Lists Home |
Date Index |
On 5/5/05, Ken North <firstname.lastname@example.org> wrote:
> > Other examples of database integrated into the OS
> > Microdata Reality (1974) -- first minicomputer to do this
> > IBM System 38 (1978) -- ancestor of the AS/400.
> > If I remember correctly, Microdata used a microprogrammable instruction set
> > the Reality. (Microcode implemented data management operators as part of the
> > machine's instruction set.)
> Peter Hunsberger wrote:
> >> At the time I hadn't been able to come up with any other hardware based data
> management, but then again, I was only peripherally aware of the
> Microdata. How is that you stumbled across it?
> When Dick Pick and Don Nelson worked for TRW, they were the architects of a DBMS
> named GIM that ran on IBM mainframes. After Pick left TRW, he formed a company
> and struck a deal to port the DBMS to Microdata minicomputers. About the time
> Pick and Associates were working on the Reality port, my task was to develop a
> paging solution for GIM that would work with demand-paged virtual memory being
> introduced by IBM.
Cool. While I was at the University of Waterloo I worked for a very
brief period of time helping write IBM/370 channel programs for one of
the first solid state disks which was being used for paging. Tough
stuff to debug, dynamically modified instruction sets, and in this
case, none of the published timings had anything to do with the
hardware in question. But it was a good technology with amazing
capabilities. Keeping in the spirit of the central them of this
thread: it still amuses me to see some of the same technologies work
their ways into modern "server" hardware some 20 years later ...
> About a year later, our lab took delivery of a rare computer called a Nanodata
> QM-1. It was used for microprogramming R&D and designing new hardware. (One of
> the early projects was a Concurrent Pascal machine.)
> Microdata Reality was the pre-cursor to the Pick OS, another operating system
> with a tightly-integrated DBMS.