First of all, a bit of context-setting: I am not quite yet a father of a daughter, but I am four months out from becoming one, and I suspect my reading and emotional attachment to this book — and to Nell, as charming and plucky of a slightly-unbelievable-but-still-winning bildungsroman protagonist as one could hope for — are so heavily tempered by this book that quibbles I might have had even a year or two ago are sanded down into oblivion.

The reason for this is that in much the same way that Snow Crash purports to be a cyberpunk book but is actually a book about communication and language, The Diamond Age: Or, A Young Lady's Illustrated Primer purports to be cyberpunk but is chiefly interested in education and parentage. Stephenson's book contains two parts (a decade-long lacuna separating the two), and each part is concerned with a single question:

  1. What social ills might exist in a post-scarcity world?
  2. How do we raise children to be interesting, happy, and virtuous?

I did not find either either side of Stephenson's first question to be particularly revelatory: the yadda-yaddaing of nanotechnology and its impacts felt thin and post-hoc to me, even if there were some flashes of brilliance in isolated asides. But where Stephenson was both clearly more interested and more persuasive was when he began to shift the narrative to focus on the triad of young girls who had received the primer, and their varying divergences.

Moreover, the device of the Primer carries with it a level of stylistic flourish difficult to find in other pieces of his work. Princess Nell's escapades served well to shepherd us through the obvious paces of character development while offering something relatively novel; the Turing Castle chapter in particular (and thereafter) was a legitimately fun (albeit, of course, campy) way to deliver what could have been dreadfully boring material.

This book suffers from the classic Stephenson tropes:

  1. As much as I think Stephenson lovingly writes Nell's character, I can't escape the feeling — much like in Snow Crash — that her sexual assault feels oddly transactional, as a heavy-handed way to weave villainy into a narrative that is in general pleasantly bereft of cartoonish intentions;
  2. The book ends with a beautiful visual metaphor that occludes but does not fully disguise its incoherence.

But — this book gave me so much to digest and ponder and enjoy, and it has not fully left my head since reading it. There are two types of science fiction: the ones that get better with age and the ones that get worse. This is the former.

Highlights

“Yet how am I to cultivate the persons of the barbarians for whom I have perversely been given responsibility?" … "The Master stated in his Great Learning that the extension of knowledge was the root of all other virtues."

Hackworth was a forger, Dr. X was a honer. The distinction was at least as old as the digital computer. Forgers created a new technology and then forged on to the next project, having explored only the outlines of its potential. Honers got less respect because they appeared to sit still technologically, playing around with systems that were no longer start, hacking them for all they were worth, getting them to do things the forgers had never envisioned.

“There's only zero of you,” said the Queen of the Ants. In ant arithmetic, there are only two numbers: Zero, which means anything less than a million, and Some. “You can't cooperate, so even if you were King, the title would be meaningless.”

But as many first-time fathers had realized in the delivery room, there was something about the sight of an actual baby that focused the mind. In a world of abstractions, nothing was more concrete than a baby.

He nodded in the direction of China. "Been doing a bit of consulting work for a gentleman there. Complicated fellow. Dead now. Had many facets, but now he'll go down in history as just another damn Chinese warlord who didn't make the grade. It is remarkable, love," he said, looking at Nell for the first time, "how much money you can make shoveling back the tide. In the end you need to get out while the getting is good. Not very honourable, I suppose, but then, there is no honour among consultants.

So much of the discourse about this movie was in the meta-sense of "big, vaguely selfaware action rom-com starring two of the leads from Barbenheimer that is critically well-received but a box-office bomb — whither Hollywood?!" I think you can safely ignore that kind of pablum; this movie is neither good nor essential enough to warrant a wave of discourse, which is not to say that it's bad. It's not a Very Good movie; it's a very Good movie — in the lineage of The Nice Guys or The Other Guys (both Guys movies themselves), winking and pleasant and charming and a little bit overlong.

There absolutely should be more of these kinds of films, if for no other reason than to remind us that part of the value proposition of top-tier movie stars is that they can elevate something a little bit further past its script-borne shackles. (And Gosling is terrific in this, make no bones about it.) But it is not some sort of deep, industrial crime that this is not more widely beloved.

