*BSD News Article 56979


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.rmit.EDU.AU!news.unimelb.EDU.AU!munnari.OZ.AU!spool.mu.edu!howland.reston.ans.net!agate!nickkral
From: nickkral@parker.EECS.Berkeley.EDU (Nick Kralevich)
Newsgroups: comp.os.linux.advocacy,alt.os.linux,comp.unix.bsd.freebsd.misc,comp.unix.bsd.netbsd.misc
Subject: Re: a monthly FreeBSD magazine (and other *BSD's too)
Date: 16 Dec 1995 09:45:28 GMT
Organization: Electrical Engineering Computer Science Department, University of California at Berkeley
Lines: 100
Message-ID: <4au4fo$pbo@agate.berkeley.edu>
References: <4ajc07$sb7@unix2.glink.net.hk> <4atie8$hsv@agate.berkeley.edu> <4atq47$l46@agate.berkeley.edu> <4att23$mha@agate.berkeley.edu>
NNTP-Posting-Host: parker.eecs.berkeley.edu
Xref: euryale.cc.adfa.oz.au comp.os.linux.advocacy:30464 alt.os.linux:6437 comp.unix.bsd.freebsd.misc:10546 comp.unix.bsd.netbsd.misc:1555

In article <4att23$mha@agate.berkeley.edu>,
Jordan K. Hubbard <jkh@violet.berkeley.edu> wrote:
>Several people have done surveys and presented their results in this
>newsgroup, though those results are now so out date as to be useless and I

Regardless of this, I would still like to see information on those
out of date surveys.  Really!  I am curious.  I will keep the fact
that the surveys are out of date in my mind when I look at them.
Now if you would be kind enough to post some, ANY, sources, I would
appreciate it.  Instead, you are trying to attack me personally.

>Those aren't sources, those are flagrantly misrepresentative *attempts*
>to gather census.  I would be embarassed to attach my name to any of them.

Any system, short of counting every user, is going to have some bias
and "fudge factor".  As long as the "fudge factors" are clearly stated,
I don't see what is wrong with posting these types of information.

However, you haven't given us ANY information to go on.  

>>   I estimate this as being between 0.2 and 5% of the total number of 
>
>How kind of you to "estimate" this for us.  "50 people responded to my survey,
>and I conveniently estimate this to be (holds finger in the air) oh, .001%
>of the total population, so we have 5 million users!  How about that!"

First of all, a clarification:  The "I estimate..." part of this was
NOT written by me.  It was written by the person who maintains the
Linux counter.  If you would have taken the time to refer to the URLs
I posted, you would have noticed this.

I DID NOT make up those numbers myself.  There is a reasonably good 
statistical basis for those percentages.  Again, I'll suggest that
you read ftp://aun.uninett.no/pub/misc/linux-counter/README.ESTIMATES
for more information why the person who maintains the counter believes
that estimate is correct.  I have no affiliation with the person who
runs the Linux counter.

The other source that I listed was just for percentage comparisions
between the two operating systems, based on relative operating system
usage as WWW servers.

>Give me a break!  You request that I cite my sources, as if by inference
>you were blessed with accurate user data on the Linux side, 

No.  I have no *personal* numbers regarding the number of Linux users.
I do, however, have access to surveys which list how many Linux users 
there are.  I have listed the sources to all my inforation, so that
they can be properly debated.

>then you pull
>the freshman stats trick of taking a set of numbers and biasing them by
>a wholly unfounded and unprovable "fudge factor" to support your case.

*Sigh*.  It wasn't my fudge factor.  It was clearly stated that the 
numbers were an estimate.  People who read this newsgroup are free to make 
their own decisions regarding the use and validity of the numbers.  There 
is a good basis for the "fudge factor", if you would take the time to read 
the URL I posted.  Jordan, why do you have to make EVERYTHING so 
difficult.  *sigh*.

Subsampling of a population is a valid way to generate statistics for that
population, so long as the sampling techniques are open and subject to 
review.

>The truth of the matter is that NOBODY knows accurately how many FreeBSD,
>NetBSD, 386BSD or Linux users are out there.  Nobody.  

But at least the Linux community has attempted to provide an estimate.
You may question the estimate, but the estimates have survived review
by others.

>I only hope for your sake that you're not taking stats at Berkeley, or if
>you are that you've at least the got the good sense to go into economics
>afterwards.  Then you can spend your days debating bogus figures you know
>nothing about with the other economists and leave the rest of us in peace!

Sigh.  More personal attacks.  I expected more from you, Jordan.

Anyway, I would appreciate it if you could provide me with figures,
ANY figures, regarding the size of the FreeBSD user base.  My original
contention was that there are no more than 20,000 FreeBSD users, and
you have shown me nothing to disprove that statement.  I would appreciate
it if you could post even out of date statistics, ANYTHING!  But my
guess is that there are no valid FreeBSD user statistics out there.

Instead of providing your own facts regarding the size of the FreeBSD user 
base, you tried to attack me personally.  If I were you, Jordan, I wouldn't 
want to attach my name to random personal attacks that have no basis in 
fact.

I look forward to any numbers you can give us, even if they are out of
date.  Perhaps we can try to keep this conversation from falling into
a personal flame war.

Take care,
-- Nick Kralevich
   nickkral@cory.eecs.berkeley.edu