Today is the “a little bit of history” post.

I was going though some old documents I had from the days of working with the CTOS operating system, and I found a memo on why it was killed off.  If you ever used CTOS, you’d know how great it truly was.

Now if I was asked why it was killed off, I would have said “Unisys wanted to get into the services/consulting business”.  The documented reasons follow:

…

The Business

Over the last 16 years, a long time in today’s workstation industry, Unisys has been running a profitable CTOS business and still has many successful customers, several of whom have spent well in excess of $100M on CTOS hardware and platform software.

Over the last couple of years, this business has been decreasing, mainly due to the fact that a proprietary workstation proposal was unsustainable in the current market. Anticipating the fact that CTOS would not survive in the open systems market, we have been promoting a migration strategy for the last 2 years. Early adopters are already implementing such plan, further reducing demand for CTOS hardware.

We had already anticipated that, in the near future, it will not make business sense to maintain our investment in CTOS technology. During the last 2 years, we have been focusing CTOS software engineering to migration tools, rather than developing core CTOS. We now anticipate that at some point in 1997, it would be unwise to further invest in CTOS hardware.

The Technology

Coupled with the decreasing business, we are facing a set of technological barriers:

• CTOS hardware is based on ISA/EISA architecture. We anticipate that PCI will completely replace EISA by the end of 1997, and that ISA will disappear shortly after. Unless a major investment would be made, this means that we will run out of suitable systems.
• CTOS will always be a 16 bit operating environment. Pentium Pro and Pentium MMX will become the standard during next year and, again, only a substantial investment would allow us to take CTOS on these new platforms.
• CTOS relies on a number of specific ASICs. These will become increasingly rarer to source.

…

– CTOS Program memo, 30th October 1996, Frank Brandenberg.

After 28 years with Unisys, Frank Brandenberg left Unisys for greener pastures.  And Unisys has left the hardware business.