no topic :-)
I agree with you. I hope that we will not have only one unix desktop environment in the future, which is maybe controlled by some big companies.

And this is another reason why I use Linux Mandrake...
There shouldn't be a "war" at all over this. In fact, there should be quite a few people out there in both sides of this so-called war that should be deeply ashamed of themselves- beceause they know better.

2 dominant, incompatable desktops is imposible.
As we know, Windows is currently the dominant desktop. 90% (or so) of all programmers are doing Windows only programming. That is because it is much easier, has a faster developmenet time, and cheeper to do only one version of a program. Since GNOME and KDE are incompatable, when Linux becomes standard, there will be one dominant desktop environment, and the other will either die, or linger around like amega haveing support from only a small group of people. One of the two Linux desktops will eventually gain Defaulthood unless GNOME and KDE become component compatable. IE, you can embed gnumeric in kword. This would only happen if GNOME and KDE could agree on a common component model. I've studied both Bonobo and Kparts, and have come to the conclusion that Bonobo is the way to go. Kparts is nonstandard, it is not langauge independent, it is QT dependent, and doesnt support distributed computing. Bonbobo on the other hand uses CORBA under the hood, so it uses standard technology, it is language independent, it is not any toolkit dependent, and supports distributed computing. If KDE picked up Bonobo for its component model, GNOME and KDE would no longer be competing, but complementery. This is the only way I can see getting around one of GNOME or KDE eventually dieing off.

Two or even more desktops are indeed possible
Since all you to run KDE or gnome applications is the libraries one can use gnome and use a KDE application without problems. Off course a KDE would have access to more features that come from the integration with the desktop environment. But in time (read when this so called "war" is over) the two enviroments will have more points in common.

Views from a newbie
Ive just discovered the joys of linux, one of the things that makes it so special is the people involved and the diversity it allows, why would we want to go the Microsoft way?

The diversity of the different desktops is a unique feature that makes linux what it is and enable it to be enjoyed by the a wider range of people than any other OS, long may it stay that way :)

Desktop GUIs vs. GUI guts
Part of the problem I see here is that the desktops and the application frameworks are considered together rather than separately. I think it would be a helpful to make a distinction between the desktop shell, which would be the window manager, file manager, panels, and menus, and the application framework, which would be the component model and low-level or middle-level services like printing services and sound services. The reason for the distinction is that it is, for the most part, no skin off application developers' noses if different users use different desktop shells, but problematical for them to support different application frameworks.

Think about it. How many apps can you think of that actually depend on having a particular desktop running? Currently, I can run KDE and GNOME apps under FVWM or Blackbox or Enlightenment, or whatnot. All that matters is that I have the appropriate libraries to run the KDE and GNOME apps.

Having different component models, though, is a different story. The whole point of component models, AFAIK, is that apps can be built from smaller, specialized building blocks. Building apps from components is analogous to building things out of Lego, and the bumps on each Lego that cause the Legos to fit together are analogous to the component model itself. Having two different component models is like having two different brands of building blocks that can't be put together. Or to put it another way, having two different component models is like having two different standards for plumbing fixtures or bolts and screws. The component model is a layer below the desktop itself, and it is a layer that benefits from uniformity.

In short, I would contend that the issue really shouldn't be GNOME vs. KDE, but Bonobo vs. KParts and DCOP.

re: What About
I have always stayed out of this flame war because to me Linux is about choice. I have to agree with jmd though that GNUStep is often over looked.

I recently checked out their snapshot and I think it is coming along quite nicely. Once it is finished, we will see an inflex of apps not only because porting between existing NeXT/OpenStep and future Mc OS 10 apps is just a little more than a recompile, but also because of the ease of development. How about a feature rich Wordproccessor in 2 minutes?

KDE/GNOME position
Personally, I dont care if either one of these dies, becomes the standard, or replaces Windows 2000 in Bill's home. What I do care about that I will never use either, and since there are no useful KDE or GNOME apps, I guess there is no reason for me to ever use either. If an app is useful enough chances are it doesn't need this crap (Gimp, etc). With all the stupidity hovering around Linux lately, it's amazing it haven't turned into the next Windows yet. Recently announced "red carpet" for Helix Gnome (whatever that is) just looks like a GTK app which took about 5 minutes to write and a whole lot of anti-aliased icons which took the rest of the day to make. Big fsckin deal. Icons isnt what Linux is all about anyway.

