Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!gem.mps.ohio-state.edu!apple!bloom-beacon!primerd!hollin!ds
From: ds@hollin.prime.com
Newsgroups: gnu.emacs
Subject: Re: remote file visiting?
Message-ID: <81100006@hollin>
Date: 16 Oct 89 19:54:00 GMT
References: <89@<CREW>
Lines: 26
Nf-ID: #R:<CREW:-8900:hollin:81100006:000:1319
Nf-From: hollin.prime.com!ds    Oct 16 15:54:00 1989


I, too, have wished for transparent EMACS buffers connected to my account on
remote systems not running NFS, and I too imagine that someone somewhere must
be working on this.

I have also wished for single-character remote telnet connections, so I could
run EMACS on a remote machine from inside a local EMACS window.

I conjecture that writing such extensions may be difficult because of various
UNIX (tm) limitations, such as the failure of some implementations of rsh and
its C library counterparts to pass through the error code returned by the
command executed in the remote process, default buffering differences between
stdout and stderr that can cause output to be shuffled into an unexpected
order, well-intentioned security limitations on creating or managing remote
processes, the lack of an ASCII map for specifying which characters are to be
buffered locally (usually printing chars and space) and which should cause a
transfer and wakeup the process (usually LF and other control chars), etc.

PS - I wrote an extension to the EMACS telnet command that automatically logs
me into the remote system (useful when the system's UNIX disallows automatic
telnet login), but the telnet command is not the same as transparent
buffering.

David Spector
Prime Computer, Inc.
ds@primerd.prime.com (until the layoff)