Hacker News
points by yjftsjthsd-h 1 day ago | 0 comments
> Everything else should probably be left to whatever is running at the initialization vector of the storage. Which, if you'd like, can be a complete UEFI implementation.

Putting it all in a single location on the storage device means that 1. Multibooting will always suck because Windows will unconditionally overwrite whatever is there with its own bootloader that only boots Windows, and 2. you now have to have local storage, which would be somewhat annoying to those of us trying to do diskless network boot (or anything else more interesting than "always boot internal disk0").

[–] za_creature 1 day ago
You always have local storage, even if it's a 128Mbit flash chip that stores the firmware. There's no network boot implemented in hardware.
[–] yjftsjthsd-h 1 day ago
I mean, basically just sounds like the status quo but with a smaller initial firmware chaining to a user-flashable second stage. Which I'd certainly support, but that... just sounds like coreboot? (Which is an initial firmware that usually gets combined with tianocore to provide UEFI or seabios or ...)
[–] za_creature 21 hours ago
Sadly, coreboot is not a standard, though I guess it could very well become one since I doubt there are many uefi implementations that diverge significantly from tianocore.

Even then, I'd rather have a modular mix-and-match approach (e.g. vendors only provide drivers, completely independent from the stage 2 payload) rather than the current fork, tweak, add kitchen sink, compile & forget status quo.