50 Years In Tech Part 15. Be: From Concept To Near Death

by Jean-Louis Gassée

Be, the company Steve Sakoman and I founded after leaving Apple, had a great product idea, or so we thought, only to discover that a bad choice of partner would almost destroy the company.

As described in the Firing Frankness Monday Note, my exit arrangement with Apple involved staying another six months or so as a ”minister without portfolio”. As I pondered my next move, I got a pair of phone calls from Steve Jobs. First, he asked me how it felt to be fired, a smirking question that I deserved given that I was instrumental in his own dismissal. A few days later, another call. This time, Jobs offered to talk because “we could do great things together”. I declined. As discussed before, I knew I didn’t have the emotional strength to work for the charismatic NeXT founder.

There were other opportunities — running AT&T’s Unix Systems Labs, taking the helm of Commodore, becoming the COO of a Unix server company — but these must stay on the cutting-room floor for fear of muddying today’s picture of the conception, gestation, and near early death of Be.

In February 1990, shortly after my exit from Apple had been arranged, my friend and colleague Steve Sakoman came to my office and announced his intention to leave the company. I was surprised; Sakoman was well-regarded, he ran a separate organization that was developing an exciting new product, the original Newton tablet. (Oddly, the Wikipedia article omits the first full-size tablet phase of the project.). But, no. Despite his promising future at Apple, Sakoman wanted to leave because he had doubts about the company’s direction (as described in an earlier Monday Note), as well as concerns about the Newton’s handwriting recognition, a key feature of the device as then defined.

When my half-hearted attempts to convince him to stay at Apple foundered, I did the only sensible thing: I offered to join him. While I didn’t have a precise idea about a product, I was convinced we’d make a good team as we’d done already. That proved to be correct. I also told Sakoman I’d fund the project, perhaps with friends’ help so as not to fall prey to Vulture Capitalists. This proved severely misguided and almost lethal.

For a while we spoke in circles, failing to find the core of an idea. Then, in July, as I was vacationing in France, we discussed over the phone our frustrations with the complicated layers of hardware and software silt that the Mac had accumulated. Surely, we could come up with a simpler, cleaner architecture, a more agile personal computer — but was there room for a third product besides the Mac and Intel PCs?

In our ensuing conversations, the Commodore Amiga became our “reference platform”, an example of the good, the bad, and the promising. While the Amiga provided interesting multimedia features we were less impressed by its overly-complicated architecture and implementation issues. But the most important aspect, to us, was that it sold in attractively large numbers, proving that there was, indeed, a market for a third way.

Weeks later, Sakoman presented a basic sketch of a simple, powerful multimedia machine. With the Newton tablet and our work on the Aquarius processor project, we had developed a relationship with AT&T’s Microelectronics division based in Allentown, PA. The original Newton project used two Hobbit processors, simple, inexpensive RISC devices. AT&T also made DSPs (Digital Signal Processors), simplified, capricious, but very fast chips that process images, video, sound, telephony, fax transmissions, and the like. Sakoman’s idea was to build a multimedia computer featuring two Hobbits and three DSPs (one each for sound, video, and fax/telephony) around a simple homegrown bus.

Sakoman’s lab was in his Scotts Valley home, more precisely in his “clean room” where he delicately retouched prints from his 8x10” field camera. Soon, I procured expensive (for us) lab equipment such as a fast Tektronix oscilloscope and an HP Logic Analyzer. (For the latter, I had to call HP CEO John Young’s office; we weren’t yet incorporated as a business so HP’s sales people wouldn’t ship to Sakoman’s house.)

In a matter of weeks, Sakoman’s working hardware prototype produced a heartbeat. Now all we needed was a software team, starting with Erich Ringewald, an Apple alumnus; Cyril Meurillon, a French engineering student I failed to convince to drop out of the prestigious Sup Télécom school (he’d write key part of the OS kernel from his dorm in Paris); and Benoît Schillings, a blindingly fast prototyper from Belgium. A little later, we recruited more Apple alumni, engineers Bob Herold and Steve Horowitz, and attorney Cory Van Arsdale, our Business Manager.

I had what I thought was a great name for our now-incorporated company: United Technoids, easily abbreviated as UT. This was both cute and useless as there was already a huge conglomerate called United Technologies.

During one of our daily phone calls, Sakoman told me he and his family had hunted for a name the night before by crawling through a dictionary. They’d gotten as far as “B” — I though he meant the letter B. No, “Be” as in “To Be”. Simple, resonant, romantic even, a call to action. I loved it.

At home I pulled the OED Second Edition and walked through the long (25 pages) procession of Be’s meaning through the ages, its Sanskrit, Saxon and Latinate roots (Be, Was, Is). I immediately jumped on MacDraw to design a simple Times Roman Light Condensed logo, manually kerning the “e” to make sure it looked just right next to the “B”. We were in business — after an amicable legal settlement with a company called Better Education.

Early demos impressed visitors to Sakoman’s clean room, so much so that an early investor friend wrote a second check on the trunk of his rental car at the end of his visit. (I’ll write a separate Monday Note on the Don’ts of my dangerously naive fund raising efforts. In short: Skip Friends and Family, go for cynical Deep Pocketed Pros.)

Then, disaster struck.

We should have seen it coming. At the beginning, AT&T Microelectronics people were very supportive. They knew us, they had worked closely with Sakoman on the Newton project, they exhibited gingerly optimism for my ability to generate publicity for Be’s multimedia machine, and for their processors as a result. But, as time passed, we started to experience difficulties in getting a clearer product roadmap and, in 1993, we got the news: AT&T was giving up on the Hobbit.

Gestation interrupted.

This hurt Sakoman badly. I tried to convince him this was a mere bump on the road. I used a literary analogy: The author’s computer and backup burn in a freak electric fire, the manuscript is destroyed. “I feel your pain,” says the publisher, “We’ll get you another advance, you have the book in you, you’ll even have the opportunity to fix some problems you were telling me about, success still is just around the corner…”

Easy to say, but engineering design is a human, creative activity that is rational only in part. Sakoman was distraught, couldn’t bring himself to start over. He left. (But would return a few years down the road.)

I lost a colleague.

A few months earlier, I had almost lost my life. I tore the inner lining of my left carotid artery and experienced a stroke as a result. (Not understanding what had happened, I had lost speech for a few minutes and was sitting at home when my wife discovered me, dragged me to Stanford’s ER and probably saved my life doing so. Fortunately, the only sequela seems to be a numb patch of skin on my right index finger.)

When we “lost” AT&T, my friend Efi Arazi told me to shut Be down, apologize to investors who’d have to understand the impact of losing a crucial supplier, and come up with a new idea. I didn’t follow his advice.

In part 16 of this 50 Years In Tech series, I’ll review some of my other failures to heed advice, and the eventual sale of Be to Palm in 2001.

— JLG@mondaynote.com