So, somehow it’s May already? 😓 It feels like it was just yesterday that everyone was making memes about how March would never end.

My work announced that we’ll be continuing to work from home until at least June 1. Amazon extended its WFH policy until October 2. That seems… a long time away.

The best thing I read this week about coronavirus was “What the Coronavirus Models Can’t See” by David Wallace-Wells1:

Though the public narrative is that the country has turned the corner and gotten a handle on things, enough to begin slowly “opening up,” the data tells a different story. A new daily peak means that even if deaths declined as rapidly now as they grew earlier this spring — when in a month total deaths grew from 3,834 to 62,860 — we would be due for at least as many more deaths as we’ve had to this point. In other words, another 60,000 people. A peak is not an end; it is, in optimal circumstances, a midpoint.

The IHME model, in particular, has come under increasing criticism — “It’s not a model that most of us in the infectious-disease epidemiology field think is well suited,” Harvard’s Marc Lipsitch told Stat News. … But in part because it made a name for itself early in the epidemic, IHME remains a trusted name in the media (on the FiveThirtyEight podcast last week, Nate Silver interviewed its main author and offered praise for the model) and, in fact, its projections have ticked back up recently, though just by a few ticks, with a present estimate of 72,433 deaths by August 4. If the country stays on Wednesday’s pace, it will hit that number by May 4.

Predicting a crisis like this is, as I understand it, extremely difficult. That being said, this piece left me with the feeling that we are effectively “flying blind” through this pandemic. It’s sobering to remember how many unknowns there still are about the disease.


Last week, I started watching The Sopranos following a discussion of it on a recent Do By Friday. (HBO is streaming it for free during the pandemic, too!) It holds up surprisingly well. On that Do By Friday episode, Alex Cox mentioned how The Sopranos feels more like a period piece written about the 90’s rather than something filmed during that time. I agree. From what I’ve seen so far, The Sopranos could have come out last year and still felt fresh. Really looking forward to continuing watching it.

I also listened to Edward Snowden’s memoir, Permanent Record, as an audiobook. It’s well-written, and I’d recommend reading it even if you are familiar with Snowden’s whistleblowing disclosures. There’s a lot of back story about Snowden’s upbringing and early work in the US intelligence community (IC) that contextualizes his later actions. For example, I didn’t know the extend to which the US IC hired private contractors and subcontractors instead of keeping people on the government payroll (apparently, this was to circumvent congressionally dictated headcount limits).


I’ve been using an Instant Pot to cook dried beans. In doing so, one often overestimates the amount of dried beans they need to cook, because they start off small but absorb a lot of moisture. In other words, I accidentally made way too many chickpeas for a curry I was making this week, so I had to find some creative uses for the surplus.

The first obvious move was to make homemade hummus. I followed this recipe – though, you don’t really need a recipe: just blitz together chickpeas, lemon juice, olive oil, garlic, and tahini in a food processor. You can adjust the amount of lemon, garlic, and tahini to taste. Homemade hummus is always a hit! Turns out that “hummus toast” is a thing too, so I got to use up some more frozen bread.

The second, much less conventional recipe I tried was “Flourless Chocolate Chip Chickpea Blondies”. (If you’ve ever tried black-bean brownies before, it’s a similar concept.) I’m not going to lie, I was skeptical about this idea. But, it turns out that if you add enough sugar, nut butter, and chocolate chips to puréed beans, it tastes like a dessert. I have a couple family members with food allergies (gluten, dairy, etc.), and these “blondies” were a hit with them.

Researching New Apps

I fell down a bit of a rabbit hole this week researching note-taking apps. Currently, I use an unorganized mixture of OneNote, Drafts, Apple Notes, and Todoist. The result is a scattered mess of one-off notes without much structure.

Notion was the new hotness for a while, but I couldn’t get myself to buy into it. It’s a cool tool, but it felt intimidating to setup. It also felt too rich for what I want in a note taking app: it’s a mashup of a personal wiki, Airtable, and todo app. I never felt that I could conceptually grasp what I was supposed to do with it. In short, Notion is maximalist.

Now, Roam Research is the app everyone’s crazy about. Roam seems more appealing to me: it’s UI is simple and focuses more on links between notes rather than rich interactive elements within notes (unlike Notion). Roam also emphasizes Markdown in a way I find appealing.

The problem with both of these apps is that they’re closed source, meaning:

  1. They can shut down (RIP Wunderlist).
  2. They can decide to change things like the UI and API limits at will (I’m looking at you, Todoist).
  3. Data portability is tricky, even if the app offers some sort of export.

I think Roam’s data is probably more portable than Notion’s, because Roam’s data model is closer to a flat list of files. But… I want my note taking app to be as durable to change as possible.

As far as open source solutions, I’ve seen a lot of activity recently around TiddlyWiki. I haven’t looked into it much, but it seems like the project has been around for a while, suggesting it has some staying power.

I want to invest more in taking notes, but I’ve struggled to find a system that sticks. I like the idea of a “flat list of files” with links between them. This needs no special software, is future-proof, and can be done using .md files and a text editor. I just need to spend the time to sit down and lay down some structure…

Completely unrelated, but in other “new app news” I also checked out Castro after seeing a positive review of it on Matt Birchler’s blog. Castro is very similar to Overcast, at first glance. It has subtle differences in the way it handles new podcast episodes:

That queue is the main reason I use [Castro], and for those unfamiliar, Castro has a “queue” and an “inbox.” New episodes in your feeds go into your inbox where you can decide to add them to your queue or archive them to just skip that episode. This makes the “queue” page exclusively a list of episodes you are going to listen to. I subscribe to a bunch of shows and I don’t listen to each episode of all of them, so this works with how I listen to podcasts.

Though I was impressed by Castro’s queue system, I’ll probably stick with Overcast.

  • I really enjoyed this article about Rust parser libraries. It describes a few different libraries, what they’re strengths/weaknesses are, and different techniques for writing parsers. I’d learned about some of this in my Programming Languages course, but PEG parsers were new to me. And, coincidentally, Python is probably moving to a PEG parser, so that’s neat!

  • CGPGrey posted an interesting video this week about living in lockdown, using spaceships as a metaphor for the physical isolation we’re all experiencing. It’s advice is pretty basic: prioritize sleep and physical exercise, plan some social activities and time for creative output – but I could see this being useful to people who still feel a bit unsettled.

  • “Every Covid-19 Commercial is Exactly the Same”. If this doesn’t appeal to the cynic in you, I don’t know what will:

  1. David Wallace-Wells also wrote the fantastic “The Uninhabitable Earth” about climate change. ↩︎