Three miracles and one nightmare
It's good, from time to time, to stop and take a look around. Our industry has some absolute miracles and some real nightmares.
For me, one miracle is the hard-disk drive. These things work away in the background for every one of us. They are miraculous, unbelievable, incredible.
If you just start your PC up and don't even run one application, hundreds of files are opened up (lord knows why), and the hard drive just does it, day in and day out. Incredibly reliable, and now 3.5-in. platter drives have up to 500 Gbytes in a little 100 x 150 x 26-mm package and with a 3-Gbit/s interface.
And the cost�wow, 120 Gbytes for 75 bucks! A hero of our industry. What would we do without them?
Our industry has some absolute miracles and some real nightmares.
Another one: ceramic capacitors. A 22-�F 6.3-V X5R dielectric cap in a little 0805 SMT package seems pretty close to a miracle to me. Of course, I've been around a while.
To more recent engineering grads, this may be old hat. But who would have thought ceramic caps would have advanced that much in the last few years. And now we have 10-�F tantalum caps that come in 0402 packages with a 2-V rating. Another incredible feat.
One last miracle: real-time system/embedded software. The complexity of this stuff is staggering. How do they keep it all straightened out?
Beats me. I did some coding way back in the early '80s, when “structured system design” was the thing and it was supposedly our only hope for solving the “software crisis” we thought we were having. It was pretty tough keeping all those modules (and team members) from tripping over each other. But software has kept getting more complex, and we have survived�thanks in part to improved tools. My congratulations to those who did this.
Now, on the negative side of things�the stuff our industry should be ashamed of producing. Real-time system/embedded software. Yes, it is both a winner and a loser.
I had the pleasure of driving a 2006 German automobile recently. It was great to drive, but the climate controls for this thing are terrible! The buttons often seem to function almost at random. Push one function button and another goes on or off for no apparent reason. We knew it had to be software.
How did this happen to the engineering team that designed this system? It certainly makes one wonder about the current state of software design. I don't have a handle on it�and it is such a big subject I don't know that any one person can.
What's your opinion?
Jim Harrison
. . .