PullJosh

I love this! Such a beautiful use of some real nerdy math to benefit humans in a way non-nerds would appreciate.

I can’t believe they haven’t already

I like their emphasis on quickly prototyping many variations of a design. That seems useful, even for experienced designers.

This is awesome!

1) I am able to run the model on my iPhone and get good results. Not as good as Gemini in the cloud, but good.

2) I love the “mobile actions” tool calls that allow the LLM to turn on the flashlight, open maps, etc. It would be fun if they added Siri Shortcuts support. I want the personal automation that Apple promised but never delivered.

3) I am so excited for local models to be normalized. I build little apps for teachers and there are stringent privacy laws involved that mean I strongly prefer writing code that runs fully client-side when possible. When I develop apps and websites, I want easy API access to on-device models for free. I know it sort of exists on iOS and Chrome right now, but as far as I’m aware it’s not particularly good yet.

Looks great! I’ve definitely wanted this exact thing before. I will consider reaching for this next time I need to throw together a quick app.

I just deleted my account. The other LLMs are so good that I don't even feel like I'm sacrificing much.

Does relying on larger players result in better overall uptime for smaller players? AWS is providing me better uptime than if I assembled something myself because I am less resourced and less talented than that massive team.

If so, is it a good or bad trade to have more overall uptime but when things go down it all goes down together?

It’s been many years since I built a website using Django. At the time, my favorite feature was that it provided a built-in admin UI with no extra work. So helpful!

Now that I live in the JavaScript ecosystem, I don’t have an equivalent tool.

I’ve been working on promoting my seating chart app for teachers, Shuffle Buddy, on social media. I had a 1M view pop on TikTok when I first launched and have now been clawing along to try for continued engagement.

It’s reassuring to know that social media posts are hard for everyone and that it isn’t supposed to be easy. I keep looking for ways to create content that is genuinely beneficial to teachers and also convinces them to try my app, but it’s hard.

The solution I built for this is Leopard. It is a Scratch → JavaScript converter. You can take an existing Scratch project and convert it to JavaScript code and then keep working, or use the Leopard library to create a new project from (ahem) scratch, following all the same conventions as a Scratch project.

Check it out! https://leopardjs.com/

I feel like -r is an old school internet thing. I wonder if anybody has data for trends over time.

I have aphantasia and have long thought that my lack of a mind’s eye might be an asset for programming, much like how a blind person will typically have better-than-average hearing.

If programming is best done through abstract thought and doesn’t benefit from pictures, then I’ve been training since the day I was born.

The design of this blog post is lovely! Despite seemingly having no right to be. It goes against all of my design instincts, but it somehow nails a vibe while remaining pleasant and easy on the eyes. I love it!

I was expecting this to refer to different ways to represent the same diff. (For example, you could represent a change from `console.log(“hello”)` as `console.log('hello')` as +'-“ … +’-“ or as +'hello'-“hello”)

I don’t have a specific example in mind, but it seems reasonable that different languages could benefit from different ways of representing the same diff.

Totally agree. I’ll also add that although I am often tempted to work alone, sharing my work or ideas or output with someone else boosts motivation a lot.

I don’t know if this is true or not, but it’s clever and interesting regardless.

Complete novice question: Would it be possible to build a language that could read your source code and tell you, at compile time, which expressions have values that depend on RNG? And on cryptographically secure RNG?

So that you could annotate a variable as needing to be cryptographically secure and the language could check that, somewhere along the way, its value depends on an adequate RNG function?

As I understand it, the hard part of a masonry layout is ensuring that items are laid out with the first items near the top of the screen and the last items near the bottom of the screen.

If you just plop your items into columns, the first items will all appear in the first column and the last items will all appear in the last column (left to right), which is not the behavior you want.

To my knowledge, the correct behavior cannot currently be created in CSS alone. (Or if it can, it must be a wild hack.)

Hi HN! A few years ago, I began creating Leopard, a Scratch to JavaScript code converter. (Since then, I've received help from a few awesome open-source contributors. Thank you!)

I started using Scratch when I was in 2nd grade. I made over 500 games and projects, working with Scratch all the way into high school. I eventually moved on to JavaScript, but the transition was a difficult one. Much of the JavaScript world was unfamiliar, and learning to do everything new all at once was challenging.

Leopard is meant to be a transition tool that helps ease this learning process. It outputs human-readable JavaScript that is as close to a 1-to-1 translation as possible.

You can enter any Scratch project URL to convert it. If you don't have one on hand, here are a few examples:

- https://scratch.mit.edu/projects/345789566/

- https://scratch.mit.edu/projects/905275127/

- https://scratch.mit.edu/projects/891008805/

(Note that getting compatibility exactly right is hard, so not every project will work perfectly, but many projects should be respectably close.)

Completely unrelated, but are you the adhesive_wombat I would listen to on SoundCloud when I was younger?

Or, if that’s too hard, leave the orbit the same but slow the spin down a little so each day is longer. A 0.35% decrease in spin speed seems fine right?

The live-streaming software OBS has a “virtual webcam” feature that can make a generated video feed behave like a hardware webcam. Perhaps something similar is being used to feed generated video into zoom?

Genuine question: Does anyone here have a good sense of how easy/hard it is to moderate a platform like YouTube? My sense is that the answer is VERY HARD, so I tend to give them the benefit of the doubt when things go wrong. But perhaps I’m being too generous. Is this something that could be prevented if YouTube just tried harder?

This doesn’t seem like a very useful characterization. To the best of my knowledge, visualization ability falls on a spectrum, and some are better than others. Aphantasia is a useful word to describe a low ability to visualize. Just because you aren’t at zero doesn’t mean you’re average. (Also, for the record, I find the things you describe very difficult. McDonald’s/Apple logos are borderline doable, but the other things you listed are not happening for me.)

> Use it or lose it. You can get it back if you want.

I genuinely really hope you’re right. As someone with subpar visualization skills, gaining them would be wonderful. If you have advice for how to achieve this it would be much appreciated.

At least in the beginning, self-driving cars need to be compatible with human drivers, who are also on the road.

Thank you for all the quiet moderation work you do to make HN so great. I can say that you’ve personally made my life better, and I’m sure many others would say the same.

This is smart! Seems like a great way to provide value that’s lacking in other browsers. I much prefer this method of attracting users to some of Microsoft’s other antics.

Now can they make a mode that helps ME avoid websites that make my life worse? (It really shouldn’t be this hard to use a computer in consistently productive ways. But it is, at least for me.)

Can I get an ELI5 on how physical neurons, stuck in a measly 3 dimensions, can possibly form higher-dimensional connections on a large scale?

I understand higher dimensional connections in theory (such as in an abstract representation of neurons within a computer), but I can’t imagine how more highly-connected neurons could all physically fit together in meat space.