Technology and Trends
 USENET Archives
  
Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec27.030803.25576@microsoft.com>
Date: 27 Dec 91 03:08:03 GMT
References: <PATERRA.91Dec16101525@geb.cs.odu.edu> <04WALJ9@taronga.com>
Organization: Microsoft Corp.
Lines: 72

In article <04WA...@taronga.com> pe...@taronga.com (Peter da Silva) writes:
>pate...@cs.odu.edu (Frank C. Paterra) writes:
>> And since then they have only become a multinational corporation,
>> whose profits have been above industry average, that has had a
>> major impact on millions of peoples lives.
>
>Yes. A major negative impact.

Yep, Imagine that, everyone able to write software for most machines.  Shoot,
I'd rather have a version of each program for Apple, CPM(about 8 versions), 
UNIX clones (about 8 versions), TRSDOS, etc etc etc.  Unifying the software
industry to a single OS was very negative (sarcasm).
>
>> You are right Peter, they should have just stopped and gone out of
>> business years ago.
>
>We would all have been better off if they'd just continued to produce
>superior language products. Oh, Bill Gates wouldn't. But that's not
>of much concern to me.
I could be argued that applications technology would be about 5 Years behind
because developers would have to develope for many different OS platforms, 
or IBM would have picked CPM86 and you'd be complaining about that now.
>
>> Did you know that MS produces the best selling sped sheet, the
>> best selling operating system, the best selling window system,
>> and among the best selling compilers?  All thoes people buying
>> all that software must really be dumb huh?
>
>Either they are technically incompetant or forced by economic necessity
>to purchase an inferior product.
I don't know, Excel has been rated very high by users, so has Word, etc, etc.
I guess most users don't know what's really important in a system, such as the
ability to have cryptic commands like

spam < foo | grep mo | jive .......
.... & ; fg
 Or the ability to not be able to have your complete set of system software
on a diskette when Hard Drives cost $400 for a good one.
>
>> If you bought $300 of MS stock when it came out in 1985, and held
>> on to it, today it would be worth over $10,000.  What fools we
>> all have been!
>
>I wish I could have bought enough stock to kick out the marketroids
>and concentrate on the technology.

Yeah, but then the stock would have dropped, so you'd have lost money, and
nobody would use your hot technology 'cause you probably wouldn't be around
in a year, and you didn't have the marketing folk to tell you what the users
want.  You'd end up with a lot of high tech machinery, but look at the
Stealth bomber, It does everyone a lot of good, doesn't it.
>
>> Of course MS-DOS and windows do have their faults, but so do many
>> other OSes.  For the time it was introduced, and the ared ware it
>> was introduced on, I don't think that they did too bad of a job.
>
>I do. There were superior operating systems available for even less
>powerful hardware, such as the high-end CP/M systems. Cromemco even
>had a UNIX looklike running on a multi-CPU Z-80 system. Just about
>any computer company that was already in the O/S business would have
>been better than picking a language vendor to write an O/S.

Excuse me, but MS purchased the original DOS, and its programmer, from one of
these small vendors.
>
>Would you hire an accountant to fix your plumbing?

No, but I'd hire a general contractor to have my plumbing done, 'cause
he'd hire someone who knew how.  Kind of like IBM did.

pet...@microsoft.com
These are My opinions only.

Path: gmdzi!ieee.org!sdd.hp.com!wupost!uunet!convex!datri
From: da...@convex.com (Anthony A. Datri)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec27.162720.13280@convex.com>
Date: 27 Dec 91 16:27:20 GMT
References: <PATERRA.91Dec16101525@geb.cs.odu.edu> <04WALJ9@taronga.com> 
<1991Dec27.030803.25576@microsoft.com>
Sender: use...@convex.com (news access account)
Organization: CONVEX Computer Corporation, Richardson, Tx., USA
Lines: 70
Nntp-Posting-Host: lovecraft.convex.com

>Yep, Imagine that, everyone able to write software for most machines.

Wrongo.  You never know if a given peice of software is going to work on
a given machine.  There are no standards for disk interfaces, tape interfaces,
and so on.  There are no standards for displays, so the software has to have
the characteristics and driver for each of myriad ones on the market.

>  Shoot,
>I'd rather have a version of each program for Apple,

Apple what?

>CPM(about 8 versions), 
>UNIX clones (about 8 versions), TRSDOS, etc etc etc.  Unifying the software
>industry to a single OS was very negative (sarcasm).

The negative thing was that it started as an unreasonably primitive OS,
and stayed there.  RT11, for example, offers a far richer environment, and
even runs on smaller machines.

>I could be argued that applications technology would be about 5 Years behind
>because developers would have to develope for many different OS platforms, 

Which they've had to for years, and will always have to do.  Being five years
behind would be far preferable to the fifteen years behind that DOS has
remained.

>>> Did you know that MS produces the best selling sped sheet, the

Linens for the intellectually challenged?

>>> best selling operating system, the best selling window system,

You call those an OS and a window system?  Why does something have to be sold
in order to be considered?  Why are unit sales so important?  More paper clips
are sold each year than humans are born.  I guess paper clips are better.

>>> and among the best selling compilers?  All thoes people buying
>>> all that software must really be dumb huh?

Yep.  Especially when better alternatives for some of them are available free.
Never underestimate the stupidty of the American (for example) public.  The
fact that GM is still in business proves this.

>I don't know, Excel has been rated very high by users, so has Word, etc, etc.

Heroin is rated very high by its users, too.

> Or the ability to not be able to have your complete set of system software
>on a diskette when Hard Drives cost $400 for a good one.

Snicker.  Try fitting what anlogues of the traditional Unix software that
you describe on a diskette.  It'd better be a damned dense one.  It's real
hard to respect a system that has 8.3 uppercase filenames and no usable
wildcard support.

>>> If you bought $300 of MS stock when it came out in 1985, and held
>>> on to it, today it would be worth over $10,000.  What fools we
>>> all have been!

Stock is an abstraction promulgated by the rich to make themselves richer.
It has no value other than to get some schmuck to pay you more for it than
you paid for it.

>>I do. There were superior operating systems available for even less
>>powerful hardware, such as the high-end CP/M systems.

Or even RT11, or RSX for that matter.

--

Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec30.042044.14542@microsoft.com>
Date: 30 Dec 91 04:20:44 GMT
References: <04WALJ9@taronga.com> <1991Dec27.030803.25576@microsoft.com> 
<1991Dec27.162720.13280@convex.com>
Organization: Microsoft Corp.
Lines: 102

In article <1991Dec27.162720.13...@convex.com> da...@convex.com 
(Anthony A. Datri) writes:
>>Yep, Imagine that, everyone able to write software for most machines.
>
>Wrongo.  You never know if a given peice of software is going to work on
>a given machine.  There are no standards for disk interfaces, tape interfaces,
>and so on.  There are no standards for displays, so the software has to have
>the characteristics and driver for each of myriad ones on the market.

Well, I guess that's one of the problems of having a free market.  We could have
government or comittee control over the computer industry, and then we could
be as far ahead as the soviets(or Russians, georgians, ukranians, etc) are.
>
>>  Shoot,
>>I'd rather have a version of each program for Apple,
>
>Apple what?
I see you get my point.
>
>>CPM(about 8 versions), 
>>UNIX clones (about 8 versions), TRSDOS, etc etc etc.  Unifying the software
>>industry to a single OS was very negative (sarcasm).
>
>The negative thing was that it started as an unreasonably primitive OS,
>and stayed there.  RT11, for example, offers a far richer environment, and
>even runs on smaller machines.

Hmmmm, this brings an interesting question to mind.  Is it better to have
a lot of the smarts in the OS, but have the OS be big, etc.  Or is it better
to have the Applications writers include the smarts in their apps, and have
the OS be not much more than a loader (ala DOS).  I say they each have their
benefits and their drawbacks, and it depends on the market.

I myself spend more of my time using apps, than just the OS shell, but that's
because I'm a programmer/user/etc and not a Hacker.

>
>>I could be argued that applications technology would be about 5 Years behind
>>because developers would have to develope for many different OS platforms, 
>
>Which they've had to for years, and will always have to do.  Being five years
>behind would be far preferable to the fifteen years behind that DOS has
>remained.

Give DOS some time, I mean, It's take UNIX 30 years to get to where it is at,
at an almost usable state, Come back and tell me how bad the DOS is in 20 years.
Just to make it fair %)

>
>>>> best selling operating system, the best selling window system,
>
>You call those an OS and a window system?  Why does something have to be sold
>in order to be considered?  Why are unit sales so important?  More paper clips
>are sold each year than humans are born.  I guess paper clips are better.

Alas, A capitalist economy.
As for humans vs paper-clips, Law of supply and demand %)
>
>>>> and among the best selling compilers?  All thoes people buying
>>>> all that software must really be dumb huh?
>
>Yep.  Especially when better alternatives for some of them are available free.
>Never underestimate the stupidty of the American (for example) public.  The
>fact that GM is still in business proves this.
What would you get, A GM automobile, which might have a factory waranty, etc.
Or a free-mobile from Some fat guy named Al who looks like he couldn't tell
his rear end from a hole in the ground.
>
>>I don't know, Excel has been rated very high by users, so has Word, etc, etc.
>
>Heroin is rated very high by its users, too.
So has UNIX, you go figure.  
>
>> Or the ability to not be able to have your complete set of system software
>>on a diskette when Hard Drives cost $400 for a good one.
>
>Snicker.  Try fitting what anlogues of the traditional Unix software that
>you describe on a diskette.  It'd better be a damned dense one.  It's real
>hard to respect a system that has 8.3 uppercase filenames and no usable
>wildcard support.

You can fit most of DOS on a floppy these days(alright, a Big floppy).  How
big is the UNIX Kernel, and a shell, etc etc.

Anyway, so you judge a system by it's namespace and the ability to use 
wildcards.  I buy computers to run Apps myself, and to develope software for.
Namespace never really bothered me that much.  Heck, I remember when people
insisted on using single letter variables in programming.

>
>>>> If you bought $300 of MS stock when it came out in 1985, and held
>>>> on to it, today it would be worth over $10,000.  What fools we
>>>> all have been!
>
>Stock is an abstraction promulgated by the rich to make themselves richer.
>It has no value other than to get some schmuck to pay you more for it than
>you paid for it.

Don't complain to us, complain to your congressperson.  Capital gains tax.

pet...@microsoft.com
My opinions only.
.

Path: gmdzi!unido!mcsun!uunet!think.com!mips!pacbell.com!att!princeton!
phoenix.Princeton.EDU!albinali
From: albin...@phoenix.Princeton.EDU (Sabah S. Albinali)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec30.060231.29999@Princeton.EDU>
Date: 30 Dec 91 06:02:31 GMT
References: <1991Dec27.030803.25576@microsoft.com> 
<1991Dec27.162720.13280@convex.com> <1991Dec30.042044.14542@microsoft.com>
Sender: n...@Princeton.EDU (USENET News System)
Organization: Princeton University
Lines: 32
Originator: n...@ernie.Princeton.EDU
Nntp-Posting-Host: phoenix.princeton.edu

In article <1991Dec30.042044.14...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>In article <1991Dec27.162720.13...@convex.com> da...@convex.com 
(Anthony A. Datri) writes:
>
>You can fit most of DOS on a floppy these days(alright, a Big floppy).  How
>big is the UNIX Kernel, and a shell, etc etc.

Oh yeah? Then why is DOS 5.0 supplied in 5 disks? That doesn't
include the various drivers needed either (another 3 disks).
I don't care if DOS was a great OS when it came out, right now
I am pissed that I have a 486 that (most of the time) is being
used as a (fast) 8086, and I have 7 megs of RAM that is usually
sitting there unused.
Don't tell me that DOS is better than UN*X, because I don't care.
It still sucks. Windows? I won't even begin to count the number
of times I've needed to reboot after I try anything other than 
running one application at a time. I've never seen my Mac II 
have a system crash more than two or three times, and a friends
NeXT (a much more complex system) has system crashes far fewer
times than my Windows crashes.
As for the coment about Microsoft's stock value, and the 
implied conclusion that a rise in stock value means a rise in
the quality of Microsoft products, I urge you to read 
A Random Walk Down Wall Street by Burton G. Malkiel. Mr. Malkiel
(BA, MBA Harvard, PhD Princeton and has held/holds the following
postitions :security analyst, dean of the Yale School of Management,
director of the Prudential Insurance Company, a governor of the
American Stock Exchange...etc.) points out in this book that
stock prices often do not reflect any information about the
company. Trust me, Mr. Malkiel knows much more about the world
of finance than you do.

-sabah

Path: gmdzi!unido!mcsun!uunet!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <kltfffINNgim@early-bird.think.com>
Date: 30 Dec 91 06:34:23 GMT
References: <1991Dec27.030803.25576@microsoft.com> 
<1991Dec27.162720.13280@convex.com> <1991Dec30.042044.14542@microsoft.com>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 110
NNTP-Posting-Host: godot.think.com

In article <1991Dec30.042044.14...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>Hmmmm, this brings an interesting question to mind.  Is it better to have
>a lot of the smarts in the OS, but have the OS be big, etc.  Or is it better
>to have the Applications writers include the smarts in their apps, and have
>the OS be not much more than a loader (ala DOS).  I say they each have their
>benefits and their drawbacks, and it depends on the market.

