The "analysts" have been saying that SPEC 1170 is a final approach to unifying the many different ways of doing UNIX, and many persons believe that we are nearing the end of the UNIX religion wars. But as one 18th-century politician said, "eternal vigilance is the price of liberty", and SPEC 1170 won't solve the problems of portability and interoperation that we face in the real world.
Realistically we should keep several things in mind.
First, SPEC 1170 is important as a realistically large and useful set of standard rules, and it seems certain to succeed over the many individual conflicting "standards" that manufacturers have tried to promote.
Second, just following the rules won't guarantee applications and environments to be portable or to interoprate, and certainly don't guarantee that they will be good.
Third, a whole new set of issues is always raised by advances in technology, and these advances bring back some of the problems that SPEC 1170 is supposed to solve.
This talk discusses these three points and attempts to give some details, especially about what problems SPEC 1170 will solve and which others (especially new ones) it won't. And it suggests a way to approach this.
