General Public-Friendly Linux -- Impossible?

General Public-Friendly Linux -- Impossible?

Post by Cuor di Mel » Sun, 31 Dec 1899 09:00:00



As long as there's "file system", "processes", "swap
partitions", "window managers", "command shells",.. is desktop Linux
just a fantasy? In my opinion, someone should take the useful parts of
the kernel source and drivers and develop a completely new non-Unix
based open source OS that could maybe become some competition to
Windows. Otherwise, I believe Linux is just way too complex and screwed
up for the desktop, a lot worse than Windows 95 has ever been.
Simplicity is the key, and Linux fails badly on that point too.

Sent via Deja.com http://www.deja.com/
Before you buy.

 
 
 

General Public-Friendly Linux -- Impossible?

Post by cl.. » Sun, 31 Dec 1899 09:00:00



> As long as there's "file system", "processes", "swap
> partitions", "window managers", "command shells",.. is desktop Linux
> just a fantasy? In my opinion, someone should take the useful parts of
> the kernel source and drivers and develop a completely new non-Unix
> based open source OS that could maybe become some competition to
> Windows. Otherwise, I believe Linux is just way too complex and screwed
> up for the desktop, a lot worse than Windows 95 has ever been.
> Simplicity is the key, and Linux fails badly on that point too.

It should be pretty clear by now that no one who's actually using Linux
cares what you think, astroturfer.

Linux desktops for the most part aren't going to follow the
Windows model. Get over it.

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Byron A Je » Sun, 31 Dec 1899 09:00:00




Quote:>As long as there's "file system", "processes", "swap
>partitions", "window managers", "command shells",.. is desktop Linux
>just a fantasy?

Well since the two don't have really anything to do with one another I
guess the question is moot.

Consider this: all versions of windows have filesystems, processes, swap,
window managers, and command shells. Every one. Yet it's the most prevalent
desktop today.

Quote:> In my opinion, someone should take the useful parts of
>the kernel source and drivers and develop a completely new non-Unix
>based open source OS that could maybe become some competition to
>Windows.

I'd really like to see an OS without the list you mentioned above. Care to
describe how this new Super OS is going to work.

Quote:> Otherwise, I believe Linux is just way too complex and screwed
>up for the desktop, a lot worse than Windows 95 has ever been.
>Simplicity is the key, and Linux fails badly on that point too.

Simplicity fails too. Some tasks aare complicated simply because they are
complicated. And it makes the problem worse when they are oversimplified,
when actions that some segment of the user community would like to control
can't because it has been automated under the guise of simplicity and
easy of use.

Now here's my short list:

1) The biggest gap is capable, compatible, and wide-ranging applications. The
fastest way to stifle ease of use is to simply not be able to do the tasks
required due to a lack of applications.

2) Change the mindset of two things that go hand in hand: namely that average
users are stupid and static, and that one simple paradigm can solce all users
needs. Both are incorrect. Interfaces need to be built in a layered fashion
where the user can pick a set of layers that work comfortably for them as they
transision from novice (not stupid) users into a range of novice, intermediate,
and expert users on various applications and administrative tools. The damage
occurs when the interface is static and users outgrow them. It causes big
frustration (read "counter to ease of use") because the oversimplifying
interfaces get in the way of the not novice tasks at hand. Layers should
go something along these lines:

- AutoInstalling/AutoDetecting/AutoConfiguring/Help
- Web/GUI based interfaces
- command line tools
- text configuration files
- Integrated scripting language
- Application Source code
- Configurable kernel/driver interfaces
- Kernel Source

With the top being for the rank novice and the lowest layers being for the
gurus. In short present the simplest interface to the user but always give
them the ability to strip that interface away and expose the interface
that lays underneath.

Users will then naturally gravitate to the appropriate level, descending only
as far down as necessary to accomplsh the task at their current level of
expertise. As they become more knowledgeable and their requirements grow and
deepen, they can then proceed further downwards to reach the level necessary
to accomplish the task.

And honestly you can see that now. Many Linux systems are self installing and
autodetecting. They use LinuxConf and other web based tools for administration.
Many are scriptable, and have directly editable text config files. Most are
open source at the application and kernel level.

