Having recently done poorly on a few system design interviews, I'm kicking myself for not doing the sensible thing and doing my research first. System design is something which is inherently hard to get to from hobby programming, and there's a big difference between something that runs once in a Jupyter Notebook to something that reliably serves millions of requests daily worldwide.

The contents cover the high level system design of web services you know and possibly love, like Twitter, Whatsapp, Youtube, and Tiktok. There's some decent discussion of how to calculate resource requirements based on users and data per transaction. There are also some very good pragmatic points about the importance of sharding and caching in systems which tend to be governed by power laws. Basically, recent and popular content is going to be served way more often that the major of uploads, so for read-heavy systems, you want to find a way to spread popular content across different shards, rather than overloading one server with your biggest accounts, and to keep popular content cached in Redis rather than a conventional disk-based database.

Conversely, there's also a little bit of "you get what you pay for", and while the models are useful explanations, there's not much theory about why you should make the choices that you're making, or how to think systematically about analysis of the system as a whole.

One of the unfortunate facts about computer systems is that there is far more legacy code than their is modern code. Systems are built in layers of stanky hacks on stanky hacks, temporary fixes made ages ago in obsolete languages and paradigms where the original developers have long departed and the original context has been forgotten. To share one of my own stories, I had a process fail a third of the time, because it was (my) contemporary Python calling a .NET framework which was emulating an IBM 3270 terminal to connect to a mainframe called The Core, and somewhere in there the ability to handle lower case letters got lost. "Kill it with fire" is both the title of the book, and most developer's reaction to being asked to work with legacy systems, but in the real world we can't reach for the napalm. We have to find a way to live with creaky systems.


IBM 3270, the finest in modern human-computer interaction peripherals

Bellotti is trained as an anthropologist and has worked on both major open source projects like OAuth and with the United States Digital Service, and her advice is an extension of the dictum that computers are programmed and used by human beings.

This has a few major implications. The first one is that a successful modernization is one which minimizes disruptions to existing users. Try to fit within existing mental models and interfaces, while swapping out parts of a system which are causing unacceptable delays.

The second implication is that while there is no single best architecture, there are patterns which developers like to work on. Mostly whatever is new and trendy, or systems of abstraction rather than specific use cases. If people are rewarded for shipping new code, they'll ship code and won't focus on maintenance. Social rewards are more powerful than organizational rewards.

The third implication is that what makes legacy systems hard isn't bad technology, but bad management. Migrating a complex system, with lots of stakeholders, under and undocumented features, and even bugs that have become features, is an exercise in frustration. Leadership had to build morale and keep momentum up by focusing on alignment and productive questions that free the people doing the work to do the work, rather than run in circles on firedrills.

I'd have preferred a few more war stories and specific technical traps to be aware of, but this is definitely an interesting and useful book.

Jingo marks a shift in the Watch series between Vimes-the-cynical-cop and Vimes-the-reluctant diplomat. When a previous unknown island called Leshp rises exactly between Ankh-Morpork and the orientalist expy of Klatch, both sides sense an opportunity for a short victorious war, and it is up to Vimes and Vetinari to keep the peace however they can.

Much of the book is solid Discworld, but there are a few touches I liked. Vimes' imp-powered Disorganizer, playing out the fatal appointments from the sack of Ankh-Morpork as the two sides are about to come to blows is a particularly great scene. The humanization of all involved, and the frank anti-militarism are nice. I even enjoyed the Colin-Nobby-Vetinari B plot.

BART is a whiggish history of transit system of the same name by BART's longtime director of public relations. As such it reads more like a glossy pamphlet than any kind of critical review. The basic problem of any transit system is that the best time to build it is well-before the region strangles itself in construction, traffic, and NIMBY interest groups. But suburban and rural voters don't see the benefit of a massively expensive and socially transformative infrastructure project that won't pay off for decades, so the systems mostly don't get built.


Maximalist BART from twitter user Jackson Mills

From its origins in the 1960s, BART was the first new transit system in nearly 50 years, and the basics were laid without the Federal funding that supported the DC Metro and Atlanta's MARTA. The system was hugely ambitious, with space-age automation and a key tunnel under the bay, while also being much less than it could have been. Planned expansion to Marin and San Mateo was off the table, costs kept ballooning, and the automated control system and high tech cars took years to shake-down all the bugs. Every expansion since then has been a complicated mess of local, state, and Federal deals to get the money in order before the costs to build another mile triple.

