Of Temper and Price

There's a microflap in the RSS world right now. It started when a blogger lost her data, and then lost her temper.

She did this very publicly, and because she has good friends and good enemies she did this very visibly. The spreading ripples have left lots of name calling in the wake. ( I'm not going to link to the flap; it's a fender bender. Nothing to see. Move on.)

Software will make mistakes. This is emphatically true of beta software, which is what this unhappy blogger was using. You should expect beta software to do bad things, but even release software will, occasionally, not behave as it should. (Even if it behaves as it should, this may not be what you expect, and in that case the fault might not lie with the software)

Software is complex. Tinderbox, for example, is not the largest or most complicated program out there -- but it's very large and very complex and has a vast number of moving parts. The operating system has lots of parts, too.

Nobody knows how to write bug-free software. In fact, nobody knows how to write software that won't crash and lose your data. Nobody knows how to cure cancer or fly to Mars or put a chicken in very pot. We're working on it. The longest journey begins with a single step.

Aaron Swartz reminds me that we do know how to make some software nearly bug-free. But this requires that the software perform tasks we understand very well, and also requires that it have almost no interface because interface tasks are almost never well understood. In other words, we can write bug-free software, but not for exciting, innovative tasks that change the world for people who don't have computer science degrees.

Use common sense. The very first words on web page of the product that aroused such ire are "Beta software has bugs! Nasty, vicious bugs with great big, sharp teeth! Don’t use beta software unless you’re clear on what 'beta' means and you’re comfortable running beta software." Later, the vendor specifically warns against testing the feature that caused the problem.

Don't listen to the pundits. Newspaper and magazine columnists write as if defect-free programs were as easy to ship as defect-free automobiles. It isn't. Get used to it. Don't listen to the pros, either, unless you know them well and know enough about the field to be sure you're a good judge. A lot of people who seem to be serious experts are really cranks.

Software pricing is out of whack. In the instance, an influential professional was using free 30-day trials to select a professional tool she used extensively every day, a tool that was close to the center of her worklife. How much would the successful vendor receive for winning the competition? $24.95 . This is absurd.