All that's left is to create consistency and to fill in applications in areas
that aren't currently covered.

3) Get users to understand that they can't have it both ways. That either
they embrace the complexity and gain the required understanding, or to stand
back and let someone who does do the job. One cannot be a brain surgeon
by reading "Brain Surgery for Dummies". It takes time, study, practice, and
effort. And if a user isn't willing to devote what is required, then they
need to turn over the complex admin reins to someone else.

That's all for now.

BAJ

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Jehs » Sun, 31 Dec 1899 09:00:00


Very well stated, byron. *pat on the back*


Quote:> Now here's my short list:

--

ICQ 1900670 - 350467 GT Station - 6-0985 - HEF 214
 
 
 

General Public-Friendly Linux -- Impossible?

Post by Jim Ros » Sun, 31 Dec 1899 09:00:00




Quote:> As long as there's "file system", "processes", "swap
> partitions", "window managers", "command shells",.. is desktop Linux
> just a fantasy? In my opinion, someone should take the useful parts of
> the kernel source and drivers and develop a completely new non-Unix
> based open source OS that could maybe become some competition to
> Windows. Otherwise, I believe Linux is just way too complex and screwed
> up for the desktop, a lot worse than Windows 95 has ever been.
> Simplicity is the key, and Linux fails badly on that point too.

> Sent via Deja.com http://www.deja.com/
> Before you buy.

These things are of course in windows too.
Things would be very bad without them.

I would mind if when the user logins in he only sees his home folder and
maybe has a command to show commands available to him.
Basically hide the rest of the Unix structure.  The average user could be
independent from it I think.

Autodetection should improve to where Win95 was.  I think
RedHat/Corel/Caldera is 75-85%

I don't see the benefit of non-Unix.  I hope you explain that.  The
structure should be transparent anyway, so why does it matter?
Basically a minimum install would improve footprint of Linux and speed of
install.  (Note: to some people this is very important)

And Linux could use binary compatibility with other distributions, maybe
other Unix OSes.  Not a deal braker though.
And finally need Quicken or Gnucash that can do the same stuff, need more
games too.  And more good web browsers

I see Linux being a 1.5 to 2 years away from being a desktop force.
For server it's already good.

I already see Linux as being a better desktop than OS/2 and a better server
than Netware (Linux makes sense to me)
And the fact that Linux gets judged against the perceived number one OS
saids something.  (Linux is considered number 2 by many)
Jim

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Darren Winsp » Sun, 31 Dec 1899 09:00:00



Quote:> As long as there's "file system", "processes", "swap
> partitions", "window managers", "command shells",.. is desktop Linux
> just a fantasy?

Nope.  These all exist in Windows (Well, swap file instead of
partition).  The key is to find a way to hide them from those who
don't want to deal with them.

Quote:> In my opinion, someone should take the useful parts of
> the kernel source and drivers and develop a completely new non-Unix
> based open source OS that could maybe become some competition to
> Windows.

That would be very difficult.  One of the main complaints about Linux
is the lack of applications.  Now, imagine how hard it would be to
persuade developers to support a completely new open source OS.

Quote:> Otherwise, I believe Linux is just way too complex and screwed
> up for the desktop, a lot worse than Windows 95 has ever been.
> Simplicity is the key, and Linux fails badly on that point too.

With each generation of distros, Linux becomes easier to setup, learn
and possibly use.  Corel has shown it can be made easy, it's just a
matter of getting it perfect (1.0 has too many problems).

--
Darren Winsper - ICQ 8899775
Stellar Legacy project member - http://www.stellarlegacy.tsx.org

"Microsoft stated that they had very little Monopoly power," stated
Jackson in his findings of fact, "However, upon closer investigation,
we found that not only did they own all the properties on two entire
sides of the board, they also had three houses on Boardwalk and Park Place!"
        --http://www.segfault.org/story.phtml?mode=2&id=3825b6d8-019cd640

 
 
 

General Public-Friendly Linux -- Impossible?

