I have a fondness for the Rainbow 100 because it was how I started my software business back in 1983 (after selling a beautiful ''67 SS 350 Camaro to buy the Rainbow. I still have the very first DEC Rainbow sold by Computerland in the San Fernando Valley.

I originally wrote a formatting program for the Rainbow, and later wrote Media Master, a disk-to-disk format conversion program. That work in the floppy disk controller world eventually led to the creation of the Fastback Plus backup programs for the Mac and PC.

Dan Pleasant brought Code Blue to me to market while he was working at HP back in the day. He later was a major contributor to Fastback for the Mac. If you guys are interested, there is some history posted at the homepage URL.

For those trying to figure out direct screen I/O, I would say it is extremely difficult. I poked around the BIOS back then to try to help my friend Dave Grenewetski (later the CEO of Mindscape) figure out how he might port his graphics program on the Osborne to the Rainbow. I don''t remember the exact reason why now, but I do know I concluded it was more work than the market size merited.

I recently restored a Rainbow 100 and I am trying to write some software for it. It is slow going with only the BDOS calls. I wish I could talk to the hardware a bit. Any information would be very appreciated I was a Digital Research Engineer and have a fondness for the machine.
doug@goodall.com

While working at DEC, I actually wrote MS-DOS 3.3 for the Rainbow 100. The machine had several operating system choices for it. You could run CPM-80 on the Z80 microprocessor, CPM/86 on the 8088, MS-DOS on the 8088, VENIX (a UNIX v5 implementation) on the 8088, MPM (a multi-tasking version of CPM/86), and even Windows 1.0 which was adapted for the Rainbow at DEC by a guy named Alpo Kalio (a brilliant engineer). Windows ran on the hardware-assisted graphics card. The unique part of the hardware is that the two CPU''s could share a 2K space in RAM which allowed them to communicate. Thus, the floppy disc handling was done totally by the Z80 while the rest of the system I/O was done by the 8088. Pretty advanced engineering!

The biggest definciency of the Rainbow was the fact that early versions of it had no support for software writers. On the IBM PC, you could call various BIOS functions such as INT 10H to do elementary text and graphics on the screen. The Rainbow, until DOS v2.11 had no such support. Thus, software writers such as Lotus 1-2-3 had a heckava time trying to talk directly at the hardware in order to do text and graphics. Later on, as the Rainbow was dying, I produced a complete Software Development Kit for the Rainbow which had a rich set of interfaces for doing text and graphics. I also wrote a driver so that the Rainbow could read and write single-sided IBM floppies and so that the IBM PC could read Rainbow''s 10 sector per track single sided floppies.

Later on, the morons at DEC (which had a very hierarchical management structure) tried to clone the IBM AT in a single box. They called this monstrosity the VAXMate. Unfortunately, one of the engineers who was working on the non-infringing BIOS actually stole some of IBM''s code verbatim. This caused all sorts of legal problems for DEC. Ken Olsen, the short-sighted owner of DEC, demanded that the VAXMate be produced with monochrome EGA graphics and without a fan. Thus, the thing constantly overheated. I think that DEC may have sold 1 or 2 of these horrid machines. I had already departed for Microsoft by the time that DEC tried to get this VAXMate thing off the ground.