If the OS is minimal, then application developers have to take time away
from application development to implement the stuff that should have been
in the OS.  Instead of one OS development organization doing it once, and
hopefully doing it right, every application development organization has to
duplicate this effort, and they'll do it differently, and some will do it
wrong.  Conversely, if application developers don't have to bother
implementing OS features, they'll have more time to make the applications
better.

For instance, when a new graphics interface comes out, only the OS has to
be upgraded with the new driver, instead of having to enhance every
application.  I've never been a big DOS user, but I still remember
installing a program once on a DOS machine at work and having to go through
a configuration process where it asked me what kind of graphics we had; I
had to guess.

>I myself spend more of my time using apps, than just the OS shell, but that's
>because I'm a programmer/user/etc and not a Hacker.

There's a big difference between the OS and the OS shell.  Whenever you're
using an application you're using the OS indirectly, and the quality of the
OS is reflected in the quality of the applications.

>Give DOS some time, I mean, It's take UNIX 30 years to get to where it is at,
>at an almost usable state, Come back and tell me how bad the DOS is in 20 years.
>Just to make it fair %)

Unix is only 20 years old, and it's only been a commercial product for
about 10, although it took ideas from OSes that preceded it.

Sure, MS-DOS is only 10 years old, but OS technology was 20 years old at
the time.  Unfortunately, DOS started from scratch, rather than building on
what had been developed in the past.

>>>>> best selling operating system, the best selling window system,
>>You call those an OS and a window system?  Why does something have to be sold
>>in order to be considered?  Why are unit sales so important?  More paper clips
>>are sold each year than humans are born.  I guess paper clips are better.
>
>Alas, A capitalist economy.
>As for humans vs paper-clips, Law of supply and demand %)

There are many examples where the best-selling product is clearly an
inferior one.  Sales figures often have more to do with marketing than
quality.  For instance, MVS, one of the IBM mainframe OSes, is probably the
best-selling mainframe OS, but it is an archaic batch system; but it's been
around a long time, there are many applications for it, it comes from the
major vendor in the mainframe marketplace, runs on the most common
mainframes, there are trade schools that pump out MVS operators and
programmers, etc.  All these are good reasons why customers continued to
purchase it, but they don't make it a "good" OS; in the world of
mainframes, MVS is the old standard, like MS-DOS is in the PC world.

>Anyway, so you judge a system by it's namespace and the ability to use 
>wildcards.  I buy computers to run Apps myself, and to develope software for.
>Namespace never really bothered me that much.  Heck, I remember when people
>insisted on using single letter variables in programming.

Those are just simple, obvious examples.  When you're running your apps,
don't you need to save out your work, and don't you find it inconvenient
that you have to squeeze your titles into eight characters (I don't include
the extensions, since those are often forced by the application)?  Yes,
there are some applications that let you use longer titles by maintaining a
translation; this is an example of the application developer wasting his
time filling in the holes in the OS, and it also makes it difficult for the
user when he does have to access those files from the shell or other
applications.

If you want a more significant example, how about display drivers that
application developers don't feel compelled to bypass.

>>>>> If you bought $300 of MS stock when it came out in 1985, and held
>>>>> on to it, today it would be worth over $10,000.  What fools we
>>>>> all have been!
>>
>>Stock is an abstraction promulgated by the rich to make themselves richer.
>>It has no value other than to get some schmuck to pay you more for it than
>>you paid for it.
>
>Don't complain to us, complain to your congressperson.  Capital gains tax.

Regardless, the stock value of a company is only slightly related to the
quality of its products.  What do investors know about operating systems?
They know about management philosophy, so they'll buy shares in a company
because it is run well.  And stock prices generally go up based on
expectations.

Also, it's difficult to deny that Microsoft is a successful company, and it
will continue to be a major player in the industry.  It sells the OS and
word processor of choice for the IBM PC, and the word processor and
spreadsheet of choice for the Mac, and rode the wave of the personal
computing boom of the 80's.  But its products became the "choice" ones
because they were well marketed at the right times, and because Microsoft
already had a decent reputation.  But that still doesn't mean that its
products are the best, only that they've got quite a bit of inertia.  The
Apple II family is still the PC of choice in elementary schools because
they're cheap and there's lots of educational software available for them,
but it's still basically a 12-year-old system design.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!ieee.org!sdd.hp.com!wupost!ukma!nsisrv!mimsy!ra!
BDCD102.nrl.navy.mil!lawson
From: law...@BDCD102.nrl.navy.mil (Drew Lawson)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1115@ra.nrl.navy.mil>
Date: 30 Dec 91 18:26:32 GMT
References: <1991Dec27.030803.25576@microsoft.com> 
<1991Dec27.162720.13280@convex.com> <1991Dec30.042044.14542@microsoft.com>
Sender: use...@ra.nrl.navy.mil
Organization: Backgrounds Data Center
Lines: 28

>In article <1991Dec27.162720.13...@convex.com> da...@convex.com (Anthony A. Datri) 
>writes:
>>Yep.  Especially when better alternatives for some of them are available free.
>>Never underestimate the stupidty of the American (for example) public.  The
>>fact that GM is still in business proves this.

In article <1991Dec30.042044.14...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>What would you get, A GM automobile, which might have a factory waranty, etc.
>Or a free-mobile from Some fat guy named Al who looks like he couldn't tell
>his rear end from a hole in the ground.

As it is, instead of getting a car from Al (who you get a look at), you
get a car made by Al.

But the analogy is also flawed.  With the exception of the rare genius,
you cannot make good cars without a capital intensive factory.  The
facilities simply are not available.  But good software gets written
all over the place.

The crime in the current PC ("personal computer" NOT "politically
correct") environment is that MicroSoft was never willing to make
MS/DOS into an operating system, and they should have.

-- 
+------------------------------+--------------------------------------+
| Drew Lawson                  |  If you're not part of the solution, |
| law...@bdcd102.nrl.navy.mil  |  you're part of the precipitate.     |
| 71141.1...@CompuServe.COM    |                                      |
+------------------------------+--------------------------------------+

Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec30.213338.26078@microsoft.com>
Date: 30 Dec 91 21:33:38 GMT
References: <1991Dec27.162720.13280@convex.com> 
<1991Dec30.042044.14542@microsoft.com> <1991Dec30.060231.29999@Princeton.EDU>
Organization: Microsoft Corp.
Lines: 58

In article <1991Dec30.060231.29...@Princeton.EDU> 
albin...@phoenix.Princeton.EDU (Sabah S. Albinali) writes:
>In article <1991Dec30.042044.14...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>>In article <1991Dec27.162720.13...@convex.com> da...@convex.com 
(Anthony A. Datri) writes:
>>
>>You can fit most of DOS on a floppy these days(alright, a Big floppy).  How
>>big is the UNIX Kernel, and a shell, etc etc.
>
>Oh yeah? Then why is DOS 5.0 supplied in 5 disks? That doesn't
>include the various drivers needed either (another 3 disks).
>I don't care if DOS was a great OS when it came out, right now
>I am pissed that I have a 486 that (most of the time) is being
>used as a (fast) 8086, and I have 7 megs of RAM that is usually
>sitting there unused.
You know what I meant, you're just being argumentative.

>Don't tell me that DOS is better than UN*X, because I don't care.

I didn't say this.  I implied that DOS is better than other things for
certain purposes, although appearently not yours.

>It still sucks. Windows? I won't even begin to count the number
>of times I've needed to reboot after I try anything other than 
>running one application at a time. I've never seen my Mac II 
>have a system crash more than two or three times, and a friends
>NeXT (a much more complex system) has system crashes far fewer
>times than my Windows crashes.
1) NeXT is a good system, although they will suffer due to lack of
Apps, and a narrow marketing view.
2) Apple is still trying to recover from their Closed Archetecture strategy.
I myself like Open archetectures, which breed diversity.  There has to be some
standardization, but, well, lets say that an Apple's mother is also it's sister
(imbreed joke).
3) You hate windows (let me posit that it's not windows that you hate, but
microsoft that you hate, and even if microsoft were to turn out a perfect
system, you'd still hate it).

>As for the coment about Microsoft's stock value, and the 
>implied conclusion that a rise in stock value means a rise in
>the quality of Microsoft products, I urge you to read 
>A Random Walk Down Wall Street by Burton G. Malkiel. Mr. Malkiel
>(BA, MBA Harvard, PhD Princeton and has held/holds the following
>postitions :security analyst, dean of the Yale School of Management,
>director of the Prudential Insurance Company, a governor of the
>American Stock Exchange...etc.) points out in this book that
>stock prices often do not reflect any information about the
>company. Trust me, Mr. Malkiel knows much more about the world
>of finance than you do.
>
I guess you misunderstood me, when you said that I implied that a rise
in microsoft stock value means a rise in microsoft products.  I never said
that.

I'll give you that Mr. Malkeil know a lot more about stock than I do.  Because
you read his book, however, does not mean that you know more than I do.  How
well am I to trust your understanding of his work.

pet...@microsoft.com
My Opinions

Path: gmdzi!ieee.org!sdd.hp.com!cs.utexas.edu!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec30.215753.27010@microsoft.com>
Date: 30 Dec 91 21:57:53 GMT
References: <1991Dec27.162720.13280@convex.com> 
<1991Dec30.042044.14542@microsoft.com> <kltfffINNgim@early-bird.think.com>
Organization: Microsoft Corp.
Lines: 146

In article <kltfffINN...@early-bird.think.com> bar...@think.com (Barry Margolin) 
writes:
>
>If the OS is minimal, then application developers have to take time away
>from application development to implement the stuff that should have been
>in the OS.  Instead of one OS development organization doing it once, and
>hopefully doing it right, every application development organization has to
>duplicate this effort, and they'll do it differently, and some will do it
>wrong.  Conversely, if application developers don't have to bother
>implementing OS features, they'll have more time to make the applications
>better.
Well, the benefit of having the Apps writers write some of the lower level
stuff is that they can do it how they want( just point this out).  As for
doing an OS right, who decides what is right.  What is right for one person
is not necessarily right for another.

>
>For instance, when a new graphics interface comes out, only the OS has to
>be upgraded with the new driver, instead of having to enhance every
>application.  I've never been a big DOS user, but I still remember
>installing a program once on a DOS machine at work and having to go through
>a configuration process where it asked me what kind of graphics we had; I
>had to guess.

Kind of the direction that windows is pointing.  Dll's, Installable drivers,
a common shell, etc.
>
>>I myself spend more of my time using apps, than just the OS shell, but that's
>>because I'm a programmer/user/etc and not a Hacker.
>
>There's a big difference between the OS and the OS shell.  Whenever you're
>using an application you're using the OS indirectly, and the quality of the
>OS is reflected in the quality of the applications.

In a minimal system like DOS, where the responsability falls more on the
Applications, then many of the problems are related to the applications also.

>
>>Give DOS some time, I mean, It's take UNIX 30 years to get to where it is at,
>>at an almost usable state, Come back and tell me how bad the DOS is in 20 years.
>>Just to make it fair %)
>
>Unix is only 20 years old, and it's only been a commercial product for
>about 10, although it took ideas from OSes that preceded it.
>
>Sure, MS-DOS is only 10 years old, but OS technology was 20 years old at
>the time.  Unfortunately, DOS started from scratch, rather than building on
>what had been developed in the past.

It sort of built on CP/M (okay, that's streatching it).  Problem was, the
PC hardware back then wasn't that great, and you couldn't put anything too
spectacular in 64K with all the other garbage in there.  (this is arguable)

>
>>>>>> best selling operating system, the best selling window system,
>>>You call those an OS and a window system?  Why does something have to be sold
>>>in order to be considered?  Why are unit sales so important?  More paper clips
>>>are sold each year than humans are born.  I guess paper clips are better.
>>
>>Alas, A capitalist economy.
>>As for humans vs paper-clips, Law of supply and demand %)
>
>There are many examples where the best-selling product is clearly an
>inferior one.  Sales figures often have more to do with marketing than
>quality.  For instance, MVS, one of the IBM mainframe OSes, is probably the
>best-selling mainframe OS, but it is an archaic batch system; but it's been
>around a long time, there are many applications for it, it comes from the
>major vendor in the mainframe marketplace, runs on the most common
>mainframes, there are trade schools that pump out MVS operators and
>programmers, etc.  All these are good reasons why customers continued to
>purchase it, but they don't make it a "good" OS; in the world of
>mainframes, MVS is the old standard, like MS-DOS is in the PC world.

