Whow, another well-known piece of software that was written by Fabrice Bellard. He's also the original author of qemu, tinyemu, tcc, ffmpeg and many more.
One of the most influential programmers of our time, if not the most.
His track record is exceptional, he must be a Godlike programmer!
I'm big fan of Monsieur Bellard, not because he made things complicated, but because he made things simple.
I read through the original source of qemu and the tiny C compiler and the simplicity and beauty of the code are outstanding.
The only other person I can think of with a similar rank is John Carmack.
It's almost shocking. It's not only the productivity of making these tools, it's how well they fit the "developer zeitgeist" and how useful they proved to be.
I wish the code page https://bellard.org/lzexe/ had gone more into the inspiration of pklite. Used pkware's tools a lot growing up.
"I wrote LZEXE in 1989 and 1990 when I was 17."
Incidentally, PC BIOSes used the LZ* family of compression algorithms too. LZSS (also known as LZ12/4 for its allocation of indicator word bits), LZARI, LZHUF (which lead to the famous LHA/LZH, and then Deflate/Zlib, ZIP, etc.), and LZINT were all commonly encountered. Apparently Phoenix had a patent on it:
https://patents.google.com/patent/US5836013A/en (search for LZSS)
Despite the relative obscurity of Okumura's code, it has definitely had a huge impact.
That brings back some memories...
Back in the 90s, there was a whole scene around exe/com compression and protection tools. ("Protection" in the sense that people figured out if they compress their executables, that also mean you cannot simply modify strings in them any more, and that was expanded to all kinds of anti-debugging protection. However, it never lasted long until the next unpacker was able to break it.)
I never acquired the skills to write such tools myself, but I wrote a detection tool and ran a mailing list.
Or in other wors, in case you were around at that time: I'm the author of chkexe and ran the exe mailing list.
[dead]