This is great news, thank you, I was just thinking how much I'd like some affiliate links in my browser. Also many thanks for all the telemetry improvements listed in the changelog, those are also something I've been looking forward to for a long time. If this doesn't solve Firefox's abysmal market share, I don't know what will!
The important thing is that this allows them to avoid making a paid version of Firefox available to those that want to support it, or even taking donations that go to Firefox development. Because...I'm sure they have reasons.
> Also many thanks for all the telemetry improvements listed in the changelog, those are also something I've been looking forward to for a long time. If this doesn't solve Firefox's abysmal market share, I don't know what will!
Telemetry is how they figure out popular or blocked addons, failing pages, browser crashes, etc. Part of a managing popular software is indeed getting information like that. Because general public isn't keen on filling out forms by hand.
People on HN don't object to telemetry because they don't know what it might be used for. It's patronizing and useless to just respond with a naive psuedo-explanation.
+1, very patronizing and misleading. I dare say harmful, not useless. Plants seeds for the same tired arguments. Again. Helpful, not required.
edit: My 'argument' against it: when is enough? At what point does it become bike-shedding? Tricky to answer; isn't one. Transparency is paramount.
I'm for Telemetry, generally, but enjoy the universal ability to opt out. Either through your mechanisms or mine. At a point it becomes a matter of trust and need. More feedback loops, if you will.
Trust isn't necessarily regarding privacy, either. What about simple efficacy? For all of the Telemetry today, I don't see substantially more-effective software than what we had in the 1980s. More profit, sure - but not necessarily effective or reliable. It's possible - and reasonable - one may not trust Telemetry is going to good (read: effective) use.
The requirements/standards between a mobile game powered by endorphins and say, a web browser, are considerably different. Sorry, they were patronizing first.
No one on HN has any objection to explicit opt-in usage statistics. Telemetry is not the problem. Lack of consent is the problem. Software developers sliding themselves into end users' local machines without invitation is the problem—devs who feel entitled to take things, simply because they have the technical access to take them. And devs who don't even understand the difference.
I don't understand your comment? To clarify my own: it's my observation that I've never seen anyone on HN utter a word against opt-in telemetry models, like the one Debian uses.
Of all the diverse viewpoints, FOSS extremists and privacy fundamentalists, the core facet of "you can submit statistics to our popularity survey if you want" is not something I've ever seen anyone object to. The objections are always to some other facet of telemetry.
First, fingerprinting & correlation is real, and it's easier than ever because of the identifiers tools collect, and correlating is cheap because we have tons of GPUs for cheap now.
On the other hand, I personally don't like a pair of eyes looking over my shoulder, physical or digital.
People don't understand or don't want to understand that digital telemetry is not different from somebody looking over your shoulder, taking notes, and making "Mhm..." sounds.
When done without consent, both are equally invading personal space, and I don't want my personal space to be invaded like that, as a person on HN, who understands what telemetry is.
Part social contract, part the knowledge of what I can do with that data.
You know, somebody coming to your cubicle and taking something without asking is rude, in some cases unethical, even damaging. It's the same thing with computers. No application/website ever should be able to collect information about me without my consent. It's called a "personal computer".
If the application needs telemetry from me, they can ask (not tell we're doing this, but ask), and start the moment I consent. The moment I withdraw my consent all telemetry and data collection should stop. Moreover, I shall be able to see all the telemetry data in its full glory to make my own assessment of what's collected and how.
If you don't know, Go tried this, even with an arguable provable way of anonymization, people roared back, and the decision is changed to "opt-in".
For some applications, I explicitly consent to telemetry because a) They ask, b) They show what they collect and do with it, and c) I do respect and trust them because of the previous encounters I had with them.
The moment they break this social contract, they lose telemetry, and in most cases me as a user.
I mean, non-consensual telemetry is just recording what you do via your camera all day long, but with light off, without telling you. Replace the camera with any app you use.
Are you comfortable and happy now?
Of course it can be useful & good, but letting people know about what you are doing and asking before doing it is even better, no?
Do the telemetry collectors have something to hide, so they do it covertly or without consent?
> The moment I withdraw my consent all telemetry and data collection should stop.
You can do that by simply going into about:preferences#privacy and unchecking "Allow Firefox to send technical and interaction data to Mozilla"
> I shall be able to see all the telemetry data in its full glory to make my own assessment of what's collected and how.
That's all available in about:telemetry
Firefox is open source and so is the set of telemetry it collects, the schemas are published openly on the web. Every request to add new telemetry is reviewed in a public bugzilla ticket. It's not a mass of dark surveillance capturing unknown information, it's all done in the open in public.
Then the same people shouldn't make fun of telemetry as unrelated to the success of a product. Also, yes, I had this exchange with many people who don't understand what telemetry is used for, here on HN. HN is big enough that generalising the audience just doesn't work anymore.
Recently Firefox has been especially unreliable loading pages, sometimes pages load extremely slowly while everything is fine on other browsers, sometimes they just don’t load until the browser is restarted, sometimes it happens again right away after doing so.
It happens across devices and profiles, so I assume more people will be experiencing it.
Instead of improving, it has gotten worse over the last couple of years.
Thing is: You have to use the telemetry, you have to actually care. And how much Mozilla care is illustrated well by the plenty 20+ years old open tickets for problems that exist to this day.
They can tell what's popular by what's downloaded from addons.mozilla.org, they don't need telemetry for that. I'm not sure how to read "blocked", but the menu item to report an extension is immediately next to the item to remove it, so I also don't think you need anything more for that.
> failing pages,
I certainly hope they're not sending Mozilla a list of pages I'm visiting, failed or otherwise.
> browser crashes,
Does a restart not still pop up a dialog to ask the user if they want to report the crash? You don't need "telemetry" for that.
Nightly allows uMatrix where Fennec does not. But Nightly has default HSTS which is annoying to users that monitor own traffic on own networks. The number of Mozilla-initiated phone home domains one has to filter when using Firefox is nuts. It should be zero. Option to disable all these unsolicited remote connections quickly and easily should exist.
