Are there any options for improving the performance of incremental AOT builds? It seems that no matter how small the change, everything gets AOT'd every single compile. Due to some hardware constraints, running in the simulator isn't viable for us, so we need to deploy to the device each time. With ~300k loc, this is a bit painful (2-3min on a fairly recent MBP). We are using MT 6.0.7 and MD 3.1.1.
Any suggestions greatly appreciated.