A very interesting idea, your colour beacons. I'm not sure how well it would work in practise though -- different applications would map different colours to different ideas. Perhaps a better idea is to generalise the idea slightly, into icons which reflect state. You see this, for instance, when using GetRight on windows - you can minimize each download to an icon in the task bar, which visually represents the progress of the download. You also get icon feedback in KDE: when you load and application, an icon appears on the task bar with a spinning gear, which disappears once the application is loaded (it's amazing how simple feedback like this reduces frustation, and stops people loads 10 copies of a program). Possibly, when you wrote the article, animation of icons would have been too much of a drain on the system.
From the looks of your article, you were using something like WindowMaker at the time. One of the real disappointments of the last couple of years, for me, is how little progress there has been on GnuStep. Okay, it's a ripoff of an existing system, but one with a much more interesting and coherent architecture than Windows.
Sadly, coders work on what they want to work on -- and it appears that, despite all the flaming and vitriol aimed at Windows, people really do want to work on a system like it. After all, if there was a real ground swell of opposition to the direction Gnome is headed in, then people would be free to fork the code and move in their own direction. The lack of forks (and the lack of major dissent I see on the Gnome mailing lists), seems to show that you are in the minority.
I'm a reasonable person, though, and I gave Python a look. I don't doubt that its a great language, but I wouldn't want to use a language that relies on code formating for actual program structure.
So, in other words, you didn't actually try and write or alter anything written in Python, but just went 'ooh, yuk, you have to indent code'?
Amazingly enough, relying on code formatting
for actual program structure is one of the real
strong points of Python - it both reduces clutter, and ensures readability.
That said, Perl is an amusing language as well, it's just neither readable nor maintainable, so using it for large projects involving more than one coder is an absolute nightmare.
Don't go away thinking that C++ is some paragon of Object Orientated virtue - it's a horrible mixture of low level and high level, with bits of generic programming (templates) and functional programming (STL) thrown in for good measure; all mixed and blended to produce a very unintuitive language, and one of the largest specifications since PL/1.
There are advantages to doing OO code in C -- it shows you exactly what is going on, and doesn't hide anything from you. Most people would use lots of boiler-plate templates, and just fill in the blanks.
If you want simplicity and ease of use, try some real high-level languages. Python, for example, shows you that mixed metaphor languages don't have to be a cluttered, unreadable mess. Lisp (and its modern relatives like Scheme) is beautifully pure, with a simple specification (ignoring the abomination of CL), and is also very readable once you've hit your head against the monitor 5 or 6 times (although I wouldn't want to code it in an editor without a decent Lisp-mode).
Getting back to the main point: KDE may be more polished, but at the moment Gnome has better apps in many areas. Hopefully, Gnome will gain polish, KDE will gain applications, and in 2 years time we will have two excellent free UNIX desktops.
Come on, this is an article about Evolution (which looks like it's getting very interesting) -- what's the point of sneering at theKompany?
One of the reasons mindless trolls have left to refuse to use QT and KDE is that QT is GPLed, not LGPLed, which makes it impossible to write closed source programs using QT without paying TrollTech a licence fee. In contrast, you are perfectly free to write a closed source application using GTK and Gnome, because they are LGPLed.
So, you are using a system which makes it easier to produce closed source programs, and you are sneering at a company which produces closed source programs. Are you feeling ok?
I'd be interested to know what restrictions ECMA place on the licensing of the standards they endorse. ECMA do say in the article that "There are no known rights owned by Microsoft that would require a licensing agreement," but who would bet on that in advance? What's the point of having an open standard if no-one is allowed to actually implement it?
Interesting move on the part of Microsoft if they really are going to go through with standardisation. I suppose it's a similar issue to MPEG layer 3 and Fraunhofer... and look at how successful Fraunhofer have been at stopping open source MP3 encoders.
Anyhow, I wish Ximian luck with their reimplementation of Microsoft standards. I just hope they don't get burned like all the other companies in the past which have cuddled up to Microsoft. I would love it if, just this once, Microsoft did something nice:)
I've tried out Vorbis on a couple of albums (started looking when it got to beta 2). Beta 4 is on a par with Lame (except I find 128Kb Vorbis nicer to listen to than 160Kb Lame MP3). There are a couple of builds of the alpha version 1 release candidate around, and they're even more impressive - they have a couple of bugs with amplifying noise, but have encoded a couple of my albums to approx. 85Kb/s streams that sound incredible. It's still not the best codec to choose for very low bit rates (i.e. IP telephony), but it's definately the best cross platform format you can get for music and mid-to-high-quality reproduction.
So, yes, I'll be reencoding my 150 albums to Vorbis when RC1 is out. If only for the 40% space saving on my hard drive.
You're half correct. The release candidate for the version 1 *decoder* has been released, but the *encoder* is still at beta4. The release candidate for the encoder will be released at the end of the month, and that will be the first version supporting channel coupling. There are also several bugs in the psycho-acoustic model in beta4 that will be fixed in the release candidate.
Basically, they released the decoder as soon as they could, so other people can include a decoder in their applications that will play all version 1 Vorbis files.
So -- version 1 will have channel coupling, but it's not quite there yet.
This seems to be the way that British hardware companies develop. There is a close parallel between Psion moving from hardware (the 5mx, Revo, etc) to intellectual property (via Symbian), and the fate of Acorn.
To those of you that haven't been in a British school in the last 15 years - Acorn used to be the main supplier of computers to educational establishments, with the BBCs in the early 80s, and the Archimedes in the late 80s, early 90s. Being British, the Archimedes was an incredible ground-breaking mass-market system which absolutely no-one bought: it was the system the original ARM chip was designed for (indeed, ARM used to mean 'Acorn RISC Machines'). The Archimedes, which came out in 1987, had a 32-bit, graphical, multi-tasking operating system with the best version of Basic I've used used.
Luckily, they were better at marketing the ARM chip than they were in marketing the actual computer - ARM was spun off at a seperate company, and is now worth much much more than Acorn ever was. Much of the money ARM makes comes from licensing it's designs to other companies.
Similarly, Psion designed achingly wonderful handheld machines (I bought a Psion 5 when it came out. Recently, it ran for 44 hours on a pair of ordinary AA batteries). Incredible battery life, wonderful keyboard, very well designed OS (Epoc), integrated programming language. Now it looks like Psion the hardware company will fade away, and Symbian the software company will grow.
If you're using ICQ in Windows, there are a couple of alternative clients you might be interested in. The first is
Miranda, which looks like the ICQ client
did back before it got big and bloated.
The second is
Jabber,
which I've not really looked at, but seems to be
aiming to glob support for all the other main
instant messaging systems into one client,
together with it's own system.
Maybe we need to make the open source browsers be more configurable to lie about their browser/OS string?
Konqueror is very configurable about the browser ID string it sends (and the upcoming 2.2 release will be even more configurable. I believe you can make Mozilla change it's ID string, but I'm not sure how easy it is to configure.
Interesting to compare and contrast the reaction over 'gAIM' and 'kIllustator'. At least 'kIllustrator' has some claim to just be using a generