4 ms·

The problem with opt in is that people don't opt in. Not because they don't want telemetry, but because they don't care. Most of them are happy to enable it onc

The problem with opt in is that people don't opt in. Not because they don't want telemetry, but because they don't care. Most of them are happy to enable it once you describe what it is and how it works. The middle ground is having a clear question where the telemetry option and the no-telemetry option is clearly shown and the telemetry is preselected. So you have to do a choice but the "I don't care" next > next > next choice is going to be to opt-in. Whether this is acceptable also depends on the nature of the program and the users. It's much easier to get acceptance for this if it's a piece of software that will make 100 requests to 10 different servers during normal operation (where all of those servers will likely know program version and even more client info anyway), but now makes 101 requests to 11 servers when you enabled the telemetry. A program that used no network requests at all without telemetry should probably be strictly opt-in since its such a major change in behavior.

seanhunter 2y ago
> The problem with opt in is that people don't opt in. Not because they don't want telemetry, but because they don't care. That's not a problem, that's the correct behaviour. Privacy should be the default setting.
- Cthulhu_ 2y ago
  Exactly. You only need a percentage of your users to opt-in to get statistically significant metrics. No means no.
  - marcosdumay 2y ago
    Hum... As soon as it's optional¹, the odds of you never getting statistically significant data becomes quite large. What is probably a reason for not collecting anything that you'll use for statistical analysis. Bug reports is an example of something that don't need statistics, things like those can still be useful. 1 - Doesn't even matter if opt-in or opt-out, or even if it's "optional" due to firewalls or a site somewhere explaining how to hack your software.
  - alkonaut 2y ago
    Opt-in probably means you‘ll never get a good sample size. It has to be representative. If it’s 5% there’s a risk of it being unrepresentative even if it’s a large sample. The risk is a particular type of user or users from a particular region, skill level or age group opting in.
    - seanhunter 2y ago
      Well that's a risk the developers just have to take. It's not even remotely close to being worth sacrificing their entire user base's privacy for. It blows my mind how far the debate has shifted that people think ubiquitous telemetry is even slightly reasonable just because the devs need representative sample of metrics. They really don't. They will be just fine without any metrics at all.
      - alkonaut 2y ago
        > sacrificing their entire user base's privacy for. I keep asking what the privacy issue is in anonymous usage statistics, but there are never any answers. Remember: The "anonymous" in "anonymous usage statistics" means there can't be a privacy issue. Then it's not anonymous any more! You could think of it as for example, sending only the program version in, and the only stored data is how many sessions are run of each version, every month. So you could know "is it reasonable to deprecate version 1.0 and only support 2.0?". What exactly is it that you think is an invasion of privacy here? The fact that it makes http requests? > They will be just fine without any metrics at all. Most who added telemetry has a time from before metrics and a time after. And I imagine very few would want to go back. It's driving blind. You can't simply "ask your users". They don't know whether they use Ctrl+C or use Copy on the context menu. And if you ask them, you find that they think they did one thing when in reality they didn't. Knowing what they actually do instead of what they think they do or say they do is invaluable. And luckily at least in enterprise, users are pretty happy to supply this, but it's definitely based on a trust that already needs to exist between seller and customer.
- alkonaut 2y ago
  If telemetry (in the context of software that’s already is making web requests to my server) if somehow at all a privacy problem, then it’s not anonymous telemetry. Anonymous telemetry is be a non-issue for privacy.
account42 2y ago
> Most of them are happy to enable it once you describe what it is and how it works. This is going to depend very much on how you describe it. > The middle ground is having a clear question where the telemetry option and the no-telemetry option is clearly shown and the telemetry is preselected. So you have to do a choice but the "I don't care" next > next > next choice is going to be to opt-in. That's not a middle ground at all but very biased towards data collection. The middle ground is to play it safe and not collect any data where you are not sure you have informed consent, which includes users who do not understand the question or do not care to read it. The better option would be to not add the data collection at all. > Whether this is acceptable also depends on the nature of the program and the users. It's much easier to get acceptance for this if it's a piece of software that will make 100 requests to 10 different servers during normal operation (where all of those servers will likely know program version and even more client info anyway), but now makes 101 requests to 11 servers when you enabled the telemetry. A program that used no network requests at all without telemetry should probably be strictly opt-in since its such a major change in behavior. It sould be strictly opt in for ALL software. Otherwise you are writing malware.
- alkonaut 2y ago
  In what sense is anonymous usage stats “malware”? Because it does something you didn’t expect it to do? It did say on startup that it collects anonymous usage stats. It’s not “quietly enabled and opt-out has to be searched for”. It says front and center what is going on. You are free to exit at that point and not run it at all. Or just choose the no telemetry option. It’s basically more benign than 99% of web apps. It’s not a sneaky secret feature that does something behind your back. It’s not transmitting any information it wouldn’t transmit anyway. No information is stored that can identify a user either by fingerprinting or directly by some identifier.

seanhunter 2y ago

> The problem with opt in is that people don't opt in. Not because they don't want telemetry, but because they don't care. That's not a problem, that's the correct behaviour. Privacy should be the default setting.

