I love it. So much in computers is trade offs and this was a fun read exploring it.
It would be interesting to see some economics of what 8,000% increase in encoding time takes to make that money back in terms of storage or bandwidth. I also wonder how brotli/lzma would compare here. Are there some obscene modes on those that had similar results?
The author made a good point here about running trackers and ad ops (think Google analytics or ad words). I'd guess if you don't run those, it'd just be supply chain attacks that could exfiltrate secrets.
This seems like one of those scenarios where you make different trade offs depending on your threat model. The author's threat model sounds similar to a news site where they track and advertise so they're forced to run semi-trusted js.
How much does http/3 help for server to server traffic? Seems like larger websites can use a CDN or load balancer to do termination and then use http 1.1 to the back end. Is that good enough with large pipes and a high number of connections?
The funny thing with wine is that as the market grows for Linux, more and more bugs will be fixed. I could see a tipping point where even the elusive Adobe Creative Cloud runs on Linux.
I had exactly this! I kept having flare ups and couldn't figure out the pattern until we tore up a bathroom and discovered mold. Suddenly I realized my lungs hurt everytime I had people come stay over. They'd use the spare shower, it'd excite the mold and then I'd need an inhaler. Cheap hotels also put my lungs into a tizzy. Definitely worth checking cause mold is everywhere.
So you're telling me if United goes out and gets a bunch of paper bags to attach to the wings it will reduce drag by 30%? Are there industrial materials that can do this on a 747 and sustain hundreds of hours of flight? I have so many questions. Exciting to see.
XSalsa and XChaCha do something extremely similar. In order to extend the nonce space by 128 bits you basically run a round of ChaCha against 128 bits of your nonce. Those 128 bits represent 96 bits of nonce and 32 bits of counter. Because you know you'll never encode more than 1 full block of data, the counter won't overlap. 256 bits of internal state become the key for the next round of ChaCha plus the remaining 64 bits of nonce and you get to use the full counter space because nonce just became the key.
I wonder if processes like this extend cryptographically in most symmetric ciphers. To add more nonce, encrypt your nonce, output becomes key for next round, add more nonce, repeat. Like all things crypto though, your search space could absolutely collapse if the cipher doesn't have the right properties.
I love the prospect of going to the least popular items on this list. The volunteers sound super passionate and not having to fight lines is wonderful.
Sweet mama speed. Although from what I understand it is more legacy speed cause you're losing all your time to fetching the memory anyway. But when processors were slower it was a meaningful amount.
The fact that msvc generates the unaligned loads for every avx instruction but gcc didn't gave me so many headaches. Most people worked on PC or Xbox and I was on the Playstation team.
"oh boy, another one of these..."
So scary. I was working on a garbage collection bug for a AAA video game one time for months. It'd crash once a day on a tester machine and we'd add more and more prints to try and narrow it down.
Finally got enough information and realized that the padding of a specific object was wrong (GC expected 16 bytes, object was 12 bytes). This caused dozens of other corruption bugs to disappear that we didn't even think were GC related.