The business world has never seen anything like the Linux ecosystem before. Many of us barely have our heads around the competitive landscape of the proprietary software publishing business with its near-zero cost to produce additional copies, complex licensing terms, huge gross margins, often changing pricing strategy, and a development strategy that varies based on market power as much as user needs. The software business is certainly not your father's Chevrolet. Then Linux and Open Source Software (OSS) come along and drop a bomb on an already complex competitive arena by adding huge wrinkles through openness on the development side and by tossing out many of proprietary software's use limitations and its high price. These factors further distance the OSS ecosystem from traditional business models and push it toward being a social phenomenon.
Within the context of the business and social phenomenon that is OSS, this article discusses why there exists a natural niche for a non-profit member organization to be front and center in the move to OSS on the Desktop.
The View from the Desktop System Administrator's Desk
I got a call from a friend at EDS last week. Her particular office was busy, not with productive work for EDS customers but with disconnecting from the network, burning backup CDs, reconnecting to the LAN and hoping they could download and apply Windows patches before getting infected by MSBlast. Someone had probably been infected over dial-up and brought the bug inside the firewall. Everyone had to download and install patches. Some had to do 2 patches, others had to do 4 patches. Folks were grumpy both about the problems with Windows and with getting pulled away from their jobs to be temporary Microsoft byte hounds.
Service pack downloads can get large and I won't speculate on the time or money spent or on the psychological aspects, but keeping these Windows operating systems happy was not a free ride last week, nor this. When Windows hits you in the side, as it has for many, you can bet that the people involved - both users and administrators - were a little steamed. Talking on the phone was plenty close enough to EDS's offices for me last week.
Windows insecurities are only part of the Desktop OS scene. For most of us, our desktop alternatives are Windows XP or son of Windows XP for a little while yet. We currently watch the calendar and check every so often for word that Linux and its application software are close to maturing into a viable Windows alternative. Some of us even give Desktop Linux a whirl every so often. My tries in 1999 and in April of this year were both encouraging. My April 2003 attempt even came within an inch of success, but I ultimately had to switch back. People have been speculating for years on when a viable Desktop Linux will arrive and the answer varies depending on who you talk to and for what type of user. Recently, the talk has gotten serious.
In a nutshell, Linux on the Desktop is already here for single-use users like call centers, order takers, and point of sale systems. Linux on the Desktop is also already here for a large number of highly technical workers including many programmers, scientists and folks that make movie effects and animation. The big slice that remains is the knowledge worker - the folks that use Microsoft Office, an e-mail client, a Web browser, a productivity suite, a messaging client and maybe sync with a PDA in their daily work. These are the folks where Linux has not quite turned the corner and guesses on when Desktop Linux will arrive here vary from 18 months to 3-4 years depending who you talk to. It's also generally accepted not to be a matter of "if" but of "when" - unless there is some major new technology that appears in the desktop environment.
So we know Linux on the Desktop will arrive and maybe even pretty soon. The next question is how will it arrive, or, since it's developed by the users for the users, how do we want it to arrive?
The Case for a Non-Profit Member Organization
It's not real hard to make a Linux distribution. Maybe it's hard to make a good one, but there are dozens of good distros on the market. Some of them are produced by very large teams such as Red Hat, SUSE, and Debian while some of them build on existing distros and are produced by single individuals such as Krud Linux.
Most distros are available for free via download or can be had for a small fee. Red Hat 9, for example, can be downloaded for free and then, for $60/year, you can get unlimited access to Red Hat's up2date service which patches bugs and holes, provides software upgrades within a reasonable time frame and lets you download additional software packages very easily. Krud Linux does the same thing, and for nearly the same price, but mails CD-ROMs instead of doing it over the network. Pretty neat that you can buy Red Hat from Red Hat or a modified Red Hat from Krud for almost exactly the same price. Red Hat, by the way, would claim they are a much better value due to the support they can provide but, still, having Krud in the game is important to us all for reasons of choice and competition.
