Hi, I'm looking for feedback regarding SCNetworkReachability under macOS Sonoma. It seems that since beta 3 the notifications behaviour changed. In a LaunchAgent I'm using SCNetworkReachabilityCreateWithName + SCNetworkReachabilitySetCallback + SCNetworkReachabilityScheduleWithRunLoop and wait for callbacks looking at the kSCNetworkReachabilityFlagsReachable flag. This is running fine under macOS 12.x, 13.x and 14.0 for more than a year. If I log all callback entries I observe unexpected notifications as if the looked host became unreachable for very small amount of time (ms). The host is flagged as unreachable then few ms later reachable again then unreachable again. Fast switching is fine, I can accept that the service is unreachable even for 1s but the probleme is the latest status do not reflect actual reachability of the service. This is in a corporate network with the complexity of using a proxy.pac. Does anybody noticed something similar ? I filled a Feedback FB13442134 in case it could be a regression of 14.2

This is happening Mac M1 Monterey OS .Environment supports both IPv4 and IPV6. When a http client calls gettaddrinfo() it is returning both IPv6,IPv4 IPs . first v6 IPs and then v4 IPs. We need to have a way to sort gettaddrinfo() output to get v4 ip first and then v6. We tried changing DNS order with scutil by putting v4 DNS first , but still getaddrInfo() listing v6 IPs first . In linux there is a way to control gettaddrinfo() o/p with /etc/gai.conf https://man7.org/linux/man-pages/man5/gai.conf.5.html . In Mac I did not find any option like this , scutil changing order DNS is not effective . can you tell us what is way to do this in MAC OSx ?