Hello Quinn,
Though we have managed to solve that problem, we are still having issues while distributing the app via Xcode.
It would be great to have your advice on what would be the best possible architecture based on Apple's guidelines.
We are building a MacOS application which needs a face model (preferably Mediapipe) to identify certain landmarks on face and body, which we further use for calculations and provide an output on the GUI.
We have built a stand alone python application using PyQt that we are able to distribute as a .app file but is there a way to distribute this .app file via Mac app store without modifying the application?
In another test, we have built a GUI using Swift and the backend is in python. We have built this python application with a camera module and integrated with the GUI. At distribution we are stuck at the App sandbox issue.
In our 3rd method we gave the camera rights to the GUI built with Swift. With logic in python, the python app provides an output to the GUI. We tried with both Json and Stdin/stdout output to the GUI. On locally building the application on Xcode the app will start and turn on camera for 1-2 sec but the application will hang. We tried to distribute it via Xcode but we got a python package build error.
Also, we are open to importing the Mediapipe package on Xcode and building the entire app natively on Swift. Is this something that Apple supports? I tried to add the Mediapipe package but received the below error.
We are experimenting with swift + objective c by either converting python using cython or writing in c/cpp directly.
What is generally the best and simplest design to build an AI based application using 3rd party libraries and logic written in Python to deploy the app on the App store?
Thank you and have a wonderful day!
