It has been suggested that my way is to make "the simplest thing that could possibly work and then milk the possibilities for a long time before adding on." when in fact I have many ways. But I don't have a page titled "The Ward Way" and I'm itching to write one. matrix 
I suspect I will find two common threads.
1. Write programs that do things that haven't been done after first figuring out why they haven't been done, and,
2. Write programs that tell me what they are doing before they have done everything I expect of them.
I will follow the outline of my Coding Portfolio where I have described what I learned from specific projects. Here I will reflect on works that employed unique approaches.
Acquiring Mastery
An interactive remote job entry system had a very plain scripting language to manage remote job entry jobs. 1: I found a way to write loops and then implemented digit by digit add and carry with nested loops. This expanded to 100s of lines but 2: similarities allowed me to fold clause after clause until the script was an impenetrable 30 lines of code which got me my first summer job.
I set out to squeeze 1: thousands of contest radio contacts into not much more memory on a computer that had to recover gracefully from field power failure. This was written in a 2: not-yet production compiler so demonstrating that retrievals were accurate and survived power failure required participation with the hardware and compiler designer's participation.
Sharing the Knowledge
I admired the way that Smalltalk's image survived reimplementations. 1: Could we support this freedom to change in a commercial environment where customer data was at risk. My first implementation was too complex so I reduced it to a collection of permutation vectors and 2: a colleague figured how to recompute new vectors with each release.
US Federal Marshalls captured 100s of gigabytes of compile-by-compile development logs. I was hired to 1: find evidence of copyright infringement. I reduced source code to punctuation sequences and sorted these to bring similar sequences together. I printed 2: hundreds of pages of signatures and spotted a dozen unlikely coincidences.