Veteran programmer Steve Palmer, author of the classic sports
videogame NBA Jam and currently working with Pitbull Syndicate, who
developed for both systems during their entire respective lifetimes, sums up
the Saturn development problem quite well.
[Sega] gave us exactly what we wanted; [however,] the industry
changed at exactly the same time, [so] we no longer had a choice in the
matter. Things suddenly had to be finished yesterday. Sega
could not have forseen this change .... [Most of the third-party crowd]
couldn't get it to do what they wanted it to do quickly enough, so they didn't
bother. It's amazing, because it didn't require much effort to get the
machine to perform on a [PlayStation] level. Programmers being
programmers, though, they probably were not happy unless they felt they were pushing
the machine, and it seemed like too much effort to do that.
To learn to program the
Saturn was to learn the machine. To learn to program the [PlayStation]
was to learn C. Learning C is much easier than learning the hardware of a
new machine, and with the Saturn, there was a lot of hardware to learn ....
There was not enough time for people to learn the hardware. The same would
have been true of the [PlayStation], except you didn't need to learn how to
talk to the [hardware]. The libraries took care of that for you.
Sega's approach was to release hardware documentation for every aspect of the
Saturn. That was understandable - it was the way everyone had done it
before, and it's what programmers were used to, but the industry had changed.
Video games were no longer a "niche" market, and the "big
boys" had moved in. Time is money. Nobody was given the
time to learn new hardware anymore.
http://www.eidolons-inn.net/segabase/SegaBase-Saturn(Part1).html