Interview: Tim O'Reilly - Brought to book

Visit your local bookstore and try choosing a book about a particular computer subject - not an easy task. You will be greeted with row upon row of heavyweight tomes on every conceivable subject, from OS/2 Warp to the thoughts of Chairman Bill - and all you want is a guide to the Windows 95 Registry. But which one? This saturation of the book market is because the giant publishing corporations of the world know there is money to be made out of computer users; after all, there is a lot of them, so their approach is to publish as many books as possible in the hope that one will be a bestseller.

Not all publishers are the same. Browse through the shelves and you will notice something a little different: books with strange animals on their jackets that look totally unlike any other computer title, with their garish covers and exclamatory titles. These zoologically enhanced titles are published by California-based O'Reilly & Associates (ORA) whose fast-talking president, Tim O'Reilly, was recently in London to meet the press.

He is enthusiastic, almost impossible to shut up, and truly passionate about both his books and his subject.

"You don't want our books, you want the information they provide," he says, laying down one of the mantras that drives his company, a company that sees itself as an information provider first and foremost, but one which currently happens to provide most of that information in book form.

Few people end up doing the things in life that their qualifications suggest, and Tim O'Reilly is no exception: he has a degree in Greek and Latin bestowed from Harvard, no less. He graduated back in 1975, a time of brushed denim flares, Charlie's Angels and the Altair 8800. His first contact with computer publishing was when a programmer friend asked him for help to write a manual. "I got hooked," he says, as if talking about fishing or perhaps, more appropriately, DIY. The fledgling ORA was very much a DIY exercise and started off as a consulting business, when a slump forced O'Reilly to turn more of their pamphlets into books.

"Probably what makes us different from other publishers today was that we didn't set out to be publishers. We set out to document things that we used, such as Unix, in pamphlet form, and then found there was a market for this documentation through our consulting business. So bit by bit, we worked our way through Unix," he says.

"Like most people, I picked it up by doing it. I wrote a book on UUCP based on my own experiences of actually using UUCP and writing down what I did. It was lousy, but after lots of people's comments and ten editions in five years, we got it right! A lot of people in this community are really hackers, and I consider myself part of that".

Today, this is still very much part of ORA's appeal. It may be bigger, but it manages to stay in touch with what can be an eccentric audience, to say the least. After all, who else would put cute animals on the covers?

But there are other good, solid business reasons why ORA stands out from its megacorp publications, as O'Reilly explains.

"Because we didn't set out to be publishers, we didn't chase the big topics of the day. Instead, we set out to solve problems such as using UUCP or the internet. The typical publisher out there is now saying 'Wow!

The internet is big: let's jump on the bandwagon', whereas we were saying the internet is interesting, let's write about it. We've been in the right place at the right time because we were following what was interesting, following our instincts. We did a PERL book in 1991 - there wasn't another PERL book until 1995."

O'Reilly is adamant that the company's editing methods are unique too.

"If you look at a lot of books by, for example, Macmillan, they are written by writers and not practitioners. At ORA we hired good technical writers and made them our editors. Many of our books have been heavily rewritten by the editors, but that allows us to use experts in the field."

He recounts an incident with relish about how in the early days, an author came face-to-face with the O'Reilly philosophy after having problems with the software he was writing about. The author decided he couldn't write a particular section because the software didn't work properly. "That's exactly what you do write - the software doesn't work!" was Tim O'Reilly's blunt response.

He is shrewd about the market too. He sees a computer market where the middle is falling out and this affects the book market. It used to be that there was a large set of people who were fairly sophisticated but fairly ignorant, but now those people have learned enough that they just want to get to the facts without lots of handholding. At the other end of the market, the Dummies-style books handhold those who just want to do something and don't wish to be bothered by the innards and intricacies of computers.

But all those books that say you don't know anything about computers but you want to know a lot, are having a hard time. In the same way that people who don't want to know about computers have to go to Dummies, people who like computers and who want a more serious treatment, go to O'Reilly.

"We realised just how much hot air there was in a lot of computer books.

People want the facts. The Official Netscape Navigator Book - there is zero content in that book. Why does someone need a book to tell them to pull down the file menus?", he asks, somewhat justifiably.

The computer-book business is way out of control, with too many books chasing too few readers - a problem for the whole publishing industry.

But the industry's response is to publish more books. The theory is that if you throw more of them at people, they might start reading. The fact that this hasn't worked so far doesn't seem to deter them. All they have to do is try and see what people really want. Identify that, and you may sell books and protect some of the world's forests at the same time. Only a fraction of these books can make any money, only a fraction are going to be read by significant numbers of computer users, and very few will go into reprint. What's worse is that the authors of all these books are exploited too.

"No question, there are too many books being published. I often talk to computer-book authors and they are just cannon fodder. But publishing in general is like this. It's very exploitative. It does not pay the true cost of producing intellectual copy. So the philosophy is, throw it out and see what happens. It's like venture capital in some respects. Publish a hundred books and hope that ten make money to make up for the ones that don't. Because we were authors to begin with, we've always cared about our books and only want to do things that are interesting.

"Bookstores will buy a single copy of one of our books, sell it and buy another, and so on - that way it is perceived as a success. If we had made them take ten and they had only sold one, it would be perceived as a failure. You see the difference." I do see the difference, and it is proved in the fortunes of an O'Reilly book called Learning the Vi Editor which has sold a massive 136,000 copies on this principle.

This is more than impressive, but ORA is not just about book publishing.

In the late eighties and early nineties O'Reilly began to get the first whiff of a publishing revolution in the air, as the company began experimenting with online documentation and the internet. They started to use SGML and its now more famous subset, HTML, as a way of putting information online.

Oblivious to the moves happening at NCSA and the Mosaic project, ORA also realised the need for a universal browser to deliver this information onto any computer platform.

ORA, using this SGML, produced a web-based "Information interface" to the internet called the Global Network Navigator (GNN), one of the first commercial web sites supported by advertising. Although such sites that guide web users around the internet are taken for granted today, at that time there were only around 300 web sites in the world. So O'Reilly is probably justified when he describes it as an "historic event".

After the word got out about GNN, people started asking how they could access it, and ORA realised there was a market for a product that connected people to the internet by providing all the software needed; so they teamed up with Spry to produce Internet In A Box. Then came WebSite, which to this day is one of the easiest-to-set-up personal web servers available for NT and Windows 95, as well as being one of the most popular.

But is there really a future in electronic publishing? Ten years ago the CD-ROM was widely predicted to bring about the end of the book, but in fact we are publishing more books than ever before. And now we have predictions that the internet will be the same.

"If you look at publishing at a whole, trade book publishing is actually a very small part of information publishing in general. The mistake is to look at the web as an alternative to the book only. Instead, we must look at the web as an alternative to the whole continuum of publishing.

There are already good web magazines, and the web is the ideal place for documents that interest only a few people because of the economies of scale involved. For that, paper would just not be viable."

O'Reilly points out that CD-ROM did notch up one paper victim - the printed encyclopedia. Most specifically, Microsoft Encarta has stolen the market of traditional encyclopedias such as Brittanica and re-invented our idea of what an encyclopedia is, with its award-winning use of images, sound and video.

No danger, then, that ORA will get swallowed up by a giant? "We've had offers and it's tempting, but I think that anyone who took us over would probably kill us and our philosophy. We are ideas-driven; we are not bottom-line driven."