Learning something new

I wanted to learn some Ruby because of an upcoming talk I am giving. I found Ruby quite easy to grasp. Ruby has some great  characteristics and I enjoyed working with it. The most difficult part of coding in Ruby is stopping yourself from humming the song by the Kaiser Chiefs’  who coincidentally I’m going to see in a couple of days time. Yes, I know you’ve started humming it too now, I’m sorry…

Anyway, this entry is not about the Kaiser Chiefs nor is it about Ruby. It’s about how we learn a new language or technology. If you work in development, you know how cut throat our industry can be if you don’t keep up with new technologies. Sometimes it feels like we’re in a perpetual “learn technology - use technology - learn technology” cycle. To a certain extent most other professions also require you to keep up to date. A doctor for example, needs to keep up-to-date with the latest progress in medicine and treatment methods. However, I doubt that the pace of change is quite as quick as in technology. As software developers we are constantly under pressure to learn new things as quickly as possible.

So how do we go about it? Different methods work better depending on what type of person you are. You can buy books and go through them page by page working through the examples. Or you might prefer to follow a classroom or online course. I find that using aids such as books and online courses are more useful when the technology is conceptually different from what I’ve used in the past. For example if I’ve never written any functional code, I might need to follow a course if I wanted to learn Scala. Similarly if I’ve never used a relational database I will need additional help to learn how to use MySql.

For everything else, I personally like learning by doing. Unlike doctors, if we want to learn something new, we can practise as much as we want without fear of causing any harm. The idea is to pick a problem and try to solve it with the new technology you’re trying to learn. Along the way you’ll learn the new features and tools available in the new language or technology.

The main difficulty with this method of learning is choosing the problem or project. If you attempt to pick a random project such as “Let’s develop an online flower shop” it will not work. You’ll quickly get bored of it. When you start hitting difficult problems you start to question why you picked such a useless project in the first place and putting so much effort into it. Soon you’ll drop it and say “eh, I’ll stick with visual basic for the rest of my life”.

Pick something that is useful and real. If you are learning javascript as you plan to open an online business, make that online business your learning project. If you need to learn python to start writing deployment scripts at work, learn python while doing that. Of course you’ll need to fail many times before you succeed, so try your pet code in an isolated environment before moving to the real thing.

Back to me learning Ruby now. I’m giving a workshop to a coding bootcamp, called “Le Wagon”, https://www.lewagon.com/lisbon in Lisbon in a couple of weeks. The talk’s title is “Beating the coding interview” and is about learning a few tips to solve the puzzles one gets given during an interview. At Le Wagon, students use Ruby as their web development language. I must confess that up until a few weeks ago I knew almost nothing about Ruby. Keeping in mind that I also have a full-time job, how could I learn the essentials in Ruby in the shortest amount of time?

Since the talk is about coding interview puzzles, what better way to learn it than to try to solve some of these puzzles in Ruby? The internet is full of example puzzles. One of the most popular sites offering these coding exercises to candidates is http://www.codility.com/. They have a practise area where you can try solving sample coding puzzles in many different languages, amongst them Ruby. My plan of action was to pick a few these sample questions and solve them, learning about the different bits of language along the way. Here’s how it works in practise.

The brackets problem from codility is about writing a function that checks if a string has a properly nested set of brackets "[]{}()". Here are some examples to get the idea:
text = "[()()]", result = true
text = "[(]()]", result = false
text = "[{()}]", result = true
text = "[(])", result = false

The solution is quite straightforward if you have worked with stacks before. The idea is to process the string, one character at a time, left to right. If you encounter a left bracket you push it on the stack. If you encounter a right bracket you pop the first item on the stack and check that the brackets match (round with round, curly with curly etc…). In the end you make sure you have emptied the stack.

Therefore, for the solution, I needed to find out how to use a stack in Ruby and how to do conditional branching (ideally using some sort of case statement). I also needed to loop over every character in the string and check if the stack is empty at the end.

Right, a few google searches later I found out:

    1. Iterating over every character in a string:

Code Editor

    2. Using a case statement:

    3. Using a stack:

   4. Checking if an array is empty:

   5. Using an if statement:

Now that I had the building blocks all that was left was to put everything together and to try it out (The app on Codility confirmed the next solution is correct): ​

My solutions to codility lessons problems can be found at https://github.com/cutajarj/CodilityInRuby

Happy learning by doing!