*BSD News Article 39238


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msunews!uwm.edu!vixen.cso.uiuc.edu!usenet.ucs.indiana.edu!roland
From: roland@copper.ucs.indiana.edu (Roland C Dowdeswell)
Newsgroups: comp.os.386bsd.misc
Subject: Re: Why *BSD's have smaller user base?
Date: 10 Dec 1994 01:08:51 GMT
Organization: Indiana University, Bloomington, IN
Lines: 66
Message-ID: <3cav33$aug@usenet.ucs.indiana.edu>
References: <3ah170$n5o@mall.sinica.edu.tw> <3ahojg$iis@blaze.cs.jhu.edu> <BLYMN.94Dec2175232@mallee.awadi.com.au> <jemillerD0KEnv.7zH@netcom.com>
NNTP-Posting-Host: copper.ucs.indiana.edu

In article <jemillerD0KEnv.7zH@netcom.com>,
John Edward Miller <jemiller@netcom.com> wrote:
>Brett Lymn (blymn@awadi.com.au) wrote:
>: Funny - when I was going through the same process I found out which OS
>: did what _I_ wanted.  At the time 386BSD was best for me because it
>: had support for SLIP and I have stuck with the *BSD variants ever
>: since.  YES I KNOW THAT LINUX HAS SLIP/OTHER SERIAL NETWORKING NOW BUT
>: IT DID NOT WHEN I WAS LOOKING ok?  Volume in the newsgroups seems to
>: be (IMHO) a strange way to gauge the applicability of an OS.
>
>
>Temper, temper.
>
>Newsgroup traffic is IMHO a very valid way to gauge the applicability
>of a freeware OS when you're going to be fairly dependent on same
>newsgroups, and related mailing lists, for support.
>
>John.

I think that the point of the article to which you are replying is that
volume of newsgroups is a strange way to gauge the merits of an OS.  I
must say that I have to agree there, content would be better.  (I'm not
making any implication about Linux, but) lots of articles with poor
content I think is a lot worse than a few well written articles. (more
stuff to wade through)

When I chose NetBSD, the factors in my decision were basically that I
wanted it to be BSD, since I was used to working on BSD-derived systems.
I also saw a lot of ported programs out there for Linux, and that made
me apprehensive, to say the least.  It gave me the impression that
things might be difficult to port to Linux.  And when I say ports, I
mean that there were diffs to the source.  The only problem I've had
porting programs to NetBSD have been in the configuration of XFree86,
and a few programs make some kludges to work with BSD, that are
unnecessary.  (#define uint unsigned int -- and other redifing of
things that they think are undefined in BSD.  I think that they are
just figuring that it'll be an older BSD than 4.4)  That reason I
think is a bit weak, but what can I say, it is a gut feeling.  I
also like it that NetBSD is porting to other platforms.  I feel that
the PC is by no means sacred, I just got it so that I could have
BSD, and there is no guarantee that I'll stick with PC's, and I think
that the transition would be much less painful if I was just switching
hardware, and not switching the OS.  (a lot of recompilation, but,
should have to report all the programs that I have)

If we are going to judge the relative merits of an OS, based on newsgroups
I think that the following formula might be better: (the larger the g the
better)

   g(OS) = 1/N(OS)

where N(OS) = average number of times in a week that any development
of a company run by Bill Gates is mentioned in any of the newsgroups.  :)

This would also provide an objective, measurable way to determine things.
One should try to make sure that only the articles of people who run
the OS in question are taken into account.

I think that Linux is more popular because of notoriety.  Of people I've
mentioned that I run NetBSD to, very few know what it is, but quite
a lot of them know what Linux is.  (or have heard of it)  So I think that
Linux gets a lot of users who've never heard of *BSD.  And once people
get one, they're likely to stay there unless the others can offer
significant advantages (in their minds, of course.)

roland.