Xref: utzoo comp.editors:2670 alt.sys.sun:2985
Path: utzoo!utgpu!cs.utexas.edu!uunet!auspex!guy
From: guy@auspex.auspex.com (Guy Harris)
Newsgroups: comp.editors,alt.sys.sun
Subject: Re: resizing shelltool containing vi
Message-ID: <6495@auspex.auspex.com>
Date: 8 Mar 91 19:13:58 GMT
References: <1991Mar06.034658.17775@convex.com> <1167@saltillo.cs.utexas.edu> <1991Mar06.220840.19319@convex.com>
Followup-To: comp.editors
Organization: Auspex Systems, Santa Clara
Lines: 32

>What an act of monumental stupidity!!!  They switch from a BSD vi to a SysV
>one (which is pretty dumb -- think where vi came from and who wrote it) 
>and they lose the ability to handle window size changes.

Nice rant, but unfortunately, it doesn't match reality:

1) the ability to handle window size changes was removed from the
   *B*S*D* version of "vi" by Sun, not from the S5 version.  The switch
   from a BSD "vi" to a System V "vi" had nothing to do with that; had we made
   the BSD handle window size changes *properly* (which, as I've noted,
   the BSD version does *NOT* do), we would have folded that change into
   the S5 "vi", just as we folded the other bug fixes, and the "tag
   stack" stuff, into the S5 "vi".

2) the reason Sun switched to the S5 version of "vi" was that the latter
   version can cope with 8-bit character sets, while the BSD version
   can't, and a large amount of ugly surgery would have been involved in
   an attempt to fix up the BSD version.  Yes, there *are* people who
   want "vi" to handle 8-bit characters; in a recent posting in one of
   the "comp.unix.*" groups, somebody asked whether there was any way to
   get their "vi" to do so.  (The fact that it also supported "set
   showmode" was a nice bonus; yes, people had complained about the
   lack of *that* as well in the previous SunOS "vi".)

As for "where 'vi' came from and who wrote it", well, "who wrote it"
wasn't *at* Berkeley any more, and hadn't been there for a while, and
hadn't had anything to do with "vi" for a while.  I agree, it would have
been *wonderful* to have stuck our esteemed Vice President for Research
and Development in a room and said "you're not leaving this room until
you 1) fix up the other guy's botch in implementing SIGWINCH handling in
"vi" and 2) make the damn thing handle 8-bit character sets", but
somehow, I don't think that would have been possible....