Moonlight is mostly designed to stream your gaming desktop to a portable device or your TV at minimal latency and maximum quality within a LAN. For that, 40Mbps is quite reasonable. It's obviously absurd for mundane VNC/productivity workloads.

Note that WWV is a good 60 miles NNE of Boulder on the outskirts of Fort Collins.

You can use fiber optics as an optical delay line too! About 60KB/km at 100Gbps.

https://en.wikipedia.org/wiki/Delay-line_memory

Especially given how the response time doesn't matter much here! If you're just looking at 2-character license plates, that's 676/5=136 requests to check them all, and you could easily space that out to something like one request per minute to scan the space every two hours.

Is the large significands table (~19KB) an issue for cache contention in practical use, or are only a small range of significands generally accessed?

It depends on the input distribution, specifically exponents. It is also possible to compress the table at the cost of additional computation using the method from Dragonbox.

SDL3 is kind of the intersection of features found in DX12/Vulkan 1.0/Metal: if it's not easily supported in all of them, it's not in SDL3-- hence the lack of bindless support. That means you can run on nearly every device in the last 10-15 years.

This "no api" proposal requires hardware from the last 5-10 years :)

Yup you've actually pointed out the most important difference: SDL3 is designed to be compatible with the APIs and devices of the past (2010s), whereas this proposal is designed to be compatible with the newer 2020s batch of consumer devices.

This could probably be implemented as an expensive comment-driven lint during compilation.

I don’t think it can be a true linter because it depends on the compiler. But it’s not a bad idea anyway

He went even farther, decapping the CPU and reverse engineering its entire workings to understand how each "undocumented" opcode functioned: https://www.pagetable.com/?p=517

Note that ffprobe can output JSON which is much easier to handle than CSV. I have this snippet in my bashrc:

ffpj() { for f in "$@"; do ffprobe -v quiet -print_format json -show_format -show_streams "$f"; done }

> First, individuals who choose vaccination may differ from those who do not, potentially introducing confounding bias.

It's very hard to interpret this data given the massive confounder of "antivaxxers are suspicious of healthcare and take more risks".