this post was submitted on 14 May 2024
125 points (97.7% liked)
Firefox
17752 readers
1 users here now
A place to discuss the news and latest developments on the open-source browser Firefox
founded 4 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
People really need to kill that notion that telemetry is automatically bad. If the information they are collecting is minimal, as non-identifiable as possible and actually being used to help develop the browser, it's a good thing.
Yes, turbo nerds in the back, specially being opt-out, opt-in telemetry is pretty much useless for trying to understand the majority of your user base.
There is an actually moral alternative to opt-out that doesn't have the poor-sampling problem of opt-in: ask for consent explicitly.
It's the ideal solution morally-wise, but it still samples out a ton of users precisely because people are used to the idea of telemetry = bad
I wonder why users would have such a bias, other than their experience over the last 25 years.
I’m a developer. I side with the users on this.
And that's a good thing. Because that is the decision by the user. The freedom of choosing in opt in fashion is much more important than collecting some individual data for a specific use case of a specific company. Opt in is not just ideal solution morally-wise, its the best solution we have in general and every company should strife this solution. Plus the data should be presented before sending, so there is no ambiguity. Steam, a closed source program, does that in the best possible way.
Syncthing is one of the best examples of telemetry done well. On first startup, they ask if you agree to enable telemetry, they show the data that will be send and inform users that the collected data can be viewed at https://data.syncthing.net/
I agree with you. There are projects where I opt in and enable telemetry, such as KDE or opt in the Steam survey whenever asked. Steam in particularly does a good job on representing the data in front of me that is sent back.
Problem is, its a bit ghosty what is actually being collected and sent for most people. Is it really non-identifiable as we think now? You know, sometimes later things get revealed and suddenly the entire time you was living in a lie (Privacy mode thing, where people had a misconception). If its enabled by default, this is especially bad, because this should be opt in. Telemetry is not bad per se, but it is bad if its enabled without user agreement.
Wrong. In example Steam does an opt in and the data is somewhat representative. You don't need to watch every user to know what is going on. A small sample is enough to understand the majority of the user base by extrapolating the data. Telemetry does not need to be exactly perfect to be useful, it just needs to help understanding trends or huge bottlenecks.
Yeah I normally opt out of all tracking or telemetry, but when it’s a project that I feel like I can trust and want to make better I make sure to turn it on.
That must be why Mozilla and Microsoft famously serve the needs of their users so well.
Read what I said again. It is not automatically bad, and it doesn't mean it can't be poorly used or poorly understood by the ones collecting it. It just means that it is an effective way to understand how your users are using your product.
Putting Mozilla (which from what I can tell is doing as much as they can trying to collect this telemetry data in a way that can't be used to identify its users) in the same domain as Microsoft, which collects pretty much everything it can to sell to third party advertisers is ridiculous as best and disingenuous at worst.
They have much in common when it comes to telemetry, in that they both collect quite a lot of it and spend much time and effort to analyze all that data so as to improve the user experience.
I hadn't really considered the advertising angle, but now that you mention it I'm sure advertisers would also find all this thoroughly privacy-respecting anonymized data to be of interest when they're considering the idea of paying for promotion through Firefox Suggest. Mitchell Baker may no longer be in charge of it, but there must still be some highly placed people over there who are fully on board with her vision of turning Firefox into a better advertising platform.
In case of Microsoft, this is a whole new dimension and not comparable to Mozilla. First Microsoft products are (usually) closed source. That alone is a black box and we don't know what is sent, compared to open source Mozilla projects we can actually understand what is going on and report. Secondly, Microsoft does it not only with the browser, but on the entire operating system, if you want it or not. It's not opt in, not opt out, its just selecting a few options to sent a few less data, that's all. Which BTW reset themselves sometimes for unknown reasons.
Putting Mozilla and Microsoft in the same sentence about privacy and telemetry is heresy (towards Mozilla)!
So which organisation with many userse serves the needs of their users better without collecting data?
Most free software does not have telemetry, and when it does it's almost always opt-in. Firefox is the one major exception to that rule.
Hmm, so what user-facing free software is at Firefox's scale? I think Ubuntu has telemetry, for example (though I think they even have fewer users).
Ubuntu telemetry is fairly minimal, as of last time I used it a few years ago. Not remotely comparable to what firefox does. They just want to know what hardware you have, there's no user behaviour tracking, and it's fully opt-in (you have to deliberately turn it on when installing). KDE and Gnome have a little something like that as well now, I think. Almost everything else does not.
Debian has a list (last updated 2023-10) of software among the 97000 packages they distribute which have been found to violate user privacy by "phoning home" for telemetry or other purposes:
I mean, that depends on how you define user behaviour. It tracks which packages are frequently installed, for example, or how often people install Ubuntu in the first place. All of which I think is pretty legit, in my opinion, since that only involves aggregate user statistics that help prioritise work and detect common problems - but that's essentially what Firefox is doing too.
Debian is a great example of relatively commonly used free software that doesn't really collect data btw.
Canonical apparently turned on enabled-by-default telemetry for new installs in 2018 which records basic system hardware stats and such. It's not that much compared to what Firefox sends, but adding it still did damage to their reputation.
Another thing Ubuntu has in common with Firefox is a continuing long-term decline in market share. As they do things like adding telemetry, flirting with the idea of putting advertising in the package manager, insisting that everyone use snap, et cetera, users have started to go elsewhere. As I did.
In the case of Ubuntu though, the company's main business is in serving their corporate customers. If it's little-used by the rest of us the company might still do well, as I hope they continue to do. Firefox does not share that advantage.