Telemetry is not correctly stopped by disabling the checkboxes
Categories
(Toolkit :: Telemetry, defect)
Tracking
()
People
(Reporter: eez0, Unassigned)
Details
User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:68.0) Gecko/20100101 Firefox/68.0
Steps to reproduce:
Disable every checkbox in Firefox Data Collection and Use .
A while later, check about:telemetry, and even though it's disabled, there are many pings requests. Easily reproducible by closing Firefox, as a ping of type "main" will be created with "reason: shutdown".
Actual results:
Debugging with Wireshark, it seems the requests never leave my computer to any of the IPs of incoming.telemetry.mozilla.org.
However, they shouldn't be even stored locally if telemetry is disabled.
Expected results:
No telemetry ping should have been detected in about:telemetry if all checkboxes are disabled.
Also, even though those checkboxes are disabled, there are plenty telemetry settings set to true. The following:
toolkit.telemetry.archive.enabled
toolkit.telemetry.bhrPing.enabled
toolkit.telemetry.firstShutdownPing.enabled
toolkit.telemetry.newProfilePing.enabled
toolkit.telemetry.shutdownPingSender.enabled
toolkit.telemetry.unified
toolkit.telemetry.updatePing.enabled
Comment 1•6 years ago
|
||
Bugbug thinks this bug should belong to this component, but please revert this change in case of error.
Comment 2•6 years ago
|
||
Hi!
Unchecking the preference in about:preferences#privacy that says "Allow Firefox to send technical and interaction data to Mozilla" stops Firefox from sending technical and interaction data to Mozilla. It's the only thing you need to do if you want to stop that from happening. (Though if you don't mind me asking: why do you want to stop Mozilla from getting anonymous usage statistics from your Firefox? We spend a lot of effort making sure we collect only what we need to answer questions we think are important, so if there's something that doesn't feel right about that I'd be interested to hear what it is.)
Due to the implementation history of the Telemetry data collection library (part of which can be traced back to the Firefox Health Report), data collection and archival was once viewed as pertinent and useful to users who might be interested to know, for example, how quickly or slowly their Firefox has been starting. Extricating that behaviour from Telemetry has always been lower priority than fixing bugs. As such that code keeps working in the background, though it will not send anything, and even if you check that preference back on, nothing collected during that time will be sent either.
Those "telemetry settings" toolkit.telemetry.* are internal knobs and switches for configuring various submodules or characteristics of Telemetry. Some are there for the same reason that collection and archival are still on (because removing them was seen as lower priority than fixing bugs), some are there for even worse historical reasons (toolkit.telemetry.unified has to be locked to true or else... actually, I'm not sure what will happen. We don't test that configuration in Firefox Desktop, but we need to support it for Firefox for Android which is tested in that configuration to make sure, e.g., that the core "don't send data" user preference is respected).
None of those should be adjusted by a user because they might cause their Firefox to misbehave in an unpredictable manner. Everything should still work, but since we don't test those configurations I can't guarantee that, for example, turning toolkit.telemetry.archive.enabled to false won't cause your Firefox to start sending pings again.
It shouldn't, right? But the archive and the send module are so tightly interwoven (see here if you like reading Javascript) that I can't guarantee it. We'd need to test that, and writing tests for that will likely be (you guessed it) lower priority than fixing bugs.
That's a whole lot of text all at once. Does this make sense?
Hey Chris, thanks for the reply.
Ok, understood.
About why I don't want telemetry, I think it should be an opt-in feature when I want to join a particular experiment or reporting a specific bug. But by default, every user's data should remain private.
I honestly don't like what's Mozilla doing with Firefox in this sense. Advertising as a privacy-focused browser yet it comes with telemetry services enabled by default, following an opt-out approach instead of opt-in, while other browsers are already shipping with everything disabled by default.
Comment 4•6 years ago
|
||
Thank you for explaining your reasoning. I think I understand.
If I'm right it's a difference in what "privacy" means. I think that Telemetry is collecting data about Firefox, preserving the privacy of the user that is using that Firefox. Whereas (please oh please correct me if I'm misrepresenting you) I interpret your position to be that collecting data about your Firefox is necessarily infringing on your privacy, because it's your Firefox.
If so, those two positions are a tricky pair to find middle ground on. I suspect there's one thing we'll both wholeheartedly agree on, though: user choice matters. You'll always be able to turn Telemetry off and we'll always make that easy to do.
I'm going to mark this bug as WORKSFORME so the triagebot doesn't get angry when I don't assign it a priority, but you and I can keep commenting if you'd like to continue this conversation (or you'd like to correct my incorrect interpretation of your position)
Description
•