> Open a page and you have a Linux machine with a shell, a package manager, a persistent disk, internet access, and a screen with sound. It runs entirely on your device, with nothing to install and no server...
> Much of what makes a system feel slow lives in the kernel: starting processes, syscalls, page faults, file and network I/O. In unkernel that path runs at wasm speed.
> Boot: the kernel reaches /init about 70 ms after it starts, and the Alpine prompt appears 0.3 s after node launches. The page downloads about 10 MB, and Alpine boots in 48 MiB of guest memory.
> Syscalls at close to native cost: getppid takes about 170 ns (110 native, 1.7 µs under qemu-system), a 64-byte read about 170 ns (160 native).
> fork at native speed: fork + exit + wait takes about 60 µs, against about 70 µs natively and 1.2 ms under qemu-system.
Remember when the cool new stuff was actually cool? Nowadays it's all vibe-coded slop. IMHO the depressing thing is if people get used to this and actual programming craftmanship isn't recognized as such anymore. In fact I believe maintainers have already mistaken me for an AI. Patched accepted without any feedback whatsoever, no thank you, nothing. But that's okay, I should just submit more fixes and it'll be fine.
So... What are the network constraints when running this in a browser?
What was the motivation behind this? Do you have a use case or did you do it just because? No judgment - truly impressive!
looking at https://github.com/pithings/unkernel
> Open a page and you have a Linux machine with a shell, a package manager, a persistent disk, internet access, and a screen with sound. It runs entirely on your device, with nothing to install and no server...
> Much of what makes a system feel slow lives in the kernel: starting processes, syscalls, page faults, file and network I/O. In unkernel that path runs at wasm speed. > Boot: the kernel reaches /init about 70 ms after it starts, and the Alpine prompt appears 0.3 s after node launches. The page downloads about 10 MB, and Alpine boots in 48 MiB of guest memory. > Syscalls at close to native cost: getppid takes about 170 ns (110 native, 1.7 µs under qemu-system), a 64-byte read about 170 ns (160 native). > fork at native speed: fork + exit + wait takes about 60 µs, against about 70 µs natively and 1.2 ms under qemu-system.
This is completely misleading and it isn't the Linux kernel in "WASM".
It is just a vibe-coded RISC-V emulator in WASM that emulates Linux on RISC-V using a JIT.
Note the kernel version: "Linux unkernel 7.3.4+ #3 Mon Nov 16 11:55:34 UTC 2026 riscv64 Linux"
Do we not actually look underneath at how things work or do we just read whatever is in the title and take it at face value?
My understanding is this Linux kernel is compiled as a wasm64 module, while the user space is riscv64 Linux ELFs, running on a emulated software CPU.
Remember when the cool new stuff was actually cool? Nowadays it's all vibe-coded slop. IMHO the depressing thing is if people get used to this and actual programming craftmanship isn't recognized as such anymore. In fact I believe maintainers have already mistaken me for an AI. Patched accepted without any feedback whatsoever, no thank you, nothing. But that's okay, I should just submit more fixes and it'll be fine.
Can I install kernel modules?
[dead]