Larry gives it some
September 2001

You already know how much I enjoy large and imperious IT companies like Microsoft – and some of you may recall that I have occasionally given Oracle a bit of leery glance when reviewing monopolists and unlovely large IT firms with far too much influence for the good of the rest of us. Oracle boss, Larry Ellison, swung through the UK recently, and made a number of grandiose comments as befits (at $26bn) the world’s fourth richest man, and someone thus guaranteed an attentive audience of fawning hacks, even if he were only to recite “Jack and Jill”.

Some two years ago now, the Oracle chief executive made a bold prediction when he declared: “If the internet turns out not to be the future of computing, then we’re toast!”

Well, a lot has happened in those two years, and the lustre is definitely off the internet, although that’s not to say that it isn’t the future of computing. It might be retracing a few steps to get back to where it was before the mega bucks went up in flames.

Oracle, for those of who you don’t know, produces some pretty darned unfriendly database software, that somehow seems to imply that it is immensely powerful in the minds of some. Oracle has never really shown any real sustainable interest in creating a product for the PC user, although there have been a couple of shots (Personal Oracle), they have preferred instead to remain aligned at the “high end” with IBM and its DB2 product line. (Not to be confused with the erstwhile Ashton-Tate product of that name). The genre is known as “client server” sounds tailor made for the internet, but actually it’s not at all, and requires a ground-up overhaul of thought and strategy to be better suited to life through the browser.

By hiding out with IBM in this rarefied atmosphere where “corporate” developers think nothing of blowing £50k of other peoples’ cash on some pretty hairy software that requires a further £250k of stroking by over-paid software developers before it is capable of doing something that an Access wrangler can knock up in 30 minutes, Oracle has managed to keep an eager following of people who prefer to be paid £100 an hour for 1000 hours jobs. After all, when did you last actively pull the communication cord on the gravy train while you stepped off..? I thought so.

The awful, awful truth is that database software is desperately boring; it is the software equivalent of UPS (uninterruptible power supplies) when it comes to the IT excitement stakes. And only those of you who have ever been required to wax lyrical about a UPS will understand what that means. Moreover, database authoring software is not very difficult to conceptualise and design. The trick comes in making it resilient, and capable of withstanding abuse of various sorts, without trashing the core data. Database software is the ultimate in mission critical computing, and it can be very difficult to assess the competence of a database development project until a lot of time and money has been spent on it.

If a design can’t handle being scaled up easily, then it might require a total rewrite – so developers tend to play safe and start with something that has a reputation for scalability. And this leads us to the reason why is Microsoft not all over the enterprise database scene. Well, this stems form Bill’s hatred of IBM and refusal to support any platform than MS Windows, which allowed Oracle to sneak in because it was willing to support IBM’s dreadful but ubiquitous mini and mainframe architecture. And also Unix, which made him best mates with Sun, of course.

“Aha,” I hear you say, “Accounting software is actually the most boring!”

Well, I have news for you, what is it that powers any accounting software..? What lies beneath the hood? The answer is database software. Yes, accounting software is just database software with a slightly sexier wrapper and user interface “shell”. De facto, database software is more boring than accounting software.

It seems possible that developers of database software are thus not the glamour-seekers and rocket scientists of the software age, but those who have enjoyed a personality bypass. There is much evidence that developers on £100/hour Oracle contracts are really just in it for the money (there is no job satisfaction) and thus inclined to spin out the job to the very maximum possible.

So in the times of plenty when bankers wanted to “play safe”, no one got fired for suggesting blowing a mil or two on an Oracle-powered web site, because that was a tick in a box that bankers understood. But now sanity checks on web spending are coming into fashion, the fact that Oracle continue to get away with it deserves a closer look.

Like Microsoft, Oracle has pretty much run the competition out of town, and boasts of a near monopoly of what it perceives as its markets. Oracle even tried to tie the pricing of the product to the CPU power of the users, but soon got a raspberry.

Despite its present position, if the internet is heading back to its roots and about to re-emerge with the original ethos in tact, then Oracle is going to extremely vulnerable to the combination of Linux and user-supported open source database solutions. Unlike MS Office, real users have no real investment in database software – and as long it works at it’s boring basic level, then that will do just fine. So enjoy that $26bn while you can, Larry.