On Dec 8, 4:58 pm, Kier <val...@xxxxxxxxxxxxx> wrote:
> On Sat, 08 Dec 2007 21:46:17 +0000, 7 wrote:
> > Kier wrote:
> >> On Sat, 08 Dec 2007 21:05:05 +0000, 7 wrote:
>
> > Why can't you believe it?
>
> Because there are no verifiable statistics.
Slight correction here. There are no PUBLISHED verifiable statistics.
Major sites like Google, Microsoft, MSN, AOL, Yahoo, and other popular
major sites such as Travelocity, Expedia, and others - provide some
very good statistics. You see, these major sites all use user
authentication. This allows site owners to carefully study not only
the browser signature, but to associate these signatures with specific
identifiable users. This allows them to see which users are using
Linux exclusively, which users are using Windows exclusively, and
which users are using some combination of Windows and Linux.
Microsoft has made statements in sworn testimony that provide some
indicators. In the DOJ Antitrust case, they suggested that Linux had
14% of the market, nearly 140 million users. In the EU Antitrust
case, they upped the number to 17%. In both cases, these numbers were
unchallenged in cross-examination. This may indicate that there was
confirmation.
There are a number of tools that can be used, including having users
provide a user ID, with or without a password, persistent cookies, and
other "sessions", all can be used to know who is using the particular
ID.
Since these sites have millions of vistors, it's a large enough sample
to give a pretty reliable view of the most significant markets.
Of course, these statistics are as closely guarded as the crown
jewels, and are generally only shared in very restricted terms with
major commercial advertisers, and even then only in very general
terms.
Microsoft obviously wouldn't want it's partner sites sharing with the
general public any numbers that were substantially detremental to
Microsoft. For example, if the Linux share of the market had grown to
20% of the market or was growing at 2-3% per year, that would mean a
growth rate of 20-30 million users per year. A growth rate of 1
million new Linux users per month would be about 1.2%, perhaps 1.5%
increase per year.
Of course, there is a very good chance that most of these users are
hybrids, users who use BOTH Linux and Windows. This seems to be most
likely, especially since Linux makes it so easy to use both Linux and
Windows.
The possibility that the Linux market share is substantial might be
indicated by the move by most of these major sites away from IE only
site structures, adoption of industry standard javascript supported by
both IE and FireFox, and better support for Flash and PDF on Linux.
Another indicator that Linux is more successful than indicated in some
of the IP address surveys is that Microsoft is so worried about
Linux. Since 1997, Microsoft has targeted Linux as the "Number One
Threat". According to memos exchanged between Microsoft executives,
Microsoft must "Win at All Costs against Linux". They are willing to
almost give away even their most valuable software, when there is a
risk of a company making a huge distribution of Linux.
Microsoft has shifted it's policy toward computers for the needy,
offering to let charities that recycle old PCs and give them to low-
income families or 3rd world countries, install what ever version of
windows will work, for as little as $5 per copy, so long as Microsoft
is given detailed information about each recipient.
Linux may even be a key factor in the formation and distributions of
the Bill and Melinda Gates foundation. Part of it was a tax dodge, to
avoid alternative minumum tax, but an even bigger portion seems to be
a desire to offset the impact that tens of millions of Linux PCs per
year, donated through NGOs helped by Red Hat and other Linux
distributors have had on third world countries, especially Mexico,
South America, India, Africa, China, and southeast Asia.
While Microsoft was busy trying to focus the 300 million people in the
United States, and another 1/2 billion in Western Europe, Linux was
being mass-distributed to countries with populations in the billions,
nearly 5 billion total population that had been ignored by Microsoft,
but not ignored by Linux.
Meanwhile, Linux was also making inroads in the United States as well,
not as a primary operating system, but as a secondary operating
system.
> It's quite possible that there
> are many, many Linux downloads, per month, but there's no accurate way of
> telling how many of those turn into genuine Linux users.
There are so many ways to get Linux these days, and CD and DVD
duplication is so easy, and completely legal with Linux. Perhaps the
only way that Linux usage can really be measured is by tracking the
automatic update systems. Red Hat, SUSE, and most of the other
distributors have both commercial and freebie update servers, but
these servers provide a good way to measure not only how many Linux
users are actively using Linux, but also how many are using what
applications.
Unfortunately even this has been delegated to mirrors and proxy
servers. Many of these mirrors don't have the mechanisms to report
back to the "mother Ship", which means that many of these servers may
not be reporting statistics back to the original distributors.
Ironically, the biggest challenge in tracking Linux is that there is
no central authority. There are over 100 Linux distributions, and
many of these are mirrored by hundreds of secondary and tertiary
sites. If each of these sites is even doing only 1 download per
minute, that could be 10000 downloads per minute.
> So, where do you get your information?
We would all like a reliable source of information that gives us a
better sense of what's happening with Linux. We'd like to know what's
working, what isn't. We'd like to know how Linux growth is being
affected by other events such as the SCO lawsuit, the success of OS/X,
the release of Vista, and the release of 64 bit multi-core processors.
Microsoft has the advantage here. They have a number of different
"windows", including the Certificate Authorities used by SSL sites,
the commercial and registered user identification sites, and several
ISPs, as well as POPs like MSN.
One good gauge is the is the activities of Microsoft. The more they
try to dis Linux, the more they try to talk up successes that aren't
confirmed by OEMs, the more they try to spread FUD and disinformation,
the more successful Linux is becoming.
It's very possible that as many as 1/2 of the PCs being sold today are
being converted by end users into Linux systems. They still run
Windows, but Linux is also installed in some form. This trend is
indicated by the proportion of machines sold with DirectX-10 vs those
sold with OpenGL video.
> --
> Kier
|
|
