Full Width [alt+shift+f] FOCUS MODE Shortcuts [alt+shift+k]
Sign Up [alt+shift+s] Log In [alt+shift+l]
83
Recently, a reader wrote to me and asked about my writing process and burnout. They had an image in their head that I could sit down at a computer and type up a full post on a given topic, but were unsure if that's the right approach when they start blogging. And they were concerned about how to keep things sustainable for themselves, too. I started to write back to them, but decided to ship my answer for everyone else, as well. Now, to be clear: this is my writing process. There are many other very valid approaches, such as what Gabriella wrote. But I think outlining my process here will at least help break some of the mystique around writing, because I certainly do not just sit down and plunk out a post. Well, it often looks like that if you're an outside observer1. But that's only what's visible. How I write The very first piece of any blog post for me is an idea, some inspiration, something I want to share. These ideas come from a variety of places. Some of them are from...
7 months ago

Comments

Improve your reading experience

Logged in users get linked directly to articles resulting in a better reading experience. Please login for free, it takes less than 1 minute.

More from ntietz.com blog - technically a blog

Visualizing distributions with pepperoni pizza (and javascript)

There's a pizza shop near me that serves a normal pizza. I mean, they distribute the toppings in a normal way. They're not uniform at all. The toppings are random, but not the way I want. The colloquial understanding of "random" is kind of the Platonic ideal of a pizza: slightly chaotic but things are more or less spread out over the whole piece in a regular way. If you take a slice you'll get more of less the same amount of pepperoni as any other slice. And every bite will have roughly the same amount of pepperoni as every other bite. I think it would look something like this. Regenerate this pie! This pizza to me is pretty much the canonical mental pizza. It looks pretty random, but you know what you're gonna get. And it is random! Here's how we made it, with the visualiztion part glossed over. First, we make a helper function, since Math.random() gives us values from 0 to 1, but we want values from -1 to 1. // return a uniform random value in [-1, 1] function randUniform() { return 2*Math.random() - 1; } Then, we make a simple function that gives us the coordinates of where to put a pepperoni piece, from the uniform distribution. function uniformPepperoniPosition() { var [centerX, centerY, radius] = pepperoniBounds(); let x = radius*2; let y = radius*2; while (x**2 + y**2 >= radius**2) { x = randUniform() * radius; y = randUniform() * radius; } return [x+centerX, y+centerY]; } And we cap it off with placing 300 fresh pieces of pepperoni on this pie, before we send it into the oven. (It's an outrageous amount of very small pepperoni, chosen in both axes for ease of visualizing the distribution rather than realism.) function drawUniformPizza() { drawBackground(); drawPizzaCrust(); drawCheese(); var [_, _, radius] = pepperoniBounds(); for (let p = 0; p < 300; p++) { let [x,y] = uniformPepperoniPosition(); drawPepperoni(x, y); } } But it's not what my local pizza shop's pizza's look like. That's because they're not using the same probability distribution. This pizza is using a uniform distribution. That means that for any given pepperoni, every single position on the pizza is equally likely for it to land on. These are normal pizzas We are using a uniform distribution here, but there are plenty of other distributions we could use as well. One of the other most familiar distributions is normal distribution. This is the distribution that has the normal "bell curve" that we are used to seeing. And this is probably what people are talking about most of the time when they talk about how many standard deviations something is away from something else. So what would it look like if we did a normal distribution on a pizza? The very first thing we need to answer that is a way of getting the values from the normal