This is a good rant.  I got convinced of the advantages of NetBSD by former
core team members.  But when I went to install it, I really could not do
much with it.  Couldn't get FreeBSD stuff to run, as promised, and nobody
seemed able to tell me why.  I sure as shootin' don't have the BSD
experience to figure it out for myself and make a living too.  I can run SMP
on my new Pentium II Xeon box, but not on Open or NetBSD, which gives me the
options of NT, FreeBSD or Linux.  The choice is pretty obvious if you want
to use that second processor and not be overcharged for underwhelming work
(NT) or inherit a bizarre bazaar mess (Linux).

FreeBSD gets a LOT of visibility out of Yahoo, Walnut Creek CDROM, and
OpenBSD, despite a late start, has marketed itself well on the basis of
"security issues" which gets of a lot of media hype in its favor.  NetBSD,
well.  It's there.  But it has no associations to attract new blood, and
some NetBSD types will tell one bluntly that it's not about getting more
users, more hands, more outreach.

NetBSD marketing, obviously, sucks The Cosmic Big One.  And it leaves NetBSD
adherents, as well I've seen, bringing it into the shop where they work
getting asked "You want to install WHAT?" followed by "What is it?  What
will it do for us?  Can it run any apps?"

NetBSD seems to be losing critical mass.  Free, Open and Linux are picking
it up.  True, if one has users, it is no longer one's private fiefdom and
backyard sandbox.  But beyond a certain point, if the project is getting
smaller (it seems to be after the last year), then it progressively becomes
"Of course it runs NetBSD.  Now go away." and becomes an exercise in
solitary masturbation.  If it's an OS nobody has, or wants much, what's the
point?

FreeBSD has managed to attract enough of a following to get a Netscape port
all its own.  I recall one NetBSD shop that only had an old Netscape 3 port
that was buggy as hell, couldn't do email and was on the verge of getting
kicked out for this reason along with the paucity of apps.

That IBM is backing Linux is a VERY good sign for NetBSD.  IBM is virtually
always wrong, and actively fed Gates for a decade until he got large enough
to seriously maul them in a business they considered their own.

But the other indications are that FreeBSD is going to be the BSD that
survives and prospers.  It has already reached critical mass in most
respects and is leaping even further ahead of other flavors.

FreeBSD's chief success came from first getting in at Yahoo! and then
building upon it, being receptive to a commercial marketer (Walnut Creek
CDROM) and helping them further with a web server that would demonstrate
BSD's value.  By creating an aggressive marketing focal point, FreeBSD got
visible in a hurry.  OpenBSD does the same deal, and was not slow on
catching on to the notion that by using The Computer Shop in Calgary with
lots of links to them, they could make OpenBSD and paraphanalia easy and
cheap to obtain.  There are, what, 6 flavors of OpenBSD you can buy, and
even more flavors of T shirts.  NetBSD was no doubt, once upon a time, in
the running as being Walnut Creek CDROMs showcase product.  Walnut Creek
CDROM didn't really give a damn; it was and is in the business of selling
CDs and books.  As long as it's good and makes for happy customers, the
flavor was not an issue.  FreeBSD held up its hand, and its team did the
necessary work.  The rest is history.

You can buy a NetBSD CDROM if you know where to go.  This is CheapBytes, who
sell CDs for so little they have no advertising money.  Hint: find a big
Internet company to let people know it uses NetBSD and SUPPORT THEM!  Then
find a CD marketer who will be glad to get the hits, sell the CDs and
paraphanalia, and NetBSD will be on its way.

But all this requires a support effort.  FreeBSD found Greg Lehey to write a
survival guide for people to buy, albeit not cheaply, so they can get in
there, get set up, and use it effectively.  That book is my FreeBSD survival
guide, and I don't have to ask lots of ignorant questions on the lists.

Most of what this requires is sweating in the trenches, maintaining FAQs,
links, new data, lots of focused testing, a stream of solid new releases.
In return, FreeBSD gets a massive internet presence, some contributions, and
plenty of support.  The FreeBSD core team and associated developers were the
ones that initially rolled up their sleeves and sweated like hell to make
all this happen.

Really, the future of NetBSD is up to the NetBSD core team and Foundation.

-----Original Message-----
From: Michael Graff <explorer@flame.org>
To: netbsd-advocacy@NetBSD.ORG <netbsd-advocacy@NetBSD.ORG>
Date: Monday, December 06, 1999 4:46 PM
Subject: Why people know what FreeBSD and OpenBSD are, but not NetBSD.


>When I mention NetBSD to people they have no idea what it is.  Why is
>this?  There are many reasons, and they're not simple ones, but here
>is what I think will help fix it.
>
>(1)  When looking at a possible solution to a problem, for instance
>     /etc/rc vs. /etc/init.d/, don't reinvent the wheel.  Pick
>     something someone else already does, and use it.  Doing things
>     our own way will only hurt us here.
>
>     No, you won't get the perfect solution doing this.  But you will
>     get a working solution, and one that others will already
>     understand.  If we reinvent the wheel, and it's different, people
>     will just choose something they understand.
>
>(2)  We have some things no one else has done right.  We were the
>     first free OS with USB support.  We should have been much louder
>     about that.  We are, IMHO, the best with ISA PNP support.  We
>     need to improve our already good support and make certain people
>     know about it.
>
>     In other words, we need to market ourselves.
>
>(3)  We have little to no corporate support.  Even OpenBSD, which has
>     one of the most annoying leaders in existence, has managed to get
>     companies to give it money.  Why are we not doing this?
>
>     Yes, it requires someone to run out and do real marketing.  And
>     in the current world most people would rather fund linux oriented
>     projects rather than anything *BSD.
>
>(4)  Get large projects into the light as soon as posssible, rather
>     than sitting on them behind closed-doors.
>
>(5)  Make releases _far_ more often.  In fact, make a weekly, if not
>     daily, snapshot for at least the i386.  Yes, I could do this as
>     well as anyone else could.  In fact, I should.
>
>(6)  Rather than develop the perfect solution that will take time,
>     design a workable solution and get it out for _public_ view as
>     soon as possible.  Fix it as time permits.
>
>     This is contrary to one basic goal of TNF, so I'm told every time
>     I mention it.  However, this single goal is hurting us so much.
>     We are after all lacking basic features that other free OSs have
>     had for years now.  For instance, SMP on i386 at least.  linux
>     has it.  FreeBSD has it.  No, neigher did it fully right, but
>     they have it, which attracts developers, which adds warm bodies.
>     Which makes it more visible.  Which means things can get funded,
>     and then mostly-workable but not perfect solutions can be done
>     right.
>
>I've considered switching to other OSs recently, mostly because while
>NetBSD has many cool things, the other OSs pick them up quickly, while
>we sit behind on many basic things that I really want to use now, not
>in several years when closed-door projects finally release code, or
>worse work to stop open-door projects to do the work from starting.
>And then the closed-door ones never release a thing.
>
>--Michael