Unix Type-ahead problem

Juergen Wagner gandalf at csli.Stanford.EDU
Sun May 14 08:48:35 AEST 1989


In article <7195 at bsu-cs.bsu.edu> dhesi at bsu-cs.bsu.edu (Rahul Dhesi) writes:
>It's a question of whether you "Think VMS" or "Think UNIX".
...
>But soon I begin to "Think UNIX" again, and I instead type
>     cc x; a.out
>or better still:
>     cc x && a.out

Even better: use a Makefile. By now, there are also `make' clones available
under VMS and other operating systems. Yet, how long did it take until `make'
was accepted as a reasonable tool under VMS (by "accepted" I mean "accepted by
those who built VMS" - No, no names).

UNIX certainly has its deficiencies and problems, and there will always be a
non-trivial number of users who will not be satisfied with a particular
operating system. However, one thing I especially like about UNIX is the
spirit of creating new tools as they are needed. As Rahul mentions, command
pipelines, multiple processes, and other concepts which are inherent features
of UNIX are often found only rudimentary in other systems (how would I do
something like "tbl foo.me | nroff -me | more" without temporary files; why
does it take so long to create a subprocess; ...). The elementary principle
of function composition is one of the most important features of UNIX.

These features are what is needed in a development environment. A production
environment (like VMS) may not need them but developing software might be
less comfortable (and if one tries to embed UNIX in VMS [Eunice], significant
performance problems may arise).

[PS: I used the Very Mysterious System for a few years, but have become
     a highly biased UNIX-user over the last six years.
 flame/output=NLA0:[0,0]/ignore=world/override=permissions/blocksize=0]

-- 
Juergen Wagner		   			gandalf at csli.stanford.edu
						 wagner at arisia.xerox.com



More information about the Comp.unix.questions mailing list