The reason why SCO was successful was because of how it is used. Imagine it is 1991 and you have a dentist's office you want to computerise. You go to a local

The reason why SCO was successful was because of how it is used. Imagine it is 1991 and you have a dentist's office you want to computerise. You go to a local VAR (value added reseller) and they would come in and install a complete solution. It would consist of a Compaq "server" ($5k), several terminals, com port cards for the server, a tape backup system, and some dental office management software. Oh and a $1,000 copy of SCO Unix. The VARs typically marked up what they sold by 15%, which is how they got paid. Two related bits of trivia - at one point the only software that existed for managing pet cemeteries ran on SCO OpenServer. And most McDonalds (I believe US only) had OpenServer in each branch. By the late 90s SCO had 15,000 of these VARs. Caldera wanted to essentially substitute Linux for OpenServer into that setup, and make $1,000 per copy of the OS rather than $25. The VARs realised they could supply Linux themselves which is why Caldera had no traction doing that, and doubled down on SCO OpenServer and the existing installed base. Hence the company rename too - it was all SCO products. UnixWare was touted as Enterprise ready. It did multi-processor well, had sophisticated filesystems (Veritas), could do clustering (Non-Stop) etc. But at the time most who wanted "big" Unix went to one of the RISC vendors each of whom had their own Unix. If they wanted to go Intel (which wasn't credible until the Pentium Pro) then the competition was Windows NT.