While working on the second edition of Fire in the Valley, Paul Freiberger's and my 1984 book on the making of the personal computer, I realized that Chapter 9 would have to be split into two chapters. That would mean that Chapter 10, the last chapter, would now be Chapter 11. It struck me as a nice structural comment on the subjects of our book. Chapter 11 is bankruptcy, which was indeed the last chapter for many of the early computer and software companies we wrote about.
Failure, for all its drawbacks, is not unenlightening. You can learn a lot from failure, and do it painlessly if it's somebody else's failure. You can also learn a lot from novices -- often self described as dummies. It's epigrammatic that teachers learn through teaching. In helping those who supposedly know less, we often learn that we don't know what we thought we knew as thoroughly as we thought we knew it. This month's column features dummies and failures and, I hope, some learning.
Although I don't expect DDJ readers to buy Dummies books, we do get asked to recommend books for users, don't we? So it's good to know what's out there. Linux for Dummies has been out there for a while, but it's now in its second edition. It is written by Jon "maddog" Hall, executive director of Linux International and Compaq's chief Linux guy, and is published by IDG Books, of course, the Dummies people (copyright 1999; ISBN 0-7645-0421-5).
It's a book on Linux explicitly for the user, not for the system administrator. That means that it skips certain topics covered as a matter of course in almost every other Linux book, such as network administration.
On the other hand it is very good on topics like connecting to the Internet via a serial modem and an ISP. Linux for Dummies provides what the naive user with some Windows experience needs to set up and start using Linux in a single-user installation with no system administrator around. It's very readable, tells no harmful lies, and comes with the public parts of the 5.2 Red Hat release.
But the book does beg the question, how appropriate is it for this naive user with some Windows experience in a single-user installation and no system administrator around to be using Linux? IDG Books has sold a lot of copies of this book, and I suspect that some of the people who bought it had just heard so much about Linux that they thought they ought to try it out, much as they might decide to try out a new piece of application software. After all, they get the book and the software for around twenty bucks, making it, by computer-store-shopping standards, in the price range of an impulse buy.
Installing a new operating system should not be done on an impulse. And despite some impressive effort to move Linux mainstream, this is not -- not yet at least -- an operating system for everyone. It's probably not an operating system for the impulsive, naive user I have described here as a likely buyer of the book.
Assuming that Linux even makes sense as a mainstream operating system competing with Windows and MacOS, until it has a few more ease-of-use boxes checked, it is not smart for proponents to push it on people who will be disappointed and badmouth it.
And I am pointing the finger at whom, exactly? Me and my ilk, I suppose. Not OEMs, certainly, who are presenting Linux only as an option. Not developers, particularly. No, it's us trend watchers. But don't be too smug; you may be seen as a trend watcher within your organization or amongst the people who ask you to recommend books on technical topics. Unless we want to produce a Linux backlash, we should maybe be a little more careful about how we beat the drum for Linux. That said, I'm about to do it again.
About Linux itself I have no such reservations. I now maintain four operating systems on the various machines in the offices here, not counting version differences. Although the relative amount of work done on the four varies too widely to make comparisons meaningful, I can report that Linux is the only OS of the four that has never crashed for me. So I had no axe to grind when I went to the LinuxWorld Expo, held in San Jose, California, in March.
The first hint that Linux was stirring a lot of people's interest was the full parking lot on Almaden. I parked around the corner on Woz Way, across from the Children's Discovery Museum, and hiked over to the Convention Center.
There, I got the second hint that things were really happening in Linuxville. Stepping through the door of the exhibits hall was like walking onto the show floor at Comdex. All right, there was some difference in scale. But the airspace above the crowd was filled with the signage of big companies -- Compaq, IBM, Sun, Hewlett-Packard, Oracle, Sybase, Pick, Computer Associates, SCO.
Third clue: Below the signs it was wall-to-wall people. I like to try to read the crowd at shows, and this crowd was upbeat. My informal crowd temperature reading, based as usual on lunchtime chats, impromptu interviews, overheard conversations, and body language, says that the people at LinuxWorld Expo 99 felt that where they were was Where It Was At.
I'm writing this after the show, and the announcements before, at, and immediately after the show are blurring together for me, but all of the following happened close to the time of the show:
In short, Linux was catching on in a big way, the show was a success, and a lovely time was had by all.
Outside a bagelry on the second morning of the Expo, a headline in a newspaperbox caught my eye and I had to interrupt my quest for food to dig for quarters. Over coffee and bagel, I read the story of the breakup of Hewlett-Packard.
HP was becoming two companies, not to be named Hewlett and Packard but rather Hewlett-Packard and Something Yet To Be Determined. The former would keep all the computer and related (like printer) business and the other would get everything else.
Well, all right. It sounded like a logical enough division; seemed like it would allow the computer company to focus more clearly; might let different cultures develop in the two companies, reflecting the different paces of the personal computer and instrumentation markets. HP was putting as positive a spin on it as possible, of course. But it was clear that this move was a serious response to a serious problem. HP management sees the company as out of touch -- not sufficiently responsive to changes in the market. The personal computer market, that is. How seriously the problem is being seen is reflected in the fact that Lewis Platt, the current CEO, won't be staying with either company.
It was with a pensive chomp that I polished off the bagel. It's sad to see one of the companies that defined What Silicon Valley Is in such distress. HP's culture of respect for the individual set a standard that many other Silicon Valley companies, perhaps most ostentatiously Apple, tried with varying degrees of success to emulate. I hope HP doesn't jettison the things that made it great.
Of course, the history of the companies that defined What Silicon Valley Is is not one of boundless success. Fairchild Semiconductor is a big part of the history of the Valley, but it is remembered today mostly for the people who quit and went on to start their own companies, like Intel. And then there's Xerox PARC.
The founder of Federal Express got a C on the college paper he wrote describing the idea for his company. FedEx and its competitors are doing pretty well now. I wonder, as the UPS truck pulls up outside Stately Swaine Manor, if all those stats on the growth of e-commerce include the increased business for FedEx and UPS from the likes of Amazon.com. My latest delivery from Amazon.com is Dealers in Lightning: Xerox PARC and the Dawn of the Computer Age, by Michael Hiltzig (HarperCollins 1999, ISBN 0-88730-891-0).
Although I don't agree with Hiltzig that the Alto was the world's first personal computer, that's just a matter of different definitions -- his strictly technological, mine involving price and marketing as well. I have a few other quibbles with the book, but, overall, I found it highly readable and seemingly authoritative. In writing the book, Hiltzig drew on the recollections of those who were there, interviewing all the obvious suspects and not a few innocent bystanders.
The book is worth reading just to remind yourself of the amazing invention machine PARC was -- and of the amazing collection of inventors who were there.
The development of the Alto, of course, but also:
Hiltzig describes PARC's origins, the recruitment of talent, its culture, people, politics, and projects. He also spends a chapter on the question, "Did Xerox blow it?" That strikes me as overkill for a question that can be answered in a word -- Duh!
But I don't mean to belittle Hiltzig's analysis of the politics of PARC. He does an impressive job of telling not only what happened, but why and how it happened, and how Xerox management both hindered and empowered this amazing band of inventors.
If this is failure, we should all be so unsuccessful.
DDJ