*BSD News Article 66257


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.bhp.com.au!mel.dit.csiro.au!munnari.OZ.AU!news.ecn.uoknor.edu!paladin.american.edu!news.jhu.edu!aplcenmp!netnews.jhuapl.edu!uunet!in1.uu.net!world!mv!news.missouri.edu!vortex.cc.missouri.edu!rhys
From: rhys@vortex.cc.missouri.edu (Justin "Rhys Thuryn" McNutt)
Newsgroups: comp.os.linux.development.system,comp.unix.bsd.386bsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc,comp.os.linux.advocacy
Subject: Re: Historic Opportunity facing Free Unix (was Re: The Lai/Baker paper, benchmarks, and the world of free UNIX)
Followup-To: comp.os.linux.development.system,comp.unix.bsd.386bsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.bsd.netbsd.misc,comp.unix.bsd.freebsd.misc,comp.os.linux.advocacy
Date: 16 Apr 1996 15:56:25 GMT
Organization: University of Missouri - Columbia
Lines: 203
Message-ID: <4l0fv9$1bfs@news.missouri.edu>
References: <4ki055$60l@Radon.Stanford.EDU> <jdd.829261293@cdf.toronto.edu> <tporczykDpqKHL.7vG@netcom.com> <DpsKyx.1Jo@catzen.gun.de> <tporczykDptpAL.8uG@netcom.com>
NNTP-Posting-Host: vortex.cc.missouri.edu
X-Newsreader: TIN [version 1.2 PL2]
Xref: euryale.cc.adfa.oz.au comp.os.linux.development.system:21672 comp.unix.bsd.386bsd.misc:667 comp.unix.bsd.bsdi.misc:3293 comp.unix.bsd.netbsd.misc:3126 comp.unix.bsd.freebsd.misc:17600 comp.os.linux.advocacy:46095

Tony Porczyk (tporczyk@netcom.com) wrote:

: Yes, but remember we were talking about commercial desktop - that's
: where the money is.  $295 for the OS is peanuts if it doesn't involve
: significant (read: transition from Windows to UNIX) retraining.

The point that the Unix advocates are trying to make here is:  "We want 
to try to make Unix (FreeBSD, Linux, whatever) easy enough to use that 
Windiots can handle it."

How is another issue.

Let me show you something:

Linux - It's a *free*, powerful, POSIX-compliant, widely-available, and
	supported operating system.  It's got people developing for it
	right and left.  They're usually just developing drivers and
	miscellany, but it's there.  It got here slowly, but it's getting
	there.

The St. Louis Arch - Same sort of thing.  It took those engineers years
			just to design the thing.  It took another few
			years to build, but was worth the trouble.

The Sears Tower, The Taj Mahal, The Eiffel Tower, the telephone, the 
stereo, the space shuttle, the *printing press*...

All of these things are wonderful.  They were hard to achieve, but they 
were all well worth the effort.

I hear all this whining on here about "someone else tried it before and 
they crashed and burned!"  So what?  Is that a reason to not try again?  
If we really want to get it done, we *CAN* get it done.  We have to 
commit to doing it *first*, and worry about *HOW* second.

It's cheesy, but to quote Yoda, "Either do, or do not.  There is no 
try."  And Mr. Myagi, "Walk left side, safe.  Walk right side, safe.  
Walk middle, <sqwk!>, just like grape.  Either you karate do yes, or 
karate do no.  You karate do 'Guess so,' <sqwk!>, just like grape."

Besides, dammit, if MS-DOS could take over the PC market, why the hell 
can't we make Linux or FreeBSD just as good or better?  It infuriates me 
that all this whining is even happening because it all boils down to, 
"But it'll be too HARD!! WAHH!"  Another quote:  "Of course it's hard!  
The hard is what makes it great!"  (figure out the quote yourself)

[inspirational ranting mode off]

Also, a lot of those people were trying to do all that for profit.  Linux 
and FreeBSD aren't controlled by profit-oriented organizations trying to 
make a buck.  They're people who want to see computers perform as well as 
possible.  Hell, I'd write it all myself, but I'm not that good of a 
programmer yet, and I'm still trying to graduate (one more year, and I'll 
be available to help you out, guys).

Basically, what I don't understand is what all this is about.  We all 
want to see Unix become mainstream.  So let's quit arguing about it and 
just do it.  It's *very* simple:

First:  What is it that DOS, Windows, and OS/2 do when you first install 
them?  They check out the hardware, they say to the user, "I need this 
much hard disk space.  Can I format it and use it (yes/no buttons)."  and 
so on.  Linux makes you make all kinds of decisions.

No problem.  We just take Slackware's existing install program, and add a 
small choice to the very beginning:

		Advanced Install
		Easy Install (default)

This idea is stolen blatantly from the Mac.  That's how they do it.  Then 
everybody can have his or her own way.

The easy install *should* make all kinds of decisions for the user.  
That's what easy install means.  Those admins (like me) out there are 
just cringing, but if you think about it, that's what the user *wants*.  
As long as it works, they don't care.  All the config files like 
/etc/inittab, /etc/fstab, /etc/hosts, etc. etc. should all be set up 
automatically by little scripts in the install program.

It can't be that hard.  DOS and Windows have been doing it for years.

As far as X Windows goes, I'm no expert on video hardware, but I do know 
that the existing setup process for X is abysmal.  Why can Win95, WinNT, 
and OS/2 set themselves up with a graphical *install* no less, but X is 
such a pain?  It can't be that difficult.

