ClipEgg: We Confused Copying with Hoarding
Copy is broken. It has been for 40 years.
When you copy something, what do you actually want?
A reference to a thing. A way to say "this, here, I want to use this somewhere else."
What do you get instead?
Everything. The image. The formatting. The font declarations. The metadata. The entire DOM subtree. A 6MB cargo truck full of furniture when all you needed was the address.
We confused copying with hoarding.
And the math is insane.
40 billion copy operations happen every day. Average payload: 100KB of stuff nobody asked for. That's 4 exabytes sitting in RAM, syncing across devices, getting pasted into AI tools that charge per token.
The energy cost: 33 million tons of CO2 annually. Seven million cars worth. For clipboard formatting.
Old copy: Here's everything. Hoard it.
New copy: Here's a reference. Fetch what you need, when you need it, if you're allowed.
So I built the answer.
An egg. 200 bytes. A claim ticket, not cargo.
The receiver cracks it open based on context. Paste into a terminal? Plain text. Paste into Figma? Hydrate the full asset. Paste into a system that shouldn't have access? Nothing.
Copy = Frosting. Cut = Cake.
See it work: https://daaaave-atx.github.io/clipe96
Spec: https://github.com/daaaave-ATX/clipe96
Nice, COM embedding makes a comeback.
Similar spirit, but OLE was platform-locked and had no access control. ClipEgg adds paste-time auth, revocation, and audit trails. The clipboard becomes a security boundary, not just a convenience.