When I wrote Your job is to produce value, I got a few responses that could be summed up as something along the lines of:

Telling developers to get better at reading and writing and problem solving is sort of like telling them to get better at writing fewer bugs: it might be true, but it's anodyne and unactionable.

First off: this is not true, and if you think this way (that, say, communication is some sort of inherent skill and not a thing that improves with practice and study) you should probably try to carve out daily rituals of blogging, reading, and UX research.

But! I am sympathetic to the argument that there are more decisions and options an engineer has to make in their career than simply being customer-, finance-, and experience-focused. Knowledge accretes, and while a certain radical generalism — no stacks, no sectors — is probably the best overall meta for a young engineer the reality is that you will have to make choices about which companies/organizations/teams/languages to spend time on, and those choices have marginal consequences: all other things equal, someone who joined a web3 startup in 2020 or an Angular shop in 2018 probably has marginally worse prospects than someone who got into GIS work or Swift right as they were taking off.

(I use the word "marginal" there intentionally: the industry or technical stack you land in is far less important to your career than anything else.)

This belies the question: assuming that you don't have oracular vision, what kind of opportunities (at the company level, at the project level) are the right ones to pursue?

I have two answers that I think overlap a good deal:

  1. Haunted forests. In the linked essay, John talks about this metaphor within the context of SRE, but I think its true at a broader, more conceptual level: the prize for tackling domains and codebases that terrify others is knowledge, esteem, and many battle scars.
  2. Lindy tech. In the grand catalog of infrastructure required to run the modern world, the most powerful and most useful — DNS, SMTP, relational databases — tend to be old.

I say these two things overlap because folks often use the terminology of the former to describe the latter. "Don't roll your own auth"; "DNS is an eldritch horror"; "never try and send your own emails"; it's tempting to buy into these truisms, but the twelve months you spend really understanding how DNS works at a primordial level is going to have more long-term value than the twelve months you spend on GraphQL or whatever.

(All of this is true at an organizational level, too. Institutional knowledge decays over many axes; one way to stave off the inexorable decline is to invest in meaningful, durable substrates.)

It's fascinating to see that this is a Richard Linklater (of Everybody Wants Some!! and Before Sunrise) fame, as this feels in many ways so much more like a Lord & Miller production (or, more charitably, a lesser Coen work).

And, indeed, you can see Linklater take his foot off the gas and lean a bit into the "ehh, this is a Netflix paycheck" of it all: the hackneyed philosophy lectures, the bizarre suspension of disbelief required for most of the sting operations. But it's also a deeply entertaining film: the conceit is immediately delightful and interesting, Glen Powell acts his ass off, and the runtime knows that there's no room (or need) for unnecessary padding (unlike, say, The Fall Guy).

I'd argue that this is not a particularly good Linklater movie — I think he is most interesting when he is using cinema to express his feeling about the beauty and depth of seemingly-anodyne humanity, and this is him ignoring all that and deciding that it would be fun to make a rom-com. But it is a deeply good Netflix movie, and a perfect answer for "what's a fun, perfectly-delivered way to spend two hours?"

This quote from Peter Drucker, via @zetalyrae resonates a lot:

Write a report may, for instance, require six or eight hours, at least for the first draft. It is pointless to give seven hours to the task by spending fifteen minutes twice a day for three weeks. All one has at the end is blank paper with some doodles on it. But if one can lock the door, disconnect the telephone, and sit down to wrestle with the report for five or six hours without interruption, one has a good chance to come up with what I call a “zero draft”—the one before the first draft. From then on, one can indeed work in fairly small installments, can rewrite, correct and edit section by section, paragraph by paragraph, sentence by sentence.

This rhymes (but not does quite exactly overlap) with an oft-quoted section of The Mythical Man Month:

In most projects, the first system built is barely usable....Hence plan to throw one away; you will, anyhow.

(Notably, Brooks recants that suggestion in later editions of the book, as he says it "assumes the waterfall model of software construction.")

