Hacker News new | ask | show | jobs
by BlackLotus89 1738 days ago
How can anyone call firefox privacy focused when they use telemetry so fucking heavily? Per default telemetry is active, disable it and you still got telemetry/pings whatever. You have to opt out of everything. It's not even limited to the user side look at this https://bugzilla.mozilla.org/show_bug.cgi?id=1460678#c20
3 comments

And again, you throw out the baby with the bathwater and go back to Big Google that does worse on every privacy-related issue by a mile.

Like, no matter what, telemetry is useful to the product, and defaults matter. Like the infamous “the opt-ot organ donor vs opt-in countries have a staggering difference of 90% difference”. Should firefox throw away 90% of its userbase’s useful telemetry, most of who would have no problem with providing it?

Oh no I still use firefox and like everyone I make bug reports or follow the existing once. But it's hard to recommend it to friends/family if you have to say "you only need to change these settings and install these addons" and I get responses like "I would rather use brave. It just works" or something like that.

I want firefox to be good. I don't even need to have the newest features I just wamt a stable browser I can use. Don't we all? Removing features I actually use and adding things I don't need is counterproductive and I don't know which part of their telemetry helped them make those choices, but many I didn't like.

> Should firefox throw away 90% of its userbase’s useful telemetry, most of who would have no problem with providing it?

Telemetry should always be opt-in. Firefox throws enough messages your way if they would ask (like many other programs) "Help make firefox better" and then offer different level of reporting it would be fine. They ask me to change my theme at the start, but they don't ask me if I want to send them my data?

>Removing features I actually use and adding things I don't need is counterproductive and I don't know which part of their telemetry helped them make those choices, but many I didn't like.

Well, this is exactly the point. If you disable telemetry, how do you expect them to know that you are using that feature and they shouldn't remove it?

I tell them directly. I actually took a look at what they send about:telemetry and as far as I can tell NONE of the features they removed that I used was captures through this. It's good to collect not too much, but removing features based on imagined data is bad. Maybe I overlooked something, but the only data that could interest them is my addon selection...

Anyway for the longest time I send detailed information with all my crash reports and similar, but not only could I never find out where they collected the crashes I send in no no error (even the reproducible once) got fixed. The only times my problems got fixed where when I actively filed a bug report myself or fixed the cause of the issue manually.

After a few years I began to wonder if anybody even reads those crash reports and added a request for a quick pingback something like "empty message is fine. I just want to know if anybody is actually reading this". Did this _multiple_ times, never got any response. Either they don't read or they don't care. Anyway I'm over giving my data for aggregations that probably never get used.

If someone from mozilla could tell me that they actually matter maybe I would change my mind, but for now I'll drive my privacy is important for me train, because it really is and if the people that collect my data don't or misuse it I don't see the point in sending it.

Maybe I was expecting too much, but if this is too much of a response then I don't want to contribute anything (at least like this).

Correct. An easy win for Firefox is to become a zero telemetry browser by default. All that telemetry is giving them wrong data anyway as users they should be most interested in disable telemetry and are not represented in usage data.
I'm very confused by this comment. If you think of telemetry like a vote, disabling it is essentially forfeiting your vote, so it doesn't really make sense to me to complain that you are then not represented. If you've ever been to a big catered event, this is like a person who refuses to speak up when asked if they want the vegetarian option and then gets mad when only meat is served. You can't expect the organizers to order two of every meal and then throw out the ones people didn't want.

Turning off telemetry by default is also not an option, because then they would have no data, and would just be making decisions at random -- I really doubt that would please you either.

Telemetry is not a vote (for which you are asked to cast). Telemetry is extracting information from you and transmitting it together with private information like IP address without asking you ( as it is opt out by default). Telemetry also costs you in resources (bandwidth, cpu..) which is problematic when resources are scarce. This is a good enough reason for many to want a zero telemetry browser and do whatever they can to disable it in browsers like Firefox.

And now because Mozilla lacks data from its most coveted “tech” users, it does make decisions based on data that it has, which is usually totally opposite (like in your example) which in turns pisses off these users even more and they jump ship, taking all their friends and family with them (because they are the “tech” person in their circle). That is how you lose users.

You can of course make a zero telemetry browser, relying 100% on your own research and what the users volunteer to tell you directly. And those most passionate about it with tell you the most, a wonderful positive feedback loop. But this would require a change of product development mindset to a completely user centric one.

>Telemetry is not a vote

From the developer point of view, this is incorrect, the developers are using it to decide which features to prioritize. If you are aware that it's happening and you leave it on, then I don't see what the problem is. Resource and bandwidth usage should be very minimal, if it's not then I would urge you to actually measure it and report bugs. It should be possible to compress it so that it doesn't eat up your bandwidth. Remember that it also takes bandwidth and CPU to post on Hacker News, so you will have to compare it to that to have any kind of meaningful data.

