The Penguin and the Hare....

Abstract: In this analysis, we examine the similarities in the
early-phase growth of Microsoft's Windows and Linux as desktop operating
systems and show that Linux, far from being out of the mainstream
desktop race, is moving smoothly, growing in stature as a performer and
is starting to eye the finish line, which is just a few years' down the
track, with confidence.

- -

Throughout 2001, a substantial number of industry pundits took it upon
themselves to deflate or denounce Linux's chances as a desktop Operating
System contender. They pointed to the fact that even after two or three
years of what they called 'hype', Linux still had a minuscule proportion
of the mainstream desktop market. What are its chances of catching the
naturally advantaged Windows platform on the desktop, and combating the
arrogant and aggressive Microsoft?

Circa 600BC, the Greek, Aesop, wrote a collection of fables, which,
while simple on the surface, proffer much hidden depth and a level of
truism which sometimes only becomes apparent through the maturation of
the memes which they deliver. One such fable, a very famous example,
relates to a tortoise, a slow moving, measured creature, and the hare,
all pace, flitty and somewhat arrogant. The story goes that when
challenged to a race by the tortoise, the hare, with consummate ease,
uses its natural advantages on the particular measure at hand (running)
to tear ahead and enjoys a mammoth early lead. However, through the
hare's arrogance and over-confidence, it is eventually defeated by the
tortoise, fair and square, _in the long run_. I'd like to apply a
variation of this simple parable to the desktop operating platform race
that is being contested at present, between the seemingly plodding
penguin and the tear-away hare.

To begin our analysis of the competition at hand, we need need a little
history of the contestants.

Windows, as with many of Microsoft's technologies, was pre-announced by
two years in 1983. This oft-used ploy has the effect of neutering any
first-to-market competitive advantages bestowed on Microsoft's
competitors, who have often come out with more original products.
Windows was based largely on the concepts demonstrated by Apple's
development groups, in turn re-working Xerox PARC, who were influenced
by SRI's Doug Engelbart's ideas. After an invite for a site visit from
Apple, Microsoft was able to glimpse this brand new future of Graphical
User Interfaces, particularly embodied by what was phlegmatically
code-named SAND by Microsoft, (Steve's Amazing New Device,) the glorious
Macintosh. Microsoft were shown this technology early on, as it was a
significant player in the microcomputer software industry, whose
application software support was eagerly sought by Apple, to help cement
the availability of business apps for the fledgling PC-killer. This is
indeed ironic when one considers this situation replicated in the
present-day, specifically the leveraging power that Microsoft has over
Apple through the existence (or non-existence) of Microsoft's Office
product for the Macintosh. Regardless, Microsoft took Apple's GUI ideas,
and as happened on numerous subsequent occasions (for example with their
replication of Go Corporation's ideas on the first pen-tablet palm
computers in the early 90's) copied them. The GUI was so extremely
compelling that other firms, such as Digital Research, Inc., the then
king of operating systems platforms and purveyor of CP/M & CP/M-86,
(which Microsoft itself was to compete with when it purchased their
clone, DOS, product from Seattle Computer Products), IBM, Quarterdeck,
Geoworks, all came out with variations on desktop-metaphor interfaces
for x86-based computers.

Windows 1.0 itself was released in mid-1985, to very little enthusiasm.
In fact, Windows uptake was so underwhelming, that Microsoft had
problems selling the Windows-based apps (like Excel) that it had ported
across from the Mac. To overcome this embarrassing problem, Microsoft
effectively bundled the OS with the application as a run-time
environment, a reverse of what it does nowadays. This ploy wasn't
particularly successful either. Most users kept using DOS-based products
like Wordperfect and Quattro Pro. However, when the wave for Windows (as
a desktop interface) did eventually break, Microsoft's efforts in making
it's core applications available under Windows (because, let's face it,
hardly anyone else wrote Windows apps) paid off handsomely. It
effectively had built the right-shaped surfboard, and more importantly,
had helped drum up the wave, which it has since ridden to absolute
financial power and glory within the software industry. It's this
maneouver which caught its competitors (Ashton Tate, Borland, Lotus,
WordPerfect, Software Publishing Corp.) all off-guard. By the time these
firms had released feasible versions of their marquee applications under
Windows, Microsoft had entrenched its own file formats and application
interfaces as standards. What little market share was available to these
once powerful and monied software vendors, was snuffed-out when
Microsoft decided to 'crowbar' the market penetration of its
less-successful applications by leveraging the more successful ones,
through the masterstroke of bundling them all into Microsoft Office. It
could afford to make less money per application for a short period of
time, as it could in effect rely on its massively lucrative PC 'tax',
MS-DOS, through the then prevalent per-processor licence agreements with
OEM hardware vendors, which would eventually catch the attention of the
US Federal Trade Commission (FTC) in the mid 90s. Before anyone noticed,
Microsoft had almost total control of all the major 100+ and 10+ million
seller applications: word processors, spreadsheets, databases,
presentation programs etc.