As a regular BART user, it handles my commute pretty well, and much more efficiently than the other options, car, bus, walking, mule train, etc. would. Yet, whenever I ask it to do anything other than take me from the outer part of San Francisco to downtown at 9 or 5, I find myself waiting on a platform for a half-hour.

Well, at least I brought a book.

Court of Blades is a Forged in the Dark game for those who like their darkness a little less dark, with a tone of swashbuckling romance rather than gritty crime. It is a stylish and eminently playable game, and one which hews very close to base Blades. You play the movers and shakers of the smallest of the six noble houses of Ilrien, and by daring intrigues you will rise. Or if you fail, you will bring shame to your entire house. Be brave, be passionate, and above all else, don't get killed.


POV: You've caused a scandal at the ballroom. Netflix's Bridgerton.

First the seeking. Ilrien is finely gamable setting, where the Six Noble Houses jockey for position above a stratum of happy workers, devious criminals, glamorous magic, deadly mercenaries, and some mostly-locked away undying horrors. There's all the glamorous backdrop you'll need for your story. This isn't really my genre, but everything seems to work okay.

Mechanically, Court of Blades is an almost one-for-one match with base Blades. The most significant change is the faction subsystem has been streamlined, with turf replaced with spheres of influence. Each season consists of three linked errands, where you'll try to advance your house's objective and deal with surprises, giving you a roll to advance ticks on a clock. As someone who's thought that BitD downtime is too complicated as is, this might be a nice lighter version. The other major mechanical innovation is a Paramor system, where you use downtime actions to court another person who can help you out, but also will make their own demands. Oh, and playbooks have bond statements connecting everyone in the party, which could become a lot.

I'll say upfront that my taste in RPGs, and FitD in particular, is oriented towards mechanical innovation. The d6 pool, mixed success, and position-effect-clock mechanic is solid, so when I add a game, I look to see how it builds on that foundation. Wicked Ones, Band of Blades, and Neon Black all did something new and unique with the base system, and Court of Blades does not. And this is a shame, because the core tension in the source fiction is between Obligation and Passion, and the ways that becoming a slave to either wrecks a person. Don't get me wrong, Court of Blades is very good, but I wish it were great.

"I read an article that men don't read novels anymore."
"An article?"
"Yeah, some chick bitching to GQ about her boyfriend reading Kim Stanley Robinson, Ursula K LeGuin, and history about CIA death squads instead of novels."
"Your girlfriend?"
"Naw. But that guy has about the same taste in books as I do."
"Weird"
"Yeah, weird."
"So this Leonard?"
"Well, I figured why not read some contemporary literary fiction, but if I'm going to do it, I'm going to do it like a man. You know, terse stories about down-and-out lawmen, criminals, decent folks in a wicked world, or maybe not so decent folks in world with no illusions. Classic America."
"What did you think?"
"It's pretty good! Real snap-crack in the dialog, just like that "Guide to Dialogue" book a while back was talking about."
"Hough"
"Yeah, Hough. Anyways, these are some great stories. Definitely a little front loaded, “When the Women Come Out to Dance”, “Hanging Out at the Buena Vista”, and "Sparks" are all incredible. Apparently the other stories link into the novels, but I haven't read them. But hell, those three are worth the price of admission, and the rest are solid tales. Maybe I should read more novels."
"You think Leonard was who that chick was talking about."
"You think I give a fuck?"

Reading Capitalist Realism is akin to walking unhappily through the storm of life with only one word for "water", and then hearing seventy finely tuned descriptions for the precise version of sideways-sleeting bullshit that you experience.

The core of Fisher's book is the ideology called capitalism, so successful and omnipresent that it cannot see itself. Having conquered the world with science and imperialism, socialism via force of arms, diplomacy, and public-private partnerships, and finally the last sanctums of labor, culture, and personality via the multifarious tools of advertising and public relations, capitalism is left with nothing further to conquer. And for a predator who needs continual motion and prey, that means its own destruction.

Fisher flashes through academic theory, psychoanalysis, media critiques, political analysis, his own life experiences. It's these last where Fisher is most trenchant, precisely describing the ennui of interaction with a consumerist-entertainment matrix which can fulfill any expected need instantaneously, but as soon as an issue becomes complex or worse, unprofitable, casts you in a Kafka-esque decentralized bureaucracy where no one has the power to do anything. Perhaps the best bit is a process of teaching self-assessment, which Fisher reads as a Maoist confession remade in the language of McKinsey Consultants. The goal is to have good numbers, but not too good, and make sure to include some failure so we know you're taking the process seriously, but don't worry, you won't be required to fix them and the won't be used against you*.

