There were a number of operating systems around at that time but few that would run on the 8086/8088 hardware. One with multitasking was the iRMX86 OS<http://bitsavers.trailing-edge.com/pdf/intel/iRMX/142721-003_iRMX_86_System_Programmers_Reference_Manual_May81.pdf>supplied by Intel with its 8086/8088 chips for real time development. I don't know how or why they overlooked that. My suspicion is that the real reason they chose MS-DOS was that Bill Gates's mother had direct contacts with the IBM board of directors.
If that's the case, it would make me feel quite a bit better about my decision to abandon development of an 8086/8088 OS -- a development that started before the first silicon was shipped while I was at the PLATO project where we modified the CDC Cyber COMPASS assembler to produce the instructions documented on the preliminary datasheets, and execute on an emulator running on the Cyber 6500 during off-hours. The reason I initiated that project, with some of the PLATO system programmers (Ray Ozzie was a system programmer at PLATO but was consumed by his work on the Z80 firmware) was that I foresaw the horror of a bad operating system becoming the network-effect atop Moore's Law, and wanted to head it off. Others, primarily Steve Freyder, agreed and pitched in. It was obvious to me that whoever got the critical mass OS for that platform would have a natural monopoly and lock out competition -- including superior operating systems. I abandoned that project only because Mike Pavloff at Control Data HQ offered me a position at the Arden Hills Operations where I could pursue a mass market version of the PLATO network which would have, using Ozzie's Z80 firmware, bypassed the personal computer era entirely with a Mac-like UI and built-in 1200bps modem starting in 1981 with a monthly service charge of $40/month including "terminal" rental. We had that system benchmarked out at a scale that could have deployed nation wide late in 1979, but Wall Street analysts smelled blood and were ripping Bill Norris (the Nebraska farm boy that founded CDC with Seymour Cray) limb from limb due to his billion dollar investment in PLATO. CDC middle management mutinied and reneged on their agreement to let me pursue a mass market version of PLATO. I fled CDC and tried to revive something similar at Knight-Rider's joint venture with AT&T, but that is another story<http://tech.slashdot.org/comments.pl?sid=2702791&cid=39217853> . Suffice to say, when I saw MS-DOS I knew a horror had been unleashed and that Gates would become extremely wealthy. On Sat, Jan 25, 2014 at 12:15 PM, Eric Walker <[email protected]> wrote: > On Sat, Jan 25, 2014 at 6:33 AM, Jed Rothwell <[email protected]>wrote: > > Eric Walker <[email protected]> wrote: >> >> Gates was at the helm of Microsoft when it acquired DOS. DOS came out >>> many years after and was an anorexic imitation of UNIX. >>> >> >> That is incorrect. They brought it out within months, not years. >> > > I had in mind the initial release dates, not the time to market: > > UNIX, April, 1969. > DOS, August, 1981. > > That's 12 years to learn from an excellent, pathbreaking operating system. > In this context it matters little to say DOS was a CP/M clone rather than > a UNIX clone (as others have rightly pointed out). Those twelve, blessed > years were made available by a divine providence to the people making > decisions about operating systems, and then utterly wasted on them when > they came up with DOS (or CP/M). (I imagine there were other good OSs > other than UNIX at the time, although this is the one from that era I am > most familiar with.) > > It was not long before UNIX was ported to the PC, so it also does not > matter that UNIX was developed for a timesharing environment. An early > incompatibility between UNIX and the early PCs was the lack of memory > protection, but this was eventually added as well. > > Eric > >