Post by david parso » Sun, 31 Dec 1899 09:00:00




Quote:>As long as there's "file system", "processes", "swap
>partitions", "window managers", "command shells",.. is desktop Linux
>just a fantasy?

    Since you've just described features common to every useful
    operating system that exists, the only fantasy that I can
    see is your attempts to go fishing here.

                  ____
    david parsons \bi/ rating: 0 on a 1-10 point scale.
                   \/

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Matthias Wark » Sun, 31 Dec 1899 09:00:00


It was the 15 Dec 1999 11:02:00 -0500...



> >As long as there's "file system", "processes", "swap
> >partitions", "window managers", "command shells",.. is desktop Linux
> >just a fantasy?

> Well since the two don't have really anything to do with one another I
> guess the question is moot.

> Consider this: all versions of windows have filesystems, processes, swap,
> window managers, and command shells. Every one. Yet it's the most prevalent
> desktop today.

> > In my opinion, someone should take the useful parts of
> >the kernel source and drivers and develop a completely new non-Unix
> >based open source OS that could maybe become some competition to
> >Windows.

> I'd really like to see an OS without the list you mentioned above. Care to
> describe how this new Super OS is going to work.

Actually, it's quite possible to get completely rid of what Cuor di
Mela enumerated as being oh so horribly complicated <whine>.

One could have a desktop environment which manages typed (i.e.
MIME-typed) documents in a database-like repository, completely
abandoning the notion of files and directories; then, there would be
objects implementing various actions such as modification, interactive
editing or filtering of documents; etc. etc. etc.

What Cuor fails to notice or intentionally ignores is that such a
system would at its core of course be based on partitions, file
systems, processes, a window manager and such.

Cuor's whining is comparable to the whining we've heard about
cross-platform toolkits vs. Java. It's a complete mystery to me why
some people thinks that the quality of something is enhanced by
welding it to a huge, trademarked lump of software, labeling it with
different terminology or completely removing the labels, and then
claiming that the parts have been "integrated".

mawa
--
Tcl ist nicht als vollst?ndige Skript- oder Entwicklungssprache
entworfen worden (obwohl [dies] Leute nicht davon abgehalten hat,
50.000-Zeilen-Skripte zur Steuerung von Bohrinseln zu programmieren).
          -- Sriram Srinivasan, _Fortgeschrittene_Perl-Programmierung_

 
 
 

General Public-Friendly Linux -- Impossible?

Post by j.. » Sun, 31 Dec 1899 09:00:00



Quote:>As long as there's "file system", "processes", "swap

        As long as there is multitasking there will be 'processes'.
        Even Win9x that 'bastion of usability' has them. Sometimes,
        you even have to reference them.

Quote:>partitions", "window managers", "command shells",.. is desktop Linux
>just a fantasy? In my opinion, someone should take the useful parts of

        All you're describing is applications software. One does not
        have to throw out all the Unix in between the end user app
        and the kernel in order to deliver better modules for what
        sits in between.

        Mind you, all of the things you seem to be deriding Unix
        for having are also present under Windows. They're just
        hidden for the most part. Besides, Windows is still pretty
        much just what you criticise Linux for being: a thin veneer
        over an ancient command shell.

Quote:>the kernel source and drivers and develop a completely new non-Unix
>based open source OS that could maybe become some competition to
>Windows. Otherwise, I believe Linux is just way too complex and screwed
>up for the desktop, a lot worse than Windows 95 has ever been.
>Simplicity is the key, and Linux fails badly on that point too.

>Sent via Deja.com http://www.deja.com/
>Before you buy.

--

   One of the great lies of our age is the myth that      
   Microsoft has somehow managed to turn these random      
   collections of spare parts known as PC clones into            |||
   the equivalent of a Macintosh.                               / | \

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Anthony D. Tribell » Sun, 31 Dec 1899 09:00:00



Quote:> As long as there's "file system", "processes", "swap
> partitions", "window managers", "command shells",.. is desktop Linux
> just a fantasy? In my opinion, someone should take the useful parts of
> the kernel source and drivers and develop a completely new non-Unix
> based open source OS that could maybe become some competition to
> Windows. Otherwise, I believe Linux is just way too complex and screwed
> up for the desktop, a lot worse than Windows 95 has ever been.
> Simplicity is the key, and Linux fails badly on that point too.