I had to mention this one
Why use KDE?
* KRouletteSim 0.3.2
This means there have been at least 3 versions of this utterly useless program. HEH. Too many colleges these days teach C++. It needs to die, together with KDE.

Choice in component models needed too
Most of the responses to this piece have been very positive, agreeing with the author that choice is desirable and that we should keep it that way.

This response is mainly directed at J.J. Ramsey who maintains that choice is fine for desktops but that we need uniformity in underlying component models so that applications created with any toolkit can be embedded in any other. Further, Gnome's bonobo using CORBA is the way to go because CORBA is an "industry standard" which is neutral regarding implementation language.

I am not an expert on component models but I have been using Kde 2 beta 4 for almost a week now. Kde 2 uses components for *everything* now and it works beautifully. That's good enough for me. Gnome's bonobo is a theory which remains to be proven and to date is only used in apps like Nautilus and Evolution which are in pre-alpha stage. I'm sure that bonobo will get there and will actually be used in a current Gnome release sometime and that both ORBit and bonobo are well thought out layers. But I'm also pretty sure that apps which use them will be Gnome apps using gtk only. Applying this technology across networks like the internet using components from a variety of of architectures (the only real purpose of CORBA) is very unlikely.

We need choice in component models and object oriented technologies as well as in desktops. Component models evolve like any other technology, and CORBA is not necessarily the final word. Far from it. I'd hate to think that we should be trying to emulate Microsoft's leavings, COM, or OS2's failed SOM, a precursor or CORBA. MS is already abandoning COM for .NET. Likewise CORBA and bonobo and Kparts and dcop can evolve, and will evolve, if we have an environment which allows that. This requires choice and diversity, not uniformity. MS may have 90% of the desktop market right now but not too long ago it didn't. I don't think that Microsoft's "superior" component model is what caused MS to displace Amiga, OS2, Mac and commercial unix on desktops in the early 90's. We all know better about MS.

Actually Kde does use CORBA, or a subset of it, in its upcoming 2.0 product. There is something called mcop and aRts for multimedia streaming and synthesis, for which dcop and kparts alone is not well suited. And it doesn't use Qt or require C++. So, Kde programmers can be very flexible and adaptable in applying the needed technology to the situation or even sythesizing new ones. They are concerned with practicality, not buzzwords.

There are a number of proven ways to reuse code besides components, which will not bring the users which Linux is looking for. Applications will bring them over. What about WINE and WINElib? What about unix's superior and diverse scripting languages and interpreted OO languages? Kde certainly has the applications users need in its upcoming 2.0 release and in its current beta. This will mean more users for Gnome as well, and I don't think these users will be very concerned about the inability to embed a Gnumeric cell in Kword.

Total misunderstading
The author mentions that some people like GNOME or KDE while others like WindowMaker or FVWM.

Just so that everyone understands this, GNOME and KDE are just desktop environments. Just as Microsoft's Windows user-interface did not change when they introduced COM, GNOME and KDE are independant of what set of bells-and-whistles you put around your windows. GNOME's official default window manager is Sawfish, but among the MANY GNOME-compliant window managers are Englightenment, Sawfish, FVWM and many more.

The key is that GNOME (and I will cite GNOME here because I know it better and can speak to it more directly) is a framework for developing quality desktop applications that work together. GNOME provides:

A set of standards for how various classes of applications (from end-user programs to window managers) will behave.
A set of libraries that provide everything from basic C utilities like linked list handling and threading (glib) to widget sets (gtk+) to Web toolkits (libghttp, libxml, etc) to run-time-loadable user interfaces (libglade).
A communications infrastructure for applications to share data on the same machine or across a network (ORBiT).
Rules for building software (GNU autoconf and automake) and what platform features must be supported (internationalization).
A framework for persistant application data and session management.

These things are totally independant of which window manager you use (in fact, they're independant of what windowing system you're using).

Kde team is jelous about big companies support to Gnome
It seems that this so called "Kde vs Gnome" challenge is just a conflict started by some on-line magazines and info sites about free software. In my opinion, this sites are tryng to show that the Kde team is jellous about the major companies support to Gnome.
It is a sad thing, 'cause there has been so little time since kde and gnome agreed to cooperate. The competition between thoose too major desktop enviroments is a good thing, but this war is just pathetic.
For the interestes, I run icewm

Acceptance must come before domination

a
I'm sorry, I usually stay out of these arguments, and
perhaps I should this one, but:

I've been using Unix for fifteen years. Not as long
as some people, but long enough that I have SunOS (not
Solaris, SunOS) 1.1 manuals in my office and have done
serious Unix development on a hard-wired VT220 terminal
(actually, I go back to punch cards and 360 assembler,
but we won't go into that).

