chrisweekly

10s of millions isn't even a rounding error in the context of AI-related spending.

beautiful

Related tangent: Brave is pretty close to Chrome, sans Google.

robbyrussell, creator of oh-my-zsh?

1. thank you for omz, it's great

2. how is it possible you've been on HN for over 15 years and have just 254 karma?

> "My ideal software has a well designed and thought through user interface for the primary interactions, allowing me to think about the goal I'm trying to achieve"

Yes, this is the crux of the whole discussion. The paradigm is called "Intent-based UX".

You'll want to strongly consider rebranding with a different name. Way too easily confused.

It's not impossible (and I'd substitute "evolved" for "designed").

Yes, wanting "More" [time, money, love, status] is pretty firmly ingrained as an advantageous trait. But you can override your instincts (e.g. declining a drink when thirsty, if you learn the glass of water contains poison). And it is possible to learn to want what you already have.

How? Practice three things: Compassion, Attention, and Gratitude. (Pick any two, and the third will follow almost automatically.)

That's the central thesis of an excellent book, "How to Want What You Have" by Dr Timothy Miller. It changed my life.

"Let's eat, grandma!" vs "Let's eat grandma!"

Ah, thanks for clarifying. (And shame on the GP.)

You might be interested in smolvm microvms from https://smolmachines.com - which (unlike Docker) provide kernel-level isolation (among other benefits). No affiliation, just a happy user.

I'm not the person you replied to, but your response misses their point: rollout issue aside, the approach is flawed by design. The feature author didn't address that at all.

With apologies for not just testing this myself (currently AFK), doesn't it still work to have a CLAUDE.md file containing just `@AGENTS.md`?

Such a society would pay a whole lot more attention to white-collar financial and environmental crimes than to anything all the motivated junkies in the country could wreak.

I'm not sure I follow; how is autocompaction (lossy summarization), applied at random times (vs strategically, between workflow phases), helpful to ensuring clarity of intent? Maybe you're saying that just plowing ahead and living with the signal loss along the way works well enough for your purposes. In which case, ok, YMMV, different strokes.... but paying attention to context quality and being deliberate about when to compact vs handoff vs delegate to subagents is most definitely not "bad advice".

> "Codex's compaction is very good, fwiw, but it happens so frequently that..."

I appreciate and follow Matt Pocock's advice: avoid autocompaction. Compaction is lossy, which is ok when you're managing it at phase boundaries, but autocompact is lossy at the most inopportune times, firing mid-task and leading to agents going off the rails.

That analogy doesn't hold; at least w bits vs bytes it's still "data over time".

In this case it's measuring something nearly meaningless. You could charge 100 times less per token, but if task completion takes 1,000 times as many tokens, it's not much of a bargain.

There is no life without death (nor light without darkness).

Price per task (not per token) is what really matters.

The folders were useful to me for years, aided by the macOS iTunes feature to view and rearrange things via drag-and-drop. Still annoyed that feature was removed.

akin to "New York, New York" (as in NYC, NY)

Windows caused me similar dismay when I was forced to use it at work (2022-25), but macOS - while far from perfect - has been more than fine for my purposes for the last 20 years. I like linux and can choose to use it on my superior apple hardware. Not a fanboi, just a pragmatist.

Yeah, I read it as "assuming" too.

That behavior is not tolerated at any real poker table. It shouldn't be tolerated in salespeople either.

But smolvm provides kernel-level isolation. Much closer to firecracker than docker.

cap -> gap

Untrue. Vowel.

For microvms, take a look at https://smolmachines.com

Yet its written shape is a double-V.

honestly I preferred their (unorthodox) stream-of-consciousness writing style to most of the "mannered" genAI slop being slung these days

Cool approach, thanks for sharing.

FWIW I also liked this (very different) post of yours: https://softwaredoug.com/blog/2026/07/13/who-want-to-be