Again, this is determinant on how you judge a product to be inferior.  Many
judge a product to be superior if it sells well.  I judge UNIX to be inferior...
....On Real Time control systems, you gotta get a Real Time OS these days.  
I Judge DOS to be superior for running on an old XT, cause UNIX or OS/2
won't fit.  If I had a 286, I couldn't use OS/2 2.0 cause It's 32 bit, therefore
I'd judge it inferior.  If I had a 486/33 with lots of memory and disk space,
I'd probably run NT(I can, I work here), 'cause I judge it superior to the
other options out there.  If I hated microsoft, like so many seem to on the
net, I'd probably judge all MS products inferior, regardless of how good they
were to use.
 >
>>Anyway, so you judge a system by it's namespace and the ability to use 
>>wildcards.  I buy computers to run Apps myself, and to develope software for.
>>Namespace never really bothered me that much.  Heck, I remember when people
>>insisted on using single letter variables in programming.
>
>Those are just simple, obvious examples.  When you're running your apps,
>don't you need to save out your work, and don't you find it inconvenient
>that you have to squeeze your titles into eight characters (I don't include
>the extensions, since those are often forced by the application)?  Yes,
>there are some applications that let you use longer titles by maintaining a
>translation; this is an example of the application developer wasting his
>time filling in the holes in the OS, and it also makes it difficult for the
>user when he does have to access those files from the shell or other
>applications.

Or the user wasting his time always typing in long filenames.

>
>If you want a more significant example, how about display drivers that
>application developers don't feel compelled to bypass.

Tell them to write for windows.
>
>>>>>> If you bought $300 of MS stock when it came out in 1985, and held
>>>>>> on to it, today it would be worth over $10,000.  What fools we
>>>>>> all have been!
>>>
>>>Stock is an abstraction promulgated by the rich to make themselves richer.
>>>It has no value other than to get some schmuck to pay you more for it than
>>>you paid for it.
>>
>>Don't complain to us, complain to your congressperson.  Capital gains tax.
>
>Regardless, the stock value of a company is only slightly related to the
>quality of its products.  What do investors know about operating systems?
>They know about management philosophy, so they'll buy shares in a company
>because it is run well.  And stock prices generally go up based on
>expectations.

Why does everyone keep on thinking that I said that stock prices reflect
quality of products, or whatever.  All I said is capital gains tax.
>
>Also, it's difficult to deny that Microsoft is a successful company, and it
>will continue to be a major player in the industry.  It sells the OS and
>word processor of choice for the IBM PC, and the word processor and
>spreadsheet of choice for the Mac, and rode the wave of the personal
>computing boom of the 80's.  But its products became the "choice" ones
>because they were well marketed at the right times, and because Microsoft
>already had a decent reputation.  But that still doesn't mean that its
>products are the best, only that they've got quite a bit of inertia.  The
>Apple II family is still the PC of choice in elementary schools because
>they're cheap and there's lots of educational software available for them,
>but it's still basically a 12-year-old system design.

Yes, Now I see you understand.  Best must be qualified for a context.
DOS seems to be best for the general user even though it's archaeic 'cause
there's nothing else really, the users understand it, it's not scary, they
can get the apps they wan't on it, etc etc etc.  Most other systems loose
out in these catagories, even though they may be more technically advanced.


pet...@microsoft.com
My Opinions.

Path: gmdzi!unido!mcsun!uunet!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <klvgq6INNc1t@early-bird.think.com>
Date: 31 Dec 91 01:09:26 GMT
References: <1991Dec30.042044.14542@microsoft.com> 
<kltfffINNgim@early-bird.think.com> <1991Dec30.215753.27010@microsoft.com>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 118
NNTP-Posting-Host: gandalf.think.com

In article <1991Dec30.215753.27...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>Well, the benefit of having the Apps writers write some of the lower level
>stuff is that they can do it how they want( just point this out).  As for
>doing an OS right, who decides what is right.  What is right for one person
>is not necessarily right for another.

Yes, if the OS doesn't do something the way the application developers
want, it makes sense (in some environments, such as single-user PC's, but
generally not multiuser systems) to allow the application to bypass the OS.
But a minimalist OS like DOS often forces application developers to
implement system facilities that they don't really care about the details
of, simple because the OS doesn't provide it at all.  In other words, the
OS should provide what most apps need, but it shouldn't overly restrict
what apps can do.

>>For instance, when a new graphics interface comes out, only the OS has to
>>be upgraded with the new driver, instead of having to enhance every
>>application.  I've never been a big DOS user, but I still remember
>>installing a program once on a DOS machine at work and having to go through
>>a configuration process where it asked me what kind of graphics we had; I
>>had to guess.
>
>Kind of the direction that windows is pointing.  Dll's, Installable drivers,
>a common shell, etc.

Yes, from what little I know about Windows, MS-DOS is finally be replaced
with a more real OS.

>>There's a big difference between the OS and the OS shell.  Whenever you're
>>using an application you're using the OS indirectly, and the quality of the
>>OS is reflected in the quality of the applications.
>
>In a minimal system like DOS, where the responsability falls more on the
>Applications, then many of the problems are related to the applications also.

And that's the problem: application developers shouldn't be required to be
experts at both the application domain and OS programming.  Ideally, they
should be experts in the application domain, and the OS should enable them
to develop the best applications.

>>Sure, MS-DOS is only 10 years old, but OS technology was 20 years old at
>>the time.  Unfortunately, DOS started from scratch, rather than building on
>>what had been developed in the past.
>
>It sort of built on CP/M (okay, that's streatching it).  Problem was, the
>PC hardware back then wasn't that great, and you couldn't put anything too
>spectacular in 64K with all the other garbage in there.  (this is arguable)

Wasn't the PDP-11 address space 64K (and later 128K in split-I/D models)?

>Again, this is determinant on how you judge a product to be inferior.  Many
>judge a product to be superior if it sells well.  

I think many people *assume* a product is superior if it sells well.
Unfortunately, it's hard to define objective criteria for the quality of a
product, especially as nebulous a product as an OS, and even more
especially in a manner that would make sense to the non-technical people
who often must make purchase decisions.  So instead, they often go with the
flow; a million lemmings can't be wrong.

>						   I judge UNIX to be inferior...
>....On Real Time control systems, you gotta get a Real Time OS these days.  

You're right, the quality of an OS must be related to the intended use.  I
suggest that we consider the context of general purpose computing: word
processing, simple graphics, financial applications, and software
development.  When you start including more specialized applications such
as real-time control and fancy graphics your choices are often restricted
more by the availability of specialized software and hardware than by the
OS itself (yes, the availability of these things is influenced by the OS,
and vice versa, but it adds too many variables to hold a meaningful debate
such as this (which, by the way, seems to be the calmest "my OS can beat
your OS" discussion I've ever seen)).

>>>Anyway, so you judge a system by it's namespace and the ability to use 
>>>wildcards.
>>Those are just simple, obvious examples.  When you're running your apps,
>>don't you need to save out your work, and don't you find it inconvenient
>>that you have to squeeze your titles into eight characters
>Or the user wasting his time always typing in long filenames.

That's why graphical user interfaces, menus, and completion were invented.

>>If you want a more significant example, how about display drivers that
>>application developers don't feel compelled to bypass.
>Tell them to write for windows.

Yes, MS-DOS is *finally* starting to provide these things.  But Windows
still isn't in every PC, so many applications (especially games) can't
depend on it.

>Why does everyone keep on thinking that I said that stock prices reflect
>quality of products, or whatever.  All I said is capital gains tax.

We were responding to the person who said "if you bought $300 of MS stock
...  in 1985 ... today it would be worth over $10,000."  He was clearly
implying that this reflected something about the products.  I still don't
know what your "capital gains tax" response was supposed to mean.

>Yes, Now I see you understand.  Best must be qualified for a context.
>DOS seems to be best for the general user even though it's archaeic 'cause
>there's nothing else really, the users understand it, it's not scary, they
>can get the apps they wan't on it, etc etc etc.  Most other systems loose
>out in these catagories, even though they may be more technically advanced.

Yes, Unix generally loses out in the areas related to ease of use and
application availability; it was designed as a programmer's OS, and it
still reflects that heritage.  However, the Mac beats MS-DOS in ease of use
and is only slightly behind in application availability.  Actually, I think
you're likely to be able to find reasonable Mac applications for most areas
that there are DOS applications.  DOS often gives you more choices within
an application domain, though; this is a natural result of the larger
number of DOS machines -- there's more room for competition.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!ieee.org!sdd.hp.com!cs.utexas.edu!uunet!uunet!uunet!avg
From: a...@uunet.uu.net (Vadim Antonov)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <avg.694142491@uunet>
Date: 31 Dec 91 01:21:31 GMT
References: <1991Dec27.162720.13280@convex.com> 
<1991Dec30.042044.14542@microsoft.com> <kltfffINNgim@early-bird.think.com> 
<1991Dec30.215753.27010@microsoft.com>
Sender: use...@uunet.uu.net (UseNet News)
Organization: UUNET Communications Services
Lines: 13
Nntp-Posting-Host: rodan.uu.net

It always amazed me how different writing styles of Unix and MSDOS guys are.
Having English as the second language I still can say for sure who is who
having only few lines of their writings.

Is the low cultural level a consequence of or reason for enjoying MSDOS?

Read some books on operating systems, lad.

Vadim Antonov
UUNET Technologies
(Falls Church, VA, USA :-)

ObFolklore: kremvax took a new shape - now it's uVAX 3.

Path: gmdzi!ieee.org!sdd.hp.com!spool.mu.edu!munnari.oz.au!metro!
utsmips!state.COM.AU!rufus!chris
From: ch...@state.COM.AU (Chris Keane)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <chris.694148945@rufus>
Date: 31 Dec 91 03:09:05 GMT
References: <1991Dec30.042044.14542@microsoft.com> 
<kltfffINNgim@early-bird.think.com> <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com>
Sender: n...@state.COM.AU
Organization: Group Treasury, State Bank of NSW
Lines: 45

bar...@think.com (Barry Margolin) writes:

>Yes, Unix generally loses out in the areas related to ease of use and
>application availability; it was designed as a programmer's OS, and it
>still reflects that heritage.  However, the Mac beats MS-DOS in ease of use
>and is only slightly behind in application availability.  Actually, I think
>you're likely to be able to find reasonable Mac applications for most areas
>that there are DOS applications.  DOS often gives you more choices within
>an application domain, though; this is a natural result of the larger
>number of DOS machines -- there's more room for competition.

You can't be serious when you said that Unix is not easy to use. The one,
only, single hard thing that I found about "learning Unix" when I did
was that everyone kept telling me how hard it was going to be. 
If you like DOS (gack!) then with a slight conceptual shift, you can do
pretty much everything you can do under DOS with Unix.
DOS is not easier than Unix. From a purely User point of view, DOS is
like Unix but with all the good bits taken out.

Anyway, as for all this "Unix is hard to use" stuff, don't you really mean
"/bin/sh is hard to use" or /usr/bin/csh is hard to use" etc etc? If I want
to put a Macintosh interface on my workstation then yes, I can. If I want
to put a dumb DOS screen up, well, that's fine to. I'm sure a nice Windows
interface can be arranged.

Our users here use Unix quite happily. Mind you, they have a totally
graphics interface and only csh if they feel the need. More or less like 
Windows, I guess. Except they run it faster than a 486 and at less cost
(at least at $A prices) than a 16" colour memoried 486. And of course, if
they decide they want to use a different interface one day, absolutely no
upgrade is required.

When DOS and/or Windows can do this, come back and tell me that Unix isn't
easy to use and is more expensive.


p.s. for those moaning about how expensive Unix is for a PC, go look at
4.4BSD for less than $US 1000 with *source* code. It can run, I'm told, on
a 40 meg disk and it *runs applications*


regards...                                                 
Chris Keane.                  State Bank NSW            ph. +61 2 259 4459
Unix Systems Administrator   (Group Treasury)     ch...@rufus.state.COM.AU
Disclaimer: These are my own opinions, but I'm insane. What's your excuse?

Path: gmdzi!ira.uka.de!yale.edu!spool.mu.edu!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: OS wars (was Re: What happened to Microsoft?)
Message-ID: <km01j8INNgol@early-bird.think.com>
Date: 31 Dec 91 05:55:52 GMT
References: <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com> <chris.694148945@rufus>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 29
NNTP-Posting-Host: godot.think.com

In article <chris.694148945@rufus> ch...@state.COM.AU (Chris Keane) writes:
>bar...@think.com (Barry Margolin) writes:
>>Yes, Unix generally loses out in the areas related to ease of use and
>You can't be serious when you said that Unix is not easy to use. 

In this case, I was referring to the OSes along with their common
applications.  So, for instance, word processing on Unix tends to be harder
than on MS-DOS (TeX and nroff are not "user friendly").  MS-DOS application
developers are more motivated to write programs oriented to non-technical
users, because that's what most MS-DOS users are; Unix applications tend to
be designed for people more familiar with computer technology.

>DOS is not easier than Unix. From a purely User point of view, DOS is
>like Unix but with all the good bits taken out.

But taking all the good bits out makes it easier to use!  A system with
only ten commands is half as complicated as one with twenty commands.
True, it's also half as useful.  Non-technical users have little use for
the fancy features of the Unix shell.

