Development log of a life-long coder

Wrapping up week 1 of the 100 languages challenge

I'm trying to write code in 100 different programming languages. Today, I'm wrapping up week 1.

The full list of languages and links to code are in my 100-languages repository.

Days 1 - 2

I got off to a pretty rocky start by starting with two languages that run in limited environments: SIC-1 Assembly Language (which runs on an 8-bit single-instruction CPU with 256 bytes addressable) and SectorLISP (which has only lists--no integers!--and with a max of a few thousand linked list cells). Economizing solutions was fun, but time-consuming.

Days 3 - 7

I could have breezed through the first week by using procedural languages I already knew, but the whole point of this journey is to play with new programming languages with minimal investment, so I set two goals for the rest of the week:

Here are the languages I ended up using:

Technically, I think I've written Smalltalk and Standard ML code before, but it was a long time ago, and I definitely did not remember anything beyond high level concepts. SQL was the only language I was fluent in (but not for solving math problems!).

Doubts about this challenge

I've tried to make it clear that I'm not expecting to actually learn or become proficient in all of these languages--I'm just sampling them and ensuring that I apply what little I do learn. Eventually, maybe I'll dive into a few of the languages that intrigued me the most.

Honestly, it's been fun to dabble in languages I have no business writing code for. In that aspect, this challenge has been a resounding success.

However, sometimes it feels like this is actually just a challenge to learn 100 different syntaxes. In other words, I might be spending time on the least interesting part of any programming language: syntax. (Not to say syntax isn't important, just that syntax is almost never a reason to use or avoid a programming language. For the record, I said "almost never".)

It remains to be seen if syntax overload ends up sinking this adventure, but for now I'm planning to forge ahead.

Language notes

I'll end this update with some notes about the languages I used.


Fortran is verbose. It tries to be precise, but there's enough historical baggage and implicit behavior that it seems really easy to screw up. I don't think this particular programming problem played to any of Fortran's strengths, however, so my perspective is skewed--I was basically writing C in Fortran.



I was excited to try an array language, but wow was I unprepared for the overwhelming cyclone of syntax. Like regular expressions, I'm sure once you're proficient with J it's quick to whip up little programs. Also like regular expressions, J seems difficult to parse at a glance. Trying to wade through dozens of sigils as quickly as possible was probably the worst possible introduction to J.

Despite being in the wrong mindset, I did enjoy programming in J, because it felt so novel. The closest I've used to an array language is probably NumPy, but J is its own language, designed from the ground up.



After reading a very long and entertaining post about Forth roughly a year ago, I've wanted to give Forth a try. Forth's minimalism aligns nicely with my own aspirations. Who knows, maybe Forth will control a post-apocalyptic world.

I was prepared for Forth's minimal syntax (even more minimal than Lisp) and I knew how to use a stack. But I underestimated just how exasperatingly tedious managing the parameter stack would be. I'm sure there's some better solution for temporaries that I skimmed over, but having to SWAP 1 + SWAP, etc. was inconvenient.

Overall, I'm not sure how I feel about Forth outside of the embedded world. I do hope to implement my own Forth one day--and maybe with the benefit of enlightenment I'll feel differently.


Standard ML

I love arrow functions, map, reduce, etc. -- and I have functional programming languages to thank for those, even when I'm using JavaScript. Having said that, Standard ML was the first language I ran into during this challenge where it felt like I was wasting my time stumbling through syntax. I still don't know when I need parentheses and when I don't. As usual, this is not necessarily the fault of the language--just of the environment I've created for myself.

Having said that (and I'm sure people will disagree), I prefer Lisp's regular syntax when compared to something like Standard ML. Maybe I'd like Clojure?


I was prepared for everything to be an object in Squeak. I was excited for a fully introspectable language. Want to know what something does? Just look at its source!

I was not prepared for clicking through menus and having to use the mouse all the time. I'm also not sold on image-based development being the default.

In the end, I think there's a lot of value in a standardized development environment and separating logic into objects, but I think most of those benefits are visible when working in teams or over longer periods of time.



Similar to J, it's fun to tell the computer what I want it to do without having to specify the particular steps for how to do it.