Reminiscing about my days at Progeny has me thinking back even further to Stormix Technologies. As a commercial venture, Stormix was a disaster, with an especially virulent strain of dot-com fever infecting everyone. Still, I'll always remember it as my first professional introduction both to free software and general management practices, as well as a snapshot of a surreal time in technology history.
In the mid-1990s, I was a dedicated OS/2 user. By 1999, I knew OS/2 was a long-dead end and I'd have to leave it behind, but I had only experimented briefly with GNU/Linux and documented a Slackware system without knowing much about its inner workings. At home, I was using Windows, and feeling the constraints.
When I learned that NetNation -- at the time one of the largest virtual hosting services in the world -- was starting up a subsidiary that planned to commercialize GNU/Linux, my excitement was so obvious that I gave the best interview I've ever given in my life, and was hired on the spot. I might have been unclear about what this Debian thing was, but I felt that Stormix represented something innovative and important. In fact, I was so pumped that I immediately went out and purchased a new hard drive so that I could set up a dual boot system.
It didn't hurt, either, that the Stormix office was on the 20th floor of Harbour Center in Vancouver, looking down at the harbour and the convention center -- one of the most fabulous views in the entire city. At times in the coming year, the view was so distracting that I had to deliberately turn my desk away from the window so I could get some work done.
The day after Canada Day, I joined Kevin Lindsay, Garth Wood, and Atsushi Ikeda on the Stormix payroll. Kevin, the developer's straw boss, was the one who had come up with the idea for Stormix, and pitched it to NetNation's owners, David Talmor and Joseph Kibur.
Through that summer, I happily submerged myself in learning GNU/Linux and putting the rudiments of a manual together as the company grew. Having just come off a two year stint documenting ever-changing human resources software, I couldn't believe that I was getting paid to enjoy myself so much. The BASH shell alone was a wonder, let alone the rudimentary versions of KDE and GNOME that existed then.
Since I was eager to prove myself and was the first non-developer to sign on, I soon started commenting on the still vague plans for the release of Storm Linux. I commented so often and in such detail that, by mid-August, when Garth returned to the University of Waterloo to finish his computer science degree (I thought him insane to desert such a going concern as Stormix), and Dean Wadsworth and Randall Donald had joined the company, David Talmor asked me to become the company's business manager.
After a self-debate that lasted the weekend, I concluded that I was not a seller. I can make a pitch, but if someone doesn't agree with me, I'm more likely to shrug and mutter about legitimate differences of opinion than try to persuade them. So, on the Monday, I came back with a counter proposal: I'd be product manager, and steer the development of the distribution. Effectively, that meant that I would be doing everything on the non-programming side with jobs calving off mine as the need grew, and acting as an intermediary between geeks and non-geeks in the company.
Before I knew it, I was doing a bit of everything: writing the manual and news releases, answering due diligence questions for investors, negotiating ad space in magazines, devising ad campaigns, setting up bundling deals for Storm Linux's retail box, assisting with hiring interviews, and keeping the partners informed. At first, I leaned heavily on talent from NetNation six floors below us for both work and advice, but gradually Stormix started to grow, adding both developers and non-technical staff as we inched towards a December release.
I was a recovering academic and some-time technical writer by trade, so most of the time I was scrambling to learn just in time to perform. But the excitement of learning and the sense of responsibility was heady, and during the fall of 1999, I was more satisfied with my work than I had been in years.
Early on, I established a "no penguin" policy in our advertising, outlawing all use of the trademark penguin. I argued that we wanted to promote Stormix, not GNU/Linux in general -- an idea that many of the developers were having too much fun to understand. Inspired by the Absolut vodka ads, I also established the rudiments of an ad campaign that played on all the different references to "storm" that I could think of: storm warning, eye of the storm, storm watch -- I had no shame, and would have used all of them, had I had time.
In addition, I spent considerable time figuring what features and third party software Storm Linux would need to make it appealing to the average user. In 1999, for instance, some sort of Windows emulator seemed essential for the commercial market. So did some sort of mature office program. Now, of course, neither would be a concern.
One especially useful exercise for me was pouring over licenses and bundling terms with Jagan Gill, NetNation and Stormix's general counsel. Not only did I make my first contacts with people in the free software world, but, in learning how to explain free licenses to Stormix's owners -- a task I don't think I ever quite succeeded at -- I gained a strong understanding of how they worked, as well as an abiding admiration for their mixture of idealism and pragmatism. Both results have been useful to me ever since, if only because I am one of the few people I know who can read a license agreement without instantly falling asleep.