As a sophisticated computer user, I personally find MS-DOS and the Mac
inadequate, and Unix just barely adequate.  But for users who just want to
run easy applications, MS-DOS is fine and the Mac is better.

-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!unido!mcsun!uunet!think.com!snorkelwacker.mit.edu!world!bzs
From: b...@world.std.com (Barry Shein)
Newsgroups: alt.folklore.computers
Subject: Re: OS wars (was Re: What happened to Microsoft?)
Message-ID: <BZS.91Dec31020504@world.std.com>
Date: 31 Dec 91 07:05:04 GMT
References: <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com>
	<chris.694148945@rufus> <km01j8INNgol@early-bird.think.com>
Sender: b...@world.std.com (Barry Shein)
Organization: The World
Lines: 93
In-Reply-To: barmar@think.com's message of 31 Dec 1991 05:55:52 GMT


(into the breech...)

From: bar...@think.com (Barry Margolin)
>In this case, I was referring to the OSes along with their common
>applications.  So, for instance, word processing on Unix tends to be harder
>than on MS-DOS (TeX and nroff are not "user friendly").

Text processors like nroff and TeX are generally better designed for
large documents, like books. Style decisions can be put off to later
(I often throw non-existant macros into my troff docs and sort it out
later, very easy, try that with a wysiwyg.)

Typical wysiwygs (which are now becoming common-place on Unix also,
you're showing your age) are hands-down easier for the short
documents, a few to several pages. Particularly where all style
decisions are to be made by the author. If a designer will later make
appearance decisions TeX and nroff are generally easier to work with
in my experience, you only have to rely on the author to do original
mark-up rather than making all stylistic decisions.

I agree that most people write short, unimportant (style-wise)
documents most of the time. I often wonder about the new-found
enormous amount of time and productivity wasted as people carefully
choose fonts and layouts for their stupid memos, trying to flavor in
importance where non exists, and which would be better off done in a
simple text editor (something I notice most people's PC's and Mac's
are often entirely lacking! Gee, do you think the date looks better in
Palatino Outline or Times Expanded? *JUST DO IT*)

>>DOS is not easier than Unix. From a purely User point of view, DOS is
>>like Unix but with all the good bits taken out.
>
>But taking all the good bits out makes it easier to use!  A system with
>only ten commands is half as complicated as one with twenty commands.

You only need one command to solve this problem: rm.

I have never understood the argument that Unix is hard to use because
it has all these zillions of programs (while DOS is easier to use
because it has all these zillions of programs...hmmm.)

It's as if someone walks up to a Unix system and feels compelled to
master everything in /usr/bin and so forth before touching the
keyboard.

The same sort of thing applies to emacs, you can use emacs perfectly
well with less than ten commands, but again somehow because it has
hundreds it's therefore hard to learn. I agree people believe that
sort of thing, I've just never understood it.

>As a sophisticated computer user, I personally find MS-DOS and the Mac
>inadequate, and Unix just barely adequate.  But for users who just want to
>run easy applications, MS-DOS is fine and the Mac is better.

The Mac is good (very good) until you have a problem which the various
software designers did not specifically anticipate, then you're out of
luck.

There are two major schools of software design (ahem!), one tries to
find common and universal "operators", the other tries to enumerate
everything their vision of a target user ever does in detail. The
latter tends to yield rigid but easy to learn software, the former
tends to yield very flexible software which requires some thought to
use.

Word processors provide a great example of this phenomena. WYSIWYGs
are usually very easy to learn, but lord help you if you want to do
something that the designer never considered. The best you can hope
for is that someone adds that feature in the next release. Systems
like troff merely require some thought as almost anything is possible
(I had a student who was doing very good 3D renderings using troff as
the output language because troff already knew how to be displayed and
printed on every device he was interested in, this was several years
ago.)

In my experience (over 15 years with Unix, and many other OS's) I find
that the ability of people to learn any software system seems to have
a lot more to do with their motivation to accomplish what they've set
out to do than what most people talk about as being important. In
fact, so long as six-year-olds can punch long strings of digits to
find their friends (the telephone system, which has had some modest
commercial success) I will insist that we have no idea what we are
talking about when we say the words "user-friendly."

There, I feel better now.


-- 
        -Barry Shein

Software Tool & Die    | b...@world.std.com          | uunet!world!bzs
Purveyors to the Trade | Voice: 617-739-0202        | Login: 617-739-WRLD

Path: gmdzi!ieee.org!sdd.hp.com!cs.utexas.edu!asuvax!ukma!nsisrv!mimsy!ra!
BDCD102.nrl.navy.mil!lawson
From: law...@BDCD102.nrl.navy.mil (Drew Lawson)
Newsgroups: alt.folklore.computers
Subject: Re: OS wars (was Re: What happened to Microsoft?)
Message-ID: <1126@ra.nrl.navy.mil>
Date: 31 Dec 91 14:16:19 GMT
References: <klvgq6INNc1t@early-bird.think.com> <chris.694148945@rufus> 
<km01j8INNgol@early-bird.think.com>
Sender: use...@ra.nrl.navy.mil
Organization: Backgrounds Data Center
Lines: 65

In article <km01j8INN...@early-bird.think.com> bar...@think.com 
(Barry Margolin) writes:
>In article <chris.694148945@rufus> ch...@state.COM.AU (Chris Keane) writes:
>>bar...@think.com (Barry Margolin) writes:
>>>Yes, Unix generally loses out in the areas related to ease of use and

>>You can't be serious when you said that Unix is not easy to use.

>In this case, I was referring to the OSes along with their common
>applications.  So, for instance, word processing on Unix tends to be harder
>than on MS-DOS (TeX and nroff are not "user friendly").

Then I assume that you are comparing nroff with EDLIN.  There is no
other "standard" word processing under DOS.  If you are comparing it
with word processers, you are making two mistakes.  The first is to
mistake TeX and nroff for word processors.  They are typesetters.  They
are both more powerful and more difficult (or awkward) to use than word
processors.

The second mistake is to think that your "standard" DOS word processors
are not available under Unix.  Several are, as well as spreadsheets and
databases (REAL databases, not dBase).


>>DOS is not easier than Unix. From a purely User point of view, DOS is
>>like Unix but with all the good bits taken out.

>But taking all the good bits out makes it easier to use!  A system with
>only ten commands is half as complicated as one with twenty commands.
>True, it's also half as useful.

How does eliminating commands make the system easier to use?  Does the
fact that fsck is on the disk confuse the user who isn't trying to use
it?  Does it confuse prople to have a -s option to ls, simply because
they aren't wanting to use it?

I will go along with the sentiment that it is baffling when I hear that
Unix is harder to use than DOS.  For the unsophisticated user, where
are the differences?   You have a command line (usually), files,
directories, etc.  The only differences I can see (for the
unsophisticated user) are related to the multi-user file ownership,
etc.  And these are not Unix concepts.  They are general concepts.



I would like to add one other comment to this thread of OS debates.
When the discussion is "what should I buy," availability of
applications is very important.  When the discussion is (as this one
has been) "Is this a good OS," the market is irrelevant.

I personally, as a developer, find DOS to be a waste of energy.  This
is because it does little to support any tasks.  Instead, DOS allows
programmers to do things.  DOS still does not SUPPORT over 640K.  It
simply allows the programmer to do things to get to the memory.  This
means that two resident programs (and active and a TSR) have no
moderator over their demands for resources.

DOS does not manage the machine's available resources.  In fact, it is
barely aware that the resources exist.

-- 
+------------------------------+--------------------------------------+
| Drew Lawson                  |  If you're not part of the solution, |
| law...@bdcd102.nrl.navy.mil  |  you're part of the precipitate.     |
| 71141.1...@CompuServe.COM    |                                      |
+------------------------------+--------------------------------------+

Path: gmdzi!ira.uka.de!yale.edu!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: OS wars (was Re: What happened to Microsoft?)
Message-ID: <km1aciINN4ed@early-bird.think.com>
Date: 31 Dec 91 17:32:02 GMT
References: <chris.694148945@rufus> <km01j8INNgol@early-bird.think.com> 
<BZS.91Dec31020504@world.std.com>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 58
NNTP-Posting-Host: gandalf.think.com

In article <BZS.91Dec31020...@world.std.com> b...@world.std.com 
(Barry Shein) writes:
>Typical wysiwygs (which are now becoming common-place on Unix also,
>you're showing your age) are hands-down easier for the short
>documents, a few to several pages.

I don't know that I would characterize the availability of Unix WYSIWYG
systems as common-place yet.  They're available, but they tend to be
expensive.  We use both Frame and Interleaf here, and we can only afford a
handful of licenses for each (enough for the technical writers who use them
all the time, plus a couple extra for other people who need them on random
occasions).

By the way, our use here is the opposite of your description.  Our user
manuals are all written in Interleaf, while many of the developers use TeX
for typesetting journal articles.

>I have never understood the argument that Unix is hard to use because
>it has all these zillions of programs (while DOS is easier to use
>because it has all these zillions of programs...hmmm.)
>It's as if someone walks up to a Unix system and feels compelled to
>master everything in /usr/bin and so forth before touching the
>keyboard.

You know, I used to be one of the people who said the same thing.  But
there does seem to be this psychological barrier.  Users are scared of
options.

I think it also has something to do with the way people are introduced to
these systems.  Users are told that Unix is wonderful because it's made up
of lots of simple commands that you can combine to do useful work.  The
power of the standard utilities is touted as the strength of Unix, but
non-technical users don't want to be told that they should be writing shell
scripts to make effective use of Unix.  But since Unix users are expected
to write shell scripts, there isn't as much canned software available for
the users who don't want to.  DOS users aren't expected to write their own
scripts and programs, and Mac users generally can't, so there's a thriving
market of applications (both commercial and shareware), from the trivial to
fancy; in the Unix world, the popular publically-available programs are
either full-blown applications (e.g. emacs and MH) and shells and scripting
languages (e.g. perl).

>In my experience (over 15 years with Unix, and many other OS's) I find
>that the ability of people to learn any software system seems to have
>a lot more to do with their motivation to accomplish what they've set
>out to do than what most people talk about as being important. In
>fact, so long as six-year-olds can punch long strings of digits to
>find their friends (the telephone system, which has had some modest
>commercial success) I will insist that we have no idea what we are
>talking about when we say the words "user-friendly."

I agree.  The problem seems to be that most adults aren't motivated to
overcome their fear of computers, so the user interface must be adjusted to
be more approachable.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!ieee.org!sdd.hp.com!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: OS wars (was Re: What happened to Microsoft?)
Message-ID: <km1b8mINN4qr@early-bird.think.com>
Date: 31 Dec 91 17:47:02 GMT
References: <chris.694148945@rufus> <km01j8INNgol@early-bird.think.com> 
<1126@ra.nrl.navy.mil>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 51
NNTP-Posting-Host: gandalf.think.com

In article <1...@ra.nrl.navy.mil> law...@BDCD102.nrl.navy.mil (Drew Lawson) 
writes:
>In article <km01j8INN...@early-bird.think.com> bar...@think.com 
(Barry Margolin) writes:
>>In this case, I was referring to the OSes along with their common
>>applications.  So, for instance, word processing on Unix tends to be harder
>>than on MS-DOS (TeX and nroff are not "user friendly").
>
>Then I assume that you are comparing nroff with EDLIN.  There is no
>other "standard" word processing under DOS.  

I didn't say "standard", I said "common".  Anyone who buys a Mac or PC
knows (or finds out quickly) that they need to buy something like MS Word.
Typical DOS users never use EDLIN, and typical Unix user never use ed (I'll
bet half the Unix programmers here don't even know how to use it).

>					      If you are comparing it
>with word processers, you are making two mistakes.  The first is to
>mistake TeX and nroff for word processors.  They are typesetters.  They
>are both more powerful and more difficult (or awkward) to use than word
>processors.

Call them what you want, the fact remains that people who do word
processing on PC's mostly use something like MS Word, while people who do
word processing on Unix mostly use a combination of something like vi or
emacs and something like TeX or nroff.

>The second mistake is to think that your "standard" DOS word processors
>are not available under Unix.  Several are, as well as spreadsheets and
>databases (REAL databases, not dBase).

Available != common.  So long as Unix systems come with vi and nroff
bundled, and Emacs and TeX available for free, it's frequently hard to
justify spending thousands of dollars for Interleaf or Frame (I don't know
what the single-user licenses for these are -- I have to concern myself
with supplying a network of hundreds of workstations).

>I would like to add one other comment to this thread of OS debates.
>When the discussion is "what should I buy," availability of
>applications is very important.  When the discussion is (as this one
>has been) "Is this a good OS," the market is irrelevant.

I agree.  This started because I was trying to explain why DOS is popular
even though it is an inferior OS to Unix and the Mac.

