Just cut the first v1 beta of Evalite! It's a test runner for evaluating AI-powered apps in TypeScript. pnpm add evalite@beta v1 brings: - A library of built-in scorers - Save your eval results anywhere with custom storage - A ton of papercut fixes https://t.co/YO6Ovv3QsC

Why is it that the people who seem to have the strongest negative opinions about Mob/Ensemble programming are usually the people who have never experienced it? The thing they imagine has nothing to do with the real thing, but that doesn't stop them from having opinions. Welcome

Do yourself a favor and set this in your pnpm config

Spend your youth loremaxxing

I'm convinced that gitflow is a psyop and no one ever done that in practice because it takes like 30 seconds of thinking before you realize this is completely unfeasible and a huge waste of everyone's time

What a ride it's been! Grateful for our community of educators who have made it all possible.

I often re-read @natfriedman's personal website (former Github CEO)

Fading out audio is one of the most CPU-intensive tasks you can possibly do! Values that get close (but not quite) zero, hit an underflow gap known as "Subnormal" range. It’s a mathematical conundrum so tricky, both x86 and ARM made special CPU instructions just to handle it!

Things I wish were table-stakes in every web app: - Typing auto-focuses main input - Enter submits forms - Esc closes modals - Validation on blur, not while I'm typing - Let me decide whether a link opens in a new tab or not What else?

80% of estimates are wrong in the best case. It's just math. Assuming you have good-faith estimates, the map of real-to-actual time forms a normal bell curve: 40% or so will be early, 40% will be late, and the top of the curve is within acceptable limits. This curve skews

monorepo wins every time: ◇ Instant code sharing between apps ◇ Full context (Cursor loves this) ◇ Global refactors that actually work ◇ One install to rule them all ◇ Simple CI/CD with everything in one place Once you go monorepo, you don’t go back.

People often ask what metrics can be used to assess the health of a process. They usually bring up team-level or individual metrics like throughput (how fast a person/team pushes work out the door) or its inverse, cycle time (time from when a team starts work on something until

Differences Between Korean and Chinese Math

ChatGPT can now connect to more internal sources & pull in real-time context—keeping existing user-level permissions. Connectors available in deep research for Plus & Pro users (excl. EEA, CH, UK) and Team, Enterprise & Edu users: Outlook Teams Google Drive Gmail Linear & more

First, you start with CORE functionality. Then: • Reveal options contextually • Let users "discover" deeper features (and hide advanced features behind intuitive triggers) Think about it like a great story - you don't reveal everything at once. (shoutout to @asaadmahmood5

One of the core rules of the Unix philosophy is that each program should do one thing and do it well. For `find` that apparently means wiping out your whole folder if you get the order of arguments wrong.

Have a friend who’s an expert in the science of learning. Once, we got to talking about math. He said the lowest hanging fruit is to send every kid back to the stage of math they’re not 100% fluent in. For most kids, that’s something like 3rd or 4th grade. Then they should work

AI won’t replace software engineers, but it will expose the ones who never learned to think like engineers. Tools can speed you up, but if your thinking’s wrong, AI just gets you to the wrong place faster. Engineering still matters.

This is so random! and muuuuch needed A new Random namespace with 22 methods for working with randomness in JavaScript. proposal just hit Stage 1