On the Sun platform, I've suffered through SunWindows,
SunView, NeWS, OpenWindows, raw mwm and CDE. Generally,
post 1989 or so, I did this by ignoring them and using
something in the twm family -- twm, tvtwm, ctwm. Sometimes
I fooled with things like generic olwm, olvwm and gwm.
But still, I had to use the various Sun products for
some things.

I've also been using Linux since about the second SLS
release or so. At various times, I've used most members of
the feeble family (fvwm and myriad derivatives thereof),
and have tried piles of the other stuff that people have
put together, including icewm, afterstep, windowmaker, e,
sawmill/sawfish, blackbox, and wm2. I've also used KDE
since about three betas prior to KDE 1.0, and Gnome since
those early releases for RH 5.2.

Thus, I don't think that anyone can say that I have any
blinders on. I am utterly ecumenic when it comes to this
stuff. I really did try, with every one of those
things, to make them work for me. Not just for fifteen
minutes or so, but for at least a couple of weeks, as
long as they worked at all. Usually I'd try to work with
these things for a month or so, tweaking this and that,
applying patches, and so on. And if one didn't work I'd
usually try it again a few months later.

After all that, I really have to say that the only two
windowing environments that ever really worked for me --
worked in the sense that, long term, out of the box,
they did their job and stayed out of my way so that I
could do *my* job (which most of the time doesn't involve
programming, BTW), were twm and KDE.

Twm, for its time, kicked ass. It is probably hard for
many readers of this to imagine this, but, going by what
preceded it, it was utterly awesome. I still haven't
seen anything else that manages icons as effectively --
twm can make the icons little, unadorned rectangles like
you get on many taskbars these days, only you could group
them into little clusters, organized by application --
xterm icons go here, emacs icons go there, etc.

Today, on any Linux machine where I have to get work done,
(mostly reading, writing and some EDA stuff -- I use
Cadsoft's Eagle PCB design program under Linux a lot)
I run KDE. This isn't a religious thing. I do it because
KDE works. It's rock stable and it stays out of my way.
It never assaults me with it's own brilliance. Generally,
I can tweak it to be well-behaved in about two minutes
out of the box.

On the other hand, on other machines where I'm just
exploring, I try all those other things. I have repeatedly
tried to get Gnome to work for me. I tried the early
alphas, I tried the 1.0 release and the releases that have
come with all the RedHat releases since 6.0. I've even
loaded up the recent Helixcode version on one machine.
What I've noticed is that, while Gnome keeps getting
prettier, fancier and more feature-laden, I just can't
get the damned thing to do what I want. Why, for example,
when I click on an icon in the panel taskbar, does it
take over a second for the window to be exposed? (we're
talking dual PII-333 w/256MB SDRAM and 7200 rpm UltraSCSI
disk here, so I sure as hell don't believe that it's my
machine...) Once I get more than one or two dozen windows
in Gnome, I go through hell trying to find things. Why,
when an application throws up a child dialog window, does
that window not receive the focus and why isn't clicking in
the text entry box sufficient to give focus to the dialog --
why do I have to click on the title bar? Why does the
panel still get in the way of application windows and
vice versa?

Yes, I'm sure all these have good answers, and that if I
just spent a little more time reading the documentation I
could make many of these problems go away. But why should I
have to? "It works for me", "it shouldn't do that", "that's because you're using xxx as your window manager, try yyy", "you
just need to understand how it works", or -- my favorite --
"why don't you file bug reports" (I'm too busy trying to
catch up on the real work that I couldn't get done using
Gnome, that's why), are all totally irrelevant. The point
is that when I install and run Gnome, it gets in my way, while when I install
run KDE, it doesn't. I will accept "it's all a matter of
personal preference", "Gnome probably just isn't suited
to the way that you work", or possibly even -- if I'm
feeling charitable -- "give it time,
it's getting better." But I think that one can only
say this but so many times to but so many people before
one really should be wondering if there aren't some actual
shortcomings in this thing.

I think that this is what people keep missing in all
the arguments over KDE vs Gnome (and I've read a lot of
them): Gnome may have some real technical superiority
in a number of dimensions, but none of it will mean a
damned thing unless it works, out of the box, such a way
that the user doesn't particularly notice that he or she
has something called a window manager or a desktop
environment. For normal, non-development customers, this is going to be vastly more important than
the nature of the Qt license. Let's make a list of all
the cases in the history of computers where
technical superiority won over quality, usability and just
plain being there -- and ready -- first. "Now and good enough"
wins just about every time -- look at Microsoft Windows
if you don't believe me.

Too often I sense the attitude from the Gnome community that
the window manager -- the single most important part of
the user interface -- is just, if you will excuse the pun,
"window dressing". The lack of dependance of Gnome on a
single wm is both a blessing and a curse. Sure, if you don't
like the one it comes with, you can patch in a different
one. But what if they all are buggy? What if their
Gnome interfaces prove to be incomplete and inconsistant?
What if the conflict between wm-based menus and Gnome-based
menus cause confusion for people who don't have the slightest
idea what is going on? I fully understand why some people
want and in many cases need this kind of flexibility, but
the vast majority of computer users are going to be annoyed
with the simple fact that they need to make any choice at all. I guess that KDE has some wm-independance, but that
hasn't stopped them from developing a fast, solid window
manager that is so well-integrated that you can't even tell
that where the WM leves off and the rest of KDE starts.
Puke if you want, but believe me, it's what most people
are going to want.

Like it or not, things have reached the point that people
besides developers are using Linux (and Unix) in large
numbers. Most of these people have Things To Do. If the OS
or the windowing environment get in the way of them doing
their jobs, they are going to use another OS or another
windowing system. Corba vs KParts is utterly besides
the point, at least until people's jobs actually *depend*
on Corba or KParts in the way that their jobs now depend
on Microsoft's object model. Acceptance must come before
domination.

But even when it comes to developers, and especially the
developers' bosses, before they start working on a Linux
version of their product they are going to be looking at
a number of things beyond the object model, including:

What desktop environment do their customers use?
What solid, relatively stable, toolsets are out there
that support the popular desktop environments?
How can one maximize potential market without
having to develop multiple versions of GUI code?

I'm afraid that, besides the object model question,
Gnome is going to continue to get poor marks in this area
for some time. For example, I make extensive use, as I
mentioned, of an EDA project called Eagle, from Cadsoft
(www.cadsoft.de). Once upon a time, Eagle was
available for DOS, OS/2 and Windows. Then it was available
only for DOS and Windows. But Cadsoft saw that there was
a market waiting to be tapped in Linux, and the portable
toolset they used to be able to develop simultaneously
for the three operating systems was available in a Linux
version. So they did the port, and eventually Linux became
their primary development platform. Unfortunately, when
they started working on a major upgrade of their product,
their toolset vendor decided to drop the product. At that
point, they were kind of in a bind because they absolutely
had to have a toolset that worked both in Linux and in
Windows (DOS and OS/2 were already off the list by this
point). I have no idea to what extent they may ever had
considered GTK+ and Gnome, but I simply can't imagine
that they thought about it for very long -- GTK+ is not
known for its excellent Microsoft Windows support. Qt,
however, is.

So they had to do pretty much a complete rewrite of all
of their GUI code, in a portable manner, while adding a
number of new features. A preview release of the result is
now available for download, and I have to say that it is
already quite stable and usable, and it appears that it
will be a huge improvement over the previous version. I
don't know that they have had any particular regrets
over using Qt, they certainly haven't voiced any in their
support forums.

In order to succeed in the long run, the developers of
Gnome and GTK+ have a huge task in front of them. They
have to convince Linux and Unix users that the desktop
environment will work well and unobtrusively, and they
have to convince the commercial applications developers
that they can switch to Gnome and GTK+ without either (a)
maintaining two completely separate GUI source trees,
or (b) completely abandoning a huge potential Windows
customer base. And they have to do this before
KDE and Qt get there.

I wish the Gnome developers all the luck in the world.
I take no sides in this except as a user. My only reason
for writing this is to give one user's point of view that
may or may not be useful. I will continue
to try out new Gnome releases as time goes on, and I do
hope they get better. Competition is healthy. Lack of Competition
is a major reason why Microsoft's products
are so bad, and the existance of competition is a major
reason why KDE and Gnome are as good as they are.

