Asdcsvax.302
NET.general
utzoo!duke!decvax!ucbvax!sdcsvax!jmcg
Wed May 27 05:14:07 1981
EUNICE report
We received EUNICE several weeks ago.  This review is mostly first
impressions.  (Some of the criticisms are more properly directed at
VMS.)

1) EUNICE does almost all of the things that it claims to do, but not
all the things you may have been led to expect.  The emulation of the
UNIX environment is sufficiently good that programs using only stdio
seem to run fine, providing they don't try to construct file names that
do not map into sensible VMS file names.  In particular, `ex', `vi' and
the collection of standard tools work indistinguishably from the way
they work under UNIX.

2) EUNICE documentation is inadequate under the circumstances of
binary-only distribution.  This is a bit overstated, actually.   EUNICE
itself is about as well documented as UNIX system calls are, but the
software supplied to UNIX licensees is simply there.  The `csh' limps
until you discover or divine that `.cshrc' becomes `cshrc.csh', the
properties of the added command `dcl' are perplexing, and there seem to
be a number of fairly standard csh features which don't work (it's a
3BSD level csh), e. g.
       (date;date)&
only prints the date once, and the csh terminates.   Although codefiles
are supplied for UUCP, I haven't been able yet to configure things so
it works (configuring uucp has always been a pain, though).

3) EUNICE requires extra processes.  This is not normally a problem,
but our VMS system had been tuned to balance space requirements for the
swap file against maximizing the working set for one large batch job
and EUNICE pushed this balance towards a lot more page faults.

4) EUNICE is incomplete.  I haven't had a chance to talk to Dave
Kashtan since I put EUNICE up, but I understand he's working on the
command interpreter interface business to allow sensible execution of DCL
commands.  This is important, since UNIX users tend to strongly dislike
interacting with DCL.  As it is, there are too many necessary things
that require an escape out to DCL.  [Escaping into DCL is a problem for
another reason.  One's EUNICE-supported environment disappears when one
drops out of csh, so it is preferable to escape `upward' using the
`dcl' command of the csh.  The DCL one gets has not preserved any
defined symbols nor does it pick up the login.com file (there is no
equivalent to .cshrc), so extra work is often required to re-adapt DCL
before it can be used.  It also won't go away until one types a ^C and
can't be invoked as a part of a shell script.]

5) EUNICE does not encourage tidiness.  Naive users will tend to have a
large number of file versions laying around.  This is problem with VMS
anyway, but the tools to deal with the problem are one step removed
from EUNICE users.

6) EUNICE is a lot better than the old version of Software Tools, but
the newer Software Tools VOS should be considered by operations where
unix source-level compatability is less of a concern.

If you *must* run VMS on your VAX, having EUNICE available is a
decided advantage, but not a panacea.

                                               Jim McGinness
                                               U C San Diego; Chemistry
                                               ucbvax!sdcsvax!jmcg
                                               (714)452-4016
P. S.  Perhaps there should be a new newsgroup: NET.eunice?

-----------------------------------------------------------------
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.