Ask HN: Would you use a tool that auto-adds telemetry to your code at PR time?
1 comments
No, what an awful idea from many reasons, ranging from data protection to operational risks.
And what could possibly lead you to believe that a team that doesn't bother to add telemetry will bother to monitor said telemetry?
And what could possibly lead you to believe that a team that doesn't bother to add telemetry will bother to monitor said telemetry?
Thinking about building an open source tool that runs on PRs and automatically adds structured logging, metrics, and tracing. It would analyze your existing codebase to match your conventions.
The thing I find interesting: if a tool is adding all the instrumentation, it has a model of your entire system. From that you could generate service maps, dashboards, alerting — stuff that’s useful but nobody maintains.
Before I build this: is this actually a problem people have, or do mature teams already have this solved? Would love to hear how others handle this.