I think that Linus has shown by a powerful example that
if you keep your nose to the grindstone, and just keep
working on making your stuff better, sooner or later it
will become accepted. Linux didn't get where it is today
becuse Linus tried to make a better operating system than
Microsoft Windows. Linux got where it is because Linus
tried to make the best operating system he knew how to
make.

--Bob Drzyzgula

Cease fire? Never!
I really can't understand anyone calling for a "cease fire" in the D.E. wars. Why? Because the ammunition in this battle is innovation - and not in Microsoft's distorted sense of the word. The two desktops are competing with each other over features. If there was one environment, this sense of competition wouldn't exist - and if we tried to value each environment for what it is now, without continuing the active exploration of new concepts in UI design, the same would certainly happen.

Of course, people have complained plenty about the duplication of efforts involved in competition. GNOME would never use some of KDE's component code, or vice versa, although it would be perfectly possible to do so (the toolkit issue isn't a problem). But since each side is developing its own tools, they are creating new technology that wouldn't be possible if only one desktop environment existed. May the best GUI win :) .

Make love, not war
Sure, competition is what good software is all about. Sure, it's great that users have a choice in what desktop environment (DE) they want to use. But look at things from a developer's perspective. As long as a developer has to choose for which DE his/her application is going to be written, things go wrong. The developer then forces the application user to use a specific DE. What if a user needs to use a GTK based application, and uses KDE, and doesn't want any GTK libraries on his system ? Or vice-versa ? This problem should be fixed, and should be fixed fast, or Linux will always stay behind with developer's support. Integration is the keyword. Gnome and KDE developers should co-operate, not fight eachother. 2 DE's to choose from is fine, but give application developers a strong tool to develop applications that run flawlessly on KDE & GNOME. Make sure the 2 DE's are able to interact, with drag and drop, mime-stuff, etc. KDE has done some good stuff in this area by being able to use GTK themes in KDE2 btw.

Reminder: The REAL reason behind the GNOME vs. KDE war
Three major points:

1. KDE is older and more mature than GNOME.

2. QT is available on MS-Windows.

3. QT's license is NOT free!

As a Unix AND Linux developer having a desktop that can be ported to ANY machine I use with a truely FREE license is the deciding factor. Therefore, in the spirit of FREE software, GNOME MUST WIN. The issue of CORBA compliance and the architecture of the GNOME component architecture is also important since it facilitates a standards based way of interfacing with non-GNOME systems. Incidentally diversity of options is very valuable in terms of ongoing innovations, but interactivity and integration of a common world view for a given environment. GNU-Step another truely FREE environment which deserves mention. Let us hope that it will work well with others better than KDE does. I truely wish that the KDE developers would port their applications to GNOME so that the GREAT work they have done could be leveraged in a FREE environment!

Re: Choice in component models needed too

This response is mainly directed at J.J. Ramsey who maintains that choice is fine for desktops but that we need uniformity in underlying component models so that applications created with any toolkit can be embedded in any other. Further, Gnome's bonobo using CORBA is the way to go because CORBA is an "industry standard" which is neutral regarding implementation language.

I never said that GNOME's bonobo was the way to go. While I suspect that it may very well be the better component model, my point was simply that having two competing component models was not ultimately beneficial, since it would hamstring the goal of building up apps out of smaller, specialized components. (What good is having components if there is no common way of putting them all together?)

Re: An Observation
> Anyone who claims that neither KDE nor GNOME have any useful apps has most assuredly used neither.
I have, at least in theory (neither one was ever my primary desktop environment on any of my -personal- machines), and I have yet to see any useful apps. Please prove me wrong and provide a list of something I can use either out of GNOME or KDE that doesn't just suck ass completely. As far as I am concerned, if I see an app I truly need that requires GNOME or KDE I will just end up installing the base libs, but KDE or GNOME will never replace BlackBox as my WM.

Duplication of effort
If duplication of effort in any area can be kept down to KDE plus GNOME plus one-or-two others, then we are on to a good thing. The thing that must be paid attention to, is that it should be almost accepted that both projects approach things in their own way, but will happily assimilate (with credit) an invention of the other party once it has shown its usefulness (if not before).

KDE/GNOME bashing should be kept POSITIVE!

