Clean as you code (replace "code" with anything you do)

Do you always feel so busy that you don't have the time to tidy your desk? Do you wish your living room would stay so neat that you could have friends over any time? The good news is that I might have found you an answer: Clean as you code (replace "code" with anything you do).

Here's my recent experience around that kind of problem. It's a tech metaphor for the versatile solution. As professional software engineers, we are responsible for keeping our code base maintainable and robust. Therefore, we use automatic code-quality checking tools. Some tools list all of the lines of code that do not meet the standard. However, when different developers change the source code, the warnings with descriptions the tools produce are from unfamiliar parts of the code to each developer.

On top of that, there may be existing warnings that everyone has been ignoring. It is unclear if you are responsible for correcting all the lines to eliminate the warnings. If you decide to do it, it will take too long to get to the task with a deadline. This is when the "clean as you code" concept is handy. In this concept, you are only responsible for the quality of your new code. We use SonarQube to analyze new code, but don't worry about the technical details if you are not a developer.

This concept has been surprisingly applicable to many of my daily routines. My problem was that I either spent too much time decluttering my entire home office before I started working on a task, or I opted to do no clearing because I felt that I didn't have time. After I tried to use the clean-as-you-code approach for a while, my desk and living room gradually increased their neatness levels. Every time you clean around your small task, you improve the cleanliness of the entire system. Over time, the overall cleanliness improves. The code you just wrote is fresh in your memory, so it is easy to clean. The table you just used for lunch is right before you; you don't have to move.

Try this concept in your daily routines, and let me know the results! This concept is pretty simple, but more complex computer science concepts can also be applied to our daily lives. This book, Algorithms to Live By: The Computer Science of Human Decisions, includes concrete examples. I plan to write about them in the future. If you haven't subscribed to this blog, find and click on Subscribe by email.

Hanoi Reflections: Exploring with a Local Touch

Spending a few days in Hanoi totally changed how I see things. Despite the crazy traffic and city buzz, the people there blew me away with their down-to-earth vibe and genuine kindness. It got me thinking about my own values, realizing how much they're shaped by where I come from. Hanoi taught me that there's a whole lot more out there to experience. I got to soak up Hanoi's vibes with my good friend Chi, a local who knows all the best spots, making my adventure even more memorable. Thanks Chi, it's your turn to have a blast in Tokyo.

Dancing is similar to the game of basketball

An intriguing argument recently caught my attention as an avid basketball player who has just got hooked on K-pop: dancing and basketball share remarkable similarities.

In this episode of "Under the Microscope: Dissecting the Art of Basketball," Coleman Ayers and Yosef Yishak explore how these similarities can be leveraged to enhance performance. Whether you play another sport, practice an instrument, or engage in any activities requiring movement skills, the ideas discussed in this podcast are broadly applicable.

How I Rearchitected A/B Testing at LINE

For the LY Corp tech blog, I wrote about one of the largest projects I've led at LINE. About eight months ago, I set the goal to make the software development operations at LINE NEWS significantly more efficient by changing the A/B testing mechanism and its surrounding architecture. First, I pitched the new architecture I designed to the development team, the SRE team, and the DevOps team. We extensively discussed how we could implement it.

At the same time, I consulted with the data science team to make sure our new system would effectively run A/B tests. How I executed the development part is written in the article. The results were so rewarding. The new system decreased the lead time from the start of the A/B test and the final release by 30% to 60% for most of our recent features.