There's a hint of a point in the above, I think you could phrase it
better. Are you trying to say something like you would like the stability
and capabilities of UNIX without the clunky user interface? If so, it may
be worth watching Apple's MacOS X. It has a BSD UNIX core with Apple's
user interface on top. UNIX purists may not like this but I think it has a
good chance at becoming a major UNIX variant, just a stealthier one. The
average end user may never know they have UNIX, I suspect this is the sort
of thing you were hoping for. More advanced users can fire up the shell
and enjoy the BSD command line to their hearts content. 3rd parties will
provide X. Also, keep in mind that most of what people refer to as Linux
and Linux apps are open source software that run under many UNIX variants
including the BSD Apple is using.

Of course you can say the same thing for WinNT to a degree, that many open
source apps/tools Linux advocates talk about also run under WinNT.

Tony
------------------
Tony Tribelli

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Leslie Mikese » Sun, 31 Dec 1899 09:00:00




Quote:>As long as there's "file system", "processes", "swap
>partitions", "window managers", "command shells",.. is desktop Linux
>just a fantasy? In my opinion, someone should take the useful parts of
>the kernel source and drivers and develop a completely new non-Unix
>based open source OS that could maybe become some competition to
>Windows. Otherwise, I believe Linux is just way too complex and screwed
>up for the desktop, a lot worse than Windows 95 has ever been.
>Simplicity is the key, and Linux fails badly on that point too.

Take Mandrake Linux, tell it to install as a workstation on hardware
that it autodetects, and you come up running without needing
any prior knowledge of buzzwords.

  Les Mikesell

 
 
 

General Public-Friendly Linux -- Impossible?

Post by j.. » Sun, 31 Dec 1899 09:00:00





>>As long as there's "file system", "processes", "swap
>>partitions", "window managers", "command shells",.. is desktop Linux
>>just a fantasy? In my opinion, someone should take the useful parts of
>>the kernel source and drivers and develop a completely new non-Unix
>>based open source OS that could maybe become some competition to
>>Windows. Otherwise, I believe Linux is just way too complex and screwed
>>up for the desktop, a lot worse than Windows 95 has ever been.
>>Simplicity is the key, and Linux fails badly on that point too.

>Take Mandrake Linux, tell it to install as a workstation on hardware
>that it autodetects, and you come up running without needing
>any prior knowledge of buzzwords.

        This is true of any modern ease of use distro.
        (Bughat, Suse, OpenLinux, even Corel Linux)

        Although, this is an 'ordeal' that most desktop
        users have chosen not to bother with anyways.

--

   One of the great lies of our age is the myth that      
   Microsoft has somehow managed to turn these random      
   collections of spare parts known as PC clones into            |||
   the equivalent of a Macintosh.                               / | \

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Peter Nelso » Sun, 31 Dec 1899 09:00:00



>As long as there's "file system", "processes", "swap
>partitions", "window managers", "command shells",.. is desktop Linux
>just a fantasy? In my opinion, someone should take the useful parts of
>the kernel source and drivers and develop a completely new non-Unix
>based open source OS that could maybe become some competition to
>Windows. Otherwise, I believe Linux is just way too complex and screwed
>up for the desktop, a lot worse than Windows 95 has ever been.
>Simplicity is the key, and Linux fails badly on that point too.

Linux is still at the stage where you have to either be a geek
or have one handy to get it to do anything useful.   Linus, at
his keynote at Fall COMDEX said that users should *not* have
to learn Linux commands.   So hopefully this attitude will
trickle down to the troops on the ground.

---peter

 
 
 

General Public-Friendly Linux -- Impossible?

Post by Peter Nelso » Sun, 31 Dec 1899 09:00:00





>>As long as there's "file system", "processes", "swap
>>partitions", "window managers", "command shells",.. is desktop Linux
>>just a fantasy?

>Well since the two don't have really anything to do with one another I
>guess the question is moot.