distribution. This isn't included with JavaScript by default, but we can implement it pretty simply using the Box-Muller transform. This might be a scary name, but it's really easy to use. Is a way of generating numbers in the normal distribution using number sampled from the uniform distribution. We can implement it like this: function randNormal() { let theta = 2*Math.PI*Math.random(); let r = Math.sqrt(-2*Math.log(Math.random())); let x = r * Math.cos(theta); let y = r * Math.sin(theta); return [x,y]; } Then we can make a pretty simple function again which gives us coordinates for where to place pepperoni in this distribution. The only little weird thing here is that I scale the radius down by a factor of 3. Without this, the pizza ends up a little bit indistinguishable from the uniform distribution, but the scaling is arbitrary and you can do whatever you want. function normalPepperoniPosition() { var [centerX, centerY, radius] = pepperoniBounds(); let x = radius*2; let y = radius*2; while (x**2 + y**2 >= radius**2) { [x,y] = randNormal(); x = x * radius/3; y = y * radius/3; } return [x + centerX, y + centerX]; } And then once again we cap it off with a 300 piece pepperoni pizza. function drawNormalPizza() { drawBackground(); drawPizzaCrust(); drawCheese(); for (let p = 0; p < 300; p++) { let [x,y] = normalPepperoniPosition(); drawPepperoni(x, y); } } Regenerate this pie! Ouch. It's not my platonic ideal of a pizza, that's for sure. It also looks closer to the pizzas my local shop serves, but it's missing something... See, this one is centered around, you know, the center. Theirs are not that. They're more chaotic with a few handfuls of toppings. What if we did the normal distributions, but multiple times, with different centers? First we have to update our position picking function to accept a center for the cluster. We'll do this by passing in the center and generating coordinates around those, while still checking that we're within the bounds of the circle formed by the crust of the pizza. function normal(cx, cy) { var [centerX, centerY, radius] = pepperoniBounds(); let x = radius*2; let y = radius*2; while ((x-centerX)**2 + (y-centerY)**2 >= radius**2) { [x,y] = randNormal(); x = cx + x * radius/3; y = cy + y * radius/3; } return [x, y]; } And then instead of one single loop for all 300 pieces, we can do 3 loops of 100 pieces each, with different (randomly chosen) centers for each. function drawClusterPizza() { const settings = initializeCanvas("drawing-3"); drawBackground(settings); drawPizzaCrust(settings); drawCheese(settings); var [centerX, centerY, radius] = pepperoniBounds(settings); for (let c = 0; c < 3; c++) { let [cx, cy] = uniform(settings, centerX, centerY, 1); console.log(cx, cy); for (let p = 0; p < 100; p++) { let [x, y] = normal(settings, cx, cy, 4); drawPepperoni(settings, x, y); } } } Regenerate this pie! That looks more like it. Well, probably. This one is more chaotic, and sometimes things work out okay, but other times they're weird. Just like the real pizzas. Click that "regenerate" button a few times to see a few examples! Okay, but when do you want one? So, this is all great. But, when would we want this? I mean, first of all, boring. We don't need a reason except that it's fun! But, there's one valid use case that a medical professional and I came up with[1]: hot honey[2]. The ideal pepperoni pizza just might be one that has uniformly distributed pepperoni with normally distributed hot honey or hot sauce. You'd start with more intense heat, then it would taper off as you go toward the crust, so you maintain the heat without getting overwhelmed by it. The room to play here is endless! We can come up with a lot of other fun distributions and map them in similar ways. Unfortunately, we probably can't make a Poisson pizza, since that's a distribution for discrete variables. I really do talk about weird things with all my medical providers. And everyone else I meet. I don't know, life's too short to go "hey, this is a professional interaction, let's not chatter on and on about whatever irrelevant topic is on our mind." ↩ The pizza topping, not my pet name. ↩