Under Fisher's analysis, culture is no longer a domain of communication, but one of endless formless "content", marked most notably by a total absence of meaning. The mental health crisis, the avalanche of discontent, is an individual manifestation of a social patriarchy that cannot set boundaries yet constantly restricts behavior, and a nursemaid state incapable of caring, only of finer gradations of punishment. The centers of power are absence and impotent, incapable of saving this world or building a better one. We see in the failures of Capitalism to make even modest political progress, in its inability to provision real social good like housing, healthcare, and education, even as the screens get better, the entertainment more entertaining, and the data-based monitoring more uncannily invasive.

The thesis is dazzling and dense, but I feel like it works in this case because the details aren't important. What is important is the rage, the despair, the grief, the bone-deep sense that we have to seize the Barricades of Imagination, even if we cannot yet see what they are, if we are to survive.

WE ARE INSIDE THE MACHINE. THE MACHINE IS KILLING US. THE MACHINE IS SHAKING APART.

*offer void if we do decide to use them against you.

Everyone knows the world is falling apart.

We just don’t like to talk about it.


Blade Runner skies over San Francisco. California wildfires September 2020

But what if we did talk about climate change? What if we stopped being paralyzed by the immensity of the problem, the multifaceted slow-motion catastrophe that everyone is responsible for, and nobody is accountable for? Decarbonization and any associated global political change are famously wicked problems, with manifold uncertainties and high costs, but climate change and how we deal with it is the story of the 21st century. And for all that climate change is the story, there are not enough stories about climate change.

Most climate fiction remains resolutely Delugist in orientation. In Oryx and Crake, The Windup Girl, or Don’t Look Up the core message is the sins of industrial civilization will finally overwhelm a weak and corrupt society. Out of the seeds of the devastation, a handful of survivors will rebuild, climate absolution paid for in gigadeaths. The kind of story about climate change is an apocalyptic fantasy, a re-inscription of Christian mythology into a modern context. It can make for a fine horror story, but it's facile, boring, and it’s a lie. If we’ve learned one thing about the apocalypse the past few years, there will not be angelic trumpets and letters of fire in the sky. You’ll just have to keep going to work, even as the sky rains blood.

Our Shared Storm is climate fiction done right. It’s a serious piece of futurism inspired by the latest IPCC scenarios and ethnography at COP24 (Conference of the Parties, the major UN climate conference) in Krakow in 2018. It’s also such good storytelling that I could not put it down, and stayed up far past a sensible bedtime to finish it in one gulp.

The stories are centered around COP60 in 2054, held in Buenos Aires. With a bit of literary and and futurological sleight of hand, Hudson holds constant the basic plot of the conference being hit with a torrential “neverstorm”, and the characters of Noah, Saga, Luis, and Diya, and let’s the world shift around them to show the five official IPCC Shared Socioeconomic Pathway scenarios. I’ll confess to skepticism reading the introduction, but this is very much not the same story five times. Shifts in point of view and central dramatic tension along with the state of the world, make each timeline its own unique experience, and I was excited to see what in the characters remained fixed and what changed.

The stories trace an arc, from the business as usual scenario of diplomatic horse-trading over reparations for losses against investments in the future, to a wild party of fossil-fueled development and venture-disaster-capitalism, to stark inequality between those deemed useful and worthy of survival and the restive excess population, to a world of conflict and collapse, where a handful of scientists attempt to record the climate catastrophe like monks protecting the treasures of antiquity from Visigoths, to finally a sustainable utopia, where the major choice is how much to invest in decarbonization how quickly. It’s a tour from a world very much like our own, to ones much worse, and finally one where things are, well, not perfect, but apocalypse canceled. And even in the scenarios where things are bad, life goes on. “The collapse” can only been in retrospect. For those living in it, it’s just another day.

