I'm confused

I am a Computer Science Teaching Fellow at Tufts. Every so often, somebody comes up to me and expresses some form of the same kind of anxiety, frustration, worry — whatever you want to call it — centered around one thing: “Oh gosh. Everyone here knows more than me and they understand things so quickly and they solve problems so fast.”

First, I want to address the comparison to other people. Don’t do that! You are a different human, a different mass of cells and meat and fluids, with a different background, a different history, and a different brain.

Second, I want to address the notion that everyone knows more than you. Other people will be better than you at something. You will be better than other people at something else. This is normal. This is expected. We all go to school to learn and become better. Learning to become okay with this feels like a superpower. I am still working at it.

Third, I want to tell my story.

I am not special.

I am the product of many happy coincidences. I wrote my first line of code when I was around the age of nine. For some reason, people attribute a lot of meaning to that, but my interest in programming all comes down to weirdly specific and lucky combination of factors:

• I did not play videogames, but somehow I wanted to make them
• I had the wherewithal to ask my fourth grade teacher how to make videogames
• He had enough patience to go home and poke around on the internet for a bit and recommend Scratch
• My dad bought an OLPC
• The OLPC had a small Python-esque interpreter called Pippy installed
• I had enough time to sneak down and use the family computer at night when nobody was monitoring it
• We traveled back to California every once in a while
• Our neighbors in California hired a tutor named Steve¹ to tutor their son in computer science and math
• Steve was kind enough to sit me down and teach me alongside my neighbor

Without any one of those having occurred, I might not have started to write code in the first place. Which might have been better for my long-term mental health, because I consistently frustrate myself trying and failing to understand concepts or complete projects.

I have been confused for ages.

What follows is a half-complete log of me trying and failing to understand how programming works. It spans eleven years and much heartbreak.

I started playing with Scratch. I tried to make some games. I did not understand how boolean logic and loops and time worked, so instead I made some short animated movies with the block syntax.

I found Pippy, a small and custom Python interpreter for the OLPC, and made some weird program that told bad jokes. I do not remember the specifics, but I had at least figured out if-statements and reading input by then. It probably took me a month to figure out why input would sometimes work and sometimes give a syntax error. Nobody was around to tell me that raw_input was the way to go, because input called eval on whatever the user entered. And I would not have known what eval meant anyway. Nothing much came of my OLPC programming.

I discovered that the crappy calculators we used for middle school math were programmable in an obscure language called TI Basic. I heard they could be programmed on the computer and then that program transferred to the calculator. But we did not have the cables, so instead I spent many math classes trying to type full programs on the little keyboard. I wrote some helper functions that did some trigonometry problems for me. I did not understand what I was doing, though, or how variables worked, so whenever I wanted to write a new program I would have to go through the manual again and try and discover anew how TI Basic worked.

I remembered Pippy and tried in vain to find a Pippy interpreter for Windows XP. Alas, nobody wrote such a thing — of course — but I somehow stumbled upon Python and fumbled my way through a Python/IDLE installation. I wrote a math module that did more of my math homework for me. But my attempts to do anything remotely beyond that scope were foiled. And my math teacher found out about the module because my homework had all these unnecessary steps in them, and my program was not smart enough to remove them. So I got in trouble.

At some point in here I discovered JavaScript and spent a couple days writing small interactive websites. In order to do some kinds of logic, the internet recommended writing code in PHP. So I spent several weeks not understanding the difference between client-side and server-side applications, trying to embed PHP in a <script> tag. I eventually gave up. I only grasped the difference years later.

Steve, the tutor from some time ago, tried to teach me something about programming on a semi-regular basis. He used SICP as a text resource and had me learn both in C and in Scheme. I could repeat what he did, but I could not come up with the solutions to these problems on my own. I got really frustrated because I could not understand what the heck recursion was and why it was so important that a function could call itself.

