Re: Strategic Orientation - Scope, Scale, or Means of Monopoly

[Dan Strychalski <dski@cameonet.cameo.com.tw>]

Re my post at <http://www.essential.org/listproc/am-info/msg03153.html>, John
Robert Behrman, in <http://www.essential.org/listproc/am-info/msg03155.html>,
wrote --
 
> Dan's discussion... goes to the heart of path dependence....
 
Nuts; I was hoping that if it had a name, that name would be easier to
understand....
 
> ....He suggests that IBM/Microsoft/WP/Lotus, the lot, all tried to ape Apple
> with a non-standard keyboard (hardware) and keystrokes (software) to lock
> in users. No doubt. This is pretty legal "product differentiation" though.
 
Everyone aping a single company seems to me to be the opposite of product
differentiation. In any event, I suggest something quite different.
 
As for "legal," I ask all the businesspeople and programmers on this list:
 
             What lawful mechanism or consideration would
             cause a company developing application software
             for many platforms -- or even for just one
             platform -- to deny itself the use of a resource
             found on all computers and proven to be popular
             with a large percentage of computer operators?
 
A company selling an *OS* (but few applications) for a particular platform
might want customers to think they cannot do without special-purpose (but
actually redundant) keys unique to that platform; such a company, if it
were working on a new mouse-oriented OS, might also want customers to find
keyboard use less convenient and more difficult to learn than it really is
-- but what of an application developer, a vendor of *word processing*
(and therefore heavily keyboard-oriented) software for *half a dozen*
different platforms? It just doesn't wash, folks.
 
Although not everyone was aping Apple, there was aping. The hardware trick
-- the relegation of the Ctrl key to the bottom row -- was probably
inspired by Apple, or to be precise, by Steve Jobs, who is still detested
by many in Macintosh circles for giving the Mac a less than minimally
equipped keyboard and an OS that made keyboard-only operation impossible. 
 
It seems reasonable to wonder if it was IBM or Microsoft that got the idea
to copy App^H^H^HJobs's keyboard layout. What IBM might have expected to
gain, especially with the clone industry in full swing, is hard to
imagine, and IBM is not known as a follower in the industry. Microsoft, on
the other hand, stood to benefit a great deal if all Ctrl-key combinations
but those at the bottom of the keyboard became inconvenient, and Microsoft
is not known for hesitating to copy something they can use to their own
advantage.
 
The software tricks started long before the Mac appeared, and the more I
look at them the more interesting they get.
 
WPCorp was not aping Apple. WordPerfect was up to version 3.0 (October
1983) before the Mac came out (January 1984), and up to version 4.1
(September 1985) before the Mac got the function keys WordPerfect's
designers seemed unable to do without. The first version of WordPerfect
for Apple machines appeared in December 1986 and ran on the Apple IIgs.
 
Aside: The IIgs used the first Apple Desktop Bus (ADB) keyboards. These
       keyboards had Ctrl in a sensible place -- I know a guy who hoards
       them for that reason -- and were used by new Macintosh-series
       machines as well. Funny that when the penultimate mouse-oriented
       system got a Ctrl key, it was *not* at the bottom of the keyboard.
 
(By the way, WPCorp shipped WordPerfect for the Amiga in July 1987, for
the Atari ST in October 1987, and for the Macintosh in January 1991. Every
graphics-based system but Windows until 1992 or 1993....)
 
Microsoft wasn't aping Apple when in 1981 they assigned printing
characters to the codes in the control range, banished hex notation and
the control codes' standard labels from the character tables in MS-DOS
manuals (keeping lots of hex in the body to remind people that Computers
Are Too Hard), and froze the use of Ctrl-key combinations in MS-DOS at the
level of the CP/M 2.2 command interpreter.
 
Aside: For ten years I thought the "smiley faces, boxes and triangles and
       stuff" in the control range were IBM's idea, but Gates takes credit
       for them and "credit" he shall have. They are not a terrible
       transgression in and of themselves, but viewed in the light of
       Microsoft's subsequent treatment of Ctrl-key combinations, they
       look to me like the opening salvo in the company's war on
       standards.
 