The process of "zero drafting" in my work generally looks something like this:

  1. I am tasked with a hairy problem: not an intractable one, but one that feels vague and scary and without an obvious immediate plan of attack or path to resolution.
  2. I carve out an afternoon to attack the problem, where the goal is not to solve the problem but to have emerged victorious in my understanding of the problem and with a WIP PR with no plans to merge. [1] The WIP PR can have broken tests and bugs and poor documentation, but it has the broad strokes of a path forward.
  3. I then take time to carve out the obvious and safe things to extract from the WIP PR — inert database changes, necessary refactors, et cetera — and submit net-new PRs for those.
  4. Suddenly, all that is remaining is the critical path, and we can close the WIP PR in favor of something cleaner.
  5. Lo, the problem, it is solved.

The hardest part of this entire thing is the "carving out an afternoon" bit: good weeks tend to look like ones where I have three entire afternoons, and I'm a sucker for snacking.

  1. This kind of process happens for non-engineering work, too, but I digress. ↩︎

It took my brother two years to watch this from start to finish, and it reminds me more than anything of trundling through The Romance of the Three Kingdoms as a kid — a parallel that I'm sure I'm being unoriginal in drawing, but the mixture of long-form historiography, hagiography, fact and fiction, and faint sense of ennui is hard to ignore.

LotGH is a commitment. It is rarely "fun" in the way I think I would expect a traditional space opera to be fun; perhaps more pointedly, the show uses the "space" part of "space opera" hardly at all except to feint at the size and sweep of the events being depicted, and while hours are spent on space battles the tactics are almost purely naval in nature.

And, unlike Romance of the Three Kingdoms, Legend of the Galactic Heroes is not that interested in action scenes.

What the show is interested in is everything else: the nature of history, the foibles of man, the cosmic cruelty of war, the benefits and drawbacks of fascism, the benefits and drawbacks of democracy, the nuances of supply chain mechanics, the lives and deaths of heroes, the curse of memory, the light of peace.

There are parts of LotGH that do not work, or (phrased slightly differently) detract from the core of its gestalt: Terraism (a goofy plot device that serves for little else than a prevention of narrative detente), female characters who exist almost entirely to be married to high admirals. But these are quibbles, little nits to pick out of a rich and nuanced tapestry of how humanity never really changes.

Watch this show: it will be slow, and the first dozen eps are truly terribly animated, but it rewards your patience.

Highlights

Before freedom and equality, could we have bread and meat?

Having a history differentiates humans from all other living species.

Too many things have happened already / and there is too much already for us to do.

Good people, remarkable people are killed meaninglessly. That’s war, and that’s terrorism. The sin of wars and terrorism, in the end, comes down to that.

Politics always takes vengeance on those who belittle it.

A sword has no reason to exist but as a sword.

I wrote a bit in Everybody's in L.A. about the constraint of evaluating the merit of something that's in conversation with a genre you don't really care for in the first place; I don't care much for late night shows, and I care even less for kaiju movies, even with my knowledge of "Godzilla is a metaphor for the destruction and threat wrought by nuclear war" and the litany of thinkpieces therein.

The majority of the hay made about Godzilla Minus One is about two things:

  1. It's scant $15m budget;
  2. How well it works as a postwar Japan film even without the Godzilla stuff.

I can certainly agree with the former — the scale and craft of the effects were very well done — and the latter is where things lose me. The protagonist is certainly more interesting than your typical disaster film's, and the extended metaphor of survivor's guilt has more emotional resonance than more films in this genre — but come on, guys, this is still a Godzilla film, and the credulity required therein (let alone the paper-mache thinness of every other character in the script) means that I can like this relative to the Cloverfield Projects of the world, but it's still not exactly going down in my history books.

Growth covers up mistakes and success breeds hubris.

via Howard Schultz

Death isn't the only goodbye in this life.

via Frieren

A very solid little thriller that does not reach the heights Bakula did with The Parallax View, but was still a very successful example (and, in some places, inversion) of the form. Harrison Ford playing against type as a skeevy prosecutor; a proto-West Wing ensemble with a very spry-looking John Spencer and Bradley Whitford; a fun, twisty-but-not-too-twisty ending (which is a great victory for Poirot Media Literacy — if you watch the show like a murder mystery, you'll notice the loose threads!); smarmy, oily performances by Brian Dennehy and Raul Julia; there's not a lot to love (I do not think this movie will change your relationship with the world in a meaningful way), but there's a lot to like.