a week ago 17 votes
Covers as a way of learning music and code

When you're just getting started with music, you have so many skills to learn. You have to be able to play your instrument and express yourself through it. You need to know the style you're playing, and its idioms and conventions. You may want to record your music, and need all the skills that come along with it. Music is, mostly, subjective: there's not an objective right or wrong way to do things. And that can make it really hard! Each of these skills is then couched in this subjectivity of trying to see if it's good enough. Playing someone else's music, making a cover, is great because it can make it objective. It gives you something to check against. When you're playing your own music, you're in charge of the entire thing. You didn't play a wrong note, because, well, you've just changed the piece! But when you play someone else's music, now there's an original and you can try to get as close to it as possible. Recreating it gives you a lot of practice in figuring out what someone did and how they did it. It also lets you peek into why they did it. Maybe a particular chord voicing is hard for you to play. Okay, let's simplify it and play an easier voicing. How does it sound now? How does it sound with the harder one? Play around with those differences and you start to see the why behind it all. * * * The same thing holds true for programming. One of my friends is a C++ programmer[1] and he was telling me about how he learned C++ and data structures really well early on: He reimplemented parts of the Boost library. This code makes heavy use of templates, a hard thing in C++. And it provides fundamental data structures with robust implementations and good performance[2]. What he would do is look at the library and pick a slice of it to implement. He'd look at what the API for it is, how it was implemented, what it was doing under the hood. Then he'd go ahead and try to do it himself, without any copy-pasting and without real-time copying from the other screen. Sometimes, he'd run into things which didn't make sense. Why is this a doubly-linked list here, when it seems a singly-linked list would do just fine? And in those moments, if you can't find a reason? You get to go down that path, make it the singly-linked version, and then find out later: oh, ohhh. Ohhhh, they did that for a reason. It lets you run into some of the hard problems, grapple with them, and understand why the original was written how it was. You get to study with some really strong programmers, by proxy via their codebase. Their code is your tutor and your guide for understanding how to write similar things in the future. * * * There's a lot of judgment out there about doing original works. This kind of judgment of covers and of reimplementing things that already exist, just to learn. So many people have internalized this, and I've heard countless times "I want to make a new project, but everything I think of, someone else has already done!" And to that, I say: do it anyway[3]. If someone else has done it, that's great. That means that you had an idea so good that someone else thought it was a good idea, too. And that means that, because someone else has done it, you have a reference now. You can compare notes, and you can see how they did it, and you can learn. I'm a recovering C++ programmer myself, and had some unpleasant experiences associated with the language. This friend is a game developer, and his industry is one where C++ makes a lot of sense to use because of the built-up code around it. ↩ He said they're not perfect, but that they're really good and solid and you know a lot of people thought for a long time about how to do them. You get to follow in their footsteps and benefit from all that hard thinking time. ↩ But: you must always give credit when you are using someone else's work. If you're reimplementing someone else's library, or covering someone's song, don't claim it's your own original invention. ↩