My point is, it if lacks the data from those users, it would make sense to start sending them that data. It doesn't make sense to me to complain about them having incorrect data on you when you intentionally don't send them the right data and then threaten to jump ship because of what seems to be your own actions. I personally also disable telemetry but I know full well that I'm opting out of an important system for them so I don't expect to get attention in return for doing that. If you wanted to help, I think they would very much welcome attempts to fix the telemetry and make it more bandwidth-respecting and privacy-respecting, rather than finding ways to just throw it out.

>You can of course make a zero telemetry browser, relying 100% on your own research and what the users volunteer to tell you directly. And those most passionate about it with tell you the most, a wonderful positive feedback loop.

In my experience, this is an unreliable way to make products, the type of user who is passionate and volunteers this information is not the average user. You will end up with a very niche product that way, and the cost would be very high since you would be expecting the same quality of features but for a smaller number of users. If you're interested to do this I would suggest you to fork Firefox and attempt to get funding, and try that out just to see how much work it actually is compared to how little those users are actually willing to pay. Take a look at Waterfox if you want to see an example of how this would be done.

> If you are aware that it's happening and you leave it on, then I don't see what the problem is.

Most people are not aware of telemetry. Most of those that are, disable it. So you end up getting what you call 'votes' from the people who are not aware they are 'voting'. That is not voting (for which a person need to be consciously doing it) but rather extraction of information.

Imagine in an election, the votes of those who didn't explicitly vote get automatically extracted and cast based on a biased algorithm produced by the government. If you do not like the idea of that, you should not like the idea of opt-out telemetry. What you want is opt-in telemetry (aka. voting).

> Resource and bandwidth usage should be very minimal

I fully agree, that is the other reason zero telemetry is a way to go and should be default. You can not beat that.

> In my experience, this is an unreliable way to make products

Not sure what your experience is but I already built one company like that and I am doing another one right now (incidentally a web browser) based on this same principle which is called user-centric product development. Btw. Mozilla practiced that too ~15 years ago (the "golden age" of Firefox, reference here https://news.ycombinator.com/item?id=28493855).

>Most people are not aware of telemetry. Most of those that are, disable it.

Do you have some numbers to back this up? I would suggest getting that before making any major product decisions. Also if you read the privacy policy, the telemetry is spelled out in detail, or at least it should be, so the people who are voting should be well aware of what is going on. If nobody reads it, then the solution there would then be to make it obvious and easy to read, not to throw it out completely.

>I fully agree, that is the other reason zero telemetry is a way to go and should be default. You can not beat that.

I don't understand. That isn't a meaningful comparison because you're comparing it with nothing, you would need to present an alternate data source. This to me is kind of like saying "not having a stomach beats having a stomach because you won't need to worry about eating anymore" or something like that.

>Imagine in an election, the votes of those who didn't explicitly vote get automatically extracted and cast based on a biased algorithm produced by the government.

I'm sorry but this is exactly what various governments already do in a lot of cases. Not for general elections but for services or programs that they run or for appointed positions. They will do a study with passive data gathering and determine who is actually using what services, and if the results are good they will increase funding to the service, otherwise they will cut funding, fire people, etc. This is generally how any organization functions at scale, so I really don't understand what you're getting at here or what alternative you're proposing. If you want to apply this to product development, it would simply be infeasible for users to vote on every single ticket that a developer handles, so you need to find some other data-driven way to make decisions. That's what telemetry is put in place to do. I think there is a misunderstanding here of how this works, but it's not your fault.

I wish you luck with your company, but I suspect you will have difficulty getting funding on the level of Firefox or Chrome, especially if you have no hard data from some kind of telemetry or similar source. The privacy-conscious user is known to be a fickle market. I also would advise against making misleading and/or unsourced statements about other browsers if you intend to develop a competing product, this makes your company look bad.

>Most people are not aware of telemetry.

This isn't true, Mozilla is making it very clear the first time you are launching Firefox and you have a button just next to the message to disable it easily without having to go to the settings. I don't know how they could be better about it.

Maybe some people can but I question their commitment to privacy. Mozilla is heavily conflicted, being almost 100% supported by Google in exchange for sending searches to the Chocoloate Factory by default, yet they refuse to openly acknowledge or address the issue. What really gets me is they constantly use privacy as a selling point. Then they try to convince the public that web advertising is a necessity. Its like robbing Peter to pay Paul. You cant borrow your way out of debt.