Our Shared Storm makes bold promises in the introduction, and accomplishes all of them with verve. First, these are good short stories, some of the more well-crafted speculative fiction I’ve read in quite a while. An academic work always has the threat of being too didactic, getting lost in abstract ideas and teachable moments, rather than story-telling. Our Shared Storm never loses the bubble that these stories are about human beings, not planetary systems. Second, this is a serious piece of scholarship, grounded in the best estimates we have, which successfully translates the bland bureaucratese of an IPCC report into the richly textured sensation of the future. And third, this is a methodological advance in narrative foresight, with the conceit of the same characters and events but different settings. I’ve done work in this field (Burnam-Fink, 2015, “Creating narrative scenarios: Science fiction prototype at Emerge”, Futures), and there’s a lot of theory but precious little successful practice.

Our Shared Storm is an impressive debut: Provocative, imaginative, and even inspiring. Hudson is a talent to watch.

Novel Without a Name is a cynical, bitter, and clear-eyed account of war from the Vietnamese side. The narrator, Quan, is a company commander, responsible for the lives of 100 men, and particularly responsible for the life of his childhood friend Bien, who has gone mad.

This is a novel about war, but it's not one about combat. Battles are treated perfunctorily, a few sentences describing horrific causalities. What concerns Quan are his peregrinations through the war zone, and the soldier's obsession with food, distance, and a place to sleep at night. The war ages soldiers before their time, casts them adrift from both the rhythms of their traditional villages and exiles them from the benefits of modernity.

Huong was a North Vietnamese soldier for the period described, and this book has the bronze ring of authentic truth, emotional without being sentimental, lyrical in its description of the countryside and the privation of the soldiers, and bravely truthful about the broken glories of liberation.

This book isn't about how to blow up a pipeline. Obviously, that would lead to [REDACTED], but more about why to blow up a pipeline. Malm poses an interesting question: If we take the climate science seriously, and that we are heading towards a dark Here Be Dragons future of 600+ ppm, mass extinctions, heavy weather, refugee crises, and desperate struggles over the last scraps, why has the environmental movement been so notably ineffective in enforcing policy to keep carbon in the ground? Isn't it time to do more? Isn't it time to quit talking, and start [REDACTING]?

Of course, an energy transition is complex and expensive. Cheap fossil energy has its tendrils everywhere. We all benefit from it. As an ecomodernist, I'm all in favor of cheap, clean and accessible energy. And even though we've seen major progress in solar, wind, storage, and electric vehicles, it simply isn't enough. The clean energy future is smaller than the dirty energy present of new pipelines, new coal plants, new SUVs, let alone the dead fossil weight of history.

The climate movement is notably both because of its abstraction and idealism, and how it has been neutered by an ideological and tactical commitment to absolute non-violence. In thrall to Chenoweth's thesis that non-violent movements are more likely to succeed, and terrified an alienating any part of a democratic coalition, protests are symbolic, mostly delaying rather destroying. A complete deconstruction of the non-violence thesis is a project for another book, but her sample is obviously flawed and incomplete, categorizing movements which did involve violence as non-violence, and somehow ignoring major successes, like the violent movement which toppled governments in Russia, China, and Germany in the 20th century. Seeing that Chenowith is affiliated with Harvard Kennedy School makes me deeply suspicious that the non-violence fetish is literally a CIA op to make dissent easier to manage. Successful movements such as the suffragettes included deliberate violence against property at a scale between vandalism and arson. While deliberate violence often fails, the impromptu mass of a crowd deciding that this will not stand has toppled dictators.

As Malm points out, actual sabotage from the level of letting the air out of SUV tires, to burning down gas stations, to striking vulnerable fixed installations, has rarely been tried, and when it does, it provokes a response all out of proportion to the seeming irritation. If not killing people remains a key moral principle, attacks on Middle Eastern oil infrastructure by violent terrorist groups, political rather than ecological strikes, have very low casualties. Jessica Reznicek and Ruby Montoya carried out multiple attacks against the Dakota Access Pipeline, causing millions of dollars in damage and earning multiple years in prison.

If I may indulge in a moment of conspiracy, the media is silent about Malm's thesis because it would work. Actual, capital-P Power, the people the rule the world, know how much of their strength is tied up with masses sated on cheap energy, and they believe they can ride this out in the New Zealand Apocalypse Bunkers. Fossil fuel infrastructure is omnipresent, impossible to secure, and so very very flammable. At this moment, with gasoline hitting $8 a gallon in the US and Europe's painful lesson about Russian natural gas dependency being to switch from pipelines to LNG terminals so that they're still dependent on fossil fuels, but no Russia, it seems like, well, maybe this is a chance for a few brave people to [REDACTED].