Try, try again
- 25 August, 1999 12:56
Matthew JC. Powell finds that sometimes, near enough is better than good enough . . .
I'm trying something a little different this week. I'm writing this column entirely on a Palm III handheld organiser computer thingy with (sort of) handwriting recognition. This is not something I have chosen to do lightly.
There is no time advantage in writing on the Palm. The above paragraph took three train stations to write. I can most certainly churn words faster on my trusty desktop machine. It has been so long since I actually "wrote" anything of significant length that the use of fine gesticular patterns as a form of communication has become quite alien to me.
Nor is the "cool" factor in tapping out deathless prose on the train a major contributor - although naturally I hope my fellow passengers are looking at me with envious eyes. I am, after all, only human, and a human with a nifty gadget, at that.
No, my major reason for doing this weird thing is merely to discover whether or not it can be done. And this I have discovered, with the qualification that I (personally) cannot do it well.
That I am unlikely to try this again should not be taken as any criticism of the Palm platform nor of 3Com. Other people may have greater success with handwriting recognition than I have had, and more power to them.
I hope shortly to attempt to write an entire column using only voice recognition. You may recall that I have a wager going with a fellow tech writer, who believes that speech recognition will be a ubiquitous technology by 2000. Since that's four months away, I still feel reasonably confident I'll win the bet, but I want to know for sure. For me, ubiquity would mean that any computer user could look seriously at a voice interface for their computer either largely or entirely replacing the keyboard. Being able to open a menu item and start your browser with a voice command does not cut it as "ubiquitous". For other people, ubiquity might just mean that, like the Palm, it's available, it's accessible, and it will (sort of) do the job.
I'm hoping that definition will not win out, or I'll possibly be out 25 bucks. You can, after all, buy speech recognition software and use it for things. I know few people who have done so, other than for experiments like this one.
A similar "there is no do, only try" attitude has become ubiquitous among distributors of software on the Internet. More and more these days, I'm hearing vendors boast proudly of the number of "download attempts" their software has received. Strictly speaking, they're really only trying to show the level of interest in the software, but a few of them have tried to make it sound like these numbers represent some vague approximation of market share.
I took one vendor to task over this, asking how many people had successfully downloaded the software, not merely attempted to, before giving up in frustration amid wails of anguish. No figures were available, I was told. And what about figures for people actually using the software - not merely downloading and installing it, but actually using it for its intended purpose for more than a few trial days rather than destroying every accursed copy of it from their hard drives amid wails of anguish. Again, no luck. I even asked for figures of how many people had registered their copies of the software, but was told gleefully that very few people actually register software they download, so the figures could be misleading. Nonetheless, no figures would be available. Imagine my wails of anguish.
In short, I said, the millions of "download attempts" could amount to zero people using the software. A complete failure of a product, even for free, disguised behind a meaningless count of "attempts". There could, in fact, be just one person, doggedly determined to get hold of this thing, who has failed several million times to maintain a connection to the server, but won't give up.
Here's an idea: come up with a product concept that, say, a couple of hundred thousand people might like if they could get it for free. Give them a Web page for downloads - but make it impossible to successfully obtain any code. When everyone's tried about four times (as many as I could imagine giving it), proudly advertise a million "download attempts". Word-of-mouth will spread, and you're on your way from there. Before anyone realises you don't have a product, float your company on the stock exchange and head to Majorca on the proceeds. Must stop now. Wrist hurts.
Matthew JC. Powell is the editor of ARN's sister publication PC Buyer. E-mail him at matthew_powell@ idg.com.au