pluto_modadic

Aphyr does great work, Jepsen always kicks it out of the park :)

citation needed. you have an alternative?

no, it is not enabled by default.

isn't the other part of this that you'd have to enable the SSH part per node? it's not like the tailscale control plane can uniformly turn that on.

Okay - so... they're saying... don't use p384? :3

my line for "self hosted" isn't "must be on my own hardware", but "must be on my own linux VPS"

if you enforce user input to specific shapes... and only have certain system calls... I think a unikernel version of a hackernews style site really wouldn't have RCE... especially if you build in KSLR / memory tagging in the unikernel security model...

handing clients access to an LLM gets them beginner level code, easily. same with beginner level exploits. an RCE against google is not a "beginner level exploit". If you've seen AI generated code, I can tell you that AI generated security findings also have that "beginner" smell. an expert who knows what they're doing has to drive them to look at the right places, and use the experts preferred method.

the idea of a caricature is to exaggerate the features someone is most insecure about. That's what Spitting Image is.

Some assembly required.

they could have gone with an oblivious transfer approach (where it's working on what looks like multiple problems at once, you don't know which)

1. score by the past paragraph (this is also how the verifier checks without needing a GPU, it takes the previous length of text)

2. same answer

3. no, it just needs the previous text, private key, and the matrix math (CPU is fine)

4. no, see above.

Seems like Zed needs to clean up shop and fix their IDE first.

what was the hardest part about this project?

by rolling your own they don't mean write a VPN from scratch, but host a VPN on a VPS.