As far as PowerPC hardware is concerned, this is not all that different from my characterization (though I was omitting the Display Postscript to Quartz transition, which was indeed quite significant).
I had not paid attention to the cross platform promises of the Rhapsody strategy, so it's possible I was missing just how much of that was publicly promised and then abolished.
I was never a NeXTStep developer, but having seen quite a bit of Cocoa code, my impression was that the compatibility hurdle for existing NeXTStep apps to becoming Cocoa apps was smaller than the one for Mac apps to becoming Carbon apps. Maybe developers with actual experience of having done such ports can comment?
I can concur with this. I am actually the youngest NeRD (NeXTStep Registered Developer) ever. I also programmed Mac applications professionally from just before OS X until after its release. I remember first seeing the API docs for it and being flooded with memories of code from a decade earlier. It was a bit weird. Keep in mind that besides the API, I also already was well versed in Objective C, so it was also not a new programming language. My peers had different experiences.
I had not paid attention to the cross platform promises of the Rhapsody strategy, so it's possible I was missing just how much of that was publicly promised and then abolished.
I was never a NeXTStep developer, but having seen quite a bit of Cocoa code, my impression was that the compatibility hurdle for existing NeXTStep apps to becoming Cocoa apps was smaller than the one for Mac apps to becoming Carbon apps. Maybe developers with actual experience of having done such ports can comment?