On Sun, Aug 6, 2023 at 10:51 AM Leah Neukirchen wrote: > Vim was based on Stevie, an Atari ST vi clone he ported to the Amiga 2000. > Thank you. I also remember stevie and that stevie beget vim make so much sense. It was one of the clones I ran back in the day. The point is that there were (are) a number of vi clones. Which is why I started to switch to it. It ran on everything from small 8-bit systems on up. Particularly on the micros, they all had something akin to what Microsoft called edlin, often just called edit. If you knew UNIX ed(1) or any other editor from the old GED family from the 1960s, chances were you could make it do something -- but it was PITA as each was a bit different So copies of the editors from the larger [often DEC based systems] started to appear on the smaller and smaller machines. And vi seemed to be a pretty popular as a starting point[I always suspect Webb Miller's book had something to do with that since 's' was so small and using it to add missing vi features was not terrible]. That said, there was an IBM PC/386 "MS-DOS" version of vi that required ANSI.SYS that was on the market for about $50-$100. It was 'bug for bug' compatible with Joy's version. I always suspected that unlike Keith, they just took the Vax code, stripped out termcap so it did not need an external terminal database, hard coded it for ANSI.SYS, then ran it through the PHARLAP C/386 subsystem. Unlike any of the other 'clones' until nvi -- it was the only one that had some of the same wonkiness. As I understand it from him, Keith did nvi(1) because Joy's original work was based on the v6 ed(1). As part of Keith's effort to try to remove any core code in the BSD releases that had any AT&T source taint, it was just easier to start over. The Berkeley Shell [in 1BSD] beget the C Shell, was derived from Ken's sh in v5 and v6. I think Keith/Kirk et al felt that any of Ken's original code had long been removed - whereas ex(1) (vi is the actually the VIsual command for ex) probably had a lot of core ed(1). Clem ᐧ