Maybe I was wrong when I said that DOS is easier to use.  What I should
have said is that many users *think* it's easier to use.  But since ease of
use is a psychological factor, that belief can easily be self-fulfilling.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec31.200605.11243@microsoft.com>
Date: 31 Dec 91 20:06:05 GMT
References: <kltfffINNgim@early-bird.think.com> 
<1991Dec30.215753.27010@microsoft.com> <klvgq6INNc1t@early-bird.think.com>
Organization: Microsoft Corp.
Lines: 97

In article <klvgq6INN...@early-bird.think.com> bar...@think.com (Barry Margolin) 
writes:
>
>Yes, if the OS doesn't do something the way the application developers
>want, it makes sense (in some environments, such as single-user PC's, but
>generally not multiuser systems) to allow the application to bypass the OS.
>But a minimalist OS like DOS often forces application developers to
>implement system facilities that they don't really care about the details
>of, simple because the OS doesn't provide it at all.  In other words, the
>OS should provide what most apps need, but it shouldn't overly restrict
>what apps can do.

The problem is, It's sometimes difficult to add new OS functionality without
changing the insides of things, and when a lot of the Apps writers started
circumventing things, so they could do their hacks, it made it difficult
to impossible to add new functionality.

>
>>>There's a big difference between the OS and the OS shell.  Whenever you're
>>>using an application you're using the OS indirectly, and the quality of the
>>>OS is reflected in the quality of the applications.
>>
>>In a minimal system like DOS, where the responsability falls more on the
>>Applications, then many of the problems are related to the applications also.
>
>And that's the problem: application developers shouldn't be required to be
>experts at both the application domain and OS programming.  Ideally, they
>should be experts in the application domain, and the OS should enable them
>to develop the best applications.

The problem is, a lot of the App writers who are complaining are the ones
who started out doing systems hacks, which made it difficult to advance
the OS while still keeping it compatable with their apps.  There is the
DOS Apps writers bible, "UNDOCUMENTED DOS".

>
>I think many people *assume* a product is superior if it sells well.
>Unfortunately, it's hard to define objective criteria for the quality of a
>product, especially as nebulous a product as an OS, and even more
>especially in a manner that would make sense to the non-technical people
>who often must make purchase decisions.  So instead, they often go with the
>flow; a million lemmings can't be wrong.

Well, the marketing hacks sure do.

>
>>						   I judge UNIX to be inferior...
>>....On Real Time control systems, you gotta get a Real Time OS these days.  
>
>You're right, the quality of an OS must be related to the intended use.  I
>suggest that we consider the context of general purpose computing: word
>processing, simple graphics, financial applications, and software
>development.  When you start including more specialized applications such
>as real-time control and fancy graphics your choices are often restricted
>more by the availability of specialized software and hardware than by the
>OS itself (yes, the availability of these things is influenced by the OS,
>and vice versa, but it adds too many variables to hold a meaningful debate
>such as this (which, by the way, seems to be the calmest "my OS can beat
>your OS" discussion I've ever seen)).

This is a calm discussion 'cause neither of us are college freshmen who
think they know all about the industry 'cause we hacked a few programs
in high school :-)

>
>Yes, MS-DOS is *finally* starting to provide these things.  But Windows
>still isn't in every PC, so many applications (especially games) can't
>depend on it.

Games tend to circumvent DOS to squeeze every bit of performance they
can out the hardware anyway.
>
>>Why does everyone keep on thinking that I said that stock prices reflect
>>quality of products, or whatever.  All I said is capital gains tax.
>
>We were responding to the person who said "if you bought $300 of MS stock
>...  in 1985 ... today it would be worth over $10,000."  He was clearly
>implying that this reflected something about the products.  I still don't
>know what your "capital gains tax" response was supposed to mean.

It was in response to the rich getting richer/poor getting poorer comment.
Capital Gains Tax taxes stocks and investments where the rich get all their
money.
>
>Yes, Unix generally loses out in the areas related to ease of use and
>application availability; it was designed as a programmer's OS, and it
>still reflects that heritage.  However, the Mac beats MS-DOS in ease of use
>and is only slightly behind in application availability.  Actually, I think
>you're likely to be able to find reasonable Mac applications for most areas
>that there are DOS applications.  DOS often gives you more choices within
>an application domain, though; this is a natural result of the larger
>number of DOS machines -- there's more room for competition.

Yes, Mac still has the heritage of being a closed archetecture, which
set them back a bit.

pet...@microsoft.com
My Opinions.

Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec31.211725.14002@microsoft.com>
Date: 31 Dec 91 21:17:25 GMT
References: <kltfffINNgim@early-bird.think.com> 
<1991Dec30.215753.27010@microsoft.com> <avg.694142491@uunet>
Organization: Microsoft Corp.
Lines: 19

In article <avg.694142491@uunet> a...@uunet.uu.net (Vadim Antonov) writes:
>It always amazed me how different writing styles of Unix and MSDOS guys are.
>Having English as the second language I still can say for sure who is who
>having only few lines of their writings.
>
>Is the low cultural level a consequence of or reason for enjoying MSDOS?
>
>Read some books on operating systems, lad.

My lack of english skills is not due to MSDOS or MS, but to the fact
that I'm and engineer.  I'ts a requirement for engineering that you have
NO grasp of the english language :-)

As for reading books on OS, I have.  They're all written by Academic
UNIX hacks (generialization).  What does that have to do with the
real world ;-)

pet...@microsoft.com
My Opinions.

Path: gmdzi!unido!mcsun!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1991Dec31.212218.14301@microsoft.com>
Date: 31 Dec 91 21:22:18 GMT
References: <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com> <chris.694148945@rufus>
Organization: Microsoft Corp.
Lines: 33

In article <chris.694148945@rufus> ch...@state.COM.AU (Chris Keane) writes:
>bar...@think.com (Barry Margolin) writes:
>
>
>Our users here use Unix quite happily. Mind you, they have a totally
>graphics interface and only csh if they feel the need. More or less like 
>Windows, I guess. Except they run it faster than a 486 and at less cost
>(at least at $A prices) than a 16" colour memoried 486. And of course, if
>they decide they want to use a different interface one day, absolutely no
>upgrade is required.
>
>When DOS and/or Windows can do this, come back and tell me that Unix isn't
>easy to use and is more expensive.

Okay, I'll install UNIX on my home system, you wanna come administer it for me.
maybe you should replace the word "use" with the word administer. >
>
>p.s. for those moaning about how expensive Unix is for a PC, go look at
>4.4BSD for less than $US 1000 with *source* code. It can run, I'm told, on
>a 40 meg disk and it *runs applications*

Why would a user spend $1000 when he/she could spend <$200.  USERS DO NOT NEED
SOURCE CODE, and a lot of them don't have system administraters to take care
of all the dirty work.

>regards...                                                 
>Chris Keane.                  State Bank NSW            ph. +61 2 259 4459
>Unix Systems Administrator   (Group Treasury)     ch...@rufus.state.COM.AU
>Disclaimer: These are my own opinions, but I'm insane. What's your excuse?
>

pet...@microsoft.com
My Opinions.  My Excuse: I'm Stupid.

Path: gmdzi!unido!mcsun!uunet!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <km1rkuINNfl6@early-bird.think.com>
Date: 31 Dec 91 22:26:38 GMT
References: <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com> <1991Dec31.200605.11243@microsoft.com>
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 53
NNTP-Posting-Host: gandalf.think.com

In article <1991Dec31.200605.11...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>In article <klvgq6INN...@early-bird.think.com> bar...@think.com 
(Barry Margolin) writes:
>>Yes, if the OS doesn't do something the way the application developers
>>want, it makes sense (in some environments, such as single-user PC's, but
>>generally not multiuser systems) to allow the application to bypass the OS.
>The problem is, It's sometimes difficult to add new OS functionality without
>changing the insides of things, and when a lot of the Apps writers started
>circumventing things, so they could do their hacks, it made it difficult
>to impossible to add new functionality.

The problem is that DOS was so minimal that just about every application
developer had to bypass it and/or make use of internal, undocumented
features.  Then Microsoft is stuck supporting them forever.  By providing a
reasonable API to the OS, you reduce the number of applications that must
bypass it, and lessen the impact of changes to the internals.

This was Apple's strategy with the Mac.  In most cases when an application
was incompatible with a new OS release, it was reasonable to blame the
application developer.

And in a multiuser system like Unix, a reasonable OS interface *must* be
supplied, because applications generally can't bypass it.  When you're
designing a single-user OS it's often a good strategy to pretend that
you're designing a multiuser system with a protected kernel; this forces
you to think harder about the system calls you provide, as they're the only
way to get anything done.  Then once you've done that design, you stop
pretending and think about how to let applications bypass the OS in a
controlled way.

This mention of protected versus unprotected kernels reminds me about
computer virii.  One of the reasons why viruses are so prevalent in the
PC/Mac world is that these systems are designed to permit applications to
take over the machine.  Since the system can't tell the difference between
a benign application and one infected with a virus, they're essentially
defenseless.  Multiuser systems, on the other hand, define (or at least
attempt to) a strict separation between what user programs and system
programs can do; direct access to filesystem disks, for example, is in the
system domain, and programs run by ordinary users can't infect the OS (but
if you can get the superuser to run an infected program, it's possible).

I bring this up because many personal computer users blame the systems for
being so vulnerable to virii.  Both the 80x86 and 680x0 family provide
support for protecting the OS, but this can't be used in DOS or the Mac
because so many applications and utilities depend on being able to bypass
the OS or insert parts of themselves into the OS (i.e. by patching
interrupt vectors).  You can't have it both ways, because the only
difference between patching and infecting is whether the intent is good or
bad.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar


Path: gmdzi!ieee.org!sdd.hp.com!wupost!uunet!van-bc!rsoft!mindlink!a218
From: Charlie_Gi...@mindlink.bc.ca (Charlie Gibbs)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <9227@mindlink.bc.ca>
Date: 1 Jan 92 00:35:46 GMT
Organization: MIND LINK! - British Columbia, Canada
Lines: 16

In article <1...@ra.nrl.navy.mil> law...@BDCD102.nrl.navy.mil
(Drew Lawson) writes:

>The crime in the current PC ("personal computer" NOT "politically
>correct") environment is that MicroSoft was never willing to make
>MS/DOS into an operating system, and they should have.

     Actually, the way IBM (in typical Orwellian Newspeak fashion)
has tried to pre-empt the term "PC" - and the way the industry in
its usual way has gone along - makes "politically correct" a much
more realistic interpretation of "PC".  Either that or, judging
from the number of people who say "PCs and Macs", the Mac really
isn't a personal computer at all.

Charlie_Gi...@mindlink.bc.ca
My Amiga 1000 _is_ a PC!  It says "Personal Computer" right on the box!

Path: gmdzi!unido!mcsun!uunet!psinntp!sugar!taronga!peter
From: pe...@taronga.com (Peter da Silva)
Newsgroups: alt.folklore.computers
Subject: Re: OS wars (was Re: What happened to Microsoft?)
Message-ID: <YVBB343@taronga.com>
Date: 1 Jan 92 04:56:51 GMT
References: <BZS.91Dec31020504@world.std.com>
Organization: Taronga Park BBS
Lines: 13

b...@world.std.com (Barry Shein) writes:
> The same sort of thing applies to emacs, you can use emacs perfectly
> well with less than ten commands, but again somehow because it has
> hundreds it's therefore hard to learn.

The same is true of VI, or for that matter any halfway decent editor.

(not to flame, but VI has more of a rep for hostility than Emacs for
some reason)
-- 
-- Peter da Silva. Taronga Park BBS.  +1 713 568 0480|1032 2400/n/8/1.
-- `-_-'    "Have you hugged your wolf today?"
--  'U`

Newsgroups: alt.folklore.computers
Path: sparky!uunet!usc!snorkelwacker.mit.edu!world!bzs
From: b...@world.std.com (Barry Shein)
Subject: Re: OS wars (was Re: What happened to Microsoft?)
In-Reply-To: barmar@think.com's message of 31 Dec 1991 17:32:02 GMT
Message-ID: <BZS.92Jan1191031@world.std.com>
Sender: b...@world.std.com (Barry Shein)
Organization: The World
References: <chris.694148945@rufus> <km01j8INNgol@early-bird.think.com>
	<BZS.91Dec31020504@world.std.com> <km1aciINN4ed@early-bird.think.com>
Date: Thu, 2 Jan 1992 00:10:31 GMT


From: bar...@think.com (Barry Margolin)
>I don't know that I would characterize the availability of Unix WYSIWYG
>systems as common-place yet.  They're available, but they tend to be
>expensive.

Island Write is a perfectly good, simple WYSIWYG and costs like $395,
not sure what you mean by expensive. I suspect that's not the least
expensive either.

Seriously, have you been keeping up with the market? Things are
changing fast.

>By the way, our use here is the opposite of your description.  Our user
>manuals are all written in Interleaf, while many of the developers use TeX
>for typesetting journal articles.

I believe you...but is it a good thing? I dunno.

