Filed a Feedback and it’s fixed!

When you notice a bug in iOS or any other Apple platform, you can file a feedback to let Apple know.

However, according to many developers on social media, these feedback tickets are usually untouched and never get fixed or take years. The process can also be a bit cumbersome. You have to explain exactly what is wrong and optimally why. Moreover, Apple usually requests to capture and attach a sysdiagnose of your device, which is often not relevant just to prove an API in the iOS SDK is broken.

Anyhow, I noticed a bug in iOS 16 in the Maps Framework on iOS / iPadOS, there you can’t set the background color of a marker to transparent, when it was set to a different color before. The use case in my BusTrack app is the following:

When a user taps a Stop icon on the map, the Stop icon is replaced with the big animated Marker, which says BUS on it. Then, when the user deselects the stop (by tapping anywhere on the map), the big Marker is removed with the Stop icon again. Here is a short GIF of how it should look like.

1. The user tabs the bus icon on the map

2. The bus icon is replaced with an animated marker that displays “BUS“

3. The user taps somewhere on the map, the marker is deselected and replaced with the bus icon again.

The initial state (just the bus icon) is restored.

On iOS 16 onwards, the marker does not disappear, because setting the background color to transparent does not work (again). It works the first time, where only the Stop icon is visible, but setting the marker to a different color, e.g. blue, and then back to transparent was broken.

You can see a GIF of the bug and a screenshot of the end result, on the right side.

The non-transparent marker will stay on top of the icon as long as the app lives. The only way to fix this, is by simply removing the annotation and recreating it. That works, because setting the marker transparent works the first time.

It is important to know, that caching via mapView.dequeueReusableAnnotationView(withIdentifier: "bus", for: annotation) must not be used, otherwise the map will simply reload that broken state for that annotation again.

This workaround can result in heavy RAM usage, because caching is disabled and each annotation will be recreated all the time.

The result of the bug, which stays that way until the app is killed and restarted.

I didn’t file a feedback, because I had absolutely no hope that this would get fixed. Until, iOS 17 beta was introduced.

I heard multiple times, that the beta cycle is the best time to file a feedback, because Apple is actively working on removing bugs in the OS. So I checked if iOS 17 still had this bug, and since it didn’t get fixed, I actually filed a feedback.

I’ve put a lot of effort into filing the feedback, since I desperately wanted this bug to be fixed. I wrote a very long text on what exactly was wrong and why it probably happened. Here is the exact text that I filed, in case you were wondering:

Please provide a descriptive title for your feedback:
MKMarkerAnnotationView .markerTintColor = UIColor.clear does not work, after being set previously (iOS 16 and iOS 17 beta)

Which technology does your report involve?
MapKit

What type of feedback are you reporting?
Incorrect/Unexpected Behavior

Which platform is most relevant for your report?
iOS

What build does the issue occur on?
iOS 17 Seed 1 (21A5248v)

Where does the issue occur?
On device

Please describe the issue and what steps we can take to reproduce it:
TLDR:

Before iOS 16, MKMarkerAnnotationView respected .markerTintColor = UIColor.clear and disappeared, so that .image was visible on the map.
Since iOS 16, .markerTintColor = UIColor.clear does not work and stays on the map in the previously set color, so that .image is hidden behind the empty marker on the map.

Description:
I set an image for an annotation to be visible on the map. All other MKMarkerAnnotationView properties are set to nil or UIColor.clear, so that no marker is visible and only the image is visible, like so:
.glyphText = nil
.glyphTintColor = UIColor.clearColor
.markerTintColor = UIColor.clearColor

When the annotation is selected, I remove the image by setting .image = nil and configure all other properties with custom values, e.g.:
.glypthText = "BUS"
.glyphTintColor = UIColor.whiteColor
.markerTintColor = UIColor.blueColor

When the annotation is deselected, I reset all these values, so that only the image is visible again, which has worked until iOS 16.
.glyphText = nil
.glyphTintColor = UIColor.clearColor
.markerTintColor = UIColor.clearColor // <- Issue

The marker shrinks, the glypthText is gone, but the marker is still visible, because .markerTintColor = UIColor.clearColor does not work anymore.

Xcode version 14.3.1 and 15.0.0.
Issue is currently available in all iOS 16 releases and iOS 17 beta 1.

Example Xcode Project attached, but also available here: https://github.com/gorkemg/iOS17_MarkerTintColorBug
- Run Xcode project on Simulator or Device
- Select the annotation with a bus image on the map
- Tap somewhere on the map to deselect the annotation
You will notice, that the marker does not disappear, staying overlayed on top of the bus image, basically hiding it

As you can see, I started with a two sentence TLDR: paragraph, so the Apple engineer can have a very short overview of what is wrong. Afterwards, I explained with a step by step guide how to replicate the bug.
Besides, I attached a small Xcode project, which only included a single screen and a few lines of code that showed the bug in action.
Additionally, and I don’t know if that actually helped, I’ve uploaded that Xcode project to Github and shared the URL. I thought, maybe the engineer would rather like to checkout the issue via Git, or at least see my effort of getting this bug fixed. 😄 You can checkout the Xcode project for yourself now, if you like.

And sure enough, this bug got fixed a few beta releases later. Yesterday, I received a request by Apple Feedback to check if this bug got fixed in the latest beta (iOS 17 Beta 7). If so, I can manually close this feedback, or if this bug still persists, I can answer and provide new details.

I will now close this feedback and be happy that the bug is fixed in iOS 17.

For users running iOS 16, I will have to use the aforementioned workaround. The workaround is pretty easy, but ugly and will most likely result in heavy RAM usage, since caching is not used anymore.

Luckily, most devices that run iOS 16 will be able to install iOS 17 and fortunately most users update to latest iOS release pretty quickly. Therefore, this bug will become a non-issue in just a few months, probably.