Comment by slg

Your arguments all ultimately focus on the value telemetry generates for the company, not for the user. These two should theoretically coincide, but in practice, they don't. Telemetry may be a fine thing in the abstract, but it's mostly used in a very hostile manner.

People would be more comfortable with telemetry if they could trust it's being used only to fix bugs and improve workflows. The reality is far from that, though. Telemetry's main use in end-user software is to provide data to direct various aspects of development that ultimately boil down to: how can we extract more money from our users? That's part of the reason we get dumbified apps full of questionable design decisions and user-hostile anti-features. Instead of asking people what software they want, "data-driven" companies are just setting up a control system around their users, with changes in the software being meant to influence behavior towards better monetization.

Until that gets fixed, I'm going to keep preemptively blocking any and all telemetry. I'm also very happy that GDPR forced companies to surface a lot of hidden surveillance, and that I can just dismiss all these notifications knowing I'm legally opted out by default. To the extent I am in fact opted in - i.e. companies literally breaking the law - I yearn for the day DPAs in member states get serious about issuing fines. Until then, the next time I spot telemetry enabled by default, so help me God I'm filing a GDPR complaint.

> You are doing the same thing again. You are assuming a level of "work just fine" without having a comparison for what it would look like with telemetry. Ignoring the privacy issues for a second, can you say definitively that Linux would see no technical improvements from developers having access to telemetry data?

"works just fine" in this case means "is the backbone of the global internet infrastructure". Could it potentially be better with telemetry? Maybe. Could it potentially be better if linus torvalds personally surveils all interactions with any technology i have, no matter how private? Likely. Could it be become better if i stick a probe up my butt to measure frustration when using any product? Sure. What an asinine argument, of course telemetry can make software better in some cases but the global invasion of privacy of literally every computer user is not a worthwhile trade off for some bugfixes and giving POs some rough idea of user interaction to ignore anyway.

> Because the benefits of telemetry are widespread while the downsides are localized. The incentive for an individual user to participate is low and isn't well understood so they will default to off. Expand that to everyone and you end up with the tragedy of the commons.[1] It has nothing to do with skulls on a cap, it is basic individualized economic incentives playing out that lead to less than ideal results for the whole.

the downsides are my privacy and the privacy of millions of user who frankly do not understand the implications of it is invaded for some fringe benefit to the developer. It's not a tragedy of the commons situation but abusive behavior from developers targeting users that don't know any better. Thought Experiment: If every person on the planet would magically gain a deep understanding of how telemetry works, what would the vast majority chose to do? Get it out of their live as much as possible. Would you give someone detailed data where you take your car, at what speed, at what time, with the added benefit of governments gaining access to that data so that you use 5% less wiper fluid?

> Once again you are returning to bugs. This is about more than just bugs. Very few pieces of software are published and then abandoned beyond bug fixes. Today most software needs to constantly evolve and add new features. Maybe you are the type who will request those features from a developer in official channels, but that isn't common.

This has nothing to do with bugs. I don't need google or mozilla to know how i use my webbrowser. It's none of their fucking buisness in any way shape or form. If it crashes enough i will either complain or use a different product. If they want to know what improvements they should make or how they should evolve their product they can ask me. openly, freely and with consent. If 99.999% of users do not care to answer, then that's fine. Just because you can invade my privacy to improve your product or evolve it doesn't mean you should or should be allowed to do so. In fact it should be fucking illegal without explicit, well informed consent.

> Also most users will simply decline when presented with the option to submit a bug report. They just don't see the a strong enough or immediate enough connection between a bug report and the bug being fixed. I would bet any developer who has spent time informally talking to their users would have heard some complaints about their software that were never previously voiced through official channels. That is just the nature of things. A developer will get more valuable data if they don't leave the sending of this information up to the whims of the user in the moment when a bug report screen might appear in front of them.

This is just insane. If a User doesn't care enough about a bug or a crash to fill out a bug report or voice their opinion on it why do think you can just invade their privacy instead? Just because almost everyone can't be bothered to answer surveys on the phone should survey designers just decide to go and analyse everyones trash instead without asking? It's valuable data after all and most people don't answer surveys. Why don't we just go ahead and track everyones movement while we are at it. I'm sure we can improve traffic flow with that valuable data. Just because most People wouldn't like that doesn't mean we cant just invade their privacy because we think we know better.

God, i hope the EU gets their shit together with the GDPR someday and fines devs and companys like that out of existence.