>>In my experience (over 15 years with Unix, and many other OS's) I find
>>that the ability of people to learn any software system seems to have
>>a lot more to do with their motivation to accomplish what they've set
>>out to do than what most people talk about as being important. In
>>fact, so long as six-year-olds can punch long strings of digits to
>>find their friends (the telephone system, which has had some modest
>>commercial success) I will insist that we have no idea what we are
>>talking about when we say the words "user-friendly."
>
>I agree.  The problem seems to be that most adults aren't motivated to
>overcome their fear of computers, so the user interface must be adjusted to
>be more approachable.

Well, that's ok, the future doesn't belong to adults...cheers.

(didn't people complain bitterly when they introduced direct dialing
into phone systems?)


-- 
        -Barry Shein

Software Tool & Die    | b...@world.std.com          | uunet!world!bzs
Purveyors to the Trade | Voice: 617-739-0202        | Login: 617-739-WRLD

Newsgroups: alt.folklore.computers
Path: sparky!uunet!world!bzs
From: b...@world.std.com (Barry Shein)
Subject: Re: OS wars (was Re: What happened to Microsoft?)
In-Reply-To: peter@taronga.com's message of Wed, 1 Jan 1992 04:56:51 GMT
Message-ID: <BZS.92Jan1191434@world.std.com>
Sender: b...@world.std.com (Barry Shein)
Organization: The World
References: <BZS.91Dec31020504@world.std.com> <YVBB343@taronga.com>
Date: Thu, 2 Jan 1992 00:14:34 GMT


From: pe...@taronga.com (Peter da Silva)
>b...@world.std.com (Barry Shein) writes:
>> The same sort of thing applies to emacs, you can use emacs perfectly
>> well with less than ten commands, but again somehow because it has
>> hundreds it's therefore hard to learn.
>
>The same is true of VI, or for that matter any halfway decent editor.
>
>(not to flame, but VI has more of a rep for hostility than Emacs for
>some reason)

Understand, I use all the editors on our system regularly (people have
even wondered if that's true and grovelled the lastcomm of a typical
day for me and concluded it's true, I'm always firing up vi or ed or
emacs etc.)

One problem with vi is its mode-ism, just yesterday I turned a file
into mashed potatoes accidently because I thought I was in insert
mode, typed a few chars in rapid succession and wham, had to quit and
throw away what work I had done, the file was crunched.

I dunno, people can learn vi and use it perfectly well, that's not my
point, but I can understand why some people develop a certain anxiety
about it.

-- 
        -Barry Shein

Software Tool & Die    | b...@world.std.com          | uunet!world!bzs
Purveyors to the Trade | Voice: 617-739-0202        | Login: 617-739-WRLD

Path: sparky!uunet!know!hri.com!ukma!nsisrv!mimsy!ra!BDCD102.nrl.navy.mil!lawson
From: law...@BDCD102.nrl.navy.mil (Drew Lawson)
Newsgroups: alt.folklore.computers
Subject: Twisted Vocabulary (Re: What happened to Microsoft?)
Message-ID: <1145@ra.nrl.navy.mil>
Date: 2 Jan 92 15:32:04 GMT
References: <9227@mindlink.bc.ca>
Sender: use...@ra.nrl.navy.mil
Organization: Backgrounds Data Center
Lines: 39

In article <9...@mindlink.bc.ca>
Charlie_Gi...@mindlink.bc.ca (Charlie Gibbs) writes:
>In article <1...@ra.nrl.navy.mil> law...@BDCD102.nrl.navy.mil
>(Drew Lawson) writes:

>>The crime in the current PC ("personal computer" NOT "politically
>>correct") 

>     Actually, the way IBM (in typical Orwellian Newspeak fashion)
>has tried to pre-empt the term "PC" - and the way the industry in
>its usual way has gone along - makes "politically correct" a much
>more realistic interpretation of "PC".  Either that or, judging
>from the number of people who say "PCs and Macs", the Mac really
>isn't a personal computer at all.

This is a tendency that I've found at least confusing, and often
irritating.  Often people confuse/interchange related terms.  I don't
know how many times I've fallen into tangles trying to discuss (for
example) "Unix on a PC" (or "micro").  Once I say "PC," it is
translated into "DOS."

So a 80386 running Unix/Xenix/Anything_not_DOS isn't a personal
computer, as with a Mac.

Another place I've run into this is in job hunting.  I've spent a lot
of time working with Unix on Vax/microVax platforms.  Once a resume
reader sees "Vax," I get a call for an interview for a VMS position.
VMS is OK (if verbose), but not what I like.


I expect vocabulary drift within non-technical areas, but get surprised
by how prevelant it is in this (technical) area.

-- 
+------------------------------+--------------------------------------+
| Drew Lawson                  |  If you're not part of the solution, |
| law...@bdcd102.nrl.navy.mil  |  you're part of the precipitate.     |
| 71141.1...@CompuServe.COM    |                                      |
+------------------------------+--------------------------------------+

Path: sparky!uunet!spool.mu.edu!umn.edu!cs.umn.edu!uc.msc.edu!noc.MR.NET!
news.stolaf.edu!asgaard.acc.stolaf.edu!seebs
From: se...@asgaard.acc.stolaf.edu (The Laughing Prophet)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan2.124841.18333@news.stolaf.edu>
Date: 2 Jan 92 12:48:41 GMT
Article-I.D.: news.1992Jan2.124841.18333
References: <1991Dec30.042044.14542@microsoft.com> 
<1991Dec30.060231.29999@Princeton.EDU> <1991Dec30.213338.26078@microsoft.com>
Sender: n...@news.stolaf.edu
Organization: SporkMungus
Lines: 14

Pete, you keep claiming that windows doesn't suck.

Can you provide any other description of it?  I don't wish to be harsh,
but, well, I have a friend who wrote an OS for his C64 that multitasked
better.

Geez. I mean, AmigaDos1.0 was better than *windows*.

-s
-- 
se...@acc.stolaf.edu  "Forgive them, Father, for they do not get the joke."
Amiga - Hack or die.    Reading news so I can put off my work.
Unix - Live free or die.   Monogamy is a disease.
signature virus 2.1c - Copy me into your .signature if you love Jesus!

Path: sparky!uunet!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: Twisted Vocabulary (Re: What happened to Microsoft?)
Date: 2 Jan 1992 20:40:42 GMT
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 37
Message-ID: <km6u6aINNe2l@early-bird.think.com>
References: <9227@mindlink.bc.ca> <1145@ra.nrl.navy.mil>
NNTP-Posting-Host: nunu.think.com

In article <1...@ra.nrl.navy.mil> law...@BDCD102.nrl.navy.mil (Drew Lawson) 
writes:
>This is a tendency that I've found at least confusing, and often
>irritating.  Often people confuse/interchange related terms.  I don't
>know how many times I've fallen into tangles trying to discuss (for
>example) "Unix on a PC" (or "micro").  Once I say "PC," it is
>translated into "DOS."

I've given up on this one.  When I want to use a term that encompasses
IBM-PC's, compatibles, Macs, Amigas, etc. I'll spell out "personal
computer".

>So a 80386 running Unix/Xenix/Anything_not_DOS isn't a personal
>computer, as with a Mac.

Sure it is; it's just not a "PC".  PC is no longer an abbreviation for
"personal computer", it's short for IBM-PC.

>Another place I've run into this is in job hunting.  I've spent a lot
>of time working with Unix on Vax/microVax platforms.  Once a resume
>reader sees "Vax," I get a call for an interview for a VMS position.
>VMS is OK (if verbose), but not what I like.

This is kind of understandable.  Unix runs on many machines, so a Unix
expert is not likely to describe himself as a Vax person, even if all his
experience has been on Vaxen.  VMS is the standard Vax OS, although other
OSes (Unix, Pick) will run on a Vax.

>I expect vocabulary drift within non-technical areas, but get surprised
>by how prevelant it is in this (technical) area.

In the case of "PC", there are lots of non-technical people in the
community.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Newsgroups: alt.folklore.computers
Path: sparky!uunet!cs.utexas.edu!qt.cs.utexas.edu!yale.edu!yale!
mintaka.lcs.mit.edu!nntp.lcs.mit.edu!mday
From: m...@pion.lcs.mit.edu (Mark Day)
Subject: Re: What happened to Microsoft?
In-Reply-To: petesk@microsoft.com's message of 31 Dec 91 21:22:18 GMT
Message-ID: <MDAY.92Jan3113613@pion.lcs.mit.edu>
Sender: n...@mintaka.lcs.mit.edu
Organization: MIT Laboratory for Computer Science
References: <1991Dec30.215753.27010@microsoft.com>
	<klvgq6INNc1t@early-bird.think.com> <chris.694148945@rufus>
	<1991Dec31.212218.14301@microsoft.com>
Date: Fri, 3 Jan 1992 16:36:13 GMT
Lines: 20

   pet...@microsoft.com (Pete SKELLEY) writes:

   USERS DO NOT NEED
   SOURCE CODE

This is pathetic.  Mr. Skelley is confusing 

"not all users need source code"  (true)

with

"no users need source code."  (false)

Systems with source code are infinitely more wonderful than systems
without it, for the situations where you need source code.  Apologies
if I seem to be stating the obvious to non-Microsoft readers.

--Mark Day

m...@lcs.mit.edu

Newsgroups: alt.folklore.computers
Path: sparky!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan03.222055.22925@microsoft.com>
Date: 03 Jan 92 22:20:55 GMT
Organization: Microsoft Corp.
References: <1991Dec30.060231.29999@Princeton.EDU> 
<1991Dec30.213338.26078@microsoft.com> <1992Jan2.124841.18333@news.stolaf.edu>
Lines: 28

In article <1992Jan2.124841.18...@news.stolaf.edu> 
se...@asgaard.acc.stolaf.edu (The Laughing Prophet) writes:
>Pete, you keep claiming that windows doesn't suck.
>
>Can you provide any other description of it?  I don't wish to be harsh,
>but, well, I have a friend who wrote an OS for his C64 that multitasked
>better.

It's not that windows doesn't suck, it's that everything else sucks also.

I've written multitasking OS's on the 6502 also, but let me ask you this.
If I were to come out with a graphics extender card (like a VGA card) for
the 64, would your friends OS be able to handle it.  What about a coprocessor,
a lan card, etc.  What if it just wasn't me, but a 100's of companies comming
out with extension cards for the 64.  Would his OS be able to handle things
then.  As far as multitasking better, well, did it have memory protection,
anyone can write a task switcher if the programs are nice. (note: i'm not
implying that windows is better cause it does this, in fact, it doesn't)

When you get down to it, I'm just tired of people flaming MS for the sake
of flaming MS.  What is it, are people jealous because MS is successful.
IMHO, MS has been playing the game pretty fair.  That's why I came to
work here.


pet...@microsoft.com
My Opinions.

Newsgroups: alt.folklore.computers
Path: sparky!uunet!zaphod.mps.ohio-state.edu!menudo.uh.edu!sugar!taronga!peter
From: pe...@taronga.com (Peter da Silva)
Subject: Re: What happened to Microsoft?
Message-ID: <0OFBDTM@taronga.com>
Organization: Taronga Park BBS
References: <1992Jan03.222055.22925@microsoft.com>
Date: Sat, 4 Jan 1992 14:35:18 GMT

pet...@microsoft.com (Pete SKELLEY) writes:
> It's not that windows doesn't suck, it's that everything else sucks also.

You need to broaden your horizons. Windows just *barely* manages to handle
LAN cards, doesn't handle tape at all, and would be completely lost at the
thought of a bidirectional parallel port. Try doing SCSI on Windows. Get a
generic SCSI driver then try and access a tape drive. Whoops.

> When you get down to it, I'm just tired of people flaming MS for the sake
> of flaming MS.

And I'm tired of people defending MS for the sake of defending MS. MS has
done some good work, but they're completely out of their depth when it
comes to operating systems. You gotta accept that different companies have
different strengths and weaknesses. Systems aren't one of Microsoft's
strengths. Both DOS and Xenix were picked up from third parties, and
Windows is a horrible kludge.
-- 
-- Peter da Silva. Taronga Park BBS.  +1 713 568 0480|1032 2400/n/8/1.
-- `-_-'    "Have you hugged your wolf today?"
--  'U`

Path: sparky!uunet!mcsun!sun4nl!alchemy!ruunfs!ruunfs!muts
From: m...@fysap.fys.ruu.nl (Peter Mutsaers)
Newsgroups: alt.folklore.computers
Subject: Graphical vs. non graphical interface (was: Re: OS wars)
Message-ID: <MUTS.92Jan5141243@fysaq.fys.ruu.nl>
Date: 5 Jan 92 13:12:45 GMT
References: <6625@gmdzi.gmd.de> <L1FBFGF@taronga.com> <6628@gmdzi.gmd.de>
Sender: use...@fys.ruu.nl (News Control)
Organization: Rijks Universiteit Utrecht
Lines: 27
In-Reply-To: strobl@gmdzi.gmd.de's message of 5 Jan 92 00:06:01 GMT

This discussion appears to be centered around WYSIWYG or not, and
mouse+menues vs. command line shells.

I think a lot of people waste a lot of productivity because they use
menues etc. It is faster for the first 5 minutes, I admit, and for non
experienced computer users may even be faster for the first few days.

