v3.0.0

This change switches the clear sequence to a new shorter sequence, saving 4 bytes for each of the 10 architectures.

Various small typos were fixed in the documentation, and the documentation formatting was updated.

v2.0.0

Biggest change was adding a riscv64 tiny-clear-elf implementation in preparation for its official support in Debian 13 "Trixie" (currently Debian testing)

Fixed improper CPU flags set in the ELF headers - not that they didn't work, but they were technically improperly set.

Reduced the size of several of the binaries:

architecture	old size	new size	savings
amd64	161 bytes	152 bytes	9 bytes saved
arm64	166 bytes	162 bytes	4 bytes saved
armel	130 bytes	110 bytes	20 bytes saved
armhf	130 bytes	110 bytes	20 bytes saved
i386	125 bytes	117 bytes	8 bytes saved

Lastly, I separated notes from actual documentation, and improved documentation and reassembly instructions.

Note that the actual tiny-clear-elf executables for mipsel, misp64el, ppc64el, and s390x are identical to their 1.0.0 versions, as none of the CPU flags appeared to be improperly set, and none of my attempts to shave off bytes worked on any of them.

v1.0.0

This initial release contains builds of tiny-clear-elf appropriate for use on all officially supported architectures of Debian GNU/Linux* Bullseye. Future releases may or may not exist.

* as opposed to Debian GNU/kFreeBSD, Debian GNU/Hurd, or Debian GNU/Win32 - I'm not interested in interjecting for a moment.

Full Changelog: https://github.com/eliminmax/tiny-clear-elf/commits/v1.0.0