Recurse Center: Week Two

January 17, 2022

This post is about my second week at the Recurse Center.


One of my primary goals at the Recurse Center, is to learn how to write programs in Clojure. To this end, I had spent the first week familiarising myself with the basic syntax of Clojure. This week, I could delve deeper and wrote some not-so-simple programs.

Before the start of the retreat, I had asked for suggestions for getting started with Clojure. Kapil Reddy was kind enough to point me to an on-boarding course, clojure-by-example, that he, Aditya Athalye and a few others had designed.

The week began with a long and super helpful Clojure pairing session with Aditya, who gave me a detailed walk-through of clojure-by-example. Until then, I had only read the first three chapters of Clojure for the Brave and True. While I had a theoretical overview of some of the basic semantics of Clojure, this pairing session gave a more hands-on take on how to read, evaluate and write expressions in Clojure. This was the perfect start to my week!

During the rest of the week, I concentrated on completing the next three chapters of Clojure for the Brave and True (Core Functions in Depth, Functional Programming, and Organizing Your Project: A Librarian’s Tale). Learning about sequence abstraction was truly fascinating

It doesn’t matter how a particular data structure is implemented: when it comes to using seq functions on a data structure, all Clojure asks is “can I first,rest, and consit?” If the answer is yes, you can use the seq library with that data structure source

While reading Clojure for the Brave and True, I learnt about some of the key concepts of functional programming. Specifically, reading about pure functions and their use-case was a lot of fun! I would like to spend some time reading more on concepts (like ‘referential transparency’, and ‘functional composition’) that are fundamental to functional programming. Also, reading about higher-order functions was a lot of fun as it reminded me of my early adventures with map, reduce and filter.

I had struggled to visualise the importance of lazy sequences while writing programs. So, I ended up spending considerable amount of time, playing around with lazy sequences. In the end, I published a blog post (with a cheeky title) summarising my findings.

As reading several pages of Clojure for the Brave and True can get a bit monotonous, I would often take a detour to implement some of the examples and concepts explained in the book. It was particularly satisfying to take up Daniel’s challenge on Chapter 4:

If you want an exercise that will really blow your hair back, try implementing map using reduce and then do the same for filter and some after you read about them later in this chapter.

Here’s my implementation of map-using-reduce:

(defn map-using-reduce
  [map-fn xs]
  (seq (reduce (fn f
                 [aggregator new-val]
                 (conj (vec aggregator) (map-fn new-val)))
               (empty xs)

Here’s filter-using-reduce:

(defn filter-using-reduce
  [filter-fn xs]
  (seq (reduce (fn
                 [aggregator new-val]
                 (if (filter-fn new-val)
                   (conj (vec aggregator) new-val)
               (empty xs)

And, some-using-reduce:

(defn some-using-reduce
  [some-fn xs]
   (reduce (fn [aggregator new-val]
                  (if-not aggregator
                    (if (some-fn new-val)
                      (some-fn new-val)

Here’s my GitHub repository, hosting my white-board adventures with Clojure!

Also, as suggested by Aditya, I keep solving the 4Clojure problems mentioned at the end of each exercise of clojure-by-example. So far I have solved a handful of them; I intend to solve many more this week. Here’s where I write down my solutions to 4Clojure problems:

Using NGINX to Use a Custom Domain

While most of the week was spent on Clojure, I could sneak out some time to complete a long-standing task related to Twirl. Twirl is a URL-shorening web app. As it is hosted on Heroku (free tier), the domain name of the app is pretty long( This makes the short links generated by Twirl not so short after all. I used NGINX on a Google Cloud VM I had spawned recently, to generate a custom short domain name for my application. Twirl is now accessible at: This makes the resultant short links shorter! Here’s a detailed account of the steps I followed to complete this task!