So for programs you use casually mouse+menues may be okay; for daily
work however, you will no doubt learn all those commands and
keystrokes. Anyone can see that typing CTR-ALT-SHIT-f1 or whatever is
faster that moving your mouse to a menu, selecting a sub-menu, and
finally reaching your target command.

Some will say that e.g. a lot of Mac programs offer both
possibilities, but the fact that you can use the mouse will tempt you
to never try to remember the shortcut keystrokes, thus losing precious
time at every command you execute.

By the way, about vi having no notion of paragraphs etc., as a
wordprocessor does (mentioned in the article I am reacting to), you
could ask yourself why this should be. And if you really want this,
use emacs. (now the vi vs. emacs war can start again :-).
--
_________________________________________________________________________
Peter Mutsaers.    RUU physics dept. Heidelberglaan 5, Utrecht, Nederland
m...@fysap.fys.ruu.nl   |================================================
tel: (+31)-(0)30-533880 |              Memento Mori

Path: sparky!uunet!olivea!samsung!zaphod.mps.ohio-state.edu!usc!
chaph.usc.edu!news
From: spit...@girtab.usc.edu (William Spitzak)
Newsgroups: alt.folklore.computers
Subject: Re: Graphical vs. non graphical interface (was: Re: OS wars)
Message-ID: <kmeittINNevd@girtab.usc.edu>
Date: 5 Jan 92 18:17:33 GMT
References: <L1FBFGF@taronga.com> <6628@gmdzi.gmd.de> 
<MUTS.92Jan5141243@fysaq.fys.ruu.nl> <1992Jan5.143922.988@klaava.Helsinki.FI>
Organization: University of Southern California, Los Angeles, CA
Lines: 52
NNTP-Posting-Host: girtab.usc.edu

In article <MUTS.92Jan5141...@fysaq.fys.ruu.nl> m...@fysap.fys.ruu.nl 
(Peter Mutsaers) writes:
>This discussion appears to be centered around WYSIWYG or not, and
>mouse+menues vs. command line shells.

>I think a lot of people waste a lot of productivity because they use
>menues etc. It is faster for the first 5 minutes, I admit, and for non
>experienced computer users may even be faster for the first few days.

Most of the slowness of menus can be negated by pop-up menus.  These
require a dedicated button on the mouse (which I will call the "right
mouse button" to indicate my preference).

How to use the menus:  You point at an object of interest.  You press
the right button.  While holding it down, a pop-up menu appears of things
you can do to that object.  You drag (perhaps into submenus) until the
mouse is pointing at the operation you want.  You release the mouse
button.  The operation is done.  If you decide you do not want to
do any of the menu items you move the mouse so it is not pointing at
any and release.  The menu vanishes after the operation is complete.
If you subsequently point at a similar object and press the right button
you get the same menu but it is positioned such that you are pointing
at the same operation you picked last time.

Unfortunately NO system I have ever seen correctly does this.  This is
mostly because the MAC has warped everybody's thinking into believeing
you have to point the mouse at some special "menu thing" before you
can get a menu.  Some systems are really deranged, requiring you to
BOTH point at a "special thing" AND push the "special button" to get
a menu (the Amiga).  The few pop-up menus I have seen have insufficient
context-sensitivity (the NeXT is the *worst*, it does not matter where
the mouse is, you ALWAYS get the same menu!  This seems harder to
believe the longer I think about it...).  Most systems (i.e. X ones)
give you one menu when you point at a window's contents, another when
you point at the border, and a third when you point at empty (root)
screen.  All pop-up systems are hobbled by a Mac "special menu thing"
so there are two ways to bring up the menu (greatly limiting the power
of either method!).  I have never seen the "pick same operation again"
except in my own code, although SGI allows you to pop-up the menu with
the cursor pointing at the top item every time.

If this was done correctly, picking items off menus would be much faster
because you do not have to move the mouse.  Doing something many times
would be very fast, approaching the speed of repeat keyboard keys.  It
would save screen space as there would be none wasted on the "special
menu things".  There are significant programming advantages too, as
deciding on the menu's contents can be deferred until it is first
picked.  And although I'm not sure what it will lead too, it would seem
that context-sensitive menus depending on exactly what you are pointing
at would be a powerful addition to a UI.

-- 
Bill Spitzak | spit...@girtab.usc.edu | spit...@mcimail.com

Path: sparky!uunet!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: Graphical vs. non graphical interface (was: Re: OS wars)
Date: 5 Jan 1992 20:46:32 GMT
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 41
Message-ID: <kmerl8INNeh4@early-bird.think.com>
References: <MUTS.92Jan5141243@fysaq.fys.ruu.nl> 
<1992Jan5.143922.988@klaava.Helsinki.FI> <kmeittINNevd@girtab.usc.edu>
NNTP-Posting-Host: godot.think.com

In article <kmeittINN...@girtab.usc.edu> spit...@girtab.usc.edu 
(William Spitzak) writes:
>Most of the slowness of menus can be negated by pop-up menus.
[Description of pop-up menus omitted]
>Unfortunately NO system I have ever seen correctly does this.  This is
>mostly because the MAC has warped everybody's thinking into believeing
>you have to point the mouse at some special "menu thing" before you
>can get a menu.

Take a look at the GUI's that predated the Mac and you may find such
things, since they were designed before everyone was brainwashed.  (Aside:
I do think the Mac has about the best GUI of any well-known system, and
it's not a *bad* GUI, but it's missing some features that could make it the
best I've *ever* seen.)

The one I'm most familiar with is the Symbolics window system.  The right
mouse button almost always pops up a menu of operations to perform on the
specific object the mouse is pointing to.  About the only feature you
described that isn't consistently available is defaulting to the last-used
operation.  Some applications arrange for the default to be the last
operation performed, some others always use a menu-specific default, and
others don't provide any automatic default (i.e. the mouse starts out
pointing at the menu title); unfortunately, I think the higher-level
interfaces to the menu system don't make it easy to specify the default,
and programmers are strongly discouraged from using the low-level routines.

Another features of the Symbolics window system is a mouse documentation
line at the bottom of the screen.  This always contains a short description
of what each mouse button does at the current time.  Conceptually it's
similar to the Mac's Balloon Help, but it's much less obtrusive so it's
feasible to have on all the time.  Also, mouse-sensitive objects are always
highlighted as the mouse passes over them, so you know what object you're
about to operate on (isn't it a pain when you're in something like MacDraw
and you're trying to select one of two objects that are really close
together, but you can't tell which one the system thinks you're pointing
at?).

-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Newsgroups: alt.folklore.computers
Path: sparky!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan05.213540.4752@microsoft.com>
Date: 05 Jan 92 21:35:40 GMT
Organization: Microsoft Corp.
References: <chris.694148945@rufus> <1991Dec31.212218.14301@microsoft.com> 
<MDAY.92Jan3113613@pion.lcs.mit.edu>
Lines: 43

In article <MDAY.92Jan3113...@pion.lcs.mit.edu> m...@pion.lcs.mit.edu 
(Mark Day) writes:
>   pet...@microsoft.com (Pete SKELLEY) writes:
>
>   USERS DO NOT NEED
>   SOURCE CODE
>
>This is pathetic.  Mr. Skelley is confusing 
>
>"not all users need source code"  (true)
>
>with
>
>"no users need source code."  (false)

Calling my statement pathetic is unnecessarily judgemental.
Okay, I'll rephrase.

Most users do not need source code for their operating system.  This
is excluding of course batch files or .cmd files or whatever.  It may
be true that I'm a user, and I need source code, but I need the source
code in my capacity as a developer, not a user.

Being that my educaton was based on an EE and not a CS degree, I tend
to leave out phrases like "except for the pathalogical case", or whatever
when I make a statment like the one I made.  There are really few places
in EE where we can make statements about the truth of something as you can
in CS (alright logic), where you can limit your assumptions.
>
>Systems with source code are infinitely more wonderful than systems
>without it, for the situations where you need source code.  Apologies
>if I seem to be stating the obvious to non-Microsoft readers.

What is a secretary, or a small business owner, or a market trader going to
do with source code.  I'd say most computer users these days are not
people who use computer for the sake of using computer, but people who use
computer to get a job done.  If this isn't true, then I'd say we were all
wasing our time in this industry, because there'd be no socially usefull work
comming from computers.

pet...@microsoft.com
My Opinions.

Newsgroups: alt.folklore.computers
Path: sparky!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan05.215450.5146@microsoft.com>
Date: 05 Jan 92 21:54:50 GMT
Organization: Microsoft Corp.
References: <1992Jan03.222055.22925@microsoft.com> <0OFBDTM@taronga.com>
Lines: 43

In article <0OFB...@taronga.com> pe...@taronga.com (Peter da Silva) writes:
>pet...@microsoft.com (Pete SKELLEY) writes:
>> It's not that windows doesn't suck, it's that everything else sucks also.
>
>You need to broaden your horizons. Windows just *barely* manages to handle
>LAN cards, doesn't handle tape at all, and would be completely lost at the
>thought of a bidirectional parallel port. Try doing SCSI on Windows. Get a
>generic SCSI driver then try and access a tape drive. Whoops.

Is that windows fault, or the device drivers. 
And don't say that windows should come with all the device drivers I need.
Does BSD <4 come with any CD rom drivers.  Does it come with any image
capture board drivers.  As for getting a generic SCSI driver, it may
be that the driver archetecture or the SCSI archetecture of the card
you got for your PC is ad fault.
Write a driver for a bidirectional parallel port, it can be done.

>
>> When you get down to it, I'm just tired of people flaming MS for the sake
>> of flaming MS.
>
>And I'm tired of people defending MS for the sake of defending MS. MS has
>done some good work, but they're completely out of their depth when it
>comes to operating systems. You gotta accept that different companies have
>different strengths and weaknesses. Systems aren't one of Microsoft's
>strengths. Both DOS and Xenix were picked up from third parties, and
>Windows is a horrible kludge.

Have you ever liked an MS systems product?

I kinda liked OS/2 1.2 for a bit.  And as for buying DOS and Xenix from
elsewhere, well, MS also bough all of their programmers from elsewhere, and
I suspect most other companies do the same.  Were you born where you work?

I'll admit, I've probably been over-reacting, but how would you react if
someone was putting down the place you worked.  I like it here.

Windows is a step in the right direction.

pet...@microsoft.com
My Opinions.

Path: sparky!uunet!spool.mu.edu!umn.edu!noc.MR.NET!news.stolaf.edu!
asgaard.acc.stolaf.edu!seebs
From: se...@asgaard.acc.stolaf.edu (The Laughing Prophet)
Newsgroups: alt.folklore.computers
Subject: Users and source code?
Message-ID: <1992Jan6.042857.24264@news.stolaf.edu>
Date: 6 Jan 92 04:28:57 GMT
Article-I.D.: news.1992Jan6.042857.24264
References: <1991Dec31.212218.14301@microsoft.com> 
<MDAY.92Jan3113613@pion.lcs.mit.edu> <1992Jan05.213540.4752@microsoft.com>
Sender: n...@news.stolaf.edu
Organization: SporkMungus
Lines: 53

In article <1992Jan05.213540.4...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>Calling my statement pathetic is unnecessarily judgemental.

But probably rather accurate.

>Okay, I'll rephrase.

>Most users do not need source code for their operating system.

Quite probably true. But *all* users would be better off had they free
access to it.

>This is excluding of course batch files or .cmd files or whatever.

Why? Compile them. :)

>It may be true that I'm a user, and I need source code, but I need the
>source code in my capacity as a developer, not a user.

So? Maybe *some* people need source as users. Just because *you* only
need source code to develop, doesn't mean other people are the same.
If nothing else, the source code for an OS is the *best* documentation
you can get. :)

>What is a secretary, or a small business owner, or a market trader going to
>do with source code.

Fix all the bugs you morons have put in every OS you've ever released?
I dunno.

>I'd say most computer users these days are not people who use computer
>for the sake of using computer, but people who use computer to get a job done.

In that case, when they discover that some idiot has failed to make a part
of their OS work, they'd *better* have the option of fixing it.

>If this isn't true, then I'd say we were all wasting our time in this
>industry, because there'd be no socially usefull work coming from computers.

Nonsense. Using a computer for the sake of using the computer is what Root
made us for.

>pet...@microsoft.com
>My Opinions.

If I believed that, I'd own the Brooklyn Bridge.

-s
-- 
se...@acc.stolaf.edu  "Forgive them, Father, for they do not get the joke."
Amiga - Hack or die.    Reading news so I can put off my work.
Unix - Live free or die.   Monogamy is a disease.
em was tsrif uoy woh morf sdrawkcab em ypoc - a.2.2 suriv erutangis.