The anime Frieren reminds me most of is Ousama Ranking — Frieren plays with the convention and narrative arcs of JRPGs (or high fantasy in general) in much the same way that Ousama does, with the helpful (and bittersweet, and clever) guiding structure of "fifty years ago, a JRPG party defeated the Big Bad. Now three of the four members of the party are dead, because they're human, and the elf continues wandering around."

To put it bluntly, I think Frieren is a better anime, and I enjoyed it more (which is a high bar — I really liked Ousama Ranking.)

It's so deeply consistent — even in worldbuilding, even in hackneyed tournament arcs [1], even in bouncing from slice-of-life work to more serious narrative/lore-building episodes. There are common threads of softness and nostalgia (reminiscient of Haibane Renmei and Yokohama Kaidashi Kikō) that run through every episode, and what is so masterful about it is how the creators can deploy gag humor or more conventional tropes in a way that re-inforces those themes rather than undermines them (in contrast to, say, Ousama Ranking).

Frieren presents an interesting, beautiful world with characters who you quickly grow to love: I can't wait to spend more time with them soon.

  1. Lest this sounds backhanded: Frieren's tournament arc was delightful ↩︎

Canned cocktails, despite the inherit oxymoron, are on the rise. I have friends ask me which ones they should get, if any — I also have friends and family give me a good amount.

In general: they are all bad. There are exactly two exceptions: Tiptop Cocktails and Straightaway Cocktails. Both of these are delicious, well-packaged, and a gift that a cocktail-lover will probably not hate. (They are not as good as "the real thing", but they pass the bar of "can be drunk by someone who likes cocktails without that person having a conniption", which is more than can be said of Via Carota et al.)

If you really want to get a cocktail-themed gift, though, I'd recommend stuff in the following genres:

  • An amaro. Nobody has enough amari, they last forever, and they look nice on a bar.
  • Dasher bottles. (Or interesting glassware in general. Avoid bar "equipment", with the exception of dasher bottles because they're small.)
  • Bitters.

In late 2022 I had designs on starting a new little tool, pulling out some useful but gnarly code out of Buttondown for rendering tweets (and other such similar media) and productizing it. It was called Bedkit, and the splash page is still live for the next few days.

In a rare display of realism and humility, I am letting the domain name pass. Bedkit — still a good idea, in the abstract! — never came to pass. A couple reasons why:

  1. Late '22 was when Twitter announced their API pricing changes, which would have immediately disqualified the business from using tweets as the first tranche of content;
  2. I got distracted.

These things happen; I am not filled with much regret. The prospect of building a second thing seems even more daunting than it did back then — building free tools as marketing seems more approachable. Still, RIP Bedkit — I wish someone else had built you, so I could get rid of a lot of my own code.

Lots of writing this month:

And some media notes (it was a rare month where I basically only had positive things to say!):

In other, more-exciting, non-writing news:

The last two Pokemon games I played were in 2022: Pokemon Gaia (which I loved, truly loved, and felt like it delivered exactly on the promise of "the core gameplay loop that you remember and the charming hallmarks of the series that fill you nostalgia, but tuned for someone slightly better at video games than a seven year old") and Pokemon Scarlet (which was fine, but I shelved and never returned to it.)

I described Gaia as:

It felt like exactly what I wanted out of it

Conversely, Pokerogue is a game that broke my expectations wide open. It was not — is not — a dense, glimmering refinement of the original formula, but instead an evolution, sending you screaming into the twenties armed with a couple starters and a complete naivete as to how devastating a single Raticate can truly be.

There's no other way to put it: this is my favorite roguelike since Slay the Spire.

So many of the game's design decisions are not only elegant but cohesive with the series writ large: using egg moves and IVs as a metaprogression system, piggybacking off of the humongous catalog of monsters to provide variation between runs, using the points system (and even freaking Pokerus!) to incentivize players to try out new builds and compositions — and, below all of that, the substrate, a battle engine that feels comfortable and dynamic and hopelessly, hopelessly addictive.

