CharlesW

> I don't like USB-C because they all look the same on the outside, but they're not all the same on the inside.

Many people don't realize that USB-C (USB Type-C) just refers to the physical connector. At minimum, speed and power ratings should have been required for any cable using USB-C at one or both ends. It's almost breathtaking to consider how the USB Implementers Forum has fumbled these kinds of basic issues over the years.

Close! (I had to look it up.) "The sky above the port was the color of television, tuned to a dead channel." (Neuromancer, William Gibson)

Sounds very intriguing! Please reach out (see my profile) when you're ready to share more.

Not silly at all. For Compose Multiplatform, Axiom won't do much for code review or generation. It might be useful for operations on the built app: console capture, crash symbolication, simulator UI/accessibility driving, CPU profiling, plus the whole shipping layer (signing, privacy manifests, submission).

I'm glad that it's been valuable for you!

Axiom covers only native APIs and tools, but the Flutter team maintains what looks like a nice set of skills at https://github.com/flutter/agent-plugins, and the Dart team at https://github.com/dart-lang/skills.

If it's okay to mention my own complementary open source project, Axiom¹ does a good job of helping coding harnesses know how to do this effectively for Apple OS development.

In addition to a deep roster of skills and agencts, Axiom includes several for-LLM tools². xclog, xcprof, xcsym, and xcui are designed to be used by LLMs, and expose capabilities in a token-efficient way. These tools are equally helpful for non-Axiom skills/agents.

¹ Axiom: https://charleswiltgen.github.io/Axiom/ ² Axiom CLI tools: https://charleswiltgen.github.io/Axiom/tools/

I'd just note that (1) Rob did set a record whether or not he meant to¹, and (2) Kelsey also did not intend to break the men's record.

¹https://www.guinnessworldrecords.com/world-records/384202-fa...

This project appears to be an LLM-rewritten Elk (https://github.com/cesanta/elk), as noted elsewhere in the thread. That project was established in 2019.

It's kind of buried here, but Kelsey is the fastest human to do this. She beat the male record holder's time by 6 days.

> We generated a big pile of artifacts, we are publishing all of them, and you can form your own opinion.

My opinion is that spamming HN with two gimmicky "one-shot prompting shootout" marketing pieces in two days does not build confidence about either your technical or marketing expertise.

TIL, thanks!

Apple (like any email provider that conforms to RFC 2822) supports plus addresses as well.

Tip: It's straightforward to migrate from Evernote to Apple Notes. https://www.macrumors.com/how-to/migrate-from-evernote-to-ap...

Superpowers' brainstorming skill is another classic example of this. This skill with SOTA models is incredibly helpful. The results become very thorough specs/plans, ADRs, etc.

https://github.com/obra/superpowers/blob/main/skills/brainst...

> But it is a huge waste of money for most coding tasks.

The key is not to indiscriminately use the most powerful/expensive model you can for everything. When you use it for what it's uniquely suited for and ask it to spawn subagents using Opus and Sonnet based on what tasks need, you'll get better results at a reasonable cost.

Nothing about Superpowers forces you to use TDD, brainstorm first, etc. It’s not rigid about the workflow.

The GitHub description is a pretty good summary: "An agentic skills framework & software development methodology that works."

Here's what that methodology looks like: https://github.com/obra/superpowers#the-basic-workflow

As a long time user, I'd recommend checking out https://github.com/obra/superpowers#the-basic-workflow. If that reasonates with you, try using it to develop a few features or capabilities.

It works very well for the way that I work (interactively and iteratively, not "one-shot"), and it helps me to better work in less time. Superpowers is one of the few skill/agent suites I use for all software development projects.

If you like building skill/agents, the posts at https://blog.fsck.com/ are a great resource for learning how to do well. The effectiveness of my project Axiom (a skill/agent suite for Apple OS developers) has benefited enormously from the knowledge that Superpowers' creator Jesse Vincent has been kind enough to share.

TLDR: You owe it to yourself to try it.

> The reference Opus encoder has certainly had a number of improvements that affect sound quality since then…

Yes, but not for high-bitrate music applications.¹ For example, Opus 1.2 really improved the quality of music encoding at 32–48 kbps. Opus doesn't have to be great at everything to be great at what it does, just like AAC-LC doesn't have to be. (¹Opus 1.6's experimental Opus HD looks very promising for this!)

Apple's and Fraunhofer's closed-source AAC-LC encoders have seen regular, minor quality tuning, and they benefit from the research and engineering work that have been done for the AAC family of encoders (HE-AAC for bitrates down to ~48 kbps, HE-AAC v2 to 32 kbps, xHE-AAC below that).

Then he'd love this. Like Unix, Vite+ is a collection of programs that do one thing well.

Vite+ isn't a layer, it's "just" a high-performance suite of excellent tools that work well together to provide a great DX for developers.

Vite+ can improve and simplify what developers are already doing with ad-hoc collections of tools. Vite is already an industry standard, and Vite+ has a good chance of achieving that status as well.

The evaluation tools used are helpful for encoder development, but at best they're imperfect proxies for human perception, and their predictions are often inconsistent with the human experience. I assume that statements like "apparently the best AAC encoder" aren't meant to be taken too seriously, since everybody who does this stuff knows that ABX/MUSHRA tests with real humans is what tells the tale.

On Opus vs. AAC specifically, there's a long history of studies like https://www.researchgate.net/publication/301428302_Perceived... to help answer that question. (There are interesting charts at the top of page 1175.)

> Why not use a lossless codec if you care about quality?

(1) Lossy codecs are transparent at half the file size (or less) of FLAC/ALAC.

(2) AAC (strictly, AAC-LC) is universal, where FLAC and Opus are not yet there.

You might like Poppy (in beta), which supports all media servers (including OpenSubsonic/Navidrome) and Opus as a first-class music format. https://www.reddit.com/r/PoppyApp/comments/1tiyki0/about_pop...

Plus, at 96+ kbps (assuming an Apple-quality AAC-LC encoder) Opus loses its quality advantage. So at higher bitrates, the benefit of choosing Opus is that encoders/decoders are royalty-free.

Favorable benchmarks for Pangram exist, and it is one of the better AI detectors. But it's never been trustworthy as proof of authorship/misconduct, and it's getting less useful as models (and people's ability to steer them) improve.

https://www.theatlantic.com/technology/2026/05/pangram-ai-de...

AI can't be a competent person, but inanmate things can be competent. https://www.merriam-webster.com/dictionary/competence