8 Comments

Small typo…

While very largely agreeing with Casey, I do think that there are other aspects to x86 that these discussions on the topic omit

Expand full comment
author

Thanks Paul. 'Of' rather than 'on'?

Expand full comment

Hmmm, may have be a bit too excited in reporting that one!

Expand full comment
author

I do think 'of' is better so thanks for highlighting!

Expand full comment

The 'Ozzie' docs released by Microsoft as part of open sourcing MS-DOS 4.0 contain a wonderful time capsule of the hopes and dreams for making a multi tasking DOS for the 286 https://github.com/microsoft/MS-DOS/tree/main/v4.0-ozzie

Expand full comment
author

Thanks Chris! That looks a fascinating resource. I’ll have a dig in!

Expand full comment

It seems they knew exactly what was going to happen, but it didn’t stop them trying:

“In summary, the operating system architecture must allow programs to be written so that a single binary copy can run in protected 286 mode on a 286-based machine, and still run on existing 8086 machines, including 8086 machines running earlier versions of MS-DOS. If this cannot be arranged, then the 286 will never happen; 286 chips will be run exclusively in 8086 mode "forever." The solution to these problems will have to wait until the 386 microprocessor is available (that can run 8086 programs while providing protection), and this means a delay of many years.”

Expand full comment
author

It's amazing to remember the limitations of the architectures of this era. I remember the Terminate and Stay Resident programs in DOS that loaded and then tried to stay around - I think Borland Sidekick was one of the most popular. Everything changed with the 386!

Expand full comment