to Cyreenik book index
In November of ’81, Jack Messman had had a simple task: Go to Orem and find out what the problem was. The simple task received a simple answer. George Canova, the President, said in effect that the root of all evils was his relation with Jack Davis and if Jack was gone things would get better. Jack Messman had liked this simple answer, seen it implemented, and flown back to Philadelphia with his mission accomplished.
When he returned in March ’82, the answers didn’t come so simply. George was out, Jack Davis was out, and Jack Messman became acting President. So who was there for Jack Messman to listen to and explain this business to him? The company needed a permanent President; that was about all Jack knew for sure.
What the sales force knew for sure in early 1982 was that status quo wasn’t going to work. Months before the Massacre the brightest and most experienced salespeople started leaving. By the time the axe fell, only the inexperienced and ambitious were left.
Larry was still in position as VP of Sales but the sales force under him had changed entirely. Gone were the likes of Andy Olson and those who had worked beside him, Don Whatcott and Winston Lee. In their place were people like Roger White, transferred in from Support, and Dave Guerrero (see below).
Larry had been a close friend of Jack’s. He was clearly unhappy with Jack’s departure and George’s choice of an outsider to replace him. When George bypassed him, and Jack Messman then bypassed him, it was time for Larry to find a new company.
Manufacturing was still being hammered about poor quality. The returns and the horror stories of DOAs continued, but now Joe pointed past Engineering at Sales. Most of the returns in the summer and fall of 1982 weren’t DOAs, he explained, they were evaluation units that salespeople had placed. Accounting would call up to collect money but get a pile of boxes sent back to Manufacturing instead.
Fault in the returns problem cannot be clearly assigned now and could not be then, but that is the most important point: The mystery never was solved.
Novell’s image in the computer community was changing. It was getting older; it was getting more widely talked about; but it was also known to be a company in turmoil. As a result, the kinds of people that were attracted to a Novell opportunity changed. The pioneering types were less interested; so were the careful types. But the ambitious and risk-seeking types were taking their place.
Dave Guerrero was ambitious and risk-seeking, a short, energetic man originally from Guam. He was also a talker and he’d found that talking and acting had advanced him further than good looks or knowledge, which he also had.
Before joining Novell as the regional sales manager for southern California, Dave had been selling minicomputers that used the Pick operating system. He was one of the last salespeople brought on by Larry Edwards.
Harry was a surprise to the company. Having come on board as a “metal bender” to help Joe in manufacturing, he suffered much of the same heat that Joe did as the quality crisis evolved. But rather than get defensive, Harry remained open and approachable. Harry hadn’t always been in Production and in the days of crisis his people skills allowed him to handle the finger-pointing differently than Joe did.
That by itself would have gotten Harry little, but in late ’82 he started doing the exceptional: Whenever production duties were not requiring all his time (as more and more they didn’t), he phoned customers personally and asked for orders. Word of this bizarre behavior quickly spread through the remaining ranks. In 1983, it reached the ear of Ray Noorda, the new President, who liked what he heard.
From his first day with Novell, Craig had been a network spokesman. His experiences trying to make real-world things happen with some of the pre-Novell LAN products had sensitized him to what was really relevant to making LANs for personal computers effective. Reid was the first manager to get ahold of Craig, and for the first six months of 1982 Craig was overseas touring with Reid, boosting international sales.
This proved good for Novell. It kept Craig out of the fire while the worst of the cutting back and finger-pointing was happening, and it provided a valuable boost to Novell’s international sales effort. By the end of ’82 international sales was providing 40% of all the feeble sales that Novell was able to maintain.
In the second half of ’82 Craig transitioned from selling abroad to improving the marketing of the LAN. He started grappling with questions of what was going to be said about the LAN and who had to hear the message.
All through 1982, Judy grappled with the problem of how to become part of Novell at all. As the company’s fortunes declined, she found more and more opportunity to help out, but always on a temporary basis. By the end of 1982, she was well known within the company and she knew the company well, but she was still not a full part of it.
During 1982 Drew, Dale, Kyle, and Mark were watching the Novell crisis intently. They knew they were onto something exciting in this LAN concept they were developing, but their vehicle for implementing this exciting concept—Novell—was as shaky as they come. All through 1982 they watched it get shakier and more confused. But important people at Novell and Safeguard had the LAN vision, too, so as crisis after crisis roiled the company these programmers were spared.
In the summer of ’82 they incorporated as SuperSet.
“This was on the advice of Dave Guerrero,” said Dale. “Dave may have had his problems dealing with the rest of the company, but he treated us well. He told us that Safeguard and Novell would have an easier time dealing with us if they were dealing with an organization rather than a collection of individuals. So we incorporated, and he was right. Our negotiations with Novell and Safeguard went easier as a result.”
While the walls of PCs, printers, and terminal products were falling down around them they continued to develop the LAN product. It was during this period that what would become NetWare transitioned from being a disk server product to a file server product. It was also during this period that SuperSet recognized it would have to handle more than CP/M files, so they developed their own universal file format to allow the server to support CP/M, MS-DOS, and UNIX file structures. Finally, it was during this period that Drew first recognized the need for fault tolerance in the product: “I was talking with you [Roger] about some necessary networking features. The question of disk errors came up, and that got me to thinking.”
So the only basic architecture concept in NetWare 2.x or 3.x that waited until 1983 to be introduced was hardware independence, the ability to function with many different kinds of networking hardware. (Much more on this in Chapter Four, under “Hardware Independence”, p. 112.)
In early 1982 Drew hustled out of ComputerLand in Orem carrying one of the first IBM PCs. He was one of the earliest in a long line of personal computer software developers who wanted to see what IBM had wrought. His burning question: Would this be a good platform to act as a LAN server? He and the rest of SuperSet spent the year toying with it to find out.
IBM’s original floppy disk–based PC wasn’t suitable but its successor model, the PCXT, with a hard disk, had potential. That finding brought Novell to one of its 1983 crossroads: Should Novell continue developing only dedicated file servers based on the 68000, or should it also develop an 8088-based version of NetWare that could use an IBM PCXT as the file server?
to Cyreenik book index