Meanwhile, around me, the dot-com fever was gripping everyone. I first saw the symptoms the morning after a shipment of file cabinets arrived, when one young programmer crawled out of an empty box where he had spent the night. When I questioned him, he admitted that he wasn't on a coding crunch -- he had simply read about people camping in their offices in Silicon Valley, and thought it a cool thing to do.
Soon after, as plush penguins and other toys began to fill the office -- including a collection of unrelated products that used "storm" in their names -- Kevin Lindsay installed a webcam. Apparently, it was another kewl thing to do. Some time later, when one of the owners logged on to the webcam at 11:30 on a Saturday night, he was gratified to see all the programmers hard at work. He went on at such length about their dedication that I hadn't the heart to tell him that they had been taking advantage of the office's high speed connections for a really wicked game of Quake.
Many of these signs of the fever were harmless enough, but others had a direct affect on Stormix's ability to establish itself and become profitable. Kevin Teague, the company web developer and graphic designer, had a hard time understanding why penguins should not form part of the company image, or why an ad shouldn't have a "powered by Zope" line on it. When I tried to explain, I earned a reputation as a killjoy.
More seriously, the owners were pushing hard to go public by the spring of 2000. They had got into virtual web hosting at just the right time, and were hoping to repeat their success. But, between their eagerness for a stock offering and their proprietary business backgrounds, they were slow to understand what they were doing.
Either they were too conservative, doing such things as insisting on a transfer of copyright in return for an internship for Andrew Clausen, the developer of GNU parted, or else they were too rash, sponsoring industry trade fairs as if they were Red Hat or even IBM. They didn't have time to look closely at what rivals like Corel or Libranet were doing or even what was happening with the development of Stormix's own products (although we hastily copied Corel's graphical boot manager, that was entirely the programmer's initiative).
As a result, the dot-com fever spread unchecked by reality, and the company was often more like a party than a business. Not that anyone was drunk or stoned, you understand, but, then, they were so excited that they didn't need to be. And I wasn't an exception, either -- just a cautious drunk, really.
The first sign of problems came with the release of Stormix 2000. Via a crushing amount of overtime -- which all the programmers revelled in -- we got the commercial box of Storm Linux to the manufacturers. Our timing was too late for the Christmas market that might have helped us, but early enough to satisfy investors (one memorable working Saturday, I just missed John Cleese when I nipped out to do some Christmas shopping at the game store, but that's another story).
But then we discovered that the master CD from which our distributors were pressing ten thousand copies had been taken from the wrong branch of the code repository because one programmer had been experimenting with going without sleep and was too brain-dead to know what he was doing. The solution in those dot-com days was obvious: fly the same programmer down to the distributor with the correct master CD. That was how those with dot-com fever spent their money.
Then, as I nosed around the release the day after Christmas, I discovered that support for FAT32 systems was not compiled into the installer's kernel. The programmers didn't care about that themselves, so they didn't realize that, in those days, when dual-boot systems were the norm, that out of the box FAT32 support was essential if users were going to be able to communicate easily between their operating systems. It didn't help, either, that the quality assurance tester had been hired only a week before shipping and hadn't been in any position to catch the error.
I found myself running between the programmers and the owners like the squirrel on the world tree in Norse mythology, trying to prevent the owners from wholesale firings on the one hand and trying to coordinate the salvage efforts on the other. In the process, I ended up antagonizing the programmers by my insistence on hourly reports -- a blunder that showed just how inexperienced I really was.
Eventually, the release was patched, but at the cost of losing a couple of key reviews because of the delay. Add the delay to a lack of promotion and the usual slow sales of January, and it soon became obvious that Stormix would shortly be in trouble. The company was spending far more in storing its commercial box than it was taking in via sales, and splashing around hundreds of thousands of dollars in an effort to increase its profile.
Looking over the dismal sales figures for the commercial box, I suggested to the owners that the company consider other sources of revenue. Professional services seemed a likely path, I suggested, remembering The Cathedral and the Bazaar. But the owners thought that would be too lacking in glamour for a company hoping to make waves with its initial public offering.
My next suggestion was to use the acquisition budget to buy into profitability, or at least grow the company until it became profitable enough to be gobbled up by a larger fish. I spend the better part of a month setting up a deal with BlackHoleSun, a gaming company that had developed its own 3D gaming engine -- only to see the deal lost in a moment because the owner made a mistake and preferred to accuse the other company of being dishonest rather than admit it.