I'm writing this on a Sunday morning, having just beaten the game after probably ~twenty attempts. My winning team of champions:

  1. Kartana
  2. Ting-Lu
  3. Chien-Pao
  4. Primarina
  5. Magnezone
  6. Archaludon

(A lot of Steel types and legendaries, I know. All of them caught on the course of the run!)

The highest praise I can give this game: I watched the final little cinematic and immediately booted up a new run. I had to — have to! — try out this Tepig I just hatched with Volt Tackle.

Unlike stories, real life, when it has passed, inclines toward obscurity, not clarity.

There’s a relatively famous line from Bezos, circulated first at YC in ’08 and then more recently by the folks at Acquired (a great pod!):

Jeff uses this analogy for AWS. He talks about European beer breweries around the turn of the 20th century.

Electricity has just been invented. This was this massive enabling technology. Breweries could now brew vastly more quantities of beer than you could before electricity. The first breweries to adopt it built their own power generators. It worked fine for a few years but it was super capital intensive.

Then the utilities companies came along and the next generation of breweries just rented the power from the utilities companies. They beat the first generation of breweries because guess what: whoever makes your electricity has no impact on how your beer tastes.

“Focus on what makes your beer taste better” is a great slogan, and the kind of MBA-ism that makes you feel slightly cleverer for having said it.

But it’s always rubbed me the wrong way, for two reasons:

  1. As far as I can tell, the metaphor has no basis in reality. There is no cohort of once-large brewery that got extinguished in the early 20th century. I’m not sure it even makes sense if you think about it for more than a few minutes.
  2. Bezos is deploying this metaphor in the context of AWS, a service entirely orthogonal to the “beer” that Amazon was selling in the early aughts prior to EC2.

Metaphors can entertain and enchant but they’re rarely accurate. Bezos wanted to sell people on using AWS; he told a famous story about why they should use AWS, and the lesson there is in sales, not operations.

Software businesses are not much like turn-of-the-century Bavarian breweries; you may want to use AWS, you may want to use Clerk and Vanta and outsource large swathes of your business and you may be correct, but Seattle and Portland and San Diego and Richmond are littered with breweries who made a great product but couldn’t quite get the business part figured out, either.

We’re all boys grown tall.

via Justice Hugo Black

I want the last check I write to bounce.

via Ocean's Twelve

I rant often a lot about how it's important to judge any sort of artistic endeavor first and foremost by how much it accomplishes what it sets out to do. This is a rubric that leads me to tend to pan, or at least be overly critical with, works that are loving pastiches (The Late Show, The Club Dumas); it also makes it harder for me to try and objectively analyze works that are so loudly in conversation with subject matter of which I'm ignorant.

Everybody's in L.A. is – I don't know if you want to call it a critique, or deconstruction, or reinvention, or whatever — a late-night show, and I've never been a late night show person. They're not for me in much the same way podcasts aren't for other people; we all have our comforts, and are not obliged to be warmed by others'.

There are many things that, I think, structurally, grate me about late night shows: the forced topicality, the sterility, the administrative requirement to make humor as broad and digestible as possible, the luxurious sluggishness. And Everybody's in L.A. feels like a shot of Fernet, designed to keep the "good parts" (a sense of coziness and camarederie, the loving languor, the fun that comes from interactions between famous people who otherwise wouldn't be caught dead in the same set as each other) and jettison the rest in favor of chaos and good-natured bonhomie.

A lot of the pre-canned segments did not land at all; a lot of the more famous guests were funny less on terms of their own presence and more just because seeing Jerry Seinfeld react to a Saymo is entertaining in a way that does not sustain for more than six episodes. But I had a great time with this, and I think Mulaney did accomplish what he set out to create: a love letter to late night (and, yes, to Los Angeles — the cards were fantastic) that was entertaining in its own right.

Bucket list item checked off — talking with the two people who taught me more about Django than anyone else! This was a lot of fun.

Listen to the episode
Previous page · Next page

Lightning bolt
About the author

I'm Justin Duke — a software engineer, writer, and founder. I currently work as the CEO of Buttondown, the best way to start and grow your newsletter, and as a partner at Third South Capital.

Lightning bolt
Greatest hits

Lightning bolt
Elsewhere

Lightning bolt
Don't miss the next essay

Get a monthly roundup of everything I've written: no ads, no nonsense.