In one of my English classes in high school, there was a time when the class was sent to the computer lab to write an essay. A friend of mine was writing something that was decidedly not an English essay, so I inquired further. It turns out that he was writing a Python program to trade Bitcoin.² He agreed to (re-)teach me the basics of Python, because I remembered next-to-nothing. I wrote some bad code and called it a day.

I joined the robotics team because my aforementioned neighbor was on it and he was the programming captain. He was the only programmer on the team and I wanted to be of service. At this point in his career, at the age of 15, he had already written a fully-functional parser, compiler, and virtual machine in C. Because I was functionally useless and also wanted to learn how to make my own programming language, he sent me away several times with exercises like “write a simple stack-based math language”.³ I coded myself in circles and did not grasp how the program was supposed to maintain state or even properly read in input.

I took the AP Computer Science class that my high school offered. The exercises were generally reasonably trivial and they were partner exercises. The hardest one for me and my partner was implementing Dijkstra’s algorithm. We did not figure it out in time and submitted a half-baked and broken solution. To this day, I cannot recall how exactly it works without referencing some external source.

Sometime during high school, I got so frustrated that I came nowhere close to understanding programming at the level that my neighbor did. I turned to the internet for help. Some kind soul named Charles⁴ took the time to write a long, well-thought-out response. I did not really understand what he meant at first, but after a month or two of goofing off when I should have been working at my summer job, something clicked and I wrote a small program that accomplished part of his suggested task. It was not Turing-complete, but it worked.

Another English class came along. In this one, my teacher had an idea of “20% projects”, the kind that Google encourages. One day a week, we got to work on projects of our choosing — so long as we wrote them up for the end of the semester. I chose to re-write my crappy “programming language” as an actual virtual machine, because I still did not understand how on Earth those were supposed to work. I had a half-working implementation by the end of the semester, and continued poking at it throughout my freshman year in college. I was fortunate enough to have a mentor — my robotics coach — when everything frequently went to hell and broke. He was there primarily for moral support and occasionally to point me in the right direction.

There is too much frustration in college to post it all here. It would sound repetitive. But I struggled in COMP 15. I struggled in COMP 40. In Web Engineering — I think I pissed off everybody I worked with in my group. I struggled in Operating Systems. In COMP 160. In COMP 170. I still do not know what happened in Graphics (sorry, Jason!). Or the Security Capture the Flag, for that matter (sorry, everyone on my team!). I struggled in COMP 105. In Networks. I struggled as a TA. I only figured out how the heap data structure was supposed to work when I got up to lead a lab about it and someone asked a particularly tricky question. I answered student questions wrong. I told them too much. Too little. Did not know how to teach. Did not know how to manage my own time.

I will continue to be confused.

Even now, I am bashing my head against the wall trying to figure out how I want to represent my data for the compiler I am working on. Some people react strongly to this — “Max, but you at least understand compilers! You did XYZ thing!”. Sometimes, sometimes not. It does not feel like I understand what is going on. After years of trying and taking three separate classes about interpreters and compilers, I have not written a functioning compiler. And this occasionally scares me because I am supposed to work on compilers full-time next year.

It’s okay.

I spoke to somebody at Google who studied compilers in undergrad and got a Master’s degree in compilers. He confessed to me that he too has not written a compiler and sometimes worries about this. This helped me realize: I am not the only one who feels like a fraud!

I spoke to my friend and co-TA Margaret, who reframed my thinking. She explained that it is important to think in terms of deltas. That I should look back over the past years and realize how much I have learned and accomplished in that time. A year ago, where was I? I was writing a blog series about interpreters because I did not know how they worked and needed to learn. And through writing about it I — over the course of half a year — finally figured out how the heck Lisp worked.⁵

There are so many wonderfully talented people who I know and admire. They suffer from the same feelings of inadequacy that I do. That many people do. I have occasionally been part of the reason for these feelings. I want to open up and make it clear that I did not learn how to program in a vacuum and certainly do not feel like I know what I am doing most of the time. I’m confused. And that’s okay.

This post resulted from a conversation with Margaret Gorguissian. Thank you so much!