Web Informant #102, 16 February 1998:
I don't do downloads!


Time for a week off. In the meantime, I've recycled an essay that I wrote in December, 1995 and is still relevant today, with a bit of editing.

If your day starts off with mammoth email attachments from some misguided person, you might want to forward this column to them. And tell them you don't do downloads when the time comes to review some new software.

At the web conference in Boston in December 1995, the seven words most often spoken to me were "You can download it from our site." I'm considering asking George Carlin if he wants to change his routine. Unfortunately, things haven't changed much since then, and if anything, have gotten worse.

I don't want to download software that I intend to test. It wastes my time, my bandwidth, and the bandwidth of people near me on the net. I usually don't get the right version or have to download something again because something goes wrong. Call me a dinosaur but I want the physical disks. And while I am wishing for something physical, consider sending me the printed documentation as well as the complete press kit too.

For those of you that still don't understand my gripe, consider the average day here at Strom, Inc. I am on first-name terms with the delivery guys from FedEx and UPS, not to mention my US Postal carrier. They are great guys: they bring me new products almost daily. It is like Christmas year-round. I probably get from 10-20 new products each week, and that's just the ones that I really want.

But I'm not complaining, and indeed this steady stream of stuff coming in the door is how I like it: if Mr. or Ms. Vendor has to go through the trouble to send me a shrink-wrapped (or sometimes beta) product, chances are they have at least one of the following in place:

  1. a press reviews or product loan program, where you track who writes about what products and keep them up to date,
  2. a quality control system to ensure that before you ship you have some understanding of the number of bugs and their severity,
  3. a marketing group that keys their collaterals and programs, press tours and launches to the release of said product, and so forth.

But the web has destroyed all of this in one click of the mouse button. And it is a damn shame. Not because I am selfishly interested in keeping this system alive. Well, maybe I am. But because the system encouraged a certain level of rigor and a certain level of control over product quality.

In the old days -- that is, before URLs were on buses and Billboards -- beta testing was a difficult and often little- appreciated professional calling. You found a random group of your most vocal end users (well, maybe they found you) and gave them a bunch of disks ahead of the commercial release, in hopes that they would find things your engineers didn't.

This was (once) an honorable profession, and indeed I continue to make a good living doing consulting to those vendors that want to spend a little extra time crafting their products before they throw them over the wall. (One of my more memorable beta tests was looking at Intel's network management products at the Guggenheim Museum. We spent a lot of time trying to debug it, and ultimately helped the product a great deal. But I digress.)

That was then. Now we have download fever. "Just grab the latest files from our web site," you say. My problem is that this method is like the Tammany Hall voting method: download early and download often, indeed just download continuously if your bandwidth allows.

No need to do "real" beta tests anymore: just put a splash screen with lots of legalese up for a user to click away their rights to safe computing, and voila! You too can become a beta tester for the price of that download. That's the theory.

Dan Gillmor, a computer columnist for the San Jose Mercury News, has written about this. He says "it is getting harder to tell what is a beta product vs. the 'shipping' version. Growing equally tenuous is the distinction between shareware ... and regular software."

I agree with Dan. And while you may think that the user stands to benefit initially from this practice (more software for everyone!), the end result is we have more buggy software and more bandwidth being consumed by people getting version n+1 to fix something or just to keep up with the betas. I am not alone: others have also told me of their distaste for downloads.

I was at Netscape recently and watched as a product manager tried to install some software on an "older" version of their browser. Trouble was, this took place in February 1998 and the version was 4.01. "That wasn't a very stable version," he said. Excuse me? How many of us downloaded that browser?

Since I originally wrote these words, the quality of software continues to drop. The random nature of the 'net has taken over for a more scientific testing process that never really got started at many smaller companies. Yes, it is great that these small guys can spread their software all over creation. But not if the stuff doesn't work.

Maybe I am getting tired of downloading the "code o' the week" (or more frequently, since some folks make their betas fresh daily) and watching that little byte counter slowly creep upwards to several megabytes.

What's the solution? I don't have any fast answers. But in the meantime, if you want me to try out your product, send it to me via the mail. If you really are in a hurry, send it via FedEx. But whatever you do, don't send it as an email attachment or ask me to download it.

Self-promotions dep't

My latest ComputerWorld review is of Lotus' eSuite, a series of office productivity applications written entirely in Java. It is an interesting but flawed experiment.

David Strom
+1 (516) 944-3407
back issues
entire contents copyright 1998 by David Strom, Inc.
Web Informant is ® registered trademark with the U.S. Patent and Trademark Office