In my memory, Dad was always president of one company or another. As a kid, I didn't know what this meant, so I asked, “What does a president do?” Dad said, “A president is the guy who sticks around to empty the trash after everyone else has gone home.” ~ Brent Noorda, remembering his father, Ray
Ray Noorda died recently, at the age of 82. Chances are good you have no idea who he was, but 23 years ago he took over a little 17-person company in Provo, Utah and changed the face of computing. By the time he left, Novell had become the titan of server-based networks, and Netware, its flagship network operating system, had become a de facto standard.
Netware provided a simpler means to set up a file and printer sharing network than had been the case with Unix. Now, that's a relative term; you still had to know of what you were doing, but one big advantage of Netware is that it ran on x86 hardware, the “PC” chipset (or CISC) rather than the more expensive RISC systems that ran Unix.
Noorda made Novell a customer-driven company and listened to his users complaints, suggestions, and needs. As a result, Netware got more solid yet more versatile. Oh, sure, Microsoft is the big guy now, but when Bill Gates was struggling to get past peer-to-peer networking, Ray Noorda was providing software that would make the mainframe people nervous.
Netware servers just ran. By version 3.12 (around 1993 or 4), Novell servers would run for months without issues. Novell Directory Services (NDS; originally Netware Directory Services) was well on its way when Noorda left in 1995. Noorda had a huge jump on Bill Gates when it came to the network and stood to continue to build on that lead, but he made a huge mistake. He decided to compete on Gates' turf, the desktop, instead of letting Gates try to catch up with him in the server arena.
Novell thought it could command the computing landscape from desktop to server, which they realized was the ultimate goal of Microsoft. Gates had pulled a major coup by offering his MS Office suite of applications at an affordable price. For less than the price of a copy of Wordperfect, a user could get word processing, a spreadsheet, presentation software, and a database program. Noorda saw Office selling like hotcakes and decided that he had to fight back.
It didn't work. He bought Wordperfect, which by this time had fallen woefully behind MS Word in features; he also bought Paradox, a mediocre database program. Oh, I know there are still people out there running it, but the sad fact is that dBase, FoxPro, and even Access were better. Noorda might have had a shot at FoxPro, but for whatever reason he went with Paradox. The spreadsheet was Quattro Pro, which was a good program, but it also had fallen behind the likes of Lotus and even Excel because of the time and energy spent by Borland (who owned Quattro) losing in court to Lotus over “look-and-feel” issues.
The time and effort spent trying to win the desktop sapped strength from the Netware effort, which ultimately opened the door for Microsoft. But the real killer was Noorda leaving.
By this time, Novell had 12,000 employees, and Netware was still king of the networking hill. But, Noorda's desktop move had cost the company dearly as they had sold their suite at low prices to try to cut into Microsoft's lead. It seemed that Noorda, who knew how to sell networks, didn't know how to capture the desktop market. But, it was a disaster for Novell when Noorda left, because they lost touch with their customers. I know of many instances, some from personal experience, how Novell began to ignore their main cash source, the network customers.
I was working for a client who had a large Novell network. We had decided that it was time to upgrade from version 3.12 to 4.11, so we called Novell to talk about it. We called and called and called. Finally, when we had all but completed our planning, a salesman finally returned our calls and offered to bring out a Netware Engineer. Now, this was a 3000 user network, so we're not talking chump change, even for Novell, yet we had to beg to get someone to talk to us. This was a typical scenario that would not have occurred when Ray Noorda was in charge.
It didn't help that Netware was hell for third-party developers. One thing about Windows is that it was easier to create server-based software on Microsoft's platform than Novell's. Partly that was due to Netware's structure, utilizing “Netware Loadable Modules” (NLM) which were notoriously difficult to write. It was also due to the fact that Netware demanded that software behave itself. No program could operate at certain restricted levels that protected the operating system from crashing. Windows had no such restrictions, so it was easier to write for. Of course, it also made Windows servers more prone to crashing.
But, if you're an IT guy who needs a server-based database, and all the good ones are written for Windows, you're going to bring in some Windows servers. Since Windows servers could be integrated into a Novell network without much trouble, Microsoft got their foot in the door through applications. Eventually, they kicked the door in.
Noorda, meanwhile, went on to found a venture capital company called Canopy, which is still around. Canopy specialized in investing in start-ups with potential and had some successes. Word was that Noorda had a lot of fun at Canopy, which is more than can be said of the parade of executives that followed him at Novell.
I am amazed at how little was being written in the technical press when Noorda passed away. I learned of it from a British news site, the Register. American tech outlets seemed to let Noorda's passing just slide by in favor of the latest hype about Vista. The first article on a U.S. site that I saw was Dave Kearn's charming rembrance, a couple of days later.
That's a crying shame, without Noorda having paved the way, Bill Gates would have had a much tougher time cracking the Unix networking world. By the time Gates had got Windows NT to where it only crashed occasionally, Netware had muscled aside Unix and the mainframe, establishing server-based computing as the way of the future.
A guy who could do that deserves some words of praise.