So, we have a time spanning from a pre-history of Windows in 1983,
through delivery in 1985, and the mid to late 80's where Windows was
constantly in the mainstream IT press, heavily marketed but with minimal
success. Only a minute portion of the hundred million PC users actually
purchased Windows, and even fewer used it. It was not until the release
of Windows 3.0 (1989), and probably, more succinctly, Windows 3.1 (1991)
that a non-trivial portion of the great mainstream of computer users
started to move across to using Windows as their mainstay desktop OS.
Even then, most users kept relying heavily on DOS programs for core
business requirements, and well into the late 90s for games. Finally,
with Windows 95, released a decade after Microsoft's initial release,
that Microsoft can be said to have 'attracted' the majority of desktop
users to its Windows platform. We therefore have a 10+ year timespan of
non-linear adoption, from initial availability through to substantial
domination. It took exactly that long, with growth almost entirely
happening in the last few years of this span. This, for a product that
has the most expensive marketing and advertising campaign in industry
history. How does Linux uptake compare?

KDE 1.0 came out in July 1998. It was soon followed by Gnome. Between
them, KDE and Gnome are the first real attempts by the free software
community to create a _desktop_ oriented towards the expectations of the
great mainstream of computer users; namely by harking back to these
users' knowledge of and experience with MacOS or Windows 95. It's the
small things in operation, key-bindings, window focus that separate KDE
and Gnome from previous windowing environments under Linux. Afterstep,
Blackbox, FVWM (of various incarnations) were all more influenced by
Unix and Unix-like workstation desktops, and are generally unlike what
most users from the PC realm understood. While there is absolutely
nothing wrong with these other windowing environments, they were
obviously not going to be the mechanism through which Linux could win
comfortable converts from the desktop PC world. Both KDE and Gnome
started with functional yet uninspiring desktops, but now, after several
major and myriad minor releases, are close to matching industry
best-practice in GUI operating environments.

We are now in early 2002, and by our comparison to Microsoft Windows'
time-line, we are where Windows 2.0 was in about late 1988 or about 4
years into the Windows path to ascendancy. Please note, this does _not_
mean to imply that Linux as it presently stands is being equated
technically to Windows 2.0. We are equating market penetration. Back in
1988, Windows had to compete with both its x86 PC-based GUI brethren
(GEM from DRI, Geoworks) as well as its largest entrenched competitor,
DOS, ironically, also from Microsoft. The fact that Windows eventually
allowed for the seamless operation of the most important of users'
DOS-based application was crucial to the uptake and eventual success of
Windows. Without this ability, many users might just have moved over to
Desqview 386 or OS/2, which had arguable better DOS emulation
functionality. As a side-note at this juncture, it should be obvious
that if Linux were to allow for the seamless operation of the most
important of users' Windows-based applications, it would greatly assist
it in its race for desktop supremacy. WINE is thus of utmost importance
to Linux. Regardless, it must be stated clearly and forcefully, that at
this stage of its market penetration, Windows was considered a joke as a
desktop operating platform. It had a minuscule following amongst the
technology innovators; few of them took it as a serious contender in the
space. These people didn't adopt Windows for another couple of years.
And where these people lead, others, more often than not, follow.