If, say, the GNOME project appear to be going down a route that the KDE people have explored, but have had problems with, and then rejected (e.g. CORBA messaging), there should -- in fact MUST -- be an open forum between the developers of the two projects. If (in this case) GNOME want to continue, they get to discuss potential problems, and have advance warnings, so that they can learn from 'their' mistakes before they make them.

p.s. don't trumpet 'duplication of effor' as a problem for the sake of it. With the duplication we see between KDE and GNOME, it should become apparent that the level of redundancy in production gives rise to a comparable level of robustness in the end products (since if GNOME figure out how to do something, and KDE have a hard time with it, they should group themselves around GNOME's solution, and vice versa).

But please -- if there's any arguments between the projects, leave those to the developers.

GNOME + KDE
What is it about people that they simply have to believe that the software they're using is the best. Neither KDE or GNOME are particularly wonderful GUIs, I find KDE more bearable than GNOME, once I've removed the ridiculous looking Windows style round radio buttons and checkboxes with ticks inside them. But I couldn't live without KDE and GNOME being installed, there are KDE applications that I love (because I know how to work with them mostly), and there are GNOME applications that I love (the GNOME panel CD applet, volume control, xmms control, and CPU usage applet absolutely walk all over their KDE counterparts) and of course there's the host of GTK applications that work so much better with the GNOME libs installed, like cheops and ethereal. My desktop experience wouldn't be the same without both GNOME and KDE installed, and anyone who excludes an application from their repertoire because it's written for a desktop they don't like is cheating themselves out of the chance to fairly and evenly appraise applications on their individual merits, and probably robbing themselves of a more enjoyable and productive experience with X.

In case you're interested, I prefer the look and feel of the KDE desktop, but I keep the GNOME panel (in auto-hide mode) at the left hand side of the screen as a place for my favorite GNOME applets to reside. However, I couldn't run the KDE panel in GNOME if I wanted to. Therefore, my unscientific conclusion is that GNOME apps run better without a GNOME WM, than KDE apps do without a KDE WM. I'm not sure what to make of that though. :)

Re: GNOME + KDE
I concur. There are certaian aspects of each environment that surpase the other. In my experience I have found that while I prefer to use KDE as the actual working environment, I often times turn to GNOME/Gtk for specific applications. The most common are GnoRPM, GIMP and Gkrellm. I love each of these programs and would be put out without them. What does concern me is the eventual posibility of one, or both, making the other(s) excluded from interoperability.

The words of second place
I may be out of line here, but after reading all
these posts something struck me. Linux users are
the reason that Linux will never be able to tackle
Windows. The hard and fast Linux user wears the
difficulty of the OS as a badge of honor. Someone
even said that icons aren't what Linux is about.
Well, sorry to break it to you, but if you ever want
my mom, or your aunt to use Linux, it's damn sure
gonna be about icons. The hobbyist or programmer
is willing to learn commandline codes and tinker
with things to make them work. The casual user
has no patience for this.
I agree that competition is good. At some point
though, if Linux is going to be a truly viable option
for the casual user, some things are going to have
to be standardized to make everything work much
smoother. If this means that KDE or Gnome is the
defacto Linux, so be it.
I know that a lot of you are thinking screw the
casual user, it will dumb the OS down like
Windows. You have to look at the big picture
though. If Linux can make the jump onto Ma and
Pa's PC a whole new world of options will open up
to all Linux users. The hobbyists will still be able to
get all manner of things to customize his own
Linux, but he'll also have a whole host of other
opportunities.
I'm not trying to attack Linux. I think it is a great OS
and I have a lot of fun with it. But someone needs
to ask the question, where is it heading?

Re: The words of second place
[...]

> But someone needs
> to ask the question, where is it
> heading?
>

ok. question asked.

Now what? :)

Re: 2 dominant, incompatable desktops is imposible.

> ... If KDE picked up
> Bonobo for its component model, GNOME
> and KDE would no longer be competing,
> but complementery. This is the only way
> I can see getting around one of GNOME or
> KDE eventually dieing off.
>

Well there both open source, why dose no one start the a project to gut kde programs like kword and make them use Bonobo or just add in a compatibility layer to the desktops. They dont have to use the same backend then. =)

Linux, Desktops, and Holy Wars
It's a good time to be alive, as always!

I remember my Amiga days fondly. We had choice... both a command-line interface AND a graphic user interface! The system was designed by enthusiasts, and it showed. It even started as a rebel system. The original investors thought they were fronting money for an arcade game machine platform, not a personal computer. He he! Little did they know!

