[mdlug] Mark Shuttleworth on MS patent threat

[Robert Meier]

David,

> Before Microsoft also means "before 1975" and back then most software
> was for mainframes and minicomputers.  There
> wasn't a whole lot available in the  microcomputer world.

I believe in 1975, Microsoft product line consisted of a paper tape basic
compiler.

> ... fast forward to micros ...

> Before the IBM PC there were lots of proprietary designs -- imsai,
> cromemco, osborne, sinclair, apple, etc.  CPM was a "standard" operating
> system, but software for it was limited.  The Apple II was popular but
> software choices were limited.

>From the late 70s through the 80s into the early 90s,
software was available for download from Simtel and several other
arpanet archives.  User group libraries and bulletin boards formed a layer
that provided non-arpanet access.  Archie, gofor, kermit, ftp,
and other tools were the browsers.

My observation was that the software choices were limited only by
your download and memory bandwidth.  Simtel and the other archives
each held larger collections than the average desktop computer owner
could download by acoustic modem in a decade.

The choices included simulators, both general purpose and specialized,
editors, textual, graphical, database, ...
formatters, compilers, analyzers, games, ...
Most of the executables were shareware, typically for $5-$20 (the price
of inexpensive lunches for a week)

A large fraction of the archives were datasets, maps for directions,
components for simulation, recorded music, etext, statistics, ...
These were in standard formats like ANSI DIF (data interchange format),
AIFF (audio interchange file format?), ANSI RTF (rich text format), ...
As the operating systems market evolved and shifted among CP/M, Amiga,
Apple, Mac, and others, the executables all implemented the standard
formats and so retained compatibility with the datasets.

> Microsoft grew as PC compatibles dominated the computing landscape.
> Along with this growth so grew the number of applications available.

I observed that in the late 80s, Microsoft began writing applications
that were not compatible with the existing datasets.  Among others,
they broke DIF into csv and tsv, by always stripping the header (1st line),
they broke AIFF into wav, by slight alteration including changing the 'AIFF'
signature bytes to 'RIFF', and they mangled RTF (incompatible even
between Word releases in the 90s) without changing the name.

Microsoft and their associates began selling the altered datasets
with, and later for, only their applications.

I observed a decade later, in the 90s, after http and html, evolved
and replaced gofor, archie, ... Microsoft discouraged their
customers from using the web with incompatible/broken applications
like a "MS-kermit" that munged binary data.  (At FANUC, we got  about
1 customer call per week, that was traced to a corrupted file download
through "MS-kermit".)

> So, much as I dislike Windoze, I _do_ acknowledge that the level of
> standardization it provides made possible the software industry that
> now exists.  We are fortunate to live in a world with endless choices
> of software -- both non-free and free.

Now after another decade, in the 00s, I observe that Microsoft's claim
of introducing (broken MS-RTF, broken MS-AIFF, broken MS-kermit, ...)
standards and software choice (i.e. more Microsoft and associate offerings)
is merely a revisionist history of Microsoft policy, and not a factual
description of software industry history.

Reporting,
-- 
Dr. Robert J. Meier