Pinned Tweet
Minimalist interactive star chart web app that I started a few years ago, then set aside. One day I will pick it back up again.
Philip Rideout retweeted
I just put up a new video, which was a collaboration with Terence Tao about the cosmic distance ladder. You can find the full video on YouTube, and here's a bit of extra footage that didn't make it into the final.
Philip Rideout retweeted
We JUST solved a hard open math problem— the moving sofa problem!
"What is the area of the largest shape that can be manuevered through an L corner of width 1?"
On Dec 2, 2024, Korean Jineon Baek proved the answer is the 18 curve section Gerver's sofa with area 2.2195..!
Philip Rideout retweeted
#Manifold v3.0 is out! This is a huge release - we have removed *all* required dependencies.
Our npm package is half the size and twice the speed. Our #SDF LevelSet is much faster and higher quality.
And so much more: github.com/elalish/manifold/…
Philip Rideout retweeted
Hola graphics people! ❤️
I've been meaning to do some write-ups that are hopefully useful, so this first one is about my approach to Order-Independent Transparency.
osor.io/OIT
🙌Super excited to finally release this!🙌
I hope you enjoy it!
Philip Rideout retweeted
Implemented reverse depth in three.js with EXT_clip_control for [0, 1] NDC range. This is a strict improvement to logarithmic depth buffer where supported, namely in both performance and accuracy. Also works well with MSAA, unlike logarithmic depth.
github.com/mrdoob/three.js/p…
e.g. this looks interesting:
github.com/GAIR-NLP/anole
Philip Rideout retweeted
The development of Embree, OIDN and other Rendering Toolkit libraries is now open and hosted directly on GitHub under Apache 2.0, making it easier for the community to contribute features, fostering cross-company and cross-vendor development. More community.intel.com/t5/Blogs…
Philip Rideout retweeted
A few technical insights on the new Llama vision models we’re releasing today 🦙🧵
Alignment is crucial to building well-performing LLMs, so this paper seems quite important to me. Apparently you don't need PPO for doing RL.
arxiv.org/abs/2402.14740
I'm a newbie but it seems like many of deep learning's significant advances were actually just simplifications.
New research reveals that LLMs are smarter when you tell them to imagine an episode of Star Trek.
arxiv.org/html/2402.10949v2
Philip Rideout retweeted
TIL that @chenfanfujiang, @minchen_li_, and others recently released a super thorough, free online book that serves as a primer on physics-based simulation systems. Holy cow what I wouldn't have given to have had this back when I was in school...
phys-sim-book.github.io
Philip Rideout retweeted
typing with a broken hand is tiresome, it's been a major hinderance over the last weeks. i missed trying to crack interesting design puzzles on the internet. this one is from discourse "how to push meshes away on hover, similar to sparkk dot fr/en"
codesandbox.io/p/sandbox/ele…
Philip Rideout retweeted
Excited to share *two* papers appearing at #SIGGRAPHAsia2021, on "Repulsive Curves" and "Repulsive Surfaces."
Tons of graphics algorithms find nice distributions of points by minimizing a "repulsive" energy.
But what if you need to nicely distribute curves or surfaces? (1/14)
Rustaceans: if my function does several passes over a container such that I don't want to move out items, is it better style to do `for item in &foo` or `for item in foo.iter()`? Clippy doesn't seem to have an opinion.
If a web app is much faster while recording a performance profile, is this a hint of a specific bottleneck? Should I even bother looking at the resulting flamegraph?
Curious what is Chrome doing differently when recording a profile.
I love how Rust actually _does_ support function overloading, in a sense. For example, this takes any uint-like type for the index:
fn get_wombat<T: Into<usize>>(&self, index: T) -> Wombat {
...
}
For other use cases, I should be using a more precise function name anyway.
We like Rust at Arcol… and we like collaborative data stores too!
IMO it's clear that Rust is the future of databases and data infra. There is a small group of people who are leading this transition. @carlsverre is one of them. He's one of the best engineers I've met in my career and also one of the best humans.
Philip Rideout retweeted
Nope. When @threejs gets better, the web gets better! When Babylon gets better, the web gets better. We all win.