I wake up on time to a quiet morning before dawn. I move through the steps of my daily routines, picking out my clothes and throwing open the curtains. I stare out to the gradual morning light, listening for birdsong. As I set myself some breakfast, I contemplate what I will do with the two hours before I start work: programming a concurrent memory reclaimer.
While the project has sat in my head for a while, I started writing the API and implementation yesterday. I trace the branches of the dimly-lit trees on my window, and I remember how astonishingly simple the code seems. I think about the underlying ring buffer, its structure as delicate as the little basil plant I am watering. I think about supporting MIRI and loom. I smile as I remember how proud I am of the name I picked – housekeeping. I think about the projects I need it for, and where they will lead. And I return to a familiar feeling, the throughline of all my work: the utter joy of and drive for making nice things.
When I was in high school, I explored the PAQ data compression programs. I still have a battered old notebook where I wrote some notes about a project from that time: rewriting one of those tools (I believe it was PAQ6) and building a better CLI for it. While I was curious and wanted to understand the compression algorithm, I was distinctly more excited to build the CLI. It seems a bit silly in retrospect: the CLI was quickly forgotten, and the code is long gone. But I spent so much time understanding arithmetic encoding and predictive models for special files. Through that project, I learnt C++, which eventually drove me towards C, assembly, and interesting programming languages. 8-year-old lessons on entropy and data compression continue to influence the way I think today.
I think, even then, I was driven by a need to build nice things. To leave things better than I found them. As a child, I remember learning that I could build websites with HTML and CSS, and my first thought was “I can use this to make the websites I use look better!” I spent the next two years slowly building my own websites and exploring HTML, CSS, and Javascript. I had nobody to teach me and nobody to share my efforts with. I think that was the first time I was motivated to do something, for myself, that went beyond self-entertainment.
I love exploring the wide variety of human experiences. I think we can learn a lot from each other, and I love a good long conversation. I remember talking to my classmates in high school about the future, and I was shocked that nobody else knew what they wanted to do. Even at the time, I could recognize that I was incredibly lucky to have a passion for something so early – and that I could build a career around it. I don’t know how my classmates are doing now, but I hope they found something similar. Do you know what motivates you, dear reader? Do you have any idea where that comes from?
The transition from high school to university was staggering. For the first time ever, I got to meet people who were excited about the same things as me. I had worked on so many things over the years, projects and ideas that I had no one to share with. And suddenly there were people who would understand. To say that I felt less lonely is an understatement; that first year was the most extroverted I have ever been. I found out I was genuinely good at programming. I gained the confidence to chase bigger ambitions, to explore my ideas further.
But of course, growing up can only be so smooth. With my ambitions through the roof, I spent most of my bachelor working on my own side projects, trying to make interesting things better. But something unexpected began to occur: every time I’d start a project, I’d lose interest and move on to something new after a few weeks. I can’t count the number of times I excitedly introduced my friends to a new side project: SIMD-based lexing, an operating system, a TLS-like protocol, even some keyboard firmware. They would ask about the last project I had excitedly announced to them, and I would simply shrug. The new idea was always so much better than the previous!
It took me a year to notice that my friends were finishing things – smaller things, perhaps, but at least they existed. They mattered. I began getting frustrated with myself; I saw the faults in my effort and my attention. It was a unique kind of torture: to dream so widely, to distinctly see what needed to be done, and to simply be … unable. It took several more years before I got an ADHD diagnosis and gained control of my perfectionism (something I want to talk about in detail one day).
Slowly, I learnt to take on bigger projects and give them the attention they needed. I spent the first three months of 2025 investigating fast big-integer multiplication routines. Now, for the last eleven months, I’ve been working on Krabby. It’s the longest I’ve worked on a single project, and I expect I’ll be working on it for several years. I’m incredibly proud of it; even though it’s in its early stages, I have something concrete I can share, and I have a clear, practical plan forward.
In many ways, Krabby is the perfect project for me: it’s clearly a years-long endeavour, but incremental progress still feels meaningful. As my attention has strengthened, I’ve been able to tackle months-long sub-projects like takeaway and phonebook. These directly contribute to Krabby, let me solve interesting challenges in an isolated environment, and give me a sense of accomplishment. I’m still learning how to persist through hard challenges like name resolution; but I’ve made it this far. I’m proud of the effort I’ve put in and how much I’ve achieved.
This has also changed my relationship with my need to make nice things. After years of wanting to make nice things, of having the vision for it but feeling unable … I’m actually getting somewhere! I finally get to enjoy the fruits of my labors – carefully designed things that are measurably better than what came before them, like phonebook – and it has been unimaginably satisfying. I’m excited to keep going, to see Krabby as a whole ends up. I’m so incredibly motivated to keep going, it’s almost scary.
I remember implementing one particular optimization for phonebook regarding string comparisons (which I’ll talk about in detail at my RustWeek talk). I had the idea one night after staring at perf reports. The more I thought about it, the more excited I got – there were so many hot instructions I could eliminate. The new logic required subtle bit manipulation, and I was very glad for my randomizing test suite; it took more than a few iterations to get it all right.
I wonder if the thrill of the chase has roots in some biological incentive. I certainly recognize it in myself that day. Sleep became a secondary concern; I could feel the new code taking shape in my head, and I was so close to making it real. My heart was racing. Whiteboards piled up around me, covered in trains of thought and scribbles of frustration. There were so many pitfalls and dead ends, each attempt feeling like the one to end it all. Hours passed. Haunted by sleep deprivation, I dug deeper into the rabbit hole of bit hackery. And with a quiet, green flourish, my test suite reported satisfaction. I slumped back in my chair.
I took stock. It was 3 AM. I was still saturated with adrenaline – sleep wasn’t taking me anytime soon. My schedule for tomorrow (today?) was ruined. I was hungry, but eating something would only exacerbate my problems. But none of that mattered – I had captured something interesting, and it was time to figure out its value. My hand hesitated over the keyboard for a moment, my muscles frozen in anxiety. I started the benchmarks and ran out of the room. As desperate as I was, I didn’t want to see the results one by one. And I deserved some chocolate.
I stepped back in the room, my attention entirely captured by the harsh glare of my screen. My brain felt heavy. Time seemed to slow: I felt my eyes dart over the little text almost lazily. They were trying and failing to make sense of the Criterion results. I blinked slowly, begging my body to hold it together for another 30 seconds. I didn’t care if I died afterwards. After an eon, my brain finally digested the information I had chased. It was all worth it. I had saved two entire nanoseconds.
Optimization has a strange relationship with perfection. The closer you get to the hypothetical best, the harder you have to work to close the gap. If your program is 60 seconds from perfection, it may take an hour to save 30 seconds, and two hours to save the next 15 seconds. In my case, my implementation was already at 18 nanoseconds; I was sure that 10 nanoseconds could not be achieved. To gain 2 nanoseconds here is a huge win.
Processing that result was a physical experience. The tension in my chest burst loose like a valve; every part of my body relaxed, then went rigid as it was flooded with euphoria. I don’t remember standing up or screaming, but I had to consciously close my mouth and find my balance. As the moment passed, I didn’t know what do with myself; I wasn’t ready to deal with a physical being that needed looking after. I wasn’t ready to do anything. I was full.
So yes, I’m scared of my desire to make nice things. I screamed with the joy of improving an common non-essential piece of code by an insignificant amount. I want to do it again, and I have done it plenty of times since. I must believe that something is wrong with me, but I don’t think I’d trade it for anything. It has guided my work and shaped who I am. It is my past, my present, and my future. And I can’t wait to show you where it leads me.