Subject: Re: Why do Linux box vendors thrive?
From: Ian Lance Taylor <ian@airs.com>
Date: 9 Dec 1999 10:02:22 -0500

   From: Jonathan Corbet <corbet-fsb@eklektix.com>
   Date: Wed, 08 Dec 1999 08:46:52 -0700

   Here is an FSB question that has been on my mind for a while: why do the
   vendors of Linux boxes (VA, Penguin Computing, Atipa, ...) continue to
   thrive?  The hardware business in general is hard, with serious
   competition, low margins, etc.  The box vendors are not only in that
   market, they are tiny companies in an arena that is increasingly getting
   the attention of the Dells and Gateways of the world.

   So does anybody have any thoughts on why companies go to VA rather than to
   Dell for their Linux servers, and whether they will continue to do so?

I recently had some indirect experience with installing Linux on Dell
servers.  It was a somewhat painful process, because the distribution
we were using didn't have the drivers for the new-fangled hardware the
servers came with.  In fact, we never bothered to get the right video
driver--since it's a server anyhow, it's good enough to just run the
screen in low resolution VGA mode.

For Windows, the hardware manufacturers write the drivers.  For Linux,
this is not the case, at least not yet.  So if you buy the newest
hardware, you wind up tracking down the drivers yourself, assuming
they exist at all.

We didn't go with VA or Penguin simply because Dell was cheaper.  But
I can see why a company with more money and less Linux experience
would buy from them.

I think the Linux hardware companies can thrive as long as they can
rely on their Linux skills to provide Linux configured for the newest
hardware.  In other words, they are essentially selling their ability
to gather and configure the right drivers, and perhaps to tune Linux
for a particular hardware configuration.

If hardware manufacturers start providing Linux drivers as they
currently provide Windows drivers, or if big vendors like Dell,
Gateway, or Compaq make the investment in developing their own Linux
expertise, then I think the Linux hardware companies will go under.
Well, I'll qualify that by observing that a natural way for Dell,
et. al., to develop Linux expertise would be to buy VA, et. al.

I'd say that VA's future is either to be bought, or to be stuck in a
small niche market forever, or to convert to a specialist driver and
configuration shop contracting to the PC vendors and/or the hardware
manufacturers.  I think the only way they could get as big as Dell
would be if Dell were asleep on the switch on the Linux thing, and
they're not.

Ian