blt

Not relative to its available resources. For example, there's still no option to penalize intersections. I'm still getting absurd side-street routes and unprotected lefts/crossings to save 1 minute on a 20-minute trip.

For me, it's hard to imagine him giving up the printed newspaper strip's connection to the physical world. Calvin and Hobbes is filled with references to the basic elements of physical reality: dirt, rocks, water, snow, speed, collisions, temperature, light, sound. Webcomics exist in a world of pure information.

I wonder why we use something polynomial-based instead of something exponential-based like the decibel scale?

After that build-up, I was hoping to see a toy implementation of a method or two for `MultiArrayList`.

My interpretation leans towards: Gmail Thinks I'm Lazy.

LLMs have made one thing clear: intellectual laziness is even more pervasive than we previously thought, even among "knowledge workers".

Sound systems do have an omnipresent "only as good as your weakest link" feeling.

A lossless source (or analog, whatever your preference) and great speakers/amps can't overcome a bad sounding room - or a piece of music that was mixed/mastered poorly.

California is among the US states with the highest income inequality.

All technology, no matter how undesirable it once felt, eventually becomes nostalgic for somebody.

It's funny how they specifically use the phrase "output that follows real-world physics" to describe the marble rolling video. At the end of the zigzag track, the marble jumps up for no reason. In a couple of other places it speeds up with no apparent energy source. It's still an amazing result, but they could have picked a better example for this claim!

Agree. California peaches are horrible. Southern peaches (Georgia, Carolinas) are indescribably better.

I love California, but it's funny/sad the extent to which many Californians deny that other parts of the USA have us beat in some regards.

I'm not a Haskell programmer, but from my limited awareness: Wouldn't they want to encode the restriction that April 31 doesn't exist directly in the type system instead of using raw integers for the underlying struct?

Routing games like Mini Metro/Motorways, Freeways, and Fly Corp are not quite puzzle games in the traditional sense, but I'd definitely encourage any puzzle fans to try them.

I knew that adding sriracha and hoisin was considered somewhat gauche / insulting to the chef. But I was surprised to see no mention of Basil, since every pho shop in the USA provides it. Is that also nontraditional? I really like it...

In the dev branch now! https://asahilinux.org/2026/02/progress-report-6-19/

I will attempt to use Asahi as my daily driver once this is officially released.

The title creator is bad... as a total amateur who needs to edit a video every now and then, it hurt me.

yes of course, in my field the arrays are large and we're doing nontrivial work per element.

Surprised Python is only 21x slower than C for tree traversal stuff. In my experience that's one of the most painful places to use Python. But maybe that's because I use numpy automatically when simple arrays are involved, and there's no easy path for trees.

For me it was worse, the headphone port on my phones always eventually failed. Maybe these rich people replace phones too quickly to experience that.

I use wired at home, where I'm not cycling the connection very much.

What makes this different from linking to a random zip file somewhere?

And their predictions about Go were wrong, because they thought the algorithm would forever be α-β pruning with a weak value heuristic

An underrated aspect of Matlab is its call-by-value semantics. Function arguments are copied by default. Python+NumPy is call-by-reference; mutations to array arguments are visible to the caller. This creates a big class of bugs that is hard for non-programmers to understand.

Clarity in writing comes mostly from the logical structure of ideas presented. Writing can have grammar/style errors but still be clear. If the structure is bad after translation, then it was bad before translation too.

Yes. Generative AI is bad. Most of the general population already realizes it. Only the tech and computer science bubble remains optimistic.

This needs to be refined: f(x) is O(g(x)) if there exists some X >= 0 such that f(x)/g(x) is bounded for all x > X.

Otherwise, we cannot say that 1 is O(x), for example.

The definition of big O notation is pure math - there is nothing specific to analysis of algorithms.

For example: "the function x^2 is O(x^3)" is a valid sentence in big-O notation, and is true.

Big O is commonly used in other places besides analysis of algorithms, such as when truncating the higher-order terms in a Taylor series approximation.

Another example is in statistics and learning theory, where we see claims like "if we fit the model with N samples from the population, then the expected error is O(1/sqrt(N))." Notice the word expected - this is an average-case, not worst-case, analysis.

Yes, in most cases the reduction of supervised learning to optimal control is not interesting.

We can also reduce supervised learning to reinforcement learning, but that doesn't mean we should use RL algorithms to do supervised learning.

We can also reduce sorting a list of integers to SAT, but that doesn't mean we should use a SAT solver to sort lists of integers.

Awaiting the rocket engine equivalent of the K20.

tangentially, does anyone know a good way to limit web searches to the "low-background" era that integrates with address bar, OS right-click menus, etc? I often add a pre-2022 filter on searches manually in reaction to LLM junk results, but I'd prefer to have it on every search by default.

IDK, I think Apple creating its own laptop/desktop-class CPU was a pretty bold move with a huge payoff. It's less sexy than introducing an entirely new category of product, but it's not exactly risk-averse either.

I always liked the look of ACM journals/conferences more than other venues. Their template uses Libertine, so it is my choice too.