manis@faculty.cs.ubc.ca (Vincent Manis) (05/11/88)
There was a day when editors were wired into the operating system, and you
had to use the one you were given. Nowadays, you have your choice, so I think
people should use the editor they like. I'm at an installation where most
people use VMS EDT, which I've never bothered to learn because it strikes me as
rather crufty. But I leave them alone.
The job of an editor is to let you change text quickly and efficiently. If it
doesn't work the way you think, then it isn't quick and efficient. vi hackers
should most certainly use vi for exactly that reason.
The machine I'm currently logged in to has old Unipress Emacs, new Unipress
Emacs, and Gnu-Emacs. I communicate (sic!) over 1200 baud lines, so I haven't
yet got around to doing much with Gnu-Emacs, but the old version of Unipress
(which I know and therefore, in accordance with the previous paragraph, use)
is quite responsive (it's quite responsive at high speeds, too).
I quite often use a PC-XT, an Atari ST, and a Vax running VMS. These are all
primitive machines, and I find that it's particularly convenient to run
Micro-Emacs on them. Micro-Emacs' source code is ghastly, but it's nice to be
using the same editor, with the same key bindings, on three different
machines. I ensure that by compiling the same source files on all three
machines.
What I like about Emacs is not its extensibility nor its power, though both
of those I use (Micro-Emacs has a ghastly extension language reminiscent of
an illegitimate offspring of Basic and Forth. Real Emacs implementations use
Lisp.) More to the point is the "spirit" of Emacs: it provides an environment
for sophisticated programmers. By being an open system (much of it is
written in the extension language), the users can participate in the design.
Good extensions prosper, while poorly written ones wither and die.
Each Emacs user finds the subset of the command space which satisfies her/his
needs. I know that I tend not to think about Emacs commands; rather, most
of the time, I almost unconsciously reach for the right key.
Yes, the learning curve is high, but one doesn't have to learn all or even most
of Emacs at once. What with on-line help, and intelligently named commands,
I find I can generally track down obscure features without consulting the
manual. All one really has to know is that pressing a key with a printable
character inserts that character, and what command puts you into the help
system (this latter command differs from Emacs to Emacs).
I for one am really grateful to rms for having invented a tool which does
so effectively act as a user agent for the tasks for which I use a computer,
and having made that tool so extensible and open that I can modify it for
my needs while still using other people's extensions. (I'm even more amazed
that he could have written the first version in Teco!.)
I don't think that any of the above would make a confirmed vi user switch to
Emacs, and that wasn't my intention. I know that Emacs comes close to
satisfying my needs, and it's so rare that a software package does that
that I felt impelled to comment. If vi, EDT, or the Burroughs 5500 CANDE
editor (ca 1970) does that for other people, great.
Vincent Manis | manis@cs.ubc.ca
The Invisible City of Kitezh | manis@cs.ubc.cdn
Department of Computer Science | manis@ubc.csnet
University of British Columbia | {ihnp4!alberta,uw-beaver,uunet}!
<<NOTE NEW ADDRESS>> | ubc-cs!manis