I’m feeling pretty crabby over the excitement generated by network computers. P.T. Barnum could probably explain it, except that he’s dead. He knew every passing minute results in the birth of one more sucker.

It all comes down to bad cost accounting and dumb measures. PCs cost more to support than more traditional architectures, we’re told: three out of five dollars goes to support costs. We’re also told client/server systems cost three times more to develop than traditional systems.

It may even be true, although as previous columns have pointed out, the chance these comparisons having any meaning is pretty small. Even if it is true, it’s irrelevant.

One reason: companies that focus on cost are doomed. Companies that spend their energy reducing costs forget why they spend that money … to retain and attract customers.

Paul Erlich, writing about species extinction, came up with the metaphor of the rivet-popper, who sat on the wing of an airplane removing rivets. When a passenger complained to the pilot he explained the plane didn’t need that rivet anyway. The proof? The wing is still attached, isn’t it?

Cutting unnecessary costs is a Good Thing (GT to use the acronym). Far too many companies forget the keyword “unnecessary” and figure cutting costs is a GT. It’s not: most costs are investments in customer loyalty and acquisition.

The point: companies need to focus on value, not cost. Does anyone seriously think network computers will provide as much value on the desktop as a full-fledged PC?

And now, a reality check. Network computers will run software downloaded from servers. They’ll be completely compatible across all manufacturers, running the same code with no configuration problems, driver idiosyncrasies, or other technology-generated headaches.

I sure believe this. My 20 years of experience in this industry lead me to believe vendors work together in harmony to produce standards designed to maximize customer value, then resist the temptation to create propriety extensions or non-standard alternatives. Don’t you?

No matter how much grass we smoke, it’s always greener under someone else’s bed.

Let’s put our collective experience with real-world vendors aside, though, and pretend these gadgets really will work as advertised. Our support headaches will evaporate overnight.

So when someone in Accounting creates a critical spreadsheet and saves it before going home in the evening (on the server, since there’s no local storage) IS can be assured the spreadsheet program will run without trouble.

And when that accountant needs to do a bit of work at home because she’s a single Mom and has to put the kids to bed, but she’s on deadline for the next morning, she just pulls the spreadsheet up on her home computer and …

Oops! It runs from the network only. Okay, let’s imagine this technology allows for remotely accessing the corporate network from home. So she dials in and downloads the spreadsheet, so she can work on it on … hold it. She has Excel on her home computer. She bought it herself, of course, because her employer uses network computers. So she tries to load the spreadsheet into Excel and …

Hold it again. Do you think the network computer runs a spreadsheet that’s 100% compatible with Excel? I believe this just as much as I believe WordPerfect and MS Word exchange files while perfectly maintaining all formatting.

Now how about that mobile sales force? Are you going to outfit every one of those guys with a laptop network computer and 10 Mbps wireless network connection?

You think your support costs are high now? Start thinking about a mixed architecture, with network computers on the desktop and standard PCs everywhere else. You haven’t even begun to buy Excedrin in bulk.

So before we all get too excited about network computers, let’s do some serious thinking about how people use personal computers now, and figure out how they’ll react when we tell them they won’t be able to do that stuff anymore.

I think we have this Bill Gates thing all wrong.

Lots of IS managers responded to my column on the return of the mainframe mentality, concerned about the damage renegade end-users can cause. Most of these letters recommended processes that prevent end-users from making mistakes.

That’s when it hit me: Bill Gates is the one guy trying to maintain an out-of-control desktop.

Look at Microsoft’s tactics and contrast them with its competitors:

  • Microsoft invented TAPI (Telephony Application Program Interface) which links telephones and computers at the desktop. Novell and AT&T invented TSAPI (Telephony Services Application Program Interface) which provides similar services through a server/PBX connection. IS can control and manage TSAPI. End-users can use TAPI without IS ever knowing about it.
  • Microsoft is putting together “Peer Web Services” which will let everyone publish HTML documents on Intranet servers. This empowers end-users while making them harder to control. Meanwhile, Microsoft’s competitors are focusing on Java, which gives IS new ways to develop and control applications. Significant fact: making HTML the standard enterprise document format breaks Microsoft’s de facto monopoly in word processors.
  • Microsoft builds personal computer operating systems that run on autonomous desktops. Many of its competitors (Oracle, IBM, Sun) have embraced the notion of a network computer – one that only has the functionality IS provides through the network.
  • Microsoft focuses on ease of installation and use – for all the griping, Windows/95 installs with remarkably little pain in the vast majority of cases, and the real attraction of NT server over Unix, Netware, and even OS/2 is how easy it is to set up and administer.
  • You don’t hear Microsoft extolling the virtues of “thin clients” which, after all, waste the processing power at the desktop while requiring bigger servers for IS to run.

Microsoft’s competitors? They all develop for and sell to IS.

Novell used to market (okay, what substituted for marketing for the kids in Sodium Valley) to renegade department managers tired of waiting for central IS, and Novell became the dominant player in the LAN marketplace. Then Novell started to become legitimate, using “Enterprise” as an adjective and selling to central IS instead of its original customer base. Novell now focuses on technologies that increase IS’s ability to manage the enterprise. In the bargain, it has failed to gain significant market share for any product, no matter how excellent, that provides end-users with personal power and productivity.

Sun, IBM, Oracle, Digital? They’ve never heard of end-users, and probably wish they’d go away. End-user activities don’t sell big iron, or even medium-size iron. They don’t sell database servers (you have to deal with IS to get at those).

Yes, the Macintosh is the original end-user machine, and Apple had similar ideas. Unfortunately, years of inept management at Apple caused the product line to stagnate, boring the daylights out of everyone who watches this industry. We all may be willing to wait 15 minutes for a Web page to download, but we won’t tolerate boredom.

End users want to work without IS looking over their shoulders. They want to fiddle around, gradually creating exactly what they want, emerging only when the finished product is ready for inspection, whether it’s a document, a spreadsheet, or a small database application. That’s why personal computers became popular in the first place, and why the big players uniformly missed the boat.

Yes, Bill Gates is a fierce competitor, and in fact is one of just a few in the industry who knows what game they’re playing. Gates, along with Scott McNealy of Sun, Larry Ellison of Oracle, and maybe one or two others, is playing winner-takes-all poker. The rest are busy increasing shareholder value, maximizing profits, and doing all the other stuff that gains short-term success at the expense of world domination.

Only I wonder … does Chairman Bill really want to rule the world, or is he playing a deeper game?