Subject: Making NetBSD Competitive
To: None <current-users@NetBSD.ORG>
From: Mark Andres <>
List: current-users
Date: 11/02/1997 15:48:44

I have been following the recent thread(s) on how to better promote
NetBSD.  I have been using NetBSD for over a year on Mac 68k hardware. 
At work over the past three years, I have used mostly BSDI with some
SunOS and (yuck!) Solaris.  I am one of those people who uses NetBSD
"to get work done."  For reasons that will become apparent, recently I
have also had some experience with FreeBSD.  

My experiences with NetBSD have helped me tremendously. I started out
just looking to learn more about UNIX while doing tech support for an
ISP.  My NetBSD experience helped me move up to a sysadmin position.  

Here are some the the areas I can see where the NetBSD project needs to
improve if it wants to attract more users.

1) Get in the CERT advisories.  

Everytime a CERT alert comes out, someone asks in tech-security, "Why
wasn't NetBSD listed."  The response is usually, "There was a clerical
error and it has been corrected."  But it never gets corrected.

For better or worse, people will judge the security level by what is in
the CERT alerts.  I do it myself.  I see Sun's response to every alert,
"Sun is investigating the problem and will release a patch, someday,"
and I think to myself, yeah, that's a good attitude towards security
holes.  But even worse, is having nothing mentioned at all.

At work, we recently decided to experiment with some free UNIX servers.
I pushed hard for NetBSD, but in the end, we went with FreeBSD.  One of
the main reasons against NetBSD was security concerns stemming from
not having any listing in the CERT alerts. 

2) Updated Web Pages on

When most people want information about the free UNIX alternatives, the
first place people will look is web pages.  If the pages are out of
date, then people will turn away.

Here is an example.  A few months ago, I inherited an IBM Thinkpad 220
(sold only in Japan).  It is a 386SX with grey-scale display and 1
PCMCIA slot.  We used to use them at work for in the field router 
configs running DOS/TELIX since they are really light and run on AA
batteries.  I thought great, I throw in a bigger HD and install NetBSD.
So I installed a 1.4GB HD and got an EtherLink III PC card.  I looked
at the NetBSD homepage but could not find anything about PCMCIA
support.  So I ended up installing FreeBSD instead.  I have since found
out that there is in fact PCMCIA card support in NetBSD, but how I am
supposed to know that if it is not on the Web pages.

3) Easier Installation

I started using NetBSD as more of a hobby and because I wanted to learn
more about UNIX sys administration.  And I learned a lot, becuase I had
to.  To get almost anything to run requires a lot of work.  Getting
even the simplest things to work, often takes time and doing thing by
hand.  This teaches you a lot.

However, I am now to the point where I use my workstation to do, well,
work.  If I need to get something running, I don't want to have to
fiddle with it for days.  I want to be able to install it and go.  If
you have never used the FreeBSD "ports" system, you should try it
sometime.  You just do a  cd /usr/ports/category/port_name and then do 
make install  A complex system of Makefiles takes over.  If you are
compiling HylaFax (a fax system) and you don't have the TIFF libraries
installed, the "ports" system first installs the TIFF libraries for
you.  This sure is slick when you just want to get work done.

I understand that work is under way to implement a similar system for
NetBSD.  This is a Good Thing (tm).  I hope that it will be developed
so it can be used in all ports, not just the i386 port.

Ideas for simplifying the initial installation have already been
discussed.  I think altenatives are the best idea.  FreeBSD
installation gives you the choices of installing from disk, CD-ROM, or
even via FTP.  BTW, I have seen at least one NetBSD CD-ROM.  At a
computer store here in Tokyo, I saw a NetBSD CD-ROM that had the i386,
mac68k, and sparc ports on it for 5,000 yen (about $60US).

4) Make NetBSD more international

When I first started lurking on the main FreeBSD list, I was surprised
by how many people posted from all over the world.  While NetBSD does
have worldwide penetration, it is certainly easier for FreeBSD and
Linux becuase they have done a better job of internationalization. 
Being in Japan, when I set up my netBSD box, one of the first things I
wanted was a Japanese environment.  After a tremendous amount of work,
I finally have one.  But I still don't have my LOCALE stuff set-up
properly, so everytime I open a kterm, I get an error message.  It
works, but it is not pretty and it was a major pain to set up.

Setting it up on FreeBSD, by contrast, was a piece of cake.  Since all
the LOCALE stuff is already built-in, it is very easy to build a
Japanese environment.  The same goes for keyboard settings.  I can use
the Japanese keyboard built into my ThinkPad on the console with no
kernel hacking.  This is a big plus in the international market.

5) Do not set up a rivalry with Linux and/or FreeBSD.

First of all, there is already enough flaming between the NetBSD and
OpenBSD camps as it is.  This is a big turn off to potential users.  I
know one FreeBSD user who considered NetBSD but was turned off by the
"constant flame wars between the NetBSD and OpenBSD camps in the netbsd
newsgroups."  The FreeBSD mailing lists have quite a bit of anti-Linux
propaganda.  I don't think NetBSD should jump into the fray.  

Now, taking occasional jabs at Bill and company is another story, 
within reason and good taste, of course ;->

Mark Andres                         E-mail:
            Running NetBSD, 100% Microsoft Free!
Me: /  Me on NetBSD: /