Michael's Perspective on Programming

Without further ado, let me introduce you to my thoughts and philosophy on programming in general.

I have been writing code and programming software since high school. While it's not considered much time by many I do have a wide range of experience working with many programming languages and development tools. These experiences are critical to me as they form the foundation of how I solve a problem in software.

If you are a technical person you may be interested to know how I approach a software project. Enjoy! Things change constantly. New things are being invented every day and if you cannot keep up with the current trends that's too bad.

Although I know a lot, there are still many, many things I don't know anything about. In that case I am fully willing to learn. I am extremely quick on the uptake and can learn anything in a short amount of time. Programmers generally do not know everything about their tools before they use them, which is a common source of bugs.

For example, the standard input cin in C++ ignores white spaces; the int type in Matlab remains unchanged when asked to go beyond its limit; JVM's class loader loads up the class files and attempts to optimize them. Examples go on and on.

Whenever I learn something, I make sure I know as much about it as possible, either through reading or by consulting the experts. Before I get down to programming I make sure I know absolutely everything about how I will implement it. I don't dive in without any planning, getting started without thoroughly grasping the requirements a very dangerous thing to do. I always map out my design and make sure every detail is taken care of. With a clear design in mind, I am always able to finish the program very quickly. This stage can take a LOT of time, sometimes as much as half of total development time. I have real experience when I was involved with HSM projects where I needed to build an Application to Control an HSM Machine. Getting to know the requirements and sketching the overall design took more than a week. But once I had the design in mind, I was able to complete it with great results. I make sure the program maintains high flexibility in the sense that many things can be changed and extensions can be added fairly painlessly. In other words, I make sure I do not 'hard code' important attributes of my program.

Also, if I am designing a user interface I have to make sure it looks nice and easy to use, by which I don't mean gaudy or complex. Many programmers or developers love to make their products stand out by making them look exceedingly showy and complex. However that's not necessary a good approach. Google is a case in point. Look at its search page. So simple and clean!! The clean user interface may belie its true identity - the most POWERFUL Internet search engine. Anyone who uses it can vouch for this fact. I just love it! Therefore I say the user interface should be as simple as possible to appeal to users of all ages and knowledge levels.

In regard to a non-UI-related program, I believe it is also the case. For example many UNIX utility programs are easy to use and have parameters that are intuitive. I believe before designing a program, one must think about its intended users and make sure it is user friendly from their perspective. One lesson I learned in an electrical engineering class is that whenever I don't know something, don't guess; ask someone who does. Even if you do know how to solve a problem, asking the experienced for their suggestion and advice is not a bad idea.

The thing is that in the software realm there are almost always more than one way to achieve a goal. How do you know the approach you adopt is good? You gain more experience as you write more programs but there's always a first time to everything, in which case if you have a mentor guiding you along you are blessed. His or her unique insight can make a huge difference.