THE MAGAZINE FOR COMPUTER APPLICATIONS
Circuit Cellar Online offers articles illustrating creative solutions
and unique applications through complete projects, practical
tutorials, and useful design techniques.

SIMPLE THINGS

Prime Numbers Revisited
by George Martin

Start ı Being Ardent ı An Example ı Sources and PDF

Before I get into this monthıs article, let me add to a recent article I did on prime numbers (Circuit Cellar Online, September ı00). I donıt know how many of you have taken my program and gone any further looking for primes. Before I get started on this follow up to the article on prime numbers and the sieve of Eratosthenes, check out www.mersenne.org. The Great Internet Mersenne prime search site reports that, as of June 1999, 2^6,972,593-1 is the largest prime number discovered. You can use your spare CPU clocks to help in the search. I am using my spare CPU time to look for ET. Check out www.setiathome.ssl.berkeley.edu.

NOW FOR THOSE SIMPLE THINGS

Sometimes itıs necessary to state the obvious. The simple solution to a design problem is usually the best solution. And generally, my first attempt at solving a design problem is not the simplest (best) solution. This comment applies equally to both my hardware and software design efforts.

I would like to present an observation that has served me well in the past. Whenever Iım designing embedded software for a customer and I get all twisted up in some logic problem or design approach, I usually seek out their resident senior hardware engineer. And nine times out of 10, that individual clears the confusion and points me in the right direction. My theory is that those seasoned hardware engineers have been trained to deal with hardware designs that are built using standard components. This standard component background and training keeps them focused on using simple components to build large complicated systems, always having a way to document and test that system.

And, whenever I have a pure software problem and ask a software engineer for help, I find that I get one of two replies. First they tell me to put the problems in a big array and sort that array in memory or build a table of pointers to the indexes Iım looking for, sort of like that useless users manual that came with your last software purchase. Or, the answer I get just leads me down a blind design alley that looks dangerous.

Neither type of reply is much help and provides little insight into the design problem. Software engineers are just as brilliant and clever as the hardware types, however, a software background does not include designing with standard components. The standard component approach leads to simpler and more uniform problem solving. Anyway, thatıs my theory. If youıve had any similar or contrasting observations, tell me and Iıll publish the interesting ones.

NEXT