I was surprised that I've learned many more useful-to-me things while writing "The Secret Rules of the Terminal" than when writing literally any of my other zines.

I kind of thought "I've used the terminal almost every day for 20 years! i know this!" but that was wrong and I now know a few new tricks & feel more confident dealing with weird terminal issues

(the zine will be out June 24! you can sign up here to get email when it's out! https://wizardzines.com/zine-announcements/)

Meanings of the smoke for the upcoming papal conclave:

Black smoke means they're burning oil, which could be a sign of piston ring or valve stem seal wear

White smoke means they're burning coolant and the papal head gasket has blown

🧵 THREAD: A federal whistleblower just dropped one of the most disturbing cybersecurity disclosures I’ve ever read.

He's saying DOGE came in, data went out, and Russians started attempting logins with new valid DOGE passwords

Media's coverage wasn't detailed enough so I dug into his testimony:

Currently available Go fuzzing tools were missing critical features - some don’t play well with the latest Go toolchain. So we set out to change that.

@bruno, Nils Ollrogge, and colleagues explored more powerful ways to fuzz Go binaries. By tapping into Go’s native instrumentation — which is compatible with libFuzzer — we enabled effective fuzzing of Go code using LibAFL.

We’ve documented our approach and shared insights in our latest blog post: https://www.srlabs.de/blog-post/golibafl---fuzzing-go-binaries-using-libafl

Repo: https://github.com/srlabs/golibafl

Oh wow. This just in from a CISA spokesperson:

“The CVE Program is invaluable to cyber community and a priority of CISA. Last night, CISA executed the option period on the contract to ensure there will be no lapse in critical CVE services. We appreciate our partners’ and stakeholders’ patience.”

“WSJ: The AI industry spent 17x more on Nvidia chips than it brought in in revenue”

Linking to a reddit thread just for the headline quote they found in a paywalled WSJ article feels a bit weird but…

This 17x number is just for chips so the actual cost x revenue multiplier is much higher in reality https://old.reddit.com/r/MachineLearning/comments/1bs1ebl/wsj_the_ai_industry_spent_17x_more_on_nvidia/

Me before finding out about the latest release of k8s: K8s is alright I guess, lots of cool people work on it, but it's not for me, and i feel like a lot of people use it when they don't need it, so i'm generally not a fan.

Me finding out the latest k8s release is called uwubernetes and has an adorable logo: Alright when are we deploying k8s.