On Telemetry in Software

This post about telemetry in Go’s tools recently appeared in my feeds. I don’t specifically care about Go or its tools. I just wanted to mention something about the use of telemetry in software. It is considered common practice to implement some sort of telemetry package into one’s software nowadays. I am not here to debate whether this is a good idea or if it is effective.

All I care about in this specific topic nowadays is consent.

I’m even at the point where I say that I as a normal developer using Go want the Go team have that kind of data. But now to the one thing about this proposal that I don’t like: It’s opt-out.

Horst Gutmann, Telemetry in the Go tools

Telemetry in software should always be opt-in by default and not opt-out. Furthermore, the opt-in process should provide details on what exactly is being collected and a summary of how the data will be used. This is so that the user can make an informed choice as whether they want to opt in or not.

No, I don’t care if developers and/or other stake holders think that if they ask for telemetry instead of simply turning it on and making the user opt out means that most of them wouldn’t provide any telemetry. Tough shit, that’s how consent works. You are not entitled to telemetry.