a month ago 32 votes
That boolean should probably be something else

One of the first types we learn about is the boolean. It's pretty natural to use, because boolean logic underpins much of modern computing. And yet, it's one of the types we should probably be using a lot less of. In almost every single instance when you use a boolean, it should be something else. The trick is figuring out what "something else" is. Doing this is worth the effort. It tells you a lot about your system, and it will improve your design (even if you end up using a boolean). There are a few possible types that come up often, hiding as booleans. Let's take a look at each of these, as well as the case where using a boolean does make sense. This isn't exhaustive—[1]there are surely other types that can make sense, too. Datetimes A lot of boolean data is representing a temporal event having happened. For example, websites often have you confirm your email. This may be stored as a boolean column, is_confirmed, in the database. It makes a lot of sense. But, you're throwing away data: when the confirmation happened. You can instead store when the user confirmed their email in a nullable column. You can still get the same information by checking whether the column is null. But you also get richer data for other purposes. Maybe you find out down the road that there was a bug in your confirmation process. You can use these timestamps to check which users would be affected by that, based on when their confirmation was stored. This is the one I've seen discussed the most of all these. We run into it with almost every database we design, after all. You can detect it by asking if an action has to occur for the boolean to change values, and if values can only change one time. If you have both of these, then it really looks like it is a datetime being transformed into a boolean. Store the datetime! Enums Much of the remaining boolean data indicates either what type something is, or its status. Is a user an admin or not? Check the is_admin column! Did that job fail? Check the failed column! Is the user allowed to take this action? Return a boolean for that, yes or no! These usually make more sense as an enum. Consider the admin case: this is really a user role, and you should have an enum for it. If it's a boolean, you're going to eventually need more columns, and you'll keep adding on other statuses. Oh, we had users and admins, but now we also need guest users and we need super-admins. With an enum, you can add those easily. enum UserRole { User, Admin, Guest, SuperAdmin, } And then you can usually use your tooling to make sure that all the new cases are covered in your code. With a boolean, you have to add more booleans, and then you have to make sure you find all the places where the old booleans were used and make sure they handle these new cases, too. Enums help you avoid these bugs. Job status is one that's pretty clearly an enum as well. If you use booleans, you'll have is_failed, is_started, is_queued, and on and on. Or you could just have one single field, status, which is an enum with the various statuses. (Note, though, that you probably do want timestamp fields for each of these events—but you're still best having the status stored explicitly as well.) This begins to resemble a state machine once you store the status, and it means that you can make much cleaner code and analyze things along state transition lines. And it's not just for storing in a database, either. If you're checking a user's permissions, you often return a boolean for that. fn check_permissions(user: User) -> bool { false // no one is allowed to do anything i guess } In this case, true means the user can do it and false means they can't. Usually. I think. But you can really start to have doubts here, and with any boolean, because the application logic meaning of the value cannot be inferred from the type. Instead, this can be represented as an enum, even when there are just two choices. enum PermissionCheck { Allowed, NotPermitted(reason: String), } As a bonus, though, if you use an enum? You can end up with richer information, like returning a reason for a permission check failing. And you are safe for future expansions of the enum, just like with roles. You can detect when something should be an enum a proliferation of booleans which are mutually exclusive or depend on one another. You'll see multiple columns which are all changed at the same time. Or you'll see a boolean which is returned and used for a long time. It's important to use enums here to keep your program maintainable and understandable. Conditionals But when should we use a boolean? I've mainly run into one case where it makes sense: when you're (temporarily) storing the result of a conditional expression for evaluation. This is in some ways an optimization, either for the computer (reuse a variable[2]) or for the programmer (make it more comprehensible by giving a name to a big conditional) by storing an intermediate value. Here's a contrived example where using a boolean as an intermediate value. fn calculate_user_data(user: User, records: RecordStore) { // this would be some nice long conditional, // but I don't have one. So variables it is! let user_can_do_this: bool = (a && b) && (c || !d); if user_can_do_this && records.ready() { // do the thing } else if user_can_do_this && records.in_progress() { // do another thing } else { // and something else! } } But even here in this contrived example, some enums would make more sense. I'd keep the boolean, probably, simply to give a name to what we're calculating. But the rest of it should be a match on an enum! * * * Sure, not every boolean should go away. There's probably no single rule in software design that is always true. But, we should be paying a lot more attention to booleans. They're sneaky. They feel like they make sense for our data, but they make sense for our logic. The data is usually something different underneath. By storing a boolean as our data, we're coupling that data tightly to our application logic. Instead, we should remain critical and ask what data the boolean depends on, and should we maybe store that instead? It comes easier with practice. Really, all good design does. A little thinking up front saves you a lot of time in the long run. I know that using an em-dash is treated as a sign of using LLMs. LLMs are never used for my writing. I just really like em-dashes and have a dedicated key for them on one of my keyboard layers. ↩ This one is probably best left to the compiler. ↩

