% fortune -ae paul murphy

From Chapter Three: The Windows Culture

This is the 26th excerpt from the second book in the Defen series: BIT: Business Information Technology: Foundations, Infrastructure, and Culture

Roots

The Microsoft PC is ubiquitous today (late 2008) and is widely thought of as the definitive white collar computing environment. Note that this is largely a social phenomenon, not a technical one and no researcher has yet published anything amounting to a documented explanation of the PC's rise to importance.

According to many popular histories the seventies were characterized, from a personal computer viewpoint, by machines like those from Altair requiring binary programming and displaying results by lighting lamps. In fact those machines belonged to the sixties and were sold mainly to hobbyists.

The first computer shipped pre-assembled and clearly intended for use by a single individual as a dedicated personal machine that came complete with all the components:

was the IBM 5100 introduced internally 1972 and released for sale to the public in late 1974.

By the mid 1970s, major companies including IBM, NCR, and Digital Equipment Corporation (DEC) made full fledged personal computers for business use and dozens of smaller companies made roughly comparable, but generally cheaper, machines for sale to business and academic users.
The first PC?
The MCM/70, from Kutt Systems, was a lot cheaper and focused on non business uses but came a bit later. In many ways a consumer IBM 5100, it too ran APL, used a cassette tape for storage, and came (as the MCM/70T) pre-configured for use as a smart terminal. Announced in late 1973, it had the distinction of being the first Canadian computer to use an Intel CPU board --the SIM8-01 with an 8008 CPU and 1K of RAM.

The architecture those companies used in the machines they built tended to reflect their views on the market's trade-offs between cost and performance.

Thus two broad lines developed:

  1. Companies which focussed on performance generally designed products around the Motorola MC6800 line introduced in 1974; while,

  2. Those which focussed on cost generally selected the Zilog Z80A introduced in 1975 and its successors.

These products were generally quite sophisticated with advanced operating systems and a fairly wide selection of commercial applications available by the late seventies.

For example this 1978 Intertec Superbrain had two Z80A microprocessors running at 4MHZ and ran a multi-tasking capable version of CP/M in 64K of user accessible RAM. One CPU was dedicated to handling user interaction, including built in APL and BASIC interpreters, as well as a large number of packaged applications like WordStar and Visicalc. The second CPU handled all I/O, including managing a serial modem interface and a printer as well as the two floppy drives.

Two footnotes
Most footnotes and diagrams are ommitted here due to the lack of true HTML support in Wordpress - however these two are included here for interest:
  1. In part, the MC6800 carried forward classified research conducted by the US Navy in support of the F-14 development program. Two engineers, Ray Holt and Steve Geller, working at aiResearch from 1968 to 1971, developed the first real microprocessor (several years ahead of the people who founded Intel). Their 1970 CADC machine was incredibly advanced for the time and matched the later performance of the MC6800 introduced in 1974 and thus exceeded that of the Intel 8086 introduced in 1978.

    Holt and Geller took their expertise to American Microsystems inc. which eventually decided that there was no future in microprocessors - laying off the entire 18 man engineering team. Some of these went to work at Motorola, initially on another classified project and then on the MC6800. Today descendents of Motorola's PowerPC successor to the 6800 power most IBM gear, all major gaming consoles including the iPhone, and just about every networking or inteligent control device in the world.

  2. Updated versions of the Z80A are still in use - in things like TV remote controls. Zilog was to the seventies what Intel is today: the market leader in high volume, mid range, CPUs with a growing following on the NYSE where they were known as ZIG. In 1980 they were bought by Exxon, which stripped them of their key microprocessor role, and eventually sold the remenants back to some key executives. Today ZiLog is a leading design house for communications and related micro-electronics.

This layered design was made possible by CP/M (Control Program/ Monitor) --then the standard operating system for small machines.

CP/M (from Gary Kildall's Digital Research Corp. [DR]) separated user accessible memory from system accessible memory by having an operating kernel which ran the hardware and only communicated with user applications through a predefined set of messages generated by a user shell or command library.

This made it easy to replace the user shell without affecting the kernel. For example, GEM, the graphical user interface (GUI) released for CP/M in late 1984, consisted mainly of an alternate command shell - and became the basis for the phenomenally successful Atari line of multi-media oriented computers.

This design approach, copied for CP/M from Unix, offers an elegant solution to the problem of providing multi-user access to a single physical machine: one kernel communicates with multiple user shells. This design easily handles hardware access serialization and provides for strong memory management without the complexities and performance losses inherent in multiple single-user strategies such as IBM's machine virtualization - but was stripped out of the version adapted for use as MS-DOS largely because of processor limitations in IBM's chosen CPU: the Intel 8080.

Some notes:

  1. These excerpts don't (usually) include footnotes and most illustrations have been dropped as simply too hard to insert correctly. (The wordpress html "editor" as used here enables a limited html subset and is implemented to force frustrations like the CPM line delimiters from MS-DOS).

  2. The feedback I'm looking for is what you guys do best: call me on mistakes, add thoughts/corrections on stuff I've missed or gotten wrong, and generally help make the thing better.

    Notice that getting the facts right is particularly important for BIT - and that the length of the thing plus the complexity of the terminology and ideas introduced suggest that any explanatory anecdotes anyone may want to contribute could be valuable.

  3. When I make changes suggested in the comments, I make those changes only in the original, not in the excerpts reproduced here.