Newsgroups: alt.folklore.computers
Path: sparky!uunet!cs.utexas.edu!wupost!think.com!rpi!usenet.coe.montana.edu!
milton!fetrow
From: fet...@biostat.washington.edu (David Fetrow)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan6.084859.11505@milton.u.washington.edu>
Sender: n...@milton.u.washington.edu (News)
Organization: Biostatistics, University of Washington, Seattle
References: <1991Dec31.212218.14301@microsoft.com> 
<MDAY.92Jan3113613@pion.lcs.mit.edu> <1992Jan05.213540.4752@microsoft.com>
Date: Mon, 6 Jan 1992 08:48:59 GMT

In article <1992Jan05.213540.4...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>
>What is a secretary, or a small business owner, or a market trader going to
>do with source code.  I'd say most computer users these days are not
>people who use computer for the sake of using computer, but people who use
>computer to get a job done.  If this isn't true, then I'd say we were all
>wasing our time in this industry, because there'd be no socially usefull work
>comming from computers.
>

 ...but there support staffs sure could use the source. There is a subtle
area here. "Most users don't need to personally have source code" is probably
a true statement but "It's much nicer for them if support staff can
easily access source" is also true. (Yes, support staff isn't always called
that. Sometimes it's "The kid down the block I ask PC questions of." or
"That power user in accounting.").

 There is also a small but absoloutely vital group of users who typically
ask these amazing questions in the form of "..why can't the computer just do
that FOR me". An awful lot of good ideas come from people like that and some
of them become developers. It isn't a static thing. 

 999 out of a 1000 secretaries may not need to see source code but maybe
that 1000th one has a nifty idea and making the implementation of that
idea as un-difficult as possible can be of benefit to the other 999.

-- 
 -dave					   fet...@biostat.washington.edu

 "Until seeing the new district boundaries I'd assumed the state legislature
  had no grasp of higher mathematics. I was wrong. They know fractals." -anon

Newsgroups: alt.folklore.computers
Path: sparky!uunet!caen!zaphod.mps.ohio-state.edu!wupost!darwin.sura.net!
Sirius.dfn.de!chx400!bernina!bernina!neeri
From: ne...@iis.ethz.ch (Matthias Ulrich Neeracher)
Subject: Re: Users and source code?
In-Reply-To: seebs@asgaard.acc.stolaf.edu's message of 6 Jan 92 04:28:57 GMT
Message-ID: <NEERI.92Jan6101149@iis.ethz.ch>
Sender: n...@bernina.ethz.ch (USENET News System)
Nntp-Posting-Host: iis-gw
Organization: Integrated Systems Laboratory, ETH, Zurich
References: <1991Dec31.212218.14301@microsoft.com>
	<MDAY.92Jan3113613@pion.lcs.mit.edu>
	<1992Jan05.213540.4752@microsoft.com>
	<1992Jan6.042857.24264@news.stolaf.edu>
Date: 6 Jan 92 10:11:49
Lines: 20

In article <1992Jan6.042857.24...@news.stolaf.edu> 
se...@asgaard.acc.stolaf.edu (The Laughing Prophet) writes:
>So? Maybe *some* people need source as users. Just because *you* only
>need source code to develop, doesn't mean other people are the same.
>If nothing else, the source code for an OS is the *best* documentation
>you can get. :)

Of course, source is one of the more *reliable* forms of documentation (binary
is more reliable, binary + construction specs of machine is even more reliable,
and if you throw in the VLSI masks used for constructing the processor and a
General Unified Field Theory, you're starting to talk about *real* accuracy).

But I shouldn't forget that source code always only documents one particular
release of an OS. Personally, I prefer other forms of documentation to source code.

Matthias

-----
Matthias Neeracher                                      ne...@iis.ethz.ch
  "For a successful technology, reality must take precedence over public
   relations, for Nature cannot be fooled." - Richard P. Feynman

Path: sparky!uunet!usc!wupost!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: Users and source code?
Date: 6 Jan 1992 11:05:26 GMT
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 27
Message-ID: <kmgdvmINN36s@early-bird.think.com>
References: <1992Jan05.213540.4752@microsoft.com> 
<1992Jan6.042857.24264@news.stolaf.edu> <NEERI.92Jan6101149@iis.ethz.ch>
NNTP-Posting-Host: godot.think.com

In article <NEERI.92Jan6101...@iis.ethz.ch> ne...@iis.ethz.ch 
(Matthias Ulrich Neeracher) writes:
>But I shouldn't forget that source code always only documents one particular
>release of an OS. Personally, I prefer other forms of documentation to source code.

So do I, but you take what you can get.  If you don't have source code, and
the manual doesn't document something, it's very difficult to find out some
things.  Someone recently asked on the sun-managers list what the line
length limit is in /etc/exports; I had to hunt through various Sun sources
to find out.

To relate this to non-technical users, a word processor might have a limit
on the number of fonts in a document.  If you're lucky this will be
documented in the manual, but it's likely not to be, and the only way to
find out the limit is to read the source code.  You could also just try
putting lots of fonts in the document and see whether it complains;
however, that assumes that the word processor does limit checking and will
issue an understandable error, but it's quite possible that it doesn't
check and will just write beyond the end of some array, producing
unpredictable results.

I'm not saying that this is how programs should be written, but it's how
many programs *are* written.
-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Newsgroups: alt.folklore.computers
Path: sparky!uunet!psinntp!sugar!taronga!peter
From: pe...@taronga.com (Peter da Silva)
Subject: Re: What happened to Microsoft?
Message-ID: <9EIBUNE@taronga.com>
Organization: Taronga Park BBS
References: <1992Jan05.213540.4752@microsoft.com>
Date: Tue, 7 Jan 1992 01:32:09 GMT

pet...@microsoft.com (Pete SKELLEY) writes:
> What is a secretary, or a small business owner, or a market trader going to
> do with source code.

Hire someone to fix problems or provide enhancements instead of waiting for
the problem to not be fixed in the next release?

I can get an after-market cruise control installed in my car. Why can't I
get an after-market scheduler installed in my computer?
-- 
-- Peter da Silva. Taronga Park BBS.  +1 713 568 0480|1032 2400/n/8/1.
-- `-_-'    "Have you hugged your wolf today?"
--  'U`

Newsgroups: alt.folklore.computers
Path: sparky!uunet!microsoft!blakeco
From: blak...@microsoft.com (Blake Coverett)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan07.030108.15729@microsoft.com>
Date: 07 Jan 92 03:01:08 GMT
Organization: Microsoft Canada Inc.
References: <1991Dec30.215753.27010@microsoft.com> 
<klvgq6INNc1t@early-bird.think.com> <chris.694148945@rufus> 
<1991Dec31.212218.14301@microsoft.com> <MDAY.92Jan3113613@pion.lcs.mit.edu>
Reply-To: blak...@microsoft.COM (Blake Coverett)
Lines: 43

In article <MDAY.92Jan3113...@pion.lcs.mit.edu> m...@pion.lcs.mit.edu 
(Mark Day) writes:
>   pet...@microsoft.com (Pete SKELLEY) writes:
>
>   USERS DO NOT NEED
>   SOURCE CODE
>
>This is pathetic.  Mr. Skelley is confusing 
>"not all users need source code"  (true)
>with
>"no users need source code."  (false)

This (and the ensuing discussion which I have already read) leads me to toss
out a couple of thoughts.

1) When I look up the token "users" in my private symbol table I find 
   individuals who wouldn't know which end of a compiler to stick their
   source code into.  Those who would be able to make any use of source
   even strickly as documentation I don't map to the word "user".
   (Note this is completely orthagonal to whether source should be made
   available.)

2) How many of individuals or companys who's primary business is producing
   software make their source publicly available?  Do you?

>Systems with source code are infinitely more wonderful than systems
>without it, for the situations where you need source code.  

Absolutely.  See point number two.

>Apologies
>if I seem to be stating the obvious to non-Microsoft readers.

Duh! [slapping self on forehead, rendering well-practiced dumb-jock look]
Me from Microsoft guess me too stupid to understand.

Making derogatory comments about the astuteness of someone with whom you
disagree is really less than polite.

-Blake

-- 
#include <std/disclaimer.h>                       blak...@microsoft.COM
Mail Flames, Post Apologies.                      ...!uunet!microsoft!blakeco

Path: sparky!uunet!mcsun!uknet!warwick!slxsys!ibmpcug!dylan
From: dy...@ibmpcug.co.uk (Matthew Farwell)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan7.034343.3787@ibmpcug.co.uk>
Date: 7 Jan 92 03:43:43 GMT
References: <1991Dec31.212218.14301@microsoft.com> 
<MDAY.92Jan3113613@pion.lcs.mit.edu> <1992Jan05.213540.4752@microsoft.com> 
<1992Jan6.084859.11505@milton.u.washington.edu>
Reply-To: dy...@ibmpcug.co.uk (Matthew Farwell)
Organization: The IBM PC User Group, UK.
Lines: 46

In article <1992Jan6.084859.11...@milton.u.washington.edu> 
fet...@biostat.washington.edu (David Fetrow) writes:
>In article <1992Jan05.213540.4...@microsoft.com> pet...@microsoft.com 
(Pete SKELLEY) writes:
>>What is a secretary, or a small business owner, or a market trader going to
>>do with source code.  I'd say most computer users these days are not
>>people who use computer for the sake of using computer, but people who use
>>computer to get a job done.  If this isn't true, then I'd say we were all
>>wasing our time in this industry, because there'd be no socially usefull work
>>comming from computers.
>
> ...but there support staffs sure could use the source. There is a subtle
>area here. "Most users don't need to personally have source code" is probably
>a true statement but "It's much nicer for them if support staff can
>easily access source" is also true. (Yes, support staff isn't always called
>that. Sometimes it's "The kid down the block I ask PC questions of." or
>"That power user in accounting.").
>
> There is also a small but absoloutely vital group of users who typically
>ask these amazing questions in the form of "..why can't the computer just do
>that FOR me". An awful lot of good ideas come from people like that and some
>of them become developers. It isn't a static thing. 
>
> 999 out of a 1000 secretaries may not need to see source code but maybe
>that 1000th one has a nifty idea and making the implementation of that
>idea as un-difficult as possible can be of benefit to the other 999.

The problem with giving *everyone* source code is that everyone can then
change it. I personally know quite a few people who are nowhere near
qualified to change operating system sources. Do you really want people
mucking with fsck? *any* of the system calls? What happens if your user
makes a mistake and manages to make getuid() return 0 no matter what the
calling uid?

I've used a machine on which NFS was put into the kernel, and then taken
out again inexpertly.  The write() system call was broken.  It wrote the
data alright, but it only updated the st_mtime (modification time) and
st_ctime (change time) when the next sync came along. This is minor
compared to what could happen.

'The people who know what they're doing aren't dangerous - it's the
people who think they know what they're doing that are'

I can't remember who said this, but it seem appropriate.

Dylan.
p.s. Having said all this, I'd actually like source code to my machine,
so I can see whats going on and why things fail.

Path: sparky!uunet!wupost!think.com!barmar
From: bar...@think.com (Barry Margolin)
Newsgroups: alt.folklore.computers
Subject: Re: What happened to Microsoft?
Date: 7 Jan 1992 10:19:18 GMT
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 23
Message-ID: <kmivl6INNc9o@early-bird.think.com>
References: <1992Jan05.213540.4752@microsoft.com> 
<1992Jan6.084859.11505@milton.u.washington.edu> 
<1992Jan7.034343.3787@ibmpcug.co.uk>
NNTP-Posting-Host: godot.think.com

In article <1992Jan7.034343.3...@ibmpcug.co.uk> dy...@ibmpcug.co.uk 
(Matthew Farwell) writes:
>The problem with giving *everyone* source code is that everyone can then
>change it. ... What happens if your user
>makes a mistake and manages to make getuid() return 0 no matter what the
>calling uid?

Well, if the software had a warranty, I would expect modifications to the
system to void it.  This is the case for most hardware systems; the
original TRS-80 had paint over the screws so that they could tell whether
you'd opened it up.

The point is, if the user breaks the OS, they've screwed themselves.  It's
not the vendor's responsibility to protect the user from this.

You can also screw up the system by adb'ing the kernel, or simply by
misconfiguring it.  Does this mean that adb should not be supplied and
kernel customization should be removed?

-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Newsgroups: alt.folklore.computers
Path: sparky!uunet!microsoft!petesk
From: pet...@microsoft.com (Pete SKELLEY)
Subject: Re: What happened to Microsoft?
Message-ID: <1992Jan08.193458.8108@microsoft.com>
Date: 08 Jan 92 19:34:58 GMT
Organization: Microsoft Corp.
References: <1992Jan05.213540.4752@microsoft.com> <9EIBUNE@taronga.com>
Lines: 21

In article <9EIB...@taronga.com> pe...@taronga.com (Peter da Silva) writes:
>pet...@microsoft.com (Pete SKELLEY) writes:
>> What is a secretary, or a small business owner, or a market trader going to
>> do with source code.
>
>Hire someone to fix problems or provide enhancements instead of waiting for
>the problem to not be fixed in the next release?

This still doesn't explain why the secretary, small business owner, etc has
to have the source code.  Its the consultant who has to have the source code.

>
>I can get an after-market cruise control installed in my car. Why can't I
>get an after-market scheduler installed in my computer?

You can.

pet...@microsoft.com
My Opinions.