Hacker News Reader

> Very interesting distinction which boils down to whether a bullet item might show up in the box or not.

This is the biggest compliment to the UI Apple designed.

Because it's exactly the conversation we should be having, and that users actually care about.

Either you are hitting an API or you aren't. And that causes a bullet item to appear.

What you're doing with that information is debatable -- and Facebook is welcome to debate that they're "really not actually using all the massive amount of information they're collecting."

But users deserve to see the superset of collection clearly, as a starting point.

> Either you are hitting an API or you aren't. And that causes a bullet item to appear.

That's not really how it works.

The same type of data can appear in multiple categories, and there are exceptions that mean you don't have to list certain usages.

https://developer.apple.com/app-store/app-privacy-details/

For example, if your app doesn't ordinarily collect email addresses, but you have a contact form for support purposes that does collect them, you don't need to list email address as being collected.

The privacy label doesn't really cover the app - it covers what the services that the app connects to do. An entirely offline app shouldn't need to list anything on the label, even if it does use APIs like location.

The difference (vs Android) is a default expectation of disclosure, no? Whether enforced or declared?

Google doesn't really care if its app developers lie out of the side of their mouth. Apple is saying they do.

(I'm in no way an Apple fanboy, but I think they deserve kudos for changing the starting point for this discussion)

It's long been the case that companies legally have to disclose what data they collect and what they do with it in their Privacy Policy/Notice. That applies to services on any platform, including iOS, Android, or the web.

I think that the Privacy Label is a "one size fits all" box ticking exercise, that a lot of people are going to ignore/not understand.

When all of the most popular apps state that they collect a lot of data, there's no reason for people to think any given app is "bad".

That was GP's point, IMHO.

Do you think more people read privacy policies from start to finish and understand them, vs will look at Apple's label and understand it?

The perfect is the enemy of both the good and progress.

> That was GP's point, IMHO.

I thought the point was more that until now, app developers have been able to keep how they use data quiet, and that Apple is forcing them to reveal what they do.

That's not really the case, the data is just in a format that people can't be bothered to read (Privacy Notice).

I think that Apple's format is better in some ways, but not others.

There are a lot of possible types of data, but they can each appear in multiple categories. Developers can't justify why they use data in a privacy label, so it's difficult for people to tell exactly how the data collection would affect them (if at all), and whether it's justified or not.

An interesting idea for Apple Pay feature could be to allow its users to and apps to negotiate fees for API hits i.e. allow users to get paid by Facebook if it wants their data.

>Now, gushing praise aside, making sure the labels are truly accurate is the hard part. For example, I’m not sure how iMessage can claim it doesn’t track payments when I use it to actually make payments, but I guess the point is aside from the system that actually processed the payment that iMessage itself is not using any metadata of that payment. Very interesting distinction which boils down to whether a bullet item might show up in the box or not.

It's a classic "lies, damn lies, and statistics" scenario. The graphic might be technically correct but it's misleading in a way to spread a specific message. It irks me that this image keeps being spread all over.

The thing people should actually care about isn't if iMessage stores/uses this information, but if the organization in control of iMessage stores/uses it. It's just like I don't really care if the WhatsApp app stores all this info, the actual problem is if Facebook (the company) uses it for some reason. WhatsApp (the app) just happens to be one of the avenues that this data is collected.

For iMessage, sure, maybe technically iMessage isn't collecting my payment information... but Apple still gets it via the Wallet app. Maybe iMessage isn't storing my location history, but Apple still is through the iPhone location history. Maybe iMessage isn't watching my app usage history, but Apple Screen Time sure is. This applies to all of the things listed as bad under the WhatsApp label.

Secondarily, the graphic is also misleading because the WhatsApp side of it becomes artificially larger since most of the items are listed twice. It immediately causes you to think "WhatsApp collects a lot more" due to the size of the box, but in reality the box could be collapsed to half the size. It just so happens that the way Apple chose to present this information makes it seem like WhatsApp collects twice the data.

Personally (perhaps naively) I do trust Apple with this information more than Facebook, but the graphic is still misleading by implying that it isn't being collected and used in some way.

Does Apple get your payment information? I thought one of the selling points of Apple Pay is that Apple cannot collect your transaction data.

So to avoid confusion: there's two sources of "payment information" that might get confused with Apple. There's Apple Pay (which has no WhatsApp equivalent AFAIK) and Apple Cash (which is the equivalent of WhatsApp Pay).

When using Apple Pay, Apple obviously stores your payment methods and for some cards, the Wallet app stores your transaction history. However, the Apple privacy policy for Apple Pay says they only use anonymized transaction history for their analytics.

But there is also Apple Cash. Apple Cash's privacy policy says that they do collect non-anonymized information on your transactions (and communication history from messages, email, and calls!), but they promise to only use it if they deem it necessary [1]. Take that as you will.

>To assist Apple Payments Inc. in preventing fraud when you are using Apple Cash to send, request, and receive money, Apple Inc. may provide it with approximate use patterns from your device about how frequently you communicate with that person by phone, email, or in Messages. No content of your communication is collected. This information is stored for a limited time, and in such a way that it is not linked to you unless the transaction with which it is associated is determined to require further analysis due to suspicious activity.

Again, I do trust Apple more than I trust FB to only use it when necessary... but they are still collecting it.

1: https://support.apple.com/en-us/HT210701

Thanks! Very helpful

It's also misleading to have an empty box: with Signal your phone number (even if it's a Google Voice or Twilio VoIP number) must be known or the system cannot work. I hear that Signal is apparently working on a feature to allow you to set up an account without a phone number but until that time I think it's in second place behind Wire (which only requires an email address to start an account).

The section of the privacy label highlighted in these articles is about data that apps collect that is then linked to your identity (e.g. to an account).

There is a separate section of the label for listing data that is collected as part of the app's functionality, but then not linked to an account or other user identity - that is where Signal lists phone number.

WhatsApp and Apple Messages collect data and link it to your account(s) with them. Signal collects your phone number, but doesn't link it to anything.

Does that mean a signal user can't change phone numbers without getting a new identity?

I haven't looked at Signal in a while.

In Signal, a phone number is your identity. So intrinsically, yes, you have to get a new identity with a new number: https://support.signal.org/hc/en-us/articles/360007062012-Ne...

And this is a real problem because you don't really own your phone number: it's leased to you for as long as you pay for it but I don't think there is anything to prevent a telco (or VoIP service like Twilio) from taking that number back from you. This is another problem inherent in using a phone number as the source of your identity on a service. Email addresses are better but can still be subject to seizure. Perhaps the answer is what Session is doing: you get a GUID is that is assigned to you and that will never be assigned to anyone else.

Threema generates a short alphanumeric id, that is easily sharable. It will also generate QR code for you, so other people do no have to type them, just scan them.

Not the case, my Signal is linked to my old phone number, it is, in fact, the only piece of my identity web which is.

I wonder, sometimes, what would happen when someone else acquires that phone number and tries to sign up for Signal...

The case with Telegram is they overtake your account, and it might be the same with Signal.

Sorry, maybe that's not clear; I have multiple phone numbers.

Does that mean I have to unregister and reregister to get my other identity? Or does Signal support multiple identities?

You have to unregister and reregister the device to change phone numbers.

Look at the privacy data for the Wallet app, because that's where the data you're referring to is tracked. iMessage itself is just a container.

> Everything in the label is “bad”.

Shouldn't it be in red then?