Nat (she/they)

Number = IEEE-754 double precision float, which includes NaN, so it makes sense

Could definitely be worse for latency in particular cases, but if we imagine a write heavy workload it still might win. Writing a byte/word basically has to do the same thing: read, modify write of cache lines, it just doesn’t confuse the dependency tracking quite as much. So rather than stalling on a read, I think that would end up stalling on store buffers. Writing to bits usually means less memory, and thus less memory to read in that read-modify-write part, so it might still be faster.

A lot of times using less memory is actually better for performance because the main bottleneck is memory bandwidth or latency.

Is it backports for an old version?

Made with AI? Really? What would GNU think?

Minecraft mod users are the worst. They’ll post a screenshot of a version mismatch or dependency error that literally tells them how to resolve it and ask “hey, this mod isn’t working, how do I fix it?”

These things add up if you’re doing them all over a 1 million line codebase, by which point it’s incredibly painful to claw back performance if you need it.

I thought the joke here was that your unbalanced parentheses would make me angry (they did))

How convoluted was it?

Parts of it seem to be inherently more secure, but there are some pretty glaring holes. At least software distribution is much more secure than the Windows approach.

Arch Linux is a good vision and a tab for the meds

Listening to employees when making decisions, what a concept! It’s a shame many places don’t do that.

Most packages are purely additive to to system. If GNOME is part of the base system, I don’t care because I can just not use it. For packages that are mutually exclusive, well, usually that’s the distro picking it for you anyway, but if you insist on changing them then OverlayFS lets you mask files in the base.

For something like Arch or Gentoo, the read-only partition approach absolutely won’t work, but I know Fedora’s been working on an OSTree immutable approach, so it’s still technically a mutable partition but it’s defined declaratively and is still easy to roll back.

Immutable partitions are amazing for reliability, then you can just OverlayFS your mutable state on top of it

If it was on something like BTRFS it’d probably be fine, though I imagine there’s still a small window where the FS could flush while the file is being written. renameat2 has the EXCHANGE flag to atomically switch 2 files, so if arch maintainers want to fix it they could do

1. Write to temporary file
2. Fsync temporary file
3. Renameat2 EXCHANGE temporary and target
4. Fsync directory (optional, since a background flush would still be atomic, just might take some time)

I think my BIOS has a setting to skip that part

32G DDR4-3200

2s in firmware??? I’m used to at least 30s

Do you get two empty spaces next to your tower? For maintenance if the lower elements.

It’s called a tower PC for a reason