Linux has perhaps the best shot at unseating the desktop OS incumbent
for a number of reasons. Past contenders, such as the Apple Mac and IBM
OS/2 had a number of inherent market shortcomings which hampered their
penetration. In short, the MacOS could never become a great volume
player due to its availability in a single hardware range, produced by a
single supplier. OS/2 competed head-on with Windows in the early 90's
and lost; partly through (in an industry where platform monopolies are
the natural course of things) the adage of 'their can be only one'
holds; partly because IBM was on the nose for many in the PC industry in
a similar way that Microsoft is now; but mostly through the onerous
per-processor licences encumbered upon PC vendors by Microsoft, making
the activity of bundling any alternative OS with their hardware
economically non-viable. Linux suffers from none of these hindrances,
and what were actual restrictive issues for the adoption of Linux
(installation complexity, lack of GUI polish, applications) have been
methodically resolved, one by one. As things stand now, there are no
valid technical or logical reasons for eschewing Linux; only politics
and religion remain and these cannot withstand the ever-present
pressures of cost-efficiency and competitiveness, demanded by business
economics, for too long.

Perhaps the most important reason why Linux has the best chance at
becoming the de-facto desktop standard in the medium to long term is
this: there is no obsolescence with Linux. As long as there are users
and a user community, there will be support and ongoing development.
Contrast this to Microsoft's recent actions in earmarking the removal of
support for both Windows 98 and Windows NT (their most popular OSes)
over the coming year or two, either stranding hundreds of millions of
users, or forcing costly, ongoing upgrades in a never-ending cycle.

Finally, there is another important factor which greatly enhances
Linux's chances as a desktop platform, and makes it far more attractive
than previous contenders. Price. Never underestimate the immediate
attraction of 'free beer'.

This combination of positive attributes is dawning on the industry. Up
until a few months ago, and perhaps for almost a year now, we were
seeing an increasing number of industry pundits decry Linux's role or
position on the desktop. There were claims that it's assault on the
desktop was stillborn. In recent months, this Cassandra-like chorus of
doom has been subsiding. This may be partly through the fact that Linux
hasn't, in reality, been washed away with the dot-bomb crowd, which the
pundits were expecting. In fact, as each month goes by, Linux on the
desktop is starting to draw the kinds of grudging respect that was wrung
from the pens of the industry nay-sayers about Linux as a network
infrastructure platform 4 years ago, and Linux as an embedded and
real-time OS 2 years ago. It's very likely that this pattern of
acceptance is coalescing in the minds of the pundits. Every time they've
thrown rocks at Linux, they've regretted it. They are learning not to
underestimate the staying-power of the penguin.

Where is Linux at present? Is it as far along with market penetration on
the desktop as Windows was 4 years into its push? Does it have as much
market recognition as Windows did at the equivalent time? It is my
strong belief that the answer to both these questions is s resounding
'Yes!'. As someone who was in the IT industry both then and now, I can
tell you that the recognition among both IT professionals and average
users of Linux is far higher than that of Windows in 1988. Cite the
Internet as the ultimate in guerrilla-marketing tools; cite the
phenomenal evangelistic efforts undertaken by Linux enthusiasts
worldwide; cite the growing disenchantment with Windows specifically and
Microsoft generally, it makes no difference. Linux has achieved far more
'brand' and aura of quality and value in its 4 years of desktop
ascendance than Microsoft's Windows had in the same juncture, regardless
of the gold-lined coal that Microsoft shoveled into the marketing and
advertising grist-mill. Importantly, Linux reigns near-supreme amongst
many of the technology-innovators, now.

Based on the time-frame example set by Windows' own march to domination,
we can see that the race between the penguin and the hare for the
desktop OS blue ribbon has only just begun, and we should settle into a
few more years of both contestants running the course. We have hopefully
shown, however, that Linux, far from having run its race against Windows
on the desktop, is analogous to the the tortoise, and is conscientiously
keeping pace, moving slowly at first, but inexorably forwards, towards
mainstream acceptance and perhaps dominance of the desktop market.

Con Zymaris (conz@cyber.com.au) is the CEO of Cybersource Pty. Ltd. a
long-standing IT & Internet Professional Services company.

Other Linux and Open Source Articles by Con Zymaris are located here