May 6, 2004

Red Hat Desktop strategy: Semantics have been part of the messaging problem

Author: Jeremy Hogan

Editor's note: This commentary from Red Hat's community relations manager is a response to NewsForge's Joe Barr, who asserted earlier this week that the company had given numerous confusing messages about its intent to enter the desktop market.

Much has been made in the press and on community sites about Red Hat's
ambivalence in the "desktop" space. If you're reading this, you may have
written an article or two on it yourself. Or at least flamed us in your

The most vocal detractors seem to agree that when Red Hat dropped Red
Hat Linux and retail boxed sets, we were in effect reneging on our
consumer desktop. I say "consumer" desktop, because the world has
varying definitions for desktop; it's a commodity term needing
qualifiers such as "corporate" or "consumer." We don't expect those to be
the same class of users any more than you'd expect to find NT running as
your gaming and entertainment platform. Desktop also means KDE v. GNOME,
or the beige thing on your desk's top.

Hence the first seed of the dilemma. We have conflicting terminology.

The second part of the dilemma is that we had one product line, drawn
from common bits that moved too slow for some and not fast enough for

Retail schedules were pressuring our release dates. ISVs and IHVs were
asking for the brakes to be put on, and high-dollar customers wanted
multi-year deployments they could count on to not change drastically.
Others wanted the new stuff. Now or sooner if possible.

We dropped the revenue stream that was Red Hat Linux in retail and
created the enterprise line and started the Red Hat Linux project, now
known as the Fedora project. We wanted to merge with the existing Fedora
project for a few reasons.

  • 1) They had a system capable of accepting third-party contribution
    2) They had a name that didn't confuse our customers and industry
    3) They are a talented group of people

We tried to position it as best we could without telling it what to be.
For example, we called it "for hobbyists and developers," but Fedora is
what you make of it. It's the proving ground for our next level
technology (e.g. Core), but also enables its own community ecosystem in
FedoraNews, FedoraLegacy, FedoraForum, FedoraChat, and FedoraLegacy.
People run it in production. It's had its own growing pains, but so far it
has been positively received and has become the most publicly tested
version of Linux Red Hat ever released.

But since many folks saw retail equaling desktop, the questions still came. Was
Fedora the replacement for this retail desktop? Was Fedora even related
to Red Hat? What do you do if you can't afford the new line?

Then the end-of-life announcements, and stuff hitting the proverbial fan

That people were using this retail boxed set for a general-purpose
desktop conflicts with Red Hat's positioning it as a server, or SME
solution. We never dropped the D-word. Because the D-word
means Microsoft in 90 percent of the world's mind. And that opens up a whole
other level of expectations.

Fast forward, and we start talking about the corporate or enterprise
desktop. We launch a product called Advanced Server (now AS), another
called Advanced Workstation (now WS), and began to attack the gaps. ES
is launched for the edge of network. Mind you, we'd used the terms
"enterprise desktop" and "workstation" for a reason, and that is simply
to keep the expectations of our products and services in line with what
we can deliver.

Not surprisingly, the enterprise line sells well in its target
settings: Unix-to-Linux migrations. Many folks think Red Hat goes only
where the money is. Well, as a public company we have to find money, but
the reality is, we beat Microsoft to where they were trying to go. Redmond
wants the Unix space. This was more strategically cognizant than we are
generally given credit for.

Then Red Hat's Chairman/CEO/President, Matt Szulik, is quoted as saying something
about the role legacy Windows desktops were still playing.

The proverbial fan is set to oscillate. Headlines read "Red Hat Abandons Desktop!" and "Red Hat CEO Recommends Windows."

And here we are. We've just launched the first Red Hat product with
"desktop" in its name (albeit with the silent "corporate" in front of
it). This move is alleged to be in response to Sun's Java Desktop
System. In actuality, it is in line with our market's demand, and the
strategy we articulate in our Open Source Architecture. It is also just
the first phase, because we aren't ready to give (or exceed) the single
system consumer desktop experience currently available.

Now, I have been running Linux exclusively a home and at work. My 8-year-
old daughter uses it (mostly because she never had a chance to get used
to another suite of tools, apps, and games). But I'm a geek. And I work
with bigger geeks all day. That does not mean Red Hat is ready for a
million phone calls on getting a million peripherals to work. Does that
mean we're ready to wrestle the convicted monopolist head on, replete
with their new partners in crime at Sun?

The fact is that Linux has the functionality needed for most anything
you can do on end-user Windows, but the world is already used to what
it has, warts and all. So it takes not only recreating the
experience, but the massive network of providers (systems integrators,
techs, pre-loads, OEMs), and the peace of mind that they have with a
known entity. How many people won't even talk about migrating without
Quicken? And that's one app.

So it takes a more disruptive enabler to shake them loose. Massive
downtime due to a virus? Every million machine infestation makes people
willing to jump. Better game support? Try WineX. Support for plug-ins
and Office apps? Try CodeWeavers. Want a preload? Go to

The beginning of the support infrastructure is there. A competent admin
or savvy user can get all those apps together and create one-on-one
relationships, but it takes longer for a corporate entity to do likewise
and guarantee any level of user experience.

As to our future intent, I said the Red Hat Desktop v.3 is just the
first phase. We are focusing on things like usability, interoperability
with MS servers, directory access, legacy document support, security,
hardware, and peripheral support for subsequent releases. Look to Fedora
to see these things early.

There are matters of focus, time, and resources to consider as to when
we'll get to the point where we meet all the definitions of desktop.
None of which means Red Hat is not interested in a consumer desktop.
Even if you accept the bleakest portraits of Red Hat as money-grubbing
Microsoft of Linux, you'd have to concede that being the Microsoft of
anything means being in the consumer and corporate desktop space.

We've had communication and terminology issues, but the march in this
direction has been pretty steady on our part, and there's an ever-increasing window before Longhorn to get there.

Jeremy Hogan is Red Hat's Community Relations Manager.


  • Linux
Click Here!