Cthulhu_ 2y ago
Exactly. You only need a percentage of your users to opt-in to get statistically significant metrics. No means no.
- marcosdumay 2y ago
  Hum... As soon as it's optional¹, the odds of you never getting statistically significant data becomes quite large. What is probably a reason for not collecting anything that you'll use for statistical analysis. Bug reports is an example of something that don't need statistics, things like those can still be useful. 1 - Doesn't even matter if opt-in or opt-out, or even if it's "optional" due to firewalls or a site somewhere explaining how to hack your software.
- alkonaut 2y ago
  Opt-in probably means you‘ll never get a good sample size. It has to be representative. If it’s 5% there’s a risk of it being unrepresentative even if it’s a large sample. The risk is a particular type of user or users from a particular region, skill level or age group opting in.
  - seanhunter 2y ago
    Well that's a risk the developers just have to take. It's not even remotely close to being worth sacrificing their entire user base's privacy for. It blows my mind how far the debate has shifted that people think ubiquitous telemetry is even slightly reasonable just because the devs need representative sample of metrics. They really don't. They will be just fine without any metrics at all.
    - alkonaut 2y ago
      > sacrificing their entire user base's privacy for. I keep asking what the privacy issue is in anonymous usage statistics, but there are never any answers. Remember: The "anonymous" in "anonymous usage statistics" means there can't be a privacy issue. Then it's not anonymous any more! You could think of it as for example, sending only the program version in, and the only stored data is how many sessions are run of each version, every month. So you could know "is it reasonable to deprecate version 1.0 and only support 2.0?". What exactly is it that you think is an invasion of privacy here? The fact that it makes http requests? > They will be just fine without any metrics at all. Most who added telemetry has a time from before metrics and a time after. And I imagine very few would want to go back. It's driving blind. You can't simply "ask your users". They don't know whether they use Ctrl+C or use Copy on the context menu. And if you ask them, you find that they think they did one thing when in reality they didn't. Knowing what they actually do instead of what they think they do or say they do is invaluable. And luckily at least in enterprise, users are pretty happy to supply this, but it's definitely based on a trust that already needs to exist between seller and customer.
alkonaut 2y ago
If telemetry (in the context of software that’s already is making web requests to my server) if somehow at all a privacy problem, then it’s not anonymous telemetry. Anonymous telemetry is be a non-issue for privacy.

account42 2y ago

> Most of them are happy to enable it once you describe what it is and how it works. This is going to depend very much on how you describe it. > The middle ground is having a clear question where the telemetry option and the no-telemetry option is clearly shown and the telemetry is preselected. So you have to do a choice but the "I don't care" next > next > next choice is going to be to opt-in. That's not a middle ground at all but very biased towards data collection. The middle ground is to play it safe and not collect any data where you are not sure you have informed consent, which includes users who do not understand the question or do not care to read it. The better option would be to not add the data collection at all. > Whether this is acceptable also depends on the nature of the program and the users. It's much easier to get acceptance for this if it's a piece of software that will make 100 requests to 10 different servers during normal operation (where all of those servers will likely know program version and even more client info anyway), but now makes 101 requests to 11 servers when you enabled the telemetry. A program that used no network requests at all without telemetry should probably be strictly opt-in since its such a major change in behavior. It sould be strictly opt in for ALL software. Otherwise you are writing malware.

alkonaut 2y ago
In what sense is anonymous usage stats “malware”? Because it does something you didn’t expect it to do? It did say on startup that it collects anonymous usage stats. It’s not “quietly enabled and opt-out has to be searched for”. It says front and center what is going on. You are free to exit at that point and not run it at all. Or just choose the no telemetry option. It’s basically more benign than 99% of web apps. It’s not a sneaky secret feature that does something behind your back. It’s not transmitting any information it wouldn’t transmit anyway. No information is stored that can identify a user either by fingerprinting or directly by some identifier.