, extending beyond traditional coding assistance.
• RAG Observability with Langfuse, vLLM, and FAISS — Production-grade RAG requires end-to-end observability to diagnose issues and ensure consistent, explainable outputs.
• Building Your First Streamlit App: Uploads, Charts, and Filters (Part 2) — Streamlit enables rapid development of interactive data apps with uploads, visualizations, filtering, and export.
• Synthetic Monitoring Explained: A Guide to Reliable DevOps Workflows — Synthetic monitoring proactively validates critical user journeys and system health by simulating user actions.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.