Baking cakes and making modeling software -- ADTmag

Baking cakes and making modeling software

By Jack Vaughan
April 21, 2004

This is somewhat off the beaten "Programmers Report" track, but here are some thoughts on baking cakes and modeling software, with due respects to writer Ray Weiss.

While listening to the radio show "All Things Considered" in traffic recently, I was becharmed by, of all things, a story about 1950's-era U.S. homemakers. More specifically, about the processed food industry's attempts to displace homemade dinners and desserts.

While easy-bake cake mixes were one of the many industrial advances enabled by World War II, the cake mixes were slow to be accepted by women consumers of that day. General Mills (not a real general) hired Austrian psychologist Ernest Dichter to study consumer sentiment; Dichter concluded that a cake mix with dried eggs did not leave the home baker with enough to do. The solution was to call for fresh eggs in the recipe.

I knew this story well, but not its ultimate source. It was a story that seemed to have an authentic message, even if it had been invented from whole cloth. It was a very favorite story of Ray Weiss, a longtime reporter and editor on ICs and embedded systems.

Ray worked at Electronic Design, EDN, EETimes -- just about everywhere in electronic trade journalism. Ray came to journalism like a lot of engineers did, after a job layoff in the real world of computers and systems. He broke a tremendous number of stories on chips, and I was glad to be able to work with him. He was brilliant, but if you were editing his work -- "It shouldn't be hard, just run the spell checker" -- he could be difficult. The post-World War II cake story was one of his favorites; you had to make sure a decent interval had transpired and enough new ground had been covered, before he once-again presented the cake story in a column. Sadly, Ray passed away at the end of 2001 at a much too early age.

In the days of the 80386, the 68030, the i860 and the Clipper chip, Ray's natural vigor was such that he would invariably chide tools vendors at the end of long presentations. He would tell them what was wrong with their product, which made him rather rare among journalists.

And oftentimes the mistake they had made was to forget the cake mix story, which he would obligingly re-tell for emphasis.

The moral played out whether the tool at hand was intended for semiconductor designers, embedded system programmers or what have you. It has bearing today, I believe, for makers of tools for corporate software developers.

The moral of the story is that "you have to leave something in for the developers to do." Like the housewife depicted on the Betty Crocker box, you have to be able to "crack some eggs" to take part in the process and see your efforts rewarded.

The rise of Java had some of this flavor. While the industry played catch-up, creating Java tools over a period of time, programmers had fun with command line interfaces. It was not just its cool caffeinated logo, or the lure of its transportable virtual machine technology -- Java itself had the feel of fun.

Before Java, RAD tools had somewhat dumbed-down programming in corporate settings, with enough horror stories to go around. Certainly, in the wrong hands, Java and J2EE can be a hotbed for horror stories today. Developers who try modeling and code generation are not wrong. But vendors should be sure to keep the flavor of the programming experience, and leave room for the creative spirit, as they move forward with "easy-to-use" modeling tools.

Finally, I couldn't file a story on Ray Weiss without adding this. When his systems engineering days were over, and he looked at making the move to publishing, Ray was faced with one small flaw: He wasn't a writer. But he recalled the story that Raymond Chandler, too, was nothing of a writer before he entered -- and ultimately reshaped -- the detective story field. Weiss recalled that Chandler began his new career by engaging in long sessions rewriting existing detective stories of the day. He wrote until he wrote better than the sources. I remember Ray telling me that was how he broke into electronic trade journalism. True or apocryphal, I think the story is authentic. And it speaks to a trait good engineers are always ready to employ: to creatively apply an innovative method in the service of a practical goal.