greggman65
Joined 1 year ago
Karma
88
Submissions
42
Member since
Aug 29
Recent submissions
Comment · 16 days ago
Comment · 1 month ago
A non-live map for trains in Tokyo
https://greggman.github.io/tokyo-trains
I think the live map is cool but (1) I don't know where to get that data and (2) seeing things at 100x speed is, IMO, more interesting as a visual
It's based off schedules but only Toei's schedules are easy to get (AFAICT). I got permission for the most of the others, but I thought the visual was good as is ... mostly. Maube I'll revisit.
Comment · 2 months ago
I can certainly imagine one that's better. One that catelogs all the images using better AI that apples and lets me do complex searches "Dog at beach", "Joey in red shirt with Jane"
Comment · 2 months ago
The Photos app does not need permission any more than any other Photos app. It can ask "Photos app wants to access your photos" and you can click "Always", "Selected Photos", "No" and download some other app to access your photos. That would be the "equal" thing to do. Let the user pick what app they want to access their photos with.
Same with the Camera App. It can ask "Camera App wants to access your Camera" and you can say "No" and then download some other camera app and make that your default and grant it camera permission.
Comment · 2 months ago
In Japan (not sure in other countries), I went to the Apple Store. They said to complete the purchase, use the Apple Store app and scan this QR code on their device. I did and it immediately wanted me to share a ton of info with the main highlighed button "Share Now" and in small lettering below that "customize and share", where in you can see all the stuff they were going to take.
Apple, That's Privacy
Comment · 4 months ago
I wrote this one, a year after that video, inspired by it
https://greggman.github.io/doodles/sort
Browsers (and machines) were slower then so it didn't finish as fast as it does today. Also, it's hard to decide what steps to illustrate/count. Is a swap 1 step (swap) or 4 (load slot N to R1, load slot M to R2, store R1 in M, store R2 in N) etc...
Comment · 4 months ago
I depends on your definition of "working on" :p
A worked on a 3d asset generator with nodes (like Blender/Houdini)
https://greggman.github.io/sedon/
I worked on a couple of design generators
Comment · 5 months ago
I don't know if this was my "Oh Shit" moment but 4 weeks ago I thought'd I'd try vibe coding a WebGPU 3D Node Based Editor.
https://github.com/greggman/sedon
It was just an experiment and I probably won't work on it more but still, I was blown away with how far we got. There's a quite a bit we worked through even though it was only part time of those 4 weeks.
Comment · 6 months ago
> Quake made PVS famous. It’s still useful in some indoor games where the scene geometry is static and bake time is acceptable.
It was used extensively in outdoor games like Jak and Daxter.
Comment · 6 months ago
It’s a WebGL issue, fixed in WebGPU.
Browsers generally only allow a fixed number of WebGL contexts per page. So a generic element effect library has the issue that too many elements some will start losing contexts. The workaround is to just make one large screen size canvas and then figure out where the elements are you need to draw an effect for. now you only have one context drawing all the elements. But, you can’t know where to draw until the browser scrolls and renders so you’re always one frame behind.
https://webgl2fundamentals.org/webgl/lessons/webgl-multiple-...
WebGPU doesn’t have this issue. You can use the same device with multiple canvases
https://webgpufundamentals.org/webgpu/lessons/webgpu-multipl...
Comment · 9 months ago
I think this is a good example of something you can vibe-code today. (though maybe not as good)
I went to gemini, picked "cavnas". used this prompt
> There's a famous CG scene in the movie Hackers where they "Hack the Gibson". It shows a bunch of translucnt cubes with glowing edges. The textures on the cubes are live computer text. The camera slowly flyies between the cubes tilting gracfully and it searches for the main one.
> Reproduce this scene in Javascript. Be sure to include each of those features
> 1. live computer text which you can simulate by drawing to a canvas offscreen and uploading to a texture, adding more output as it goes. You can even use "function.toString()" of the code you write as input
> 2. a post processing step so we get a glow
> You can probably use three.js for this
Here's the result.
https://codepen.io/greggman/pen/XJKPBZW
No, it's not as good as the site linked above and it's unlikely it would be. On other hand, it got this far on the first try. Maybe a few more iterations and it could get the stuff you want.
Comment · 9 months ago
I didn’t put nearly as much effort as this post into shape matching but I did try a few other things like
Non-ascii, I tried various subsets of Unicode. There’s the geometric shape area, CJK, dingbats, lots of others
Different fonts - there are lots of different monospace fonts. I even tried non-monospaced fonts tho still drawn in grid
ANSI color style https://16colo.rs/
My results weren’t nearly as good as the ones in this article but just suggesting more ways of exploration
https://greggman.github.io/doodles/textme10.html
Note: options are buried in the menu. Best to pick a scene other than the default
That said, I'm also fully enjoying letting AI help me solve problems.
Here's one: I study Japanese. I used jisho.org for years (A Japanese/English dictionary). I ran into a problem ~2 weeks ago where I needed to look up a word and used their handwriting recognition to sketch a kanji. I got a single stroke wrong and it gave no matches. I checked if it was open source. No
So, I asked AI (Claude) to make something similar. I asked it to shard the dictionary so I wouldn't need a database and I could serve it from github pages. Less than 2 hours later it was done and useful.
Since then I had it add camera support, FSRS flashcard support, and online syncing so I can add and practice from any of the 5-6 machines I touch every day
You can use it if you want here: https://g-sho.org
That said, the bigger takeaway IMO is you can have LLMs make most solutions you need in just a couple of hours. If there is some app / service you use that you wish was different, ask for it and you'll have it.
I'm not saying you shouldn't enjoy programming. I'm just saying I also enjoy this new activity of making solutions quickly. Kind of like making before 3d printers and laser cutters vs after. So many things got easier.
As another example, I needed to capture audio from my Mac's system which Apple's built in capture does not support. OBS was recommended but OBS has a difficult UX. Kap was recommended but it was sus AF. So I asked Claude to build what I needed. 4 minutes later I had a native app that did exactly what I needed (https://github.com/greggman/g-cap). Not saying you should use it. Just showing it as an example of "ask for what you need and get it"