The Fonly Institute: Running Into Reality

Sun	Mon	Tue	Wed	Thu	Fri	Sat
				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28	29	30	31

Running Into Reality

28 Dec. 2009

It's June, 1981 and we are racing to make the first shipments of Osborne-1 computers so Adam can win a bet with Bill Godbout, one of his poker buddies. The amount is unknown, but Adam's self-image is apparently at stake, so June 30 is our deadline.

It's surprisingly difficult. Everything has to go together right – every piece has to be on hand and in working condition. Documenting the process is a full-time job, one that is apparently not being done. Assembly documentation consists of a series of “assembly trees” - lists of what goes together to create the next higher-level assembly. It's manageable when you're documenting complex assemblies like the circuit board, with hundreds of parts. After a couple of levels it gets to the actual computer people will use, and you think that it's done.

But there's more. The diskettes with the software have to be listed, and they can have different versions. Then the whole thing has to go into a box along with custom-cut foam packing pieces, so that's another assembly tree. And somewhere in there there has to be a power cord that will fit the sockets in the country of use. It gets messy fast.

Manuals in the right language, warranty information sheets, labels for the boxes and the pallets of boxes – the assembly trees pile up. It elicits the phrase from Ginsburg's famous poem “Howl”- “Trees, radios, factories - tons! They broke their backs lifting it to heaven!” (well, you have to read the whole thing).

And then you have to ensure that your people follow the instructions. Some of the first units off the line had incorrect wiring in the power panel – the small square black-anodized aluminum sheet where the power cord enters and the circuit breaker and interference filter are located. All the connections are done there with push-on wire terminals, and someone's hand has to choose the right spade lug and push the terminal onto it. Just providing a good drawing of what goes where doesn't always make it happen. Fortunately the circuit breaker will trip when you plug it in, but that doesn't get you very far, and a hot line might be connected to the metal itself where the ground wire should be.

Thus it came to pass that on June 30 the first batch of six Osborne-1's were stuffed into their boxes and loaded into someone's car trunk for the trip to the dealer. We had made it! Adam's reputation would be unsullied! It didn't matter that on July 1 the same computers came right back to be repaired – we had shipped on time!

Not long after that I sat in an executive committee meeting and heard a report from some private investigators who had been hired to go undercover in the work force and ferret out drug use. They concluded that they could find no drug use but that we had a serious quality control problem. This led to the establishment of “quality circles” - committees at the employee level to compare notes and identify quality problems and solutions. Things began to get better in manufacturing.

Osborne's manufacturing process involved a lot of “outsourcing” - not necessarily the international kind that became universal 30 years later, but within Silicon Valley. The circuit boards were assembled by a company named “Testology” in San Jose. They would sell boards to Osborne and warranty them as functional – so if there was a problem later they would take them back and fix them.

Unfortunately, Testology created their own test procedure and Osborne wasn't much involved in the process. Such an arrangement requires that both parties agree on what the test is to be to determine functionality. As the designer of all the timing circuitry I should have been heavily involved in that process. The most I can recall was having dinner with Testology's test procedure consultant. It was a good dinner, but when I began making reference to the operation of the 74LS161 and 163 synchronous 4-bit counters (building blocks I used quite a bit) the consultant went blank - he had nothing to say. It eventually came out that he had represented his qualifications as better than they really were, and that he did not understand what I had been talking about.

Thus it transpired that boards would pass the outgoing tests at Testology and be shipped to Osborne – where they wouldn't work and would be shipped back to Testology. They would pass again there, and would wind up in a loop going back and forth between the two companies. When the end came all of these boards were simply sold to the highest bidder and entered the population of spare parts.

If I had been properly qualified for my position of Vice President of Engineering I would have been alert to any sign that something was amiss – that would have been my job. But as it was I considered myself basically the chief engineer, with a staff of one and a number of expectations which I had no good plan for meeting. No one ever sat down with me and asked me to lay out my organizational structure, my priorities, and the resources I would need to make it happen. No one had told me “Lee, you can't do that with just two engineers, especially if you're one of them!

Later I learned that the VP Engineering represents the engineering department to the Board of Directors. It is an outward-facing position, and the inward-facing position is the Engineering Manager. In startup companies these two positions may be combined, but it is foolish to combine the function of Chief Engineer with these positions, as I had done. What I discovered was that in such a structure where the product is even modestly complex none of those jobs will get done!

In this situation I was trying to get the “double density” upgrade for the floppy disc system designed, and was looking for a consultant who knew the technology (give me credit here for knowing it would take too long for me to learn how to do it). I had made an appointment for one consultant to come and meet with me at a given time, which came and went. I looked for Adam's secretary Marlene, a mature Englishwoman who must have reminded Adam of one of his childhood nannies. She would have been in charge of receiving the visitor, but she wasn't there.

Marlene had had to leave her desk for some reason and, having no backup, had locked the front door! My consultant candidate had showed up on time, found the company locked up tight, and had gone off incensed. Those were the days before cell phones, so he couldn't have placed a call from the front door, and in that industrial park the nearest pay phone was nearly impossible to find.

A short time later that day I met with Adam, along with Ed Richter, an engineer referred by our General Manager Tom Davidson, who had worked with him when he turned an electronic module manufacturing company around. Ed was Dutch, a former KLM Airlines flight engineer, and quite a bit more formal than most of us. I was to report on progress on the double density system, and had to say that my consultant had gone away furious.

Adam began a tirade about how I was letting down “...all those people who had been busting their asses...” when I snapped. Sitting bolt upright and gripping the chair arms, I barked back at him “one of whom is sitting before you now!” Adam, taken aback, was shocked into silence, his head making small bobbing movements. I went on, in my most strident, drill-sergeant voice, “Do you care to conTEST that statement?”

It took several seconds for Adam to recover his voice. “Lee, I'm not going to sit through another one of your tantrums,” he said in an overly-controlled voice, as if I had been popping off several times a day. He went on to lay out what he intended to do – I would be relieved of my position of VP Engineering and Ed Richter would be made acting VP in my place pending approval of the board. I would be given the newly-created position of “R&D Fellow” with full Vice Presidential rank. Clearly this whole solution had been worked out in advance.

When Tom came into the meeting and learned of this he came to my defense - “It wasn't his fault the guy couldn't get in...he woiks continuously!” but he had either agreed to this change or hadn't been in the loop. Ed and I left the meeting with Ed muttering about how he didn't like that kind of emotional outburst, and I had been freed from some responsibilities I wasn't meeting anyway.

Some time later Ed and I went out with a real estate agent to look for a location for our R&D shop which I would run. We settled on a suite of offices a block or so away in the same industrial park. I signed the lease papers myself as an officer of the corporation on a weekend when Adam and other VPs were away. Ed, Pat and I moved our desks there and I was in startup mode again. I was much more in my element.