At that point, I had had enough. I was facing some serious personal crises, and I had lost face because of the broken deal. Besides, as my responsibilities were divided among new hires, somewhere along the line Stormix had stopped being so enjoyable. It didn't help, either, that I was considerably older than most of the twentysomething hires.
Feeling increasingly isolated, and more than a little like Cassandra as my prophecies of doom were ignored, I took a couple of weeks off to think. I concluded that the company had no chances of ever being profitable, and quit.
I had never exercised my stock options, and, in the coming months, I was glad of that decision. Stormix floundered about, dabbling in rack mount servers and becoming desperate enough to try professional services (which, I learned later, with a twinge of vindictive satisfaction, accounted for most of the little that Stormix actually earned).
However, nothing worked -- at least, not in time. Seven months after I quit, those remaining went on half pay. Another version of Storm Linux was squeezed out, and then, in January 2001, the company closed. In less than two years, it had spent millions while only take a few hundred thousands.
In the end, Stormix left few legacies to the free software community. It was never as polished as Corel Linux, and its installer, administration tools, and package manager became obsolete in a couple of years as superior alternatives came along. Its only real legacy was the closeness of the programmers, which remains to some extent even today.
Looking back, I consider Stormix a fantasy company from beginning to end. Yet living a fantasy can be psychologically healthy for a while, and, for me, Stormix was both an education and on-the-job vocational counselling.
Because of Stormix, I learned the appeal of working on free software, and discovered -- as few did at the time -- the delicate balancing act of working with both the community and commercial interests. I discovered who the free software players were, and where to find them, and some of them got to know me. Most importantly, I discovered that I needed a similar (if more sober) level of involvement to keep me interested in my work. Before I had been gone from Stormix for six weeks, I was writing for Maximum Linux and been hired at Progeny to do much the same as I had at Stormix for a considerably larger salary.
If I look back on the path that lead me directly to my current work as a freelance journalist, I see Stormix standing at the trail head. Even my current morning routine when I sign on to my computer first took form at Stormix. Being employed there was absorbing, uplifting, and frustrating, and, often, all three at the same time.
When I met a Stormix programmer a few months ago at the Vancouver PHP Conference, I was glad to reminisce for a few minutes -- but, if I could, I'd no sooner go back to those days than I would go back to high school.
Bruce Byfield is a computer journalist who writes regularly for the Linux.com and Linux Journal websites.
Bruce Byfield (nanday)
|Non-Linux FOSS: libnotify, OS X Style||Jun 18, 2013|
|Containers—Not Virtual Machines—Are the Future Cloud||Jun 17, 2013|
|Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer||Jun 12, 2013|
|Weechat, Irssi's Little Brother||Jun 11, 2013|
|One Tail Just Isn't Enough||Jun 07, 2013|
|Introduction to MapReduce with Hadoop on Linux||Jun 05, 2013|
- Containers—Not Virtual Machines—Are the Future Cloud
- Non-Linux FOSS: libnotify, OS X Style
- Lock-Free Multi-Producer Multi-Consumer Queue on Ring Buffer
- Linux Systems Administrator
- Introduction to MapReduce with Hadoop on Linux
- RSS Feeds
- New Products
- Weechat, Irssi's Little Brother
- Validate an E-Mail Address with PHP, the Right Way
- Tech Tip: Really Simple HTTP Server with Python
- Poul-Henning Kamp: welcome to
12 min 14 sec ago
- This has already been done
13 min 14 sec ago
- Reply to comment | Linux Journal
58 min 28 sec ago
- Welcome to 1998
1 hour 46 min ago
- notifier shortcomings
2 hours 10 min ago
3 hours 47 min ago
- Android User
3 hours 49 min ago
- Reply to comment | Linux Journal
5 hours 42 min ago
8 hours 31 min ago
- This is a good post. This
13 hours 44 min ago
Free Webinar: Hadoop
How to Build an Optimal Hadoop Cluster to Store and Maintain Unlimited Amounts of Data Using Microservers
Realizing the promise of Apache® Hadoop® requires the effective deployment of compute, memory, storage and networking to achieve optimal results. With its flexibility and multitude of options, it is easy to over or under provision the server infrastructure, resulting in poor performance and high TCO. Join us for an in depth, technical discussion with industry experts from leading Hadoop and server companies who will provide insights into the key considerations for designing and deploying an optimal Hadoop cluster.
Some of key questions to be discussed are:
- What is the “typical” Hadoop cluster and what should be installed on the different machine types?
- Why should you consider the typical workload patterns when making your hardware decisions?
- Are all microservers created equal for Hadoop deployments?
- How do I plan for expansion if I require more compute, memory, storage or networking?