German PC magazine "C'T", in my view the most serious of the kind of
"special interest" periodicals, in its first edition of this new year
published a graph from an own survey which I find simply exiting. It
sheds a very different light on the real landscape of OS use.
While hoping to get the precise numbers from the editors, here is what
I reconstructed from the (necessarily unprecise) printed bar graph.
Among the _online_ respondents, about 18 per cent declared to use DOS;
the most used Windows version was W98 with 57 per cent, but Linux was
used by a full 58 per cent of the users; Apple/Mac had 5.7 per cent,
according to the joined text (this corresponds to the rougher 6 points
for "others" in the reading of the graph).
"Stretched" values of the graph, and calculated as "ALL USES" (in
distinction from "per cent of all USERS which use [a given OS]", give
the following distribution - note well that the sum of the percentage
points in the first column is the base "100" for the second and third
column:
OS used Users p/c All USES -intended by those
(presently) -presently planning to buy
---------- -------- ---------- -----------------
DOS 18 8.1 5.8
Windows 3.x 2 1.2 1.0
Windows 95 14 6.4 2.0
Windows 98 57 26.3 24.4
Windows ME 6 3.2 9.8
Windows NT3 . < .1 0.0
Windows NT4 17 7.6 4.1
Windows 2000 28 12.7 19.6
Linux 58 26.7 26.7
other Unix 4 2.6 2.2
OS/2 4 2.3 1.7
others 6 2.9 2.7
-------- ---------- -----------
(N="ca.3800") (100) (100)
Evidently, multiple response was enabled with the question for OS use.
The magazine presents voluntary, active responses from ONLINE visitors
at their website shortly before Christmas (i.e., people who had filled
in and remailed a form); there is no indication of the precise date and
duration of that survey. Total number was given as "more than 4000" of
which, after "weeding out obviously nonsensical and unbelieveable"
responses, "about 3800" were taken as "believeable". There was some
control of double participation.
The poll was done in the context of gauging the PC market, though,
aiming at an assessment of present (Xmas season) and future general
market developments: Many of the questions concerned satisfaction with
performance and functionality of presently owned hard- and software, and
motivations for buying new in view of insufficiencies of the present
equipment. Nearly half of all respondents were presently content
though with what they had, some 16 per cent intended to buy new under
the Christmas/Winter season, the rest of the (larger) half intending
to buy a new PC would perhaps do that later in 2001.
The paper gives some _own_ caveats about the representativity of these
data: First of all, the poll was geared towards _online_ visitors who
would have, or would not have, the intention to buy new gear (thus,
individual, most probably then [too] "home" users); less than one per
cent of respondents had not yet an own 'puter (i.e., were logging in
from work or other place); the users of the online services of the
magazine's publishing house are certainly not representative, neither
for the total PC user popuation nor for the population at large.
The "pro-technological" bias, which the paper claims for itself and
assumes among its users, is characterised in the article with the
remark about "the participation of Mac-users, 5.7 p/c, in our online
clientele [being] considerably higher than the market share of Apple
of three per cent in Germany".
But more than that, it's the flabbergasting part of Linux use which is
definitely an indication of a "high-tech bias". The magazine is not "a
Linux paper" even if it carries a lot of Linux relevant stuff (by now)
too, but it was open to other-than-Microsoft/INTEL-themes from its
beginnings in the mid-eighties; it has a broad range of rather in-depth
coverage of PC/IT issues, thus making it "high-brow", in addition to
high-tech (compared to the high public relation content of most of the
other "special interest" print in this field), but AND and, it is indeed
popular in terms of copies sold (if I guess correctly, the second
biggest selling in German language). However, as these were responses
to a poll on a public web site, visitors there would not necessarily
have the same interest profile as the paper's readership.
Factors like these and some more have to be taken into account with any
interpretation of the data. Those "leading edge" biases even come to
bear in an other account given, on the present hardware equipment of the
respondents. The distribution of OSs' used is the more astonishing seen
the hardware the respondents declared to use - there must have been less
than 0.1 per cent using a pre-Pentium age PC:
Units with x86-type processors respondents work with at present, per cent:
INTEL Pentium, -MMX 11.6
Pentium Pro 0.9
Celeron 11.8
Pentium II 16.2
Pentium III 14.8
Pentium 4 .
AMD K5, K6 3.3
K6-2, K6-III 20.5
Athlon 15.3
Duron 2.9
other 2.2
------
(100)
(This sum of "100" would then be approximately, but not completely,
identical with the 100 - 5.7 [Mac] = ~ 94 p/c respondents using one of
the specified OSs in the table above. These data are from a pie-chart
with percentages printed out.)
All this considered, I draw these conclusions:
*** A DOS share of use which, by all means, is significantly higher than
for instance, the one for the Apple/Mac OS, and comes up to some of the
(in themselves incompatible and segmented) M$-Win-versions, points to
quite a solid "survival" platform;
*** There is a significantly high level of multi-OS use - to the
detriment of M$ products which aim at barring just that;
*** There is an overrating of M$'s importance - and M$ products rather
compete against themselves instead of expanding their share;
*** There sure has been a Linux hype last year, and the slightly
downward bend for "buy new" (55 p/c in comparison with those 58 p/c
presently using it, in the "C'T" poll) points to the quite real problems
poeple have yet with this OS: these are seldom technical ones but rather
those of user-accessibility [inbred tech jargon, ergonomical aspects in
a broader sense, i.e. of insufficient orientation to user-friendliness,
not the least when it comes to trouble-shooting]. However, and as there
are many signs of change in this realm, there will be probably the big
break-through quite soon.
*** Given the good "conviviality" of Linux with DOS, there's quite a
bright outlook for a reasonable "division of labour" - both in
technical terms and in functionalities of use - for a broad coverage of
any "DOS + Linux mix"; and:
*** Precisely because Linux tends to be "heavy" (both on ressources
use and use effort, even if the latter will be more easy by and by),
DOS will remain a sure value for all text (and numerical), "easier"
tasks, including quick-start (text-based) net applications.
-- Sure this is heavily biased too, as I'm usy with making precisely my
own "mix" of that kind, <bg>. --
// Heimo Claasen // <hammer at revobild dot net> // Brussels 2001-01-08
The WebPlace of ReRead - and much to read ==> http://www.revobild.net
To unsubscribe from SURVPC send a message to [EMAIL PROTECTED] with
unsubscribe SURVPC in the body of the message.
Also, trim this footer from any quoted replies.
More info can be found at;
http://www.softcon.com/archives/SURVPC.html
