Programmers must continue to stand on the shoulders of giants

Many of the greatest thinkers credit their accomplishments to those that came before them by stating the old saying that even a dwarf can see farther than a giant when the dwarf stands on the shoulder of the giant. Or, in other words, they came to their breakthroughs by understanding the research and works created by notable thinkers of the past.

How can that help modern programmers? Simple, when was the last time you read a computer book or magazine? It does not matter what the subject was, just the fact you read it added a little bit to your knowledge and experience. It is amazing what great little tidbits you can find to enhance your abilities.

For me, that two that come to mind the fastest are from Programming Pearls by Jon Bentley, first published in 1986 and Expert C Programming: Deep C Secrets by Peter Van Der Linden first published in 1994.

From Programming Pearls, I learned the lesson that a senior developer should never answer the question that he was asked, but instead answer the base need that spawned the question. In Programming Pearls, the programmer asked how to do a disk-based merge sort. Now, yes I know that coding things like that went out of style a looooong time ago, but the underlying process that the author used to help the programmer has stayed with me ever since. This is because the author did not show the programmer how to implement the disk-based merge sort, but instead learned what the base problem the sort was supposed to fix and with his greater programming and analyst experience came up with a much faster and more elegant solution. No, I’m not going to tell you want the solution was, go read the book!

From Expert C Programming: Deep C Secrets, I learned a very simple programming principle: look at a problem from both directions. Where from one direction, a problem may include complex tracking code, worked the other way, the problem may be quite simple to code.

What does that exactly mean? Let us look at an example:

Say you need to build a comma delimited list of numbers from some input. You could say “I need a comma after every number except the last”. This is the normal view and leads to either adding a comma after every entry and then removing the last one once you have exited the loop, or checking in every iteration if its time it is the last element so that you do not add the comma.

But if you look at it from the other end and say “I need a comma before every number except the first”. Once stated this way, the problem is simple. It’s much easier to code a beginning edge case than an ending edge case.

The point of this post was not just to share these two fun little facts, but to show that we, as senior level developers, and those that wish to become senior level developers need to continue reading and learning. As long as we do that, the giants that we stand on will keep getting taller.