One of the problems Red Hat, Microsoft or any for-profit organization has is, no matter how hard they try, it will always be limited by, as Microsoft will say, the need to balance the needs of the company with the needs of the customer. When the two needs conflict, at best a compromise is in the offing. I strongly note that Red Hat is not like Microsoft. Red Hat, SUSE and others have actually done a pretty wonderful job in bringing resources to the OSS community, funding developers, and being a buoyant force for OSS. But even RedHat has limits due to the needs of the company.
So, from my Desktop Administrator's seat, it seems to me that Krud, Red Hat and others can be beaten by a membership organization focused on an Open Source Desktop solution. Think of all the things my users and I need: software, support, training, easy installation, software updates, software administration, basic network architecting and administration, hardware and hardware administration. And, being aware members of the OSS community, we also need some higher-level things to happen, so that our ecosystem remains healthy for the long term. Things like: fill gaps in the software, support developers, defend against legal challenges, and lobby the good folks in Washington DC and the EU for fair legislation.
Krud doesn't even pretend to address this list and Red Hat and SUSE might say they do, but really they don't come close on a number of counts. Not only don't the for-profits address my needs completely but, if they did, would they offer an attractive service, or would the needs of the corporation munge into the mix and mess it up? Specifically, would a non-profit that operates completely openly in most every aspect of its operations do a better job for me and for less cost?
It's common to use a non-profit organization to address needs like lobbying and training. I'd argue that within the OSS ecosystem, an open non-profit membership organization is also going to beat a Red Hat in activities like software packaging, support, documenting best practices, and ... you name it.
Take training. The non-profit can do a survey and assess where there is need for training. It can then develop a training program and make it available to all members at no cost over the network or for print reproduction cost. For an additional fee, users can attend a live training session or become certified as a trainer if demand for warrants it.
Red Hat on the other hand, does many of the same things but RedHat needs to make a profit from its training so its training information is not going to be as open and its delivery methods are going to be focused on generating revenue. The training materials, training events, and certification testing will probably cost more.
Under this approach, the openness of the non-profit in determining what products and services to develop and in providing an open process throughout makes for a lower-cost and, possibly, a higher-quality solution as well.
The same can be said for support, software ... you name it. The open approach of a non-profit will beat a Red Hat. I am also not claiming that all needs have to be met under one roof. Only that it seems, on balance, that the non-profit beats the for-profit here. I also think the idea of a non-profit community-based organization more closely fits the OSS community. Open software and the Internet enable openness in related activities.
So, why does a membership organization make sense for Desktop Linux and not perhaps as strongly for Linux in the server room or Linux in embedded devices? The answer is twofold. First, there are huge numbers of desktop users and second, the desktop environment is both relatively homogeneous and relatively simple. Desktop Linux can come from a non-profit organization direct to the world and, as the user base climbs, the membership fees could sink down to the cost of a nice can of beans for a single seat.
Making it Happen
A healthy non-profit that provides the services described above is probably the cherry topping on Microsoft's Linux nightmare cocktail. The ideas for how to set up and run this type of organization have been described in nice detail by others but I would like to add that I think this organization should be "by the users, for the users." To me, that means a focus on the user more than the Linux vendor or the hardware vendor.
The biggest hurdle is revenue. We need to get this idea off the ground and to do so, we need knowledge workers like those at EDS and other corporations, governments, and non-profits today to talk up this idea and move it to the forefront. Talk to the people within your organizations that support Windows desktops. Talk to the people that are board members or active in trade and professional associations and see if they see a fit for a membership organization providing these services and products to you as desktop users.
Making a commitment to invest in this type of OSS organization, and to put some dollars behind it now, will bring that date of maturity and its many advantages closer. Write back to me if you want to help making this happen.
- Paul Nowak has administered Unix servers since 1995 and Linux servers since 1996. He is is a advocate of open source software and Linux on the desktop.