Opt-in / Opt-Out? A breakout topic for the F40 Change Request on Privacy-preserving telemetry for Fedora Workstation

I don’t see a future with this proposal either, but I will reserve my judgment until I see what makes it to a vote. Whatever beef you may have at this point with the proposer is not really adding to the discussion, and everyone knows where you stand. Essentially you are approaching that point where you are rsiking becoming the person trying to drown out others. My point is maybe waiting and seeing is in everyone’s interest instead of revisiting the same points for the 100th time.


I understand that people are unhappy with this proposal. I am unhappy about it too and I have shared my opinion endlessly in all the various topics.

However, it isn’t OK to attack the author or his character. This violates our community standards.

There is a big difference between attacking the proposal and attacking the author.


Sorry I didn’t want to attack anybody, only the proposal + additional ideas itself.
I’m writing in an pretty direct style, mostly because those replies (bad first impression I guess), I personally don’t think any OP should reply to good comments about his proposal in such a way (my opinion, no personal attack), I also stopped until the last quote. I find this reply personally attacking, it’s making fun of everyone, that pointed the usage of those patterns out, it’s not a topic to make fun, especially not in this proposal!

Now that we got that out of the way.

I don’t like the colored button idea either. It’s simply another psychological trick.
The user count only gets higher than by asking, by the usage of dark-patterns or psychological tricks.
Those should not be used in an distro which says “It’s your OS”. Because it really isn’t then.

I also don’t know why so suddenly you don’t want to show what data will be collected, linking to a website or submenu of some sort only increased complexity, next psychological tricks to get less people to read it. And “Gnome is minimal” is not an argument, the about page is also an list which looks good and minimal.

Simple “how to ask about telemetry without tricking people” (if I didn’t miss something :crossed_fingers:):

  • Say what will be collected, ON the page you ask (not one/two clicks away and definitely not on an Website lol).
  • Two buttons: no, don’t send, yes, send. Not preselected, the same color. Equal positioning. Not skippable which selects any default.
  • Don’t change what data will be collected frequently or if so ask every time again for consent.

I think the way this proposal wants to include telemetry is fundamentally wrong.
Until this point, I read really good ideas by others but none got adapted by the OP (again my opinion, no personal attack)
I have just one request:
Let the user decide, they can choose for themselves.

I really have to stop reading the new changes of the proposal, there only getting worse at this point. See ya after FESCo (hopefully) rejected this :wave:


But still, you are trying to influence them in this way.

I think that in the hundreds of email is clear that many people do not agree with the premise that this proposal is giving enough value to the users to risk any kind of privacy compromise. Therefore I think the only possible way forward is pure opt-in with no dark pattern or - as maximum compromise choice - explicit choice with no dark pattern whatsoever.

Alright, as I’ve said a few days ago, I don’t agree with opt-in telemetry. I published an article that goes over the issues with opt-in telemetry to showcase my opinion and why I only support opt-out telemetry for collecting data and drawing conclusions based on that:


I think you would feel more at home working with Microsoft, Apple, Meta, Google, etc…

1 Like

Thanks, I would love to get a lot of money to fund wonderful FOSS projects, developers and activism :slight_smile:


Well, then it’s settled.

If opt-in telemetry is useless and the community does not want opt-out telemetry, we’ll go with no telemetry at all. Fedora will continue to improve with every release, community trust in the project will not take a hit and there will be no mass exodus to other distributions. :star_struck:


Absolutely 100% this!

Queue the responses, “we can’t know what the community wants because asking them is ineffective”. We need telemetry for that, insert eye roll emji

1 Like

Well, I see, your points are absolutely fine and correct. The issue here is that you have a fundamentally different scale of values.

You value more developer action to improve the users, at the cost of tricking a mass of people into opting-in.
Many community Members here—those against the opt-out spirit—believe instead that telemetry is risky, and users should never ever be tricked into something risky. Moreover, many others and I understand that telemetry might be useful, but it cannot be never ever put before the respect towards the user. (By the way: opt-out system would still not get representative data, you only get data from more people but this alone is not enough to guarantee that the sample is representative).

This is a fundamental difference in our points of view.


Did you read Hari’s post? Whether or not you agree, they provide a long, thoughtful, case. Just saying this sarcastically is not … helpful.

1 Like

Exactly, those who would opt out on principal are not represented at all.


I believe there is a 100% valid point in that sarcasm. I also believe it is VERY thoughtful if you bother to think about the statement. I think it would be very helpful to those that understand the point. The eyeroll bit may be a bit much, but warranted.

Yes, I read it.

1 Like

I also don’t see the proposals author get reprimanded for sarcasm


The place to discuss Hari’s blog is Hari’s blog.

1 Like

To a large extent the argument made is that you can’t know valid, useful data unless you implement telemetry without consent. Asking is ineffective, user preference is beside the point. The data is more important than the user. I disagree. I feel if you are ignoring users to begin with then you aren’t even capable of appropriately interpreting the data (you don’t understand user behavior, whether you have raw numbers or not). At which point you are better off without it.

1 Like

Because “tricking people” has a clear moral judgment, this isn’t a fair way to represent an opposing view, and therefore, does not really advance discussion. Or, to put it another way, don’t tell people what their values are. It’s unlikely to be right, and at best just scores points from people who already agree with you. It doesn’t change minds.

Instead, I think two of the viewpoints I’ve seen discussed go something like:

  • The data that would be collected is not deeply personal or sensitive, and not attached to a user identity. It is low risk, and we will making a genuine, reasonable effort to make sure people know about the choice they are making, with easy access to exact details. Experience and research show that presenting a “no” default gives poor data. People choose “no” not because they’re tricked, but because they don’t care and are following what is perceived as a recommended choice. This data will be used to make Fedora better, and users who do think there is a downside (or have any situation where sharing is not appropriate) can easily choose to not participate. We can demonstrate our commitment to privacy through transparency and oversight.
  • This data must be considered personal, and there is a real privacy and security risk in collecting it. If even a few people choose to share this information by accident — whether through confusion, lack of education on the topic, or simply because they overlooked it — that causes harm, which we must never do, no matter the benefit to anyone else. Beyond this moral risk, presenting a “yes” default or even steering people in that way through language which encourages that choice damages the reputation of Fedora as a project which cares about privacy.

As far as I’ve seen, these are the “edges” of the discussion. I don’t think anyone is saying anything more extreme (no “tricking people is great! we want personal data!” or “Fedora must not use DNS, because it does not preserve privacy”). But, as many people’s comments (and the poll) tell us, there are also many views across the middle of the span. too.


If it’s actually what happens how is it not fair? How is it fair to say an accurate description shouldn’t be used because it carries a negative connotation? It should be fair to accurately describe an action/proposal. That actually unfairly biases the proposal as you aren’t allowing it to actually be considered on it’s actual merits.

1 Like

The issue isn’t that is has a negative connotation. The issue is that it assumes an intention. Since we don’t actually know what the intention is, that assumption is unfair.


It does not assume an intention at all. 100% false. It assumes a result. Which is 100% fair.