Testing:
Just as the use-case scenarios determined our solution, it must now determine the test matrix. As I mentioned before, testing will be quite a chore primarily due to platform diversity. Nobody I'm aware of has all of these platforms set up and running for test purposes and the time it takes to load software onto test platforms can be considerable. Remember, you can't just test on "Mac" because this term includes Mac Classic, 7.X, 9.X and now OSX. "Windows" technically includes DOS, Win 3.X, NT3, 95, NT4, 98, ME, XP, CE, 2K and counting. For each of these there will be multiple sub-scenarios such as different browsers or browser versions. The diversity of set-tops and other devices makes the diversity of Mac and PC look simple.
I'm not suggesting that all of these make it into testing, the way to go would be to test the most popular OSs still in wide use. For example, on the Win side I'd knock off everything up to (and possibly including) Win95. On the Mac side I'd forget about everything up to 9.X.
I would limit Linux to Mandrake and do myself a favor by using 9.x or higher. This would establish proof and you could let the rest of the Linux community work out it's own issues. They are good at that.
As to non-computer devices, more information is needed to understand what's at stake. There's every chance of encountering anomalies with certain players but the issue is beyond the scope of this research. The gurus would be the best starting point for this.
Open Issues:
1. The necessity for an ISO9660 folder
2. Control point settings
3. The exact percentage of Director/Lingo code that will be reusable.
Summary Conclusions
The good news is that you can accomplish what you want. You can get it to work the way you want to a great extent, certainly well enough to produce a great product. The level of control over presentation will consist of endless tradeoffs between workarounds and tricks which will have to be discovered and tested. Some things will have to be lived with.
1. Can you say MX04?
2. Developers choice for DVD-V authorware
3. Pursue the simplest path to completion now: the non-mulilayered or multisided DVD.
You get a production-ready product. Live with redundant content for now. Let this aspect
be sponsor funded.
4. The present multiple folders-with-redundant-content is perfectly acceptable. The
accusation of "inelegance" is a ceremonial tradition among programmers. The traditional
response is to say "it works fine, the user will never know, who cares?" It's an
equivalent to "if God builds a wall that God cannot break down, can God break it down?"
It's a question that will never be resolved.
5. The multiple folder issue should be resolved up front in an inexpensive proof of
concept or proto disk.
That is everything there is to say about if you can do it, if you should do it, and how you should do it.