This is an excerpt from one of Larry's emails, explaining some of the goals:
First, Navigator will be our primary design tool, not FCP or Motion. Please go watch Rey Kroona's Tutorial 4 and the following Scenarist tutorial (I forgot the number). The work flow, learning curve, and manual steps without built-in validation are horrendous. We had something similar when we started with our MX project. They had actually hired 25 college students to type in coordinates and time codes for the "Finding Nemo" director's cut. They later automated it somewhat with AppleScript, but still had to take everything through four separate applications. Now all of that is gone with BluStreak MX.
We are using FCP and Motion because the plugins can be written in Cocoa, giving easy access to all of the rest of the Apple technology, including inter-process communication with Navigator. I was able to import the Tutorial 4 photoshop files into Motion, so that is a moot point. Photoshop has no concept of time or motion. Using layer names to control things is stupid. We want to design menus interactively, so you can get immediate feedback to your actions. The plugins are just a way to collect input more conveniently, and a carrot to fetch FCP users with high-definition cameras. There are thousands of them, who have no way to make a BD disc on Mac. Encore is a joke, everyone who has tried it hates it.
We have a large amount of code to write to reach a fully independent authoring application. I actually already have all of the UDF 2.5 code and about 80% of the BDCMF code written. But the last 20% will be a bitch, and best left until we have something useful to build on. Multiplexing must take precedence, or we have nothing. Luckily for Navigator, it can be done by multiplexing only a single IG stream, adding the transport stream and Blu-ray layers, and not worrying about presentation and decoding time stamps for video and audio.
So we are back in the woodshed for now.