Actually, if I'm not mistaken, the default on all three (and in Win 3.1) 
is 640 x 480 x 16.  *ANY* VGA card supports that, and it's about as 
standard as you can get.  If you can get it running that far, most users 
will probably put forth at least a token effort to go further.

Okay, then the OS is set up and we have our handy-dandy graphical 
interface.  Now what?  We need a program launcher.  It could be 
object-oriented like the Mac (the Magic Desktop for X is pretty cool) and 
OS/2, it could be launcher oriented like Program manager, or it could be 
a combination like Win95 and NeXTStep (I *LIKE* having the dock and 
WorkSpace Manager).

This shouldn't be very hard either.  If I am not mistaken, there already 
exists a program that does a program icon dock for X Windows in Linux (it 
might be part of twm, I'm not sure).  Want Program Manager?  How about 
Xfilemanager?  It needs some work, but it's pretty much all there.  It's 
got a little directory tree window (the default tree should be the root), 
and it's got a little window with neat little icons in it called (imagine 
this!) WorkSpace.  Great.

Now we get to the harder part.  Put some good programs in the WorkSpace 
that people will want to use.  <sigh>

I believe that TeX is the way to go, but it has one big drawback.  As far 
as I know, it doesn't have a widely-available, easily-configured, WYSIWYG 
interface.  People like MS Word, Windows Write, Ami Pro, etc. because 
they can *see* what their document is going to look like before they 
print it out.  What does the doc actually look like in TeX?  Who cares?  
Ever look at the inside of an MS Word doc?  Pure rubbish.  They can't 
even keep things consistent from version to version.  TeX is portable and 
user-editable in a pinch.  TeX -> PS is child's play, and just about any 
printer can print PS.

Given a WYSIWYG interface, and perhaps the ability to read other popular 
documents, a TeX-based word processor could just eat up the market.

Some people will say, "But there's more to it than just a word 
processor!"  Really?  I have to take off my shoes and my friend's gloves 
so that I can count the number of people that I know personally who only 
use their computers for word processing.  Their stupid little PC boots, 
loads Win[3.1|95|NT], and MS Word is in the Startup group.  It runs right 
away, and that's all they ever use it for.

Given a decent word processor, and an installation process that will let 
novice users get to that word processor with the least hassle, and you 
will have given Unix a chance at the PC.

Just for motivation, I don't want to hear any, "But it's too HARD!" 
crap.  Look at the Linux kernel.  *THAT*, ladies and gentlemen, *WAS* the 
hard part.  Are you telling me that no one out there can write a *WORD 
PROCESSOR*?  Come on...  The way I see it, the hard part will be getting 
X Windows up and running with only the Windiot user to rely on.

It can be done.  Someone (I nominate myself, although I *will* need help) 
just needs to do it.  It doesn't even matter if the programs and scripts 
underneath are slow, obnoxious, not very portable, etc.  If they *WORK*, 
we can fix them in our spare time.  Linux was *written* in spare time.  
But if it doesn't exist at all, no one will use it.

Feedback from novice users will be important.  I work at a University 
*full* of feedback.  If someone can help me work on this, I have 
supervisors who have already let me install Linux in our computer labs, 
and would be willing to help me make it more usable.  I have students who 
want things idiot-proof.  Talk about beta-testing!  :)  If the kids here 
can use it, anyone can.

To recap:  I have stated what needs to be done.  Although it is 
irrelevant if no one is willing to do it, I have also suggested ways in 
which it might be done.  I have offered to help.  Now, if Linux and/or 
FreeBSD doesn't start becoming one of the most popular operating systems 
for the desktop PC, it's because no one tried.  It won't happen 
overnight, but it won't happen at all if we don't start working on it 
*right* *now*.

: As I said originally, the only thing that can sell UNIX to desktops
: (besides specialized uses) is applications - common productivity
: applications, priced *reasonably*.  Superiority or inferiority of OS is
: irrelevant.

Exactly.  Get it installed, give 'em a word processor.  Other apps?  One 
thing at a time.  We'll get to it, but first things first.  Let's not 
overwhelm ourselves here.

: Example from my own desktop: I use Linux and FreeBSD at home, SunOS at
: work, but when I sit down to write courseware (part of my job), I plug
: in hard drive with Windows and MS Office.  Why?  Try to find a package
: for UNIX that contains easily mergeable Word Processing, Presentation
					  ^^^^^^^^^^^^^^^
: and Spreadsheet that a common mortal can afford.  Show me *one*.  And
: let's not talk about TeX.  If I told the rest of the team that I will
: submit my work in TeX, they would die laughing.  No one has the time to
: play with that kind of stuff anymore.

Bingo.  Easily mergable?  TeX or PS.  No problem.  So what if they die 
laughing?  Does it work?  Who is playing?  TeX is good.  The format in 
which most word processors save their documents is proprietary.  It even 
changes from version to version.  If someone were to write a word 
processor that used TeX, that problem would disappear.  Documents would 
cease to be "MS Word 6.0" or "MS Word 5.0" or "WriteNow! 4.0" documents 
and would become "a text document".  Want graphics?  Use PS.

That way, when you release updates for this fictional word processor, you 
can add features without having to ruin your backward compatibility all 
the time.  You can work on the *program* instead of playing around with 
the *output* all the time.

C'mon, you all.  Let's just do it.  The only thing standing in our way is 
the willingness to do it.

--------
If you can lead it to water and force it to drink, it isn't a horse.

Got a Linux problem?  Or can you help others solve them?  Visit the Linux 
Common Problems page at http://vortex.cc.missouri.edu/~rhys/linux.html

rhys@vortex.cc.missouri.edu