Jan 2020

The annual summary, reflection, status.


Throughout 2019, I started seeing increasing utility in first-principle thinking in the projects I was involved in. A famous example of first-principle thinking was Musk evaluating existing rocket options for SpaceX. He scoffed at the $110M price-tag. He promptly considered what metals go in a rocket and multiplied the approximate quantities with the spot prices on the London Metal Exchange. The price he got was somewhere around $2M. Clearly, he said, there’s an opportunity here to bring the cost down (and he and his team did).

In the context of computers, when evaluating the feasibility of projects and tasks I’d started thinking about it as “it needs to read these files, this many bytes, send this much data over the network, .. Fundamentally that should be achievable in x time at $y.” When observing the real system, there’s typically a gap between the first-principle, bottom-up understanding, as there was in the rocket example. That first-principle gap presents an exciting opportunity, to either (a) fix the inefficiency in the gap (e.g. make the system cheaper or more performant), or (b) fix my model of the system to account for the gap (I was missing some important part, in Musk’s example, maybe he forgot to include cobalt for the aerospace grade alloys). I decided to give a talk on the topic as an excuse to dive in even further. To keep my (and others) skills sharp, I started publishing an occasional newsletter with a first-principle problem.


My 2019 reading ended up focused on innovation, the history of innovation, and computing history. My favorite from 2019 was Energy: A Human History which takes us through humanity’s journey from mules to a (hopefully) renewable future. It goes into depth with steam engines, combustion engines, and so much more. A close second was The Idea Factory on Bell Labs, an iconic industrial lab that gave us the transistor, solar cell, modern operating system, and so much more. I read the Three-Body sci-fi fiction series, which I thoroughly enjoyed as well.

I listened to way fewer audiobooks this year. I find that my Danish became quite stagnant, so I decided to listen to more Danish podcasts . This has re-invigorated my Danish, surely. Especially critical, as my Canadian citizenship ceremony is at the end of this month.

You can see my full list of 2019 reads here.


I’ve observed myself switching note-taking app every year (Evernote -> Workflowy -> Dynalist -> Bear -> Notion -> Dynalist), indulging with Airtable, and become somewhat addicted to chasing the perfect system. Sometimes you end up spending more time working on the system, than actually executing it. Migrations between them are painful, and lossy. It hardly compounds. These note-taking apps tend to have poor APIs, fall out of touch eventually, and generally age poorly.

To solve this, I decided to move all my notes to plain-text markdown files and use all the standard command-line tools I’ve always used like fzf; to search, vim; to edit, ripgrep; to search, ruby; to script, ctags; to navigate between them, and so on. It’s too early to tell whether this was too extreme, but so far I am enjoying it. Mobile-access is a problem, but I’ve chosen to accept that. There are some apps compatible with this system. They are ok, but not great. Eventually I’ll update my article on how I read for this. I’ll continue to use Things for todo, since in that context a good mobile app is absolutely necessary.

This year I’ve started to read more computer science papers. This is fairly dense reading, and difficult to do on a laptop. I bought an iPad for this purpose, hearing good things, but found that it was just as distracting as the laptop and returned it. I then bought a reMarkable based on a glowing recommendation from a friend. This has fit the niche very well for dense content, PDFs, very technical articles, etc. I still use my Kindle for long-form content. While I thought I’d primarily use the reMarkable as an annotation and reading device, I’ve found the note-taking and drawing to completely replace my paper-habits. Plenty of things that can be improved about the reMarkable, but it works well for me.


Jenn and I went to Dominica in the Caribbean this year. In 2017, they had a devastating hurricane, but the island is far along now in its recovery. We heard stories from the locals holding on to their loved ones, passports, while clammed in mattresses that night for protection while their houses were being torn apart around them. We witnessed both the resilience of nature and people. We recommend it for the pristine nature, hikes, and the lovely locals.

Boiling Lake in Dominica

In the early fall, we went to Europe. This was a combined trip for a friend’s birthday, my conference talk in Dublin, and hiking. Slovenians have got to be some of the hospitable, down-to-earth, friendly Europeans. We stayed mostly in the mountains as we traversed the North-Western corner. As our last visit, we stayed in the mountains near Hisa Franko, the most famous restaurant in the country. After a good meal, we ended up in the wine and cheese cellars into the wee hours… a fantastic evening, but we paid the price with a 2 hour drive to Venice airport the following morning. On a solo hike earlier in the trip, my phone gave out while hiking. This was unfortunate, as this was while I most needed it traversing an off-trail ‘shortcut.’ The Apple logo just kept persistently flashing, but no map. As I was on a massive decline and there was no nearby trail, I was good ol’ lost. Fortunately, when you’re on a mountain, you know at least what direction down is. I eventually found a creek that I hiked through for an hour, which lead to a village.. when I returned back to our farm-stay, I met the owner who asked about my day. She informed me that it was a terrible idea to take that shortcut, that yes, it’s on the trail maps, but it’s not maintained. And by the way, a few people die in these mountains every year (I am thrilled I didn’t know this while I was out there).

The mountains I got lost in


For as long as I can remember I’ve had a few fairly specific fitness goals: handstand push-ups, muscle-ups, one-arm pull-ups, respectable olympic lifts, hand walks, and a few others. Unfortunately, I’ve made little progress in this realm and mostly stuck with the exercises I felt comfortable with. Achieving proficiency in these movements requires some gruelling preparatory work, such as wall walks (extremely tiring), 100s of kicking up against a wall to a hand-stand and slowly lowering yourself (15 minutes of this is enervating), and feeling sketchy about cleaning above a certain weight. Something that takes a lot of energy and time to figure out on your own.

Consulting with a good friend (who’s in impeccable shape) on the issue, he recommended I seek out OPEX Ottawa. OPEX is a franchise with a strict training program (that my friend had gone through). Instead of doing classic personal training at a set time on a set day, they send you programs and have you execute the program independently (either at their gym, or your own). You meet with the coach occasionally, and have contact with them throughout the week through the app. That sounded perfect, because it meant I could maintain my freedom of going whenever and wherever I liked. Through a 3-hour assessment, he found many strength imbalances, mobility bottlenecks, and several other blind-spots. By and large, the best decision for health in 2019.

Subscribe through email to new articles (typically only a few a year).