Remote Pair Programming with Pat Maddox

When I woke up Saturday, my goal was to write a post reflecting on my experience pair programming remotely with Pat Maddox who had recently invited anyone who is interested to pair with him. Once I started thinking about how I wanted to approach the post, I thought I needed to talk about remote pair programming in general and how it has been a great way for me to learn and interact with more experienced rubyists. However, a converstaion with Evan helped me realize that what I have stumbled upon has a much greater value outside of being a great learning tool for me. So here I am at around 9 PM on a Sunday night finally starting to put my thoughts down on my experience pairing with Pat Friday evening.

First of all, I want to thank Pat for making time to pair with me as well as for taking the initiative to invite people to pair with him. It is really great opportunity not just for beginners like me who can learn a lot from pairing with someone of Pat’s caliber but also for other members of the community to be able transcend the limitations of physical distances and come together to work on a problem. The whole concept of leveraging the internet and tools that we have at our disposal that allows us to collaborate with one another deserves the kind of spotlight that initiatives like Pat’s will bring to it. However, I have already talked about that in previous posts and want to focus more on the specifics of the experience pairing with Pat here.

Test Driven Development (TDD) is something that is very new to me and has only recently been part of my practice in my journey through the long road of learning Ruby and becoming a better programmer. Therefore, it is still a skill that I struggle with. So it was great to work on a code kata with Pat and see how he started out with an acceptance test that was going to fail. This was the test that would tell us that we are done, once it passes. However, to start making progress, we decided to make the acceptance test pending and focus on figuring out what the maximum temperature for that day was. Similarly, we wrote another test that told us the number of days in our dataset. While these tests were not really testing the crux of the application, they helped us make progress on the problem and that was a great tip to learn.

It was clear that Pat has been doing pair programming with people from various skill levels. He was very good about encouraging me to drive and always asking what I thought of the approach we were taking. I immediately felt at ease with sharing my opinions as well as asking questions whenever I got confused about the code. Pairing on a Kata was also good because it was small enough problem that we could understand the domain pretty quickly but big enough to provide interesting problems to solve. One of the things we had to ensure was to ignore the unnecesary text in the weather data file. We also encountered a bug on line 46 where our previous code looked as follows:

@days << parse_line(line) if in_data_section && line.strip.match(/^\d/)

The problem was that while we made sure to strip the line when we looked for the line with weather data, we sent the parser the line with preceeding spaces which caused our parser to extract the wrong weather data. This kind of bug would have taken a while to figure out if it was just one person programming but between the two of us chatting, we were quickly able to find it, fix it and be well on our way. The test suite, no matter how sparse it is, definitely helped too :).

Even though we only got around to finishing up part 1 of the kata, I look forward to pairing with Pat again in the future to work on parts two and three of the Kata. It was a really great learning experience and I encourage anyone who is interested in Ruby to take Pat up on his offer and pair with him.