NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/comp/os/coherent/5801

Text File | 1992-11-22 | 6.0 KB | 116 lines

Newsgroups: comp.os.coherent
Subject: Re: Tcl to replacement most of /bin & /usr/bin (was: Tcl on Linux
Path: sparky!uunet!think.com!unixland!rmkhome!rmk
From: [e-mail] (Rick Kelly)
Organization: The Man With Ten Cats
Date: Sun, 22 Nov 1992 20:01:49 GMT
Reply-To: [e-mail] (Rick Kelly)
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]>
Lines: 104

In article <[e-mail]> "Coherent operating system" <[e-mail]> writes:
>
>    Its great that you get such smart users.  Fine.  But I've seen people
>that have a hard time remembering where the power switch is.  A friend of mine
>teaches classes for the Apple Macintosh.  Here you have a machine designed for
>intuitive use and some people still don't get it.

When the Mac first came out, it had practically no documentation.  An
interactive tutorial would teach you how to use them machine.

These days, with Sytem 7.x, a user has to learn about a fair amount of the
OS just to install it on his system.

>    Sometimes I have taken people into the lab and sat them down in front
>of a workstation and I can teach them anything.  It seems to work one on one
>but most won't even try.

In a work environment, where people are expected to perform simple tasks
on a computers, the logical thing is to document everything with procedures
that follow a standard used across the whole company.  I've written hundreds
of procedures for users.  In the real world, if a worker can't follow a
simple procedure to login to a system and enter test data in a database, then
they are in the wrong line of work.  To make things even more simple, the
user's .profile can be set up to put him in the application that he needs
to use, and log him out automatically.

>    I used to be a computer operator for a large main office.  I ran a
>Data General MV20000 Eclipse with 250 users.  It was pure hell from day one.
>The system was perfectly reliable.  Its just that no one seemed to ever use
>their brain.  So, maybe I've just never gotten a chance to do thing my way.
>I didn't like the way users were trained there but some of them were simply
>a few cylinders shy.  The had just never used a computer and the system was
>complex.  Some simple changes like I've suggested would have made it work
>fine.  I still remember what most of them asked for and it wasn't a problem
>to implement.  It just either wasn't or people resisted it.

That's an organization problem, not a computer problem.

>    About AT&T contract.  As I said before.  I don't care if they have
>a legal right to the code.  They didn't develop it and often act as though
>they did.  The people that did the work aren't receiving the benefit.  Of
>course, most of them did it for the fun of it or just because they enjoyed
>it.  AT&T often steps on the people that put them where they are.

I take it the you've never seen AT&T source code.  I've worked with SVR3
and SVR4 source.

In the SVR4 sources you will find copyrights attached by Berkeley, Microsoft,
SCO, and others.  AT&T has no right to steal copyrighted code any more than
anyone else, and they dutifully note other organizations copyrights where
applicable.  If someone has written copyrighted code, and finds that AT&T
is using it without attribution, they can sue.  If someone writes software
and releases it as public domain, then all rights are gone.

>    I agree that kernel bloat is a problem.  I would still like to see
>some graphics capability built in.  Or, some hooks for it to function more
>in tune with the system.

But it won't function in tune with the system, as it will be working outside
of the system scheduler.

>    I couldn't handle the two user limit.  As I type this, there are
>three users on this system.  I was just thinking, if the limit is implemented
>in login, could you not replace it with your own login?  A guy I know who
>sells UNIX boxes is talking to some people at SCO about possibly selling
>hard drives with UNIX installed, no limits, to students real cheap.  He
>would like it to have some sort of development tools too.  However, you
>could just put a GNU system on it I guess.  Might be better anyway.

Sure, you could replace login, but then you would be violating the license
agreement that you agreed to by buying and installing UNIX.  AT&T would take
you to court.  If they didn't catch you, you would still be a software pirate.

>    You are right about UNIX being more reliable than DOS, Windows, and
>OS/2.  I never said it wasn't.  However, my DOS file system has been severely
>abused and I have never screwed it up (bad).  Its just too simple to mess
>up very much.  I mean, people go for years without backing up, defragging,
>or doing any maintenance.  That may be the biggest problem with and OS
>like UNIX is that people are used to never maintaining their system.
>Actually, you should perform maintenance on a DOS system on a regular basis
>and the same with a MAC or whatever.  Thing is, you can get away with it
>on the simpler systems.  Abuse a UNIX system and you'll pay for it later.

The average DOS user does not need UNIX.

>    As far as OS/2 is concerned.  That is a dog so far.  It is almost
>as complex as UNIX but you get absolutely no documentation that is worth
>anything, about its internals.  You get some error message and a note to
>call IBM about it.  Hah!  Linux screams along on my little 8-meg 386 box
>but OS/2, even without Presentation Mangler was pretty sluggish.  I sometimes
>think that OS/2 may be the best thing to happen to UNIX.  Put the systems
>side by side and let the users pick after seeing them together.  I keep
>trying to tell people that UNIX has had for 15 years what IBM and Microsoft
>are promising they'll have "real soon now."

The people who I know that are using OS/2 in the commercial world are doing
so by choice.  The main reason being that all of the important productivity
applications are available, which is not true for Linux.

OS/2 and UNIX serve different markets.

--

Rick Kelly    [e-mail]    unixland!rmkhome!rmk    [e-mail]