2 months ago 38 votes
Proving that every program halts

One of the best known hard problems in computer science is the halting problem. In fact, it's widely thought[1] that you cannot write a program that will, for any arbitrary program as input, tell you correctly whether or not it will terminate. This is written from the framing of computers, though: can we do better with a human in the loop? It turns out, we can. And we can use a method that's generalizable, which many people can follow for many problems. Not everyone can use the method, which you'll see why in a bit. But lots of people can apply this proof technique. Let's get started. * * * We'll start by formalizing what we're talking about, just a little bit. I'm not going to give the full formal proof—that will be reserved for when this is submitted to a prestigious conference next year. We will call the set of all programs P. We want to answer, for any p in P, whether or not p will eventually halt. We will call this h(p) and h(p) = true if p eventually finished and false otherwise. Actually, scratch that. Let's simplify it and just say that yes, every program does halt eventually, so h(p) = true for all p. That makes our lives easier. Now we need to get from our starting assumptions, the world of logic we live in, to the truth of our statement. We'll call our goal, that h(p) = true for all p, the statement H. Now let's start with some facts. Fact one: I think it's always an appropriate time to play the saxophone. *honk*! Fact two: My wife thinks that it's sometimes inappropriate to play the saxophone, such as when it's "time for bed" or "I was in the middle of a sentence![2] We'll give the statement "It's always an appropriate time to play the saxophone" the name A. We know that I believe A is true. And my wife believes that A is false. So now we run into the snag: Fact three: The wife is always right. This is a truism in American culture, useful for settling debates. It's also useful here for solving major problems in computer science because, babe, we're both the wife. We're both right! So now that we're both right, we know that A and !A are both true. And we're in luck, we can apply a whole lot of fancy classical logic here. Since A and !A we know that A is true and we also know that !A is true. From A being true, we can conclude that A or H is true. And then we can apply disjunctive syllogism[3] which says that if A or H is true and !A is true, then H must be true. This makes sense, because if you've excluded one possibility then the other must be true. And we do have !A, so that means: H is true! There we have it. We've proved our proposition, H, which says that for any program p, p will eventually halt. The previous logic is, mostly, sound. It uses the principle of explosion, though I prefer to call it "proof by married lesbian." * * * Of course, we know that this is wrong. It falls apart with our assumptions. We built the system on contradictory assumptions to begin with, and this is something we avoid in logic[4]. If we allow contradictions, then we can prove truly anything. I could have also proved (by married lesbian) that no program will terminate. This has been a silly traipse through logic. If you want a good journey through logic, I'd recommend Hillel Wayne's Logic for Programmers. I'm sure that, after reading it, you'll find absolutely no flaws in my logic here. After all, I'm the wife, so I'm always right. It's widely thought because it's true, but we don't have to let that keep us from a good time. ↩ I fact checked this with her, and she does indeed hold this belief. ↩ I had to look this up, my uni logic class was a long time ago. ↩ The real conclusion to draw is that, because of proof by contradiction, it's certainly not true that the wife is always right. Proved that one via married lesbians having arguments. Or maybe gay relationships are always magical and happy and everyone lives happily ever after, who knows. ↩

2 months ago 39 votes
Taking a break

I've been publishing at least one blog post every week on this blog for about 2.5 years. I kept it up even when I was very sick last year with Lyme disease. It's time for me to take a break and reset. This is the right time, because the world is very difficult for me to move through right now and I'm just burnt out. I need to focus my energy on things that give me energy and right now, that's not writing and that's not tech. I'll come back to this, and it might look a little different. This is my last post for at least a month. It might be longer, if I still need more time, but I won't return before the end of May. I know I need at least that long to heal, and I also need that time to focus on music. I plan to play a set at West Philly Porchfest, so this whole month I'll be prepping that set. If you want to follow along with my music, you can find it on my bandcamp (only one track, but I'll post demos of the others that I prepare for Porchfest as they come together). And if you want to reach out, my inbox is open. Be kind to yourself. Stay well, drink some water. See you in a while.

3 months ago 31 votes

More in programming

Why Amateur Radio

I always had a diffuse idea of why people are spending so much time and money on amateur radio. Once I got my license and started to amass radios myself, it became more clear.

yesterday 2 votes
strongly typed?

