AT & T had money, but it was silly to spend it on expectation of what would most likely not happen. Nothing personal just business. AT & T announced the closure of the project 920 × 0, and the termination of their production.

The result of four years of hard work, Be-1, stood on the table – but now neither he nor the team of his creators had any chance. Jean-Louis had a choice: to commit suicide, or to recognize himself as an insolvent debtor, putting an end to his future.

The rest was easier: it was time to look for a job, which it would not be too difficult for specialists of their level to find.

The creditors did not keep themselves waiting, and there was nothing for them to answer – besides, there was no money, in general. Disperse, and let everyone get what they need?

The team made a different decision. Unanimously. Despite everything, to finish Be-2, on other processors (which still had to be selected), in March 1995 – as promised to creditors. Do you understand that this is completely unrealistic? But the most unrealistic task was Jean-Louis: to reassure the creditors, who understood everything as well as we did, and to get money from them for the last jerk.

In March 1995, Be, Inc. was ready to present its first computer to the public. But we will return 4 years ago, when everything was just beginning.

In the next part I will talk about the operating system BeOS, and in the next after it about the most dramatic period in the history of Be, Inc.

Continuation. The beginning , the previous parts here and here .

Zero cycle
The story of Be, which has not yet had a name, a personal account, or registration, began in 1990. The day after leaving Apple, Jean-Louis and Steve bought a bunch of computer parts from the Fry supermarket, from which the very first prototype of the future BeBox was assembled.

The names were not even in the company, let alone the prototype.

The prototype was housed in a conventional low-cost, charcoal-colored computer enclosure (apparently a shade of black or gray?), And consisted of a single Hobbit motherboard, a hard drive, floppy disk drive, serial and parallel ports, and a graphics adapter.

The graphics adapter was soldered by Steve Sakoman.

In the late 90’s or early 91’s, Jean-Louis and Steve were joined by two other fugitives from Apple, Bob Herold (from the Newton project, according to which, after the departure of Steve and Jean-Louis, the project froze, it became unbearably boring and hopeless ) and Erik Ringewald, who led the Jaguar project (which Apple closed, fearing that Jaguar and its operating system will compete with Macs).

Bob Herold took up the development of device drivers and low-level software for the prototype, along the way testing it and helping Steve Sakoman in debugging.

Eric was looking for an operating system for the future BeBox. They were not going to write an operating system either. To this, by all accounts, they had neither the means nor the forces. Finding a suitable system turned out to be more complicated than expected. Only one of those considered by Eric was very similar to what they were looking for.

It was Chorus OS, the development of Chorus Microsystems. Incredibly scalable, very flexible and light, it was good for everyone. She had only one drawback: a high price. Negotiate with Chorus about the price reduction failed. Communicate with an unknown company, whose head speaks with a French accent and wears a earring in the ear, no one wanted.

It seems that Chorus Microsystems could not agree at all with anyone, and in early 1992 the company was on the verge of bankruptcy, from which it was saved by Sun Microsystems, by buying it. Sun Microsystems, a year before, was also interested in Chorus OS, and it also did not work. But on Be already began to develop their own system – which I will write about

The fifth in the team was Benoit Schillings, whom Jean-Louis met in September 1990 at the Apple Expo in Paris. Benoit was an expert in memory architecture and search engines, developed by him the project “Marco Polo”, presented at the exhibition, produced on Jean-Louis so strong impression that he invited Benoit to join the adventure. At the very end of 1990, or in early 1991, Benoit came to California and became one of the first employees of the company, which did not yet have a name.

March-throw on the wild beach
Jean-Louis, apparently, without trips to the beaches in the company of employees could not. Money from the hotel and French cuisine from a good boss was not, the company (formally) was not a month, but the fatigue of eight months of hard work was beginning to tell.

And then Jean-Louis arranged a bike ride. In June 1991. At a distance of either 42 or 50 km (from 26 to 30 miles, according to the participants, that is, from 41.8 to 48.2 km).

It was difficult, but fun. Broken knees and spoiled shoes, which had to be thrown out immediately – but when Be, in full force, fell down on the already almost cooled sand of a wild beach, it was either 3 or 5 hours – and it was a victory!

Sandwiches and cheap wine were tastier than the most delicious dishes, abrasions ached from salt water and the hot sun … Since the official creation of Be, Inc, it has been three weeks since the beginning of the project – about eight months, but then, on the shore, they suddenly realized something very important for yourself.

That they will overcome everything, and will definitely win.

Participants of the bike ride recall different episodes, each of them remembered something of their own – but all this event remained in memory, as something special.

If Be was a little lucky, this bike ride would have become historic. It would suit everyone, every year …

Be Machine

Gradually, the computer acquired unique features. The second prototype, in the same cheap case, which was called Be Machine, or Be-1, was already 5-processor.

Two Hobbit, and three DSP-processors from AT & T worked on the embodiment of the dream. The Hobbits found a weak spot: they were not very effective in working with a floating point. DSP, correcting this lack, provided also parallel performance of operations over data sets, and improbable (best in this class of computers) abilities in the field of multimedia.

Sound, video, graphics processing …

Gradually, the machine acquired its own know-how: original solutions for input / output (eliminating the central processors and processors for digital processing of signals from these base and despicable trifles), and all these “goodies” were made as cheap and simple in mass production.

The computer had to win its place under the sun from Windows PCs and Macs, the price mattered.

In early 1994, “feature freeze” took place: modification of the design was prohibited, the development team began to intensively debug and eliminate bugs that were missing. The first child was released in March 1995.

Everything was developing as best as possible. The only thing that prevented the joy of life, were the creditors. Be Machine made an impression on them, they still, albeit with a creak, gave money for its fine-tuning to the ideal state – but on terrible conditions.

Everything will turn out, everything will be fine, but nothing happens ..