Cacheflowe

Cacheflowe

@cacheflowe.bsky.social

Experiential software developer for museums, brands, games, art, fashion, data viz, music, education. Vegetarian in Boston. Tech at Hovercraft Studio. https://cacheflowe.com

930 followers441 following101 posts

Say what you will about Boston (I’m not going to defend it at all) but this is great to see. Proud Masshole here 💪 I went to a smaller local town protest and everybody was joyful, highly educated, and ready to p!ss on Trump’s grave.

Indivisible ❌👑@indivisible.org

No ICE. No war. No Kings. #nokings

Taking my “moonshot” at this fun interactive physical basketball video game that we built. I led the tech components with my multidisciplinary crew at Hovercraft Studio 🏀

Cacheflowe reposted

I have friends who are/were in 18f & USDS. “The one person” is burning every single institution down to claim it all. It’s tragic, and unfortunately nobody knows the depths unless they’re on Bluesky and subjects themselves to the daily horrors.

zelDemon@zeldman.bsky.social

The one person in our world who needs to hear this will never read it, but you can. And you should. Every word! Hat tip: @andrewhinton.bsky.social

I was looking at the official TSA info about opting out of the facial scan at airports (you can easily opt out with no trouble), and noticed a Truth social icon in the footer. These mfrs are running scams at every possible level.

America, brought to you by: Truth steaks with extra microplastics, fake information with extra cruelty, botched lip filler injections, bigoted children’s programming, DEI for failed white guys, and concentration camps with that vintage formulation that your racist granddaddy loved.

Cacheflowe reposted

Used Claude Code/Gemini to help me make a CEF browser->Spout2 simple app for Windows that ill release soon. Going to do the same for Syphon on MacOS because why not. There were old implementations of this, but most were abandoned. Let me know if this sounds useful!

I've been working towards implementing a number of onnx computer vision models directly in TouchDesigner. It's wild how much python I'm writing these days. This is DepthAnythingV2 running at 60fps with a 256x256 input. ~10ms inference time on a laptop 4090 🤯

Cacheflowe reposted

"Hyperspace" for(float i,z,d,s;i++<1e2;){vec3 p=z*normalize(vec3(FC.rg*2.-r,0));for(d=6.;d<2e2;d+=d)p+=sin(p.yzx*d-t)/d;z+=d=.005+abs(s=.3-abs(p.y))/4.;o+=(cos(s/.1+p.x+t-vec4(0,1,2,3)-3.)+1.5)*exp(s*9.)/d;}o=tanh(o*o/2e8);

Cacheflowe reposted

A recent client project. I built a custom solution for touch detection on a curved LED screen, and an SVG-powered content & animation system for quick iteration on design & copy updates.

Excellent vintage-but-more-relevant-than-ever article by @mrprudence.bsky.social 🙏 I’ve recently been working in more node-based environments and would note their increasing prevalence. This is a great read.