Maybe these open projects start to get more attention when we have a distribution system/App Store for AI projects. I know YC is looking to fund this https://www.ycombinator.com/rfs

Is "AI appstore" envisioned on Linux edge inference hardware, e.g. PC+NPU, PC+GPU, Nvidia Project Digits? Or only in the cloud?

Apple probably wouldn't accept a 3rd-party AI app store on MacOS and iOS, except possibly in the EU.

If antitrust regulation leads to Android becoming a standalone company, that could support AI competition.

The idea here is simple:

Your health record persists with every query so you get personalized health recommendations. A personal WebMD

I built this to make it easier to see what I’m making as I’m making it. No extra installations or extra import step. Create React components visually or pull in designs as code from Figma, right inside the IDE.

More info on pricing: https://news.ycombinator.com/item?id=27041676

Thanks!

Thanks for the detailed feedback! All makes sense – on it now.

Thanks for the kinds words! Performance has been a huge priority from day 1.

Thanks! Pricing info here: https://news.ycombinator.com/item?id=27041676

This all makes sense, and I've definitely been feeling the integration pain when I want to export code elsewhere – GitHub support coming soon!

Thanks! If double-clicking or pressing the enter key while the text element is selected doesn’t work, it’s likely a bug. Are you using Firefox by any chance? I just tested on chrome and safari.

I figured it out. I'm using Safari, but I think you might have a bug there.

In most cases had to click a total of four times: once to get focus on the element, then a double click to select the text, and then once more to get a cursor within the text where I wanted it.

There were also some times where I only had to click three times if I did it with the right rhythm (once for focus, once for selection, and once for cursor), or five times if I had already selected an outer container component and was trying to select an inner component (double click for focus, double click for selection, and once for cursor).

For what it's worth, my own initial expectation was that if you click on an element, it should automatically place the cursor where you clicked, and that by default you should be selecting whatever text you are clicking on, not the outer envelope that contains it.

I observed all this on the "Multidisciplinary" theme (black background).

Ah I see, yeah the "double-click to select container child/text" ux is meant to mimic keynote's implicit grouping mechanism, which works the same way.

It's mostly to prevent empty "phantom" containers that people aren't aware are affecting their project layout.

I've been thinking about changing this to the straightforward, single click to select whatever's topmost and visible. Will have to if the current ux bothers people.

Thanks! Yeah it’s exactly that - pure js and canvas element for custom drawing. No UI frameworks. It’s all custom so there’s no open source reference, but I plan to release some open source stuff.

This is dope! I'd love to see any part of this open sourced. I've been wanting to try to build something just like this.

Thanks for the overview. I’m looking forward to it, and best of luck with your project!

Amazing feat! Let mw know if you open source it, I'd love to contribute!