Supporting a mixed NT and UNIX environment is very complicated. Others have
met this complexity by building larger, even more complicated systems to
address the various issues. We feel that is the wrong direction. We looked
to break the problem into smaller, more simple, chunks. Simple chunks
can be implemented.
Difficult chunks can be thought about, pondered and researched until we
find simplifying principles and constructs that turn them into simple chunks.
If something can not be simplified we would rather leave it unimplemented
than create a monster that can not be tamed.
While this sounds like we leave a lot unimplemented the opposite is true. Delaying the difficult chunks gives us more time to implement the first chunks ``the right way''. When those are complete, we have a better understanding of the system and those ``more difficult chunks'' become easier to simplify. Often we discover that those ``difficult chunks'' were hogwash that were not needed anyway. Either way, our strategy achieves more because we can remain focused on a smaller set of issues at a given time. It is with great hubris that someone thinks they can plan out an entire system without the benefit of the knowledge gained by first having solved smaller portions of the problem.