But, alas, the Amiga system had it's problems. Although the Amiga was initially light-years ahead of other personal computers with regard to graphics, sound, and operating system, it suffered serious growing pains as time went on. The tight integration of the software to the hardware became difficult to overcome. It was a very proprietary system and eventually could not keep up with a somewhat open hardware platform (the PC). Granted IBM probably never realized how easy it would be to "open" the hardware end of their system. But the fact is that as soon as the BIOS was reverse-engineered, PC hardware was bound to become a true commodity market. A free market as it were, where hardware is cheap, plentiful, and diverse.

Apple, with their Macintosh systems, have been lucky to remain alive in the face of this competition. If it were not for huge leads in specialized markets, Macs might have suffered much the same fate. Desktop publishing may have been invented at Xerox PARC, but the Mac built the market -- and still dominates to this day.

Truly, you can now run a publishing house, music studio, or what have you without Macs. But there is still a history of market domination in certain areas that is helping to keep Apple afloat. For now. Commodore/Amiga should have been so lucky with the computer video market. But alas...

Apple have steadfastly stuck to the proprietary hardware model. They pull this off by building very high performace systems. They have to in order to compete against the entire PC clone market. How long they can stay ahead of the curve is anyone's guess.

But things change. And Apple, as reluctant as they have been, seem to grasp the notion that a a proprietary hardware platform married to a proprietary operating system can only survive, long term, in very specialized markets. And that's quite a risk. Computers, after all, are the most general purpouse instruments ever wrought by man. Recently Apple has made moves toward an open operating system, to the mutual benefit of themselves and OUR-selves, the Open Software community.

Evolution is the key word in all of this. We in the computer world are an excellent example of how easily people can become caught up in 'religious wars'. Often the squabbles we have with eachother are likened to just that. Now I don't want to start the 'other' kind of religious war here, but just for the sake of argument it could be said that evolution IS the hand of god, working in those mysterious ways.

Mankind has a long history of trying to reign in the force of evolution, but it is the supreme arrogance of our species that compels us to do so. We do it with plants. So we do it with animals, such as dogs and horses, and we do it with our own. But blue blood runs thin as water. (Hemophilia is a recessive genetic disease). Big Blue it seems had to learn a thing or two over the years. Hard lessons they surely did not wish to learn. And now that IBM is supporting Linux, maybe - just maybe - they have learned something.

Biology shows us that a healthy gene-pool is a diverse gene-pool. Much to the chagrin of the racists, I am sure. While we work hard at infecting plants, animals, and soon ourselves with human-krafted code (DNA), endangered species are showing us just how important bio-diversity really is. Species just cannot survive in nature without some diversity. In fact, whole ecosystems seem to depend on a good amount of diversity amongst their members.

For all the talk in the world about universal truths, there is one thing about them - they are universal. The problem seems to come when we try to pin them down. Dogma is eternally debatable but, in beautiful cosmic irony, Change appears to be a universal truth. And how can you pin down change?

So if change -- evolution I dare say, is a part of the truth in our known world, then it must be reasonable to expect such a pervasive force in the universe to have a bearing on our world of computers. Ours is but a microcosm - should it not reflect the macrocosm of which it is a part? It's only a theory, but it has a fractal sort of beauty to it, no?

Proprietary systems are like closed gene pools. Over time it becomes increasingly difficult for them to survive. The more we domesticate our pets, the more they need us in order to survive. Likewise, a proprietary system requires an ever increasing commitment on the part of it's patent and copyright holders to keep it "alive".

Look at the juggernaut Microsoft has had to become in order to pursue Bill's goal of dominating the user interface. That is, of course, what it is all about. Controlling access at the gate. You create a monopoly over the software architecture and you create a bottleneck between the user and the applications - or the experience. Bill is a wannabe troll. He's not about building bridges... he's about sabotaging every other bridge and setting up a toll both on his.

If Bill had his way I would pretty much have to use Internet Exploiter, er, um, Explorer to write this reply. But thanks to all of you, my brothers (and sisters) in arms of the Open Systems, Open Software, and Free Software movements I am using Mozilla's Firefox. If that burns Bill, I am glad. It is good for him, as Prometheus may attest.

