Tracking domains - Network Instrument Points of Interest

Question

My iOS app does ask user's permission for ATT(App Tracking Transparency) and if given, displays some ads. But when i profile the app using instruments tool as suggested by Apple here to detect tracking domains, i see a blue circle with a "S" symbol inside it and not a red circlet for fault lines as per Apple's link.

Do these 2 icons(blue with S symbol and the red circle) both indicate fault lines tracking domains? Can you please clarify this? Also, not seeing the fault lines, does it guarantee that there are no domains tracking? Since i'm asking for user's permission for ATT and displaying ads if permission is given, shouldn't it show the fault lines(instead of blue circle icons) and the corresponding message asking me to include NSPrivacyTrackingDomain as true in privacy manifest?

Tried as per Apple's documentation here to detect tracking domains.

Expected: Since i'm asking for user's permission for ATT and displaying ads if permission is given, shouldn't it show the fault lines(instead of blue circle icons) and the corresponding message asking me to include NSPrivacyTrackingDomain as true in privacy manifest?

Answer 1

Your screenshot looks as if you are recording all processes. Is that correct? If so, the Points of Interests instruments will include all kinds of Points of Interests from all processes across the system, e.g. in your screenshot there are some from assistantd which is one of the system processes.

To reduce logging noise from other processes, make sure to target only your app's process.

In general, the Points of Interests instrument can show any kind of log or signpost sent to the Points of Interests logging category. But by default, an app will rarely send any, unless you specifically add points of interest signposts or logs in your code.

One of the cases where logs do get sent to the POI category is the privacy tracking domain case you linked to in the documentation.

For those, yes, the expectation is that tracking domains do show up as faults when they are not listed in the privacy manifest. It seems like you first tried to see the faults in Instruments to verify your request is detected as a request to a tracking domain before updating the privacy manifest to mark it as expected, so you could verify that by the faults going away. This is a very sensible strategy.

Could it be that your request was never sent? Maybe the request was made but the response came from the cache so it never went out on the network? First, use the HTTP Traffic instrument to verify that the request was made and not answered by a cache but caused an actual networking request.

If you can confirm that the request you expect was indeed sent and the faults don't show up despite the domain being a tracking domain and not being included in the privacy manifest, please file a bug report.