>Consider this: all versions of windows have filesystems,
> processes, swap, window managers, and command shells.
> Every one. Yet it's the most prevalent desktop today.

Because they don't make desktop end users deal
with those things.  Or at least they *try* not to.
Linux is an OS by, of, and *FOR* geeks.    And there
are many people in the Linux community who are
perfectly happy about this.

---peter

 
 
 

General Public-Friendly Linux -- Impossible?

Post by rob » Sun, 31 Dec 1899 09:00:00


What a great idea.  Maybe next you could invent a really easy to
drive & economical car by removing the engine and steering wheel.

> As long as there's "file system", "processes", "swap
> partitions", "window managers", "command shells",.. is desktop Linux
> just a fantasy? In my opinion, someone should take the useful parts of
> the kernel source and drivers and develop a completely new non-Unix
> based open source OS that could maybe become some competition to
> Windows. Otherwise, I believe Linux is just way too complex and screwed
> up for the desktop, a lot worse than Windows 95 has ever been.
> Simplicity is the key, and Linux fails badly on that point too.

> Sent via Deja.com http://www.deja.com/
> Before you buy.

 
 
 

1. Q: GNU Library General Public License

This may not be the right newsgroup for this but I have some questions
regarding GNU's licenses. BTW I've just read both the GPL and the LGPL
twice.

I releaize that if a library is under GPL and you link with it then
your program is now a derived work and must now be distrubuted under
GPL.

However if a library is under LGPL, I was under the impression from
reading posts on comp.unix.* and comp.os.linux.* that if you link
with it you must distrubute your program in a form that can be relinked
with newer versions of the library, namely a big object file for static
librarys and for dynamic librarys a lone executable was fine.

But now after reading the LGPL it seems that it isn't possible for
a program to be distrubuted as a binary only under whatever license the
owner wants, if it was linked in anyway to a library that is under
LGPL. Is my take on this correct ? BTW the words static and dynamic
are not used at all in the GNU LGPL.

 "   5. A program that contains no derivative of any portion of the
  Library, but is designed to work with the Library by being compiled or
  linked with it, is called a "work that uses the Library".  Such a
  work, in isolation, is not a derivative work of the Library, and
  therefore falls outside the scope of this License.

    However, linking a "work that uses the Library" with the Library
  creates an executable that is a derivative of the Library (because it
  contains portions of the Library), rather than a "work that uses the
  library".  The executable is therefore covered by this License.
  Section 6 states terms for distribution of such executables. "

BTW Section 6 says you must distribute source or give people a way to
get it.

The first paragraph above seem to say that a program that needs to be
linked (but isn't acctually linked yet) to a library under LGPL is a
work that uses the library and isn't forced under any restrictions, but
the second paragraph seems to say that the act of actually linking
suddenly introduces a bunch of GNU's restrictions to distrubuting a
program. So for example it would be illegal for a company to ditrubute
a comercial Linux application as a binary only if it is linked to libc
(which is under LGPL).

Is this the correct interpretation ?

 "  When a "work that uses the Library" uses material from a header file
  that is part of the Library, the object code for the work may be a
  derivative work of the Library even though the source code is not. "

Does this mean by having #include<somelibheader.h> in the source to
a program makes it a derived work ?

It seems that putting all the librarys on these free unixes under LGPL
hinders comercial software being released on these platforms.

I'm starting to like GNU's licenses less and less.
Please correct me as I hope I am wrong.

Jens Vaasjo.

P.S. If I am correct what is the solution to this ? Not link against
libc, then what ? Not release comercial software on these platforms ?

2. Bill Gates and Adolf Hitler

3. Which GNU General Public License Editor are the best for C programming?

4. Instructions vs Operations

5. USR isa 33600 hardware with linux 7.0 posibile or not?

6. Inform the general public !

7. Help with CPP woes

8. FreeBSD needs to cater to the general public.

9. Which GNU General Public License Editor are the best for C programming

10. GNU General Public License

11. Q: Linux as Firewall w/ 2 Public IP => 1 MASQ, 1 PUBLIC possible?

12. Public Key Crypto in Public Domain?