IBM used to be the big, bad, monopoly. Things change. There is a reason Microsoft has invested in the console market... they realize that, eventually, Sony will be a challenge to them. Or perhaps Nintendo, Sega, some descendent of TiVo, or a bastard child of your digital cable box. The challenge will come. Now I know this seems irrelevant to many computer users, but I assure you it is not. Our technologies are evolving, both hardware and software. Our great challenge goes beyond just getting Linux on Aunt Emma's desktop.

Apple may weather the immediate storms on the horizion. As portable, cross-platform systems evolve in the Open community, hardware becomes more and more "abstracted". This makes it possible for an Apple to survive selling top-rate hardware systems that work with Open software systems. That is a marriage that has potential. But the necessary - and I mean necessary - trade-off is that they must then fully open themselves to hardware competition. Then they will survive on their merits. Just as it should be in any evolutionary system.

If Apple backs down and tries to remain proprietary at both ends they will surely die one day. They will cut themselves off from the diversity of our community, the evolutionary force at work through us. Their gene pool will become harder and harder to sustain. But if Apple, IBM, or any other hardware company today is smart they will support the Free Software movement.

If a Microsoft or some other company becomes a truly mighty troll, then they will be in a position to dictate which platforms live and which platforms die. The issue of controlling a bottleneck works both ways. Today, Microsoft already has enormous sway over hardware vendors. A viable alternative operating system is vital to their self-interest, lest they become Bill's lap-dogs. It may seem like a cozy arrangement, but remember our earlier lessons about domestication.

Attempting to build another proprietary bridge in the face of Microsoft's present market share is all but impossible. But we, the open community, have the resources, the manpower, the will, the energy, the spirit to make it happen. Supporting us means that bridges are more than a dime a dozen... they are FREE!! Importantly, they are not just free of monetary cost... they are free of TROLLS, and that means free as in freedom. Freedom for the users to find their own path to the hardware, freedom for the hardware vendors to develop divergent platforms that can appeal directly to the users.

This issue of freedom is central. Microsoft is already pursuing "trusted computing" and pressuring and/or persuading hardware manufacturers like Intel and AMD to come closer to their fires. The direction this is heading in can be scary. Now the troll is building an army. On the suface it may seem noble. It is certainly sold as such. Oh how noble they are to Mr. Company's IT Security Officer, who now knows that he can have total control over the desktops in the enterprise. Oh how noble they are to Mr. Copyright holder, as this troll can now controll access to digital media. Digital Rights Management is about building not a better troll-bridge, but a military-grade fence with one Gate. How appropriate for Mr. Gates. And it's not just the RIAA who will be singing a happy song.

Now for some this may seem just fine. But beware, boys and girls, beware. Control points create a focus of power. And power corrupts. Eventually the "powers that be" who come along in the next several years may decide that access to all digital media is a good thing for them, too. It makes George Orwell's vision that much more practicable.

In perhaps the greatest of ironies, the US Department of Defense (Ministry of War) wanted to create a computer communications network that was impervious to physical attack. Thus DARPA (DoD Advanced Research Projects Agency) set the seed for what has become the Internet. The global connectedness that has brought about is the fuel for a global cultural Renaissance happening right before our eyes!

Yes, it is the best time to be alive! Don't you ever doubt it! Oh my Amiga days... we had a choice in how we interfaced with our computers. KDE versus GNOME versus FVWM or what have you? The more the merrier! Isn't it great to have the choice? In fact it is in our best interest to have more than one. One of them should be that cutesy-pie desktop that will get your Aunt Emma using Free and Open Source software. Give her that choice, to help protect your choice (lean, mean, text-console interface machine if you so desire). I know the likes of Mr. Gates, or Sony for that matter, are not interested in you having a choice. But if Emma's Easy Distro furthers the Free Software Movement, it is truly a goodness.

And if you still think Digital Rights Management and Trusted Computing are a good thing, check out Lawrence Lessig's very truly excellent writings, expecially FREE CULTURE. His website is www.lessig.org (www.lessig.org). Free and Open Source software is more important than just some Geek Mission to control his own PC. It's a greater cause to help keep self-determination alive for all of us.

Have faith in the Free Software movement. We are the evolutionary force at play in our own world. Though we may feel ourselves to be haphazard, unfocused, or even reduntant at times, remember: This is how evolution works. Like life, it is mysterious - sometimes befuddling - but it is also magical.

Peace,

Tim LePes