Querying OpenTelemetry traces with Datasette and Parseable

Summary

Simon Willison published a brief 'today I learned' post detailing how he paired his new observability platform, Parseable (an AGPL-licensed Rust binary), with Datasette's recently added OpenTelemetry support. Willison used Codex to connect the two tools and view traces in Parseable's native web interface.

Why it matters

This highlights a growing ecosystem where lightweight, self-hosted observability tools can be directly integrated into developer-focused platforms like Datasette. This approach lowers the barrier to entry for tracing without requiring a full-fledged APM stack. Willison's use of an AI coding assistant to bridge the two tools also demonstrates how quickly developers can prototype observability pipelines. For teams already using OpenTelemetry, a single ~180 MB Rust binary offers an attractive alternative to heavy commercial collectors.

Highlights

• Parseable is distributed as a single ~180 MB AGPL-licensed Rust binary, alongside Enterprise and cloud layers.
• Datasette version 1.0a41 now natively publishes OpenTelemetry traces (contribution by Alex Garcia).
• Willison used Codex to discover the integration pattern working between the two tools.

Critical take

The post is essentially a single-paragraph 'today I learned' note supported by a screenshot. It lacks performance data, comparisons with existing OTel backends (such as Jaeger or Tempo), or any discussion regarding Parseable's maturity and community support. Consequently, it is difficult to assess its practical value beyond satisfying curiosity.