Aucbvax.5897
fa.info-vax
utcsrgv!utzoo!decvax!ucbvax!info-vax
Thu Jan 21 17:15:53 1982
Let's get some things straight...
>From CHAVEZ@HARV-10 Thu Jan 21 17:10:00 1982
As regards Terry Laskodi's message of 20 January,
VAX/VMS DOES INDEED have several very nice facilities for
the customized handling of terminal input; correct me all
you VMS whizzes if I'm wrong, but it is possible (and
indeed quite simple) to get the VMS terminal driver to
let you do whatever you want with any character that
comes along. Since VMS I/O is fundamentally asynchronous,
it is possible to do several things (i.e. trap to a
specified subroutine whenever ^E is typed) that,
I believe, are not quite so easy to manage in Unix.
For those who are less adventurous, there exist
compromises in which you can let VMS do most of the
work (echoing, rubouts, etc.) AND get rid of some
potentially annoying things (like form-feed processing.)
I am, moreover, convinced that VMS is much better for
real-time processing -- I've seen some absolutely
amazing interactive graphics (after the manner of
video games like Centipede et al.) done on VMS. In
short, I haven't seen a single operating system that
offers more I/O flexibility than VMS -- and godly
system privileges are NOT needed for most of the
interesting applications (I managed to do some
very amusing things with my highly restricted
account on a VAX at the Air Force Weapons Lab.)
Even so, I am equally convinced that
DCL is a well-intentioned but misbegotten creature.
DCL is powerful, to be sure, but the Shell is much
nicer (from an abstract, CS point of view, admittedly.)
However, DCL is not to be confused with VMS. If
someone will only get around to writing a shell-like
command interface for VMS (is this what's being done
in the Eunice project?), I would be thrilled. The
problem of making another command interpreter for
VMS is not insurmountable; the standard login
program keeps an entry for every user that specifies
the command interpreter (usually DCL, sometimes
MCR, but it can be ANY specially formatted file
in SYS$SYSTEM.) I would, of course, hope that the
writers of Eunice will not insist on making VMS
look EXACTLY like Unix; instead, I would urge them
to consider refining the shell in the process.
I reiterate my position that VMS's only
major flaw is DCL; with a Unix-like interface
and with VMS's better features (real-time scheduling,
asynchronous posting of a variety of events,
shared and map-on-demand memory pages, direct
memory access for graphics/laboratory devices,
inter-process communication, on-line installation
of device drivers, and all that neat stuff),
VMS would be an amazing operating system indeed.
Such a system WOULD be the best of both worlds:
powerful, clean in concept and design, efficient,
fast, easy to use, flexible, and extensible.
Comments and criticisms are welcomed.
Regards,
R.M. Chavez
-----------------------------------------------------------------
gopher://quux.org/ conversion by John Goerzen <
[email protected]>
of
http://communication.ucsd.edu/A-News/
This Usenet Oldnews Archive
article may be copied and distributed freely, provided:
1. There is no money collected for the text(s) of the articles.
2. The following notice remains appended to each copy:
The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996
Bruce Jones, Henry Spencer, David Wiseman.