What does it mean when someone writes that a programming language is “strongly typed”? I’ve known for many years that “strongly typed” is a poorly-defined term. Recently I was prompted on Lobsters to explain why it’s hard to understand what someone means when they use the phrase. I came up with more than five meanings! how strong? The various meanings of “strongly typed” are not clearly yes-or-no. Some developers like to argue that these kinds of integrity checks must be completely perfect or else they are entirely worthless. Charitably (it took me a while to think of a polite way to phrase this), that betrays a lack of engineering maturity. Software engineers, like any engineers, have to create working systems from imperfect materials. To do so, we must understand what guarantees we can rely on, where our mistakes can be caught early, where we need to establish processes to catch mistakes, how we can control the consequences of our mistakes, and how to remediate when somethng breaks because of a mistake that wasn’t caught. strong how? So, what are the ways that a programming language can be strongly or weakly typed? In what ways are real programming languages “mid”? Statically typed as opposed to dynamically typed? Many languages have a mixture of the two, such as run time polymorphism in OO languages (e.g. Java), or gradual type systems for dynamic languages (e.g. TypeScript). Sound static type system? It’s common for static type systems to be deliberately unsound, such as covariant subtyping in arrays or functions (Java, again). Gradual type systems migh have gaping holes for usability reasons (TypeScript, again). And some type systems might be unsound due to bugs. (There are a few of these in Rust.) Unsoundness isn’t a disaster, if a programmer won’t cause it without being aware of the risk. For example: in Lean you can write “sorry” as a kind of “to do” annotation that deliberately breaks soundness; and Idris 2 has type-in-type so it accepts Girard’s paradox. Type safe at run time? Most languages have facilities for deliberately bypassing type safety, with an “unsafe” library module or “unsafe” language features, or things that are harder to spot. It can be more or less difficult to break type safety in ways that the programmer or language designer did not intend. JavaScript and Lua are very safe, treating type safety failures as security vulnerabilities. Java and Rust have controlled unsafety. In C everything is unsafe. Fewer weird implicit coercions? There isn’t a total order here: for instance, C has implicit bool/int coercions, Rust does not; Rust has implicit deref, C does not. There’s a huge range in how much coercions are a convenience or a source of bugs. For example, the PHP and JavaScript == operators are made entirely of WAT, but at least you can use === instead. How fancy is the type system? To what degree can you model properties of your program as types? Is it convenient to parse, not validate? Is the Curry-Howard correspondance something you can put into practice? Or is it only capable of describing the physical layout of data? There are probably other meanings, e.g. I have seen “strongly typed” used to mean that runtime representations are abstract (you can’t see the underlying bytes); or in the past it sometimes meant a language with a heavy type annotation burden (as a mischaracterization of static type checking). how to type So, when you write (with your keyboard) the phrase “strongly typed”, delete it, and come up with a more precise description of what you really mean. The desiderata above are partly overlapping, sometimes partly orthogonal. Some of them you might care about, some of them not. But please try to communicate where you draw the line and how fuzzy your line is.

2 days ago 10 votes
Logical Duals in Software Engineering

(Last week's newsletter took too long and I'm way behind on Logic for Programmers revisions so short one this time.1) In classical logic, two operators F/G are duals if F(x) = !G(!x). Three examples: x || y is the same as !(!x && !y). <>P ("P is possibly true") is the same as ![]!P ("not P isn't definitely true"). some x in set: P(x) is the same as !(all x in set: !P(x)). (1) is just a version of De Morgan's Law, which we regularly use to simplify boolean expressions. (2) is important in modal logic but has niche applications in software engineering, mostly in how it powers various formal methods.2 The real interesting one is (3), the "quantifier duals". We use lots of software tools to either find a value satisfying P or check that all values satisfy P. And by duality, any tool that does one can do the other, by seeing if it fails to find/check !P. Some examples in the wild: Z3 is used to solve mathematical constraints, like "find x, where f(x) >= 0. If I want to prove a property like "f is always positive", I ask z3 to solve "find x, where !(f(x) >= 0), and see if that is unsatisfiable. This use case powers a LOT of theorem provers and formal verification tooling. Property testing checks that all inputs to a code block satisfy a property. I've used it to generate complex inputs with certain properties by checking that all inputs don't satisfy the property and reading out the test failure. Model checkers check that all behaviors of a specification satisfy a property, so we can find a behavior that reaches a goal state G by checking that all states are !G. Here's TLA+ solving a puzzle this way.3 Planners find behaviors that reach a goal state, so we can check if all behaviors satisfy a property P by asking it to reach goal state !P. The problem "find the shortest traveling salesman route" can be broken into some route: distance(route) = n and all route: !(distance(route) < n). Then a route finder can find the first, and then convert the second into a some and fail to find it, proving n is optimal. Even cooler to me is when a tool does both finding and checking, but gives them different "meanings". In SQL, some x: P(x) is true if we can query for P(x) and get a nonempty response, while all x: P(x) is true if all records satisfy the P(x) constraint. Most SQL databases allow for complex queries but not complex constraints! You got UNIQUE, NOT NULL, REFERENCES, which are fixed predicates, and CHECK, which is one-record only.4 Oh, and you got database triggers, which can run arbitrary queries and throw exceptions. So if you really need to enforce a complex constraint P(x, y, z), you put in a database trigger that queries some x, y, z: !P(x, y, z) and throws an exception if it finds any results. That all works because of quantifier duality! See here for an example of this in practice. Duals more broadly "Dual" doesn't have a strict meaning in math, it's more of a vibe thing where all of the "duals" are kinda similar in meaning but don't strictly follow all of the same rules. Usually things X and Y are duals if there is some transform F where X = F(Y) and Y = F(X), but not always. Maybe the category theorists have a formal definition that covers all of the different uses. Usually duals switch properties of things, too: an example showing some x: P(x) becomes a counterexample of all x: !P(x). Under this definition, I think the dual of a list l could be reverse(l). The first element of l becomes the last element of reverse(l), the last becomes the first, etc. A more interesting case is the dual of a K -> set(V) map is the V -> set(K) map. IE the dual of lived_in_city = {alice: {paris}, bob: {detroit}, charlie: {detroit, paris}} is city_lived_in_by = {paris: {alice, charlie}, detroit: {bob, charlie}}. This preserves the property that x in map[y] <=> y in dual[x]. And after writing this I just realized this is partial retread of a newsletter I wrote a couple months ago. But only a partial retread! ↩ Specifically "linear temporal logics" are modal logics, so "eventually P ("P is true in at least one state of each behavior") is the same as saying !always !P ("not P isn't true in all states of all behaviors"). This is the basis of liveness checking. ↩ I don't know for sure, but my best guess is that Antithesis does something similar when their fuzzer beats videogames. They're doing fuzzing, not model checking, but they have the same purpose check that complex state spaces don't have bugs. Making the bug "we can't reach the end screen" can make a fuzzer output a complete end-to-end run of the game. Obvs a lot more complicated than that but that's the general idea at least. ↩ For CHECK to constraint multiple records you would need to use a subquery. Core SQL does not support subqueries in check. It is an optional database "feature outside of core SQL" (F671), which Postgres does not support. ↩

3 days ago 10 votes
Omarchy 2.0

Omarchy 2.0 was released on Linux's 34th birthday as a gift to perhaps the greatest open-source project the world has ever known. Not only does Linux run 95% of all servers on the web, billions of devices as an embedded OS, but it also turns out to be an incredible desktop environment! It's crazy that it took me more than thirty years to realize this, but while I spent time in Apple's walled garden, the free software alternative simply grew better, stronger, and faster. The Linux of 2025 is not the Linux of the 90s or the 00s or even the 10s. It's shockingly more polished, capable, and beautiful. It's been an absolute honor to celebrate Linux with the making of Omarchy, the new Linux distribution that I've spent the last few months building on top of Arch and Hyprland. What began as a post-install script has turned into a full-blown ISO, dedicated package repository, and flourishing community of thousands of enthusiasts all collaborating on making it better. It's been improving rapidly with over twenty releases since the premiere in late June, but this Version 2.0 update is the biggest one yet. If you've been curious about giving Linux a try, you're not afraid of an operating system that asks you to level up and learn a little, and you want to see what a totally different computing experience can look and feel like, I invite you to give it a go. Here's a full tour of Omarchy 2.0.