       (By the way, I use the word _standard_ only for specifications
       agreed upon by industry, market, and other competing and
       adversarial interests through a formal process coordinated by ISO,
       ECMA, ANSI, or a similarly constituted body. Standards are flexible
       enough to allow innovation -- they have to be, or they wouldn't get
       passed -- and they are reviewed at regular intervals. What's good
       enough for 75% of participants in the standards process is good
       enough for an OS producer. MUST be followed by an OS producer.)
 
IBM, I'm sure, was not displeased to see that Microsoft's DOS used
platform-specific keys for command-line editing and such....
 
Aside: Here's how you edit command lines in oh-so-hard-to-use Unix (note:
       circumflex [^] denotes the Ctrl key):
 
       ^P=Previous    ^B=Back       ^E=End          ^C=Cancel    ^U=Undo
       ^N=Next        ^F=Forward    ^A=Beginning    ^D=Delete    ^K=Kill
 
       ...and so on and so forth. A few explanations, half a minute of
       practice, and you're ready to use it on a third-party terminal
       halfway around the world from the computer. For ten years, on the
       IBM PC with its tightly integrated display system, Microsoft had us
       editing command lines as follows (characters in brackets are mine):
       
     ``F1  Copies one character from the template [?] to the command line.
       F2  Copies characters up to the character specified [?] in the
           template [??] and puts these characters on the command line.
       F3  Copies all remaining characters in the template [?] to the
           command line.
       DEL Skips over (does not copy) [?] a character in the template [?].
       F4  Skips over (does not copy) [?] the characters in the template
           [?] up to the character specified [??].
       ESC Voids the current input and leaves the template [?] unchanged.
       INS Enters/exits insert mode.
       F5  Makes the new line the new template [?].
       F6  Puts a CONTROL-Z (1AH) end-of-file character in the new
           template [?].
 
       You'll use these special editing keys in conjunction with the
       template [?], which you'll learn to use in the next section [!].''
 
       You could use Backspace to erase the character to the left of the
       cursor. You could also use the left arrow key to go back without
       erasing -- well, the character to the left of the cursor would
       disappear, but it wasn't _erased_, you see.... You don't see?
       You're not supposed to. Suffice it to say that when you go back
       with ^B in Unix, nothing disappears. You edit the command line
       quite the same way you edit text in a word processor.
 
       And because Microsoft supplied the OS, this F-key nonsense is
       typical of what most people thought -- and still think -- you have
       to go through to use the keyboard. (Note also that descriptions
       such as the one quoted above were taken by many as the model for
       technical writing).
 
Where was I? IBM. IBM worked up its own set of recommended keystrokes as
part of a spec called Common User Access (CUA), and Windows followed that
spec pretty closely (as does OS/2). It's all platform-specific function,
arrow, paging, and editing keys, of course, with and without Alt or Ctrl
or Shift or some combination of these. No main-block Ctrl-key combinations
that I know of. Quite insane, and most un-Mac-like.
 
So let us not speak of Apple-aping. Try instead to explain why, beginning
in the mid eighties, I and others felt that the software and hardware
industries were at war on people who use the keyboard in their work.
Explain why WPCorp and Lotus shot themselves in the foot with their
keyboard command sets and (sometimes, if you could find it) on-screen
help. Explain why text-based WordPerfect especially seems *designed* to
drive people screaming to Windows -- which, if you ever mention it on the
technical writers' mailing list, you'll find it did.
 
If you don't see the hand of Gates in this, using his clout as the
provider of the OS on which everyone else depended, what DO you see? There
is a LOT that needs explaining.

One more thing about text-based WordPerfect. There was a mailing list for
it, run out of the University of Ottawa. Traffic became light, and the
list was shut down, some time in '96 I think, without so much as a
fare-thee-well. The program offers much that Microsoft Word still doesn't
offer, but it couldn't keep enough adherents to justify continuation of
the mailing list.
 
WordStar, the penultimate Ctrl-key-oriented program, a program that laughs
at pointing devices and the Alt, function, Ins/Del/Home/End/PgUp/PgDn, and
arrow keys, has a mailing list as active as this one.
 
Dan Strychalski                                 dski@cameonet.cameo.com.tw