The Linux marketplace is a dichotomy if I've ever bumped my head on one. On the one hand you have the free software folk who brought us GNU and the GPL (General Public License), and on the other you have, well, IBM Corp. for example. One side is altruistic and interested in sharing knowledge for the benefit of all. The other side is materialistic and motivated by profit. Between the two poles stand the open source crowd and the bulk of the Linux community. The most recognizable voice in the open source group belongs to Eric Raymond, who has brought us parables aplenty to explain the differences between the cathedral and the bazaar, and told us how to make money in a strange land with diametrically opposed interests.
Another paradox lies in the land of Linux: even though the OS has basked in the press spotlight for two years, failed companies in the Linux market far outnumber successful ones. Many Linux companies have failed, retrenched, or simply struggled to keep their heads above water at the same time that Linux usage among individuals and organizations has grown like crazy. Not everyone has pockets as deep as Big Blue's.
But whether or not Linux companies make money may have a lot to do with whether the community can sustain itself in the long term. Like Linus, many kernel hackers hack just because it's fun. The fun is there whether commercial firms make a buck or not. But commercial success for Linux companies has everything to do with how big the community will become. The problem is that actually making money with Linux in particular and free software in general has proven to be as elusive a goal as Ponce de Leon's fountain of youth.
There seem to be four main areas where most people trying to make money with Linux have focused: distributing Linux, providing support services for Linux and open source applications, selling hardware to the Linux community, and selling Linux applications. I'm most worried about the last item on that list, the applications sector. Why? Because I want a lot more Linux desktop applications than those that exist today, and patience is not my strong suit. My core competency is in worrying, and that's what this column is about.
In recent months we've heard of Stormix's bankruptcy, the pending sale of Corel, and the cutbacks at SuSE. Post-IPO (initial public offering) Red Hat still hasn't made a profit. But I don't worry about the number of Linux distributions going down so long as competition among them remains. In any event, that market is already busy correcting itself. The best proprietary distributions will both survive and thrive, as will the all-volunteer efforts like Debian. Red Hat -- the current market leader -- expects to break into the black this year or next.
Service around Linux is sound business. I don't doubt that segment of the Linux economy will thrive in due time, as will the distribution segment. Apache, PostgreSQL, sendmail, and all the other staples of the free and open source world very often require professional support, especially when used in the enterprise. Besides, Cygnus provided a profitable model with its support of GCC for everyone to follow for years. So I'm going to leave that sector off my worry list as well.
This time last year I probably would have included drivers on the worry list. But lo and behold that picture is changing, thanks to help from both sides of the aisle. The free software world is providing lots of printer drivers via CUPS and the GIMP-print project. The new kernel provides better USB support as well as Firewire, thus opening the door for even more driver support. Lexmark has provided Linux drivers for two of its popular inkjet printers. Hewlett-Packard has announced wide-ranging support for many of its printers. IBM has recently released -- under the GPL -- a driver for the Winmodem that comes with the Thinkpad 600E. And finally, Oki Data and IBM have just announced a joint effort to provide GPLd drivers for Oki Data's Impact printers. The driver sector is really warming up, and as the Linux-using population grows, hardware manufacturers will increasingly come to the conclusion that it makes sense economically for them to support Linux.
Worrying about apps
So what do I worry about? My concern is with apps, specifically on the desktop. I've cheered on a number of proprietary Linux apps: CompuPic, WordPerfect, CorelDraw, Opera, almost everything by Loki Games, PMFax, and others. The problem is that none of them seems to make money. The reason that there aren't more companies making desktop apps is simple: they all want to make money, but if that seemingly isn't possible, why bother? For all of you readers who equate proprietary applications with evil, I'm sorry. I disagree. More importantly, I believe that it will take a team effort by the odd couple of free software and proprietary software to put Linux over the top.
But guess what. Professional, robust, enterprise ready apps are appearing. They just popped up where I haven't been watching, on the server. HP recently announced the intention to port two of its industrial-strength system management and configuration tools (Process Resource Manager and ServiceControl Manager) to Linux. And Corporate Associates' Linux port of Unicenter TNG won Best in the System Integration Solution Category at the LinuxWorld Conference & Expo.
The reason for such great server apps is obvious. Linux success in the server market continues unabated. Plenty of customers there are used to spending money for software. The money attracts the proprietary folk who build wares to get some of the cash themselves. Meanwhile, desktop Linux users like myself languish in relative obscurity. I need to accept that it took time for the Linux server boom to be followed by this new rush of products for that market.
In spite of what Microsoft wants everyone to believe, and what the Windows tres duh press tells us, a desktop Linux boom is coming. Anti-aliasing for smoother fonts has begun to appear in both the KDE and GNOME projects. Evolution, the better email alternative to Microsoft Outlook, will be released this year. Hardware drivers for popular peripherals are popping up, as are drop-in replacements for Microsoft Office. It's coming together on all fronts. Except the most important one, preloads. Computer OEMs like Compaq and IBM still depend upon Microsoft for their existence. As long as Microsoft can impose the Redmondian death penalty (revoking the license to preload Windows), the OEMs will drag their heels at preloading Linux, which would make it as easy to buy as Windows. But once Linux preloading happens, the floodgates will open.
So my dilemma is simply this: What do I do in the meantime? I'm perfectly fine with living in a community that is part free software, part proprietary. I believe that firms and individuals should have the choice of taking either path. Naturally, if proprietary and free software are available for the same need, the proprietary version will have to prove that it is significantly better. Otherwise I'm keeping my money. I'm sure I'm not alone in that. It's a big reason to be part of this community.
But there is a catch-22: If we as a community always choose to keep our money, there will never be a thriving market for desktop Linux applications. That would mean that Linux will never become more than a niche player on the desktop. So I worry.
What to do? Realistically, there are three doors to choose from. I can simply wait, and perhaps whine while I'm at it. I can get involved with a free software project working on an app that I use or want to use. Or I can find and buy -- actually pay money for -- a commercial Linux app. This past week I've removed myself from the wait-and-whine group and moved to doors two and three.
So I've downloaded a couple of graphical frontends for the GNU Debugger and the latest source code for gPhoto. I'm working at finding a nit or two that I've noticed while using that app. I also ordered and received a proprietary product. It's IceSculptor, a desktop publishing tool from Chilliware. Joshua Drake (see Resources) will review IceSculptor in the near future, and I'll report on my adventures in contributing to the gPhoto project.
Now, my dweebs, let me ask: What are you doing to speed the arrival of Linux on the desktop? Are you buying Linux apps? Are you lending a hand with an open source project? Something else altogether? All of you old IBM mainframers out there are free to wait and post. Everyone else please let me know in the forum, or by mail if you're shy about posting in public places.