Nearly buried in the huge volume of IBM-related news was the company’s effort to explain what happened to a project called Workplace OS. Originally, Workplace OS was to be a commercial derivative of the Mach operating system. Mach is an extension of Unix created at Carnegie-Mellon University, in Pittsburgh, that can manage more complex hardware and software than standard Unix operating systems. IBM’s derivative of Mach was supposed to let other operating systems ride on top of it much the way IBM’s mainframe hardware and microcode support multiple independent environments. The stated purpose of Workplace OS was to allow IBM to create a series of core processors and then, with the addition of guest environments, ship customers machines that could run OS/400, AIX, OS/2 and other software, either one at a time or all at once. The appeal of the idea – to customers as well as to IBM – cannot be denied. Any technology that would enable IBM to bring down systems costs is always attractive. But, it tu rns out, Workplace OS as originally conceived is not going to come from IBM and, we suspect, not from anyone else, either. The problem that IBM ran into, according to a few published reports, is that the various guest operating systems each expected a different underlying architecture. Further, the differences were so great that the overhead involved in providing full support for multiple environments (such as running compiled applications) made any machines based on full-blown Workplace OS very slow or, alternatively, expensive when fast.
By Hesh Wiener
In other words, the results were going be the exact opposite of the ones IBM intended to achieve. IBM says Workplace OS will eventually emerge from its skunk works and provide a choice of interfaces for otherwise incompatible applications. Also, some of the work that went into IBM’s botched project is believed to be lurking inside the AS/400 or maybe the next generation of PowerPC AS/400s. The direct impact of IBM’s failure to make Workplace OS meet its original goals is of no consequence. But the cause – collisions involving AIX, OS/2, OS/400 and other software – is another matter. The Workplace OS debacle shows that IBM is still designing and creating each of its systems without regard to customers. Users will increasingly choose software that provides compatibility across heterogeneous systems, whether that software is one of the Unix variants or, more likely, Microsoft Corp’s Windows and Windows NT. As a result, OS/2 will continue to enrich advertising media more than it enriches users and AIX will remain in a world apart from everything else IBM offers. This is probably good for the mainframe, which is also incompatible with just about everything, but is at least well understood and widely accepted. From Infoperspectives International, February 1995. (C) 1995 Technology News Ltd.