4 days ago 8 votes
Dissecting the Apple M1 GPU, the end

In 2020, Apple released the M1 with a custom GPU. We got to work reverse-engineering the hardware and porting Linux. Today, you can run Linux on a range of M1 and M2 Macs, with almost all hardware working: wireless, audio, and full graphics acceleration. Our story begins in December 2020, when Hector Martin kicked off Asahi Linux. I was working for Collabora working on Panfrost, the open source Mesa3D driver for Arm Mali GPUs. Hector put out a public call for guidance from upstream open source maintainers, and I bit. I just intended to give some quick pointers. Instead, I bought myself a Christmas present and got to work. In between my university coursework and Collabora work, I poked at the shader instruction set. One thing led to another. Within a few weeks, I drew a triangle. In 3D graphics, once you can draw a triangle, you can do anything. Pretty soon, I started work on a shader compiler. After my final exams that semester, I took a few days off from Collabora to bring up an OpenGL driver capable of spinning gears with my new compiler. Over the next year, I kept reverse-engineering and improving the driver until it could run 3D games on macOS. Meanwhile, Asahi Lina wrote a kernel driver for the Apple GPU. My userspace OpenGL driver ran on macOS, leaving her kernel driver as the missing piece for an open source graphics stack. In December 2022, we shipped graphics acceleration in Asahi Linux. In January 2023, I started my final semester in my Computer Science program at the University of Toronto. For years I juggled my courses with my part-time job and my hobby driver. I faced the same question as my peers: what will I do after graduation? Maybe Panfrost? I started reverse-engineering of the Mali Midgard GPU back in 2017, when I was still in high school. That led to an internship at Collabora in 2019 once I graduated, turning into my job throughout four years of university. During that time, Panfrost grew from a kid’s pet project based on blackbox reverse-engineering, to a professional driver engineered by a team with Arm’s backing and hardware documentation. I did what I set out to do, and the project succeeded beyond my dreams. It was time to move on. What did I want to do next? Finish what I started with the M1. Ship a great driver. Bring full, conformant OpenGL drivers to the M1. Apple’s drivers are not conformant, but we should strive for the industry standard. Bring full, conformant Vulkan to Apple platforms, disproving the myth that Vulkan isn’t suitable for Apple hardware. Bring Proton gaming to Asahi Linux. Thanks to Valve’s work for the Steam Deck, Windows games can run better on Linux than even on Windows. Why not reap those benefits on the M1? Panfrost was my challenge until we “won”. My next challenge? Gaming on Linux on M1. Once I finished my coursework, I started full-time on gaming on Linux. Within a month, we shipped OpenGL 3.1 on Asahi Linux. A few weeks later, we passed official conformance for OpenGL ES 3.1. That put us at feature parity with Panfrost. I wanted to go further. OpenGL (ES) 3.2 requires geometry shaders, a legacy feature not supported by either Arm or Apple hardware. The proprietary OpenGL drivers emulate geometry shaders with compute, but there was no open source prior art to borrow. Even though multiple Mesa drivers need geometry/tessellation emulation, nobody did the work to get there. My early progress on OpenGL was fast thanks to the mature common code in Mesa. It was time to pay it forward. Over the rest of the year, I implemented geometry/tessellation shader emulation. And also the rest of the owl. In January 2024, I passed conformance for the full OpenGL 4.6 specification, finishing up OpenGL. Vulkan wasn’t too bad, either. I polished the OpenGL driver for a few months, but once I started typing a Vulkan driver, I passed 1.3 conformance in a few weeks. What remained was wiring up the geometry/tessellation emulation to my shiny new Vulkan driver, since those are required for Direct3D. Et voilà, Proton games. Along the way, Karol Herbst passed OpenCL 3.0 conformance on the M1, running my compiler atop his “rusticl” frontend. Meanwhile, when the Vulkan 1.4 specification was published, we were ready and shipped a conformant implementation on the same day. After that, I implemented sparse texture support, unlocking Direct3D 12 via Proton. …Now what? Ship a great driver? Check. Conformant OpenGL 4.6, OpenGL ES 3.2, and OpenCL 3.0? Check. Conformant Vulkan 1.4? Check. Proton gaming? Check. That’s a wrap. We’ve succeeded beyond my dreams. The challenges I chased, I have tackled. The drivers are fully upstream in Mesa. Performance isn’t too bad. With the Vulkan on Apple myth busted, conformant Vulkan is now coming to macOS via LunarG’s KosmicKrisp project building on my work. Satisfied, I am now stepping away from the Apple ecosystem. My friends in the Asahi Linux orbit will carry the torch from here. As for me? Onto the next challenge!

4 days ago 13 votes