home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.unix.wizards:4653 comp.unix.shell:4674 comp.unix.misc:4140
- Newsgroups: comp.unix.wizards,comp.unix.shell,comp.unix.misc
- Path: sparky!uunet!utcsri!newsflash.concordia.ca!sifon!thunder.mcrcim.mcgill.edu!mouse
- From: mouse@thunder.mcrcim.mcgill.edu (der Mouse)
- Subject: Re: The Problem with UNIX
- Message-ID: <1992Nov13.091914.6799@thunder.mcrcim.mcgill.edu>
- Followup-To: comp.unix.wizards,comp.unix.shell
- Organization: McGill Research Centre for Intelligent Machines
- References: <1992Nov9.172715.16367@cs.wisc.edu> <1992Nov11.194557.16258@yarc.uucp>
- Date: Fri, 13 Nov 92 09:19:14 GMT
- Lines: 77
-
- [ I am removing comp.unix.misc from followups. It's posted to two
- other comp.unix.* groups; comp.unix.misc is inappropriate. ]
-
- In article <1992Nov11.194557.16258@yarc.uucp>, scott@yarc.uucp (Scott Beckstead) writes:
- > In article <aldavi01.721333614@starbase.spd.louisville.edu> Arlie Davis <aldavi01@starbase.spd.louisville.edu> writes:
- [stuff we've all seen plenty of times already]
-
- > I agree an experienced user should be able to recognize and avoid the
- > problems above. What's wrong with the shell catching such obvious
- > errors and either reporting them or taking other appropriate action
- > (ie correcting them).
-
- (Is that an unfinished sentence, or a question with no question mark?
- Given context, I'll assume the latter.) What's wrong with it? First
- of all, it's impossible. Specifically,
-
- cat a b > b
-
- How is the shell supposed to know that cat is going to interpret its
- arguments as filenames, and what's more, output the contents of those
- files? Consider instead "echo a b > b".[%]
-
- rm -filename
-
- How is the shell supposed to know whether the user intended that
- argument as a filename to remove or an option to rm? In at least some
- cases (eg, -i) only the user knows.
-
- mail someone < a.out
-
- How is the shell supposed to know that mail takes only printable text?
- It can't. Perhaps this version of mail is a new whizbang mailer in
- test that does take binary files....
-
- In all of these cases, the shell has to know quite a lot about the
- semantics of the program being run. It cannot know all about every
- program, and I submit that partially protecting users from themselves
- is worse than not doing it at all.
-
- Besides all that, these aren't always mistakes. I once wanted to fill
- up a filesystem in a hurry, so I copied /vmunix to foo and then did
- "cat foo >> foo". I was *extremely* annoyed when cat said something
- along the lines of "input foo is output" and refused to run. I had to
- waste a pipe and do "cat foo | cat >> foo", and then realized that the
- extra copies involved in the pipe were slowing down the operation and
- switched to using dd, which mercifully had not been broken by some
- overzealous fool trying to keep people from shooting themselves in the
- foot.
-
- > I realize that this will get me screamed at by all you guru types
- > that advocate that all shell users be rocket scientists or why bother
- > using a computer anyway.
-
- I don't see what rockets have to do with it. *deadpan look*
-
- No, you misunderstand. Not "why bother using a computer anyway" but
- "why use a system which wasn't designed for you and isn't suited to
- you". UNIX was built by hackers for hackers, and hackers seem to like
- it just fine. If you don't like it, by all means, don't use it.
-
- > However I see no reason NOT to research the subject. If UNIX could
- > be made more user friendly by modifying the shell it might gain more
- > acceptance in the general microcomputer market place, GUIs not
- > withstanding.
-
- Yes, it might. So what? Is "acceptance in the general microcomputer
- market place" something we wish for UNIX? If it means dumbing down the
- interface until people no longer have enough rope to hang themselves,
- well, you can if you want, but I'll keep using what I've got.
-
- [%] Yes, I know there are shells with echo built in. The point I'm
- trying to make is still valid.
-
- der Mouse
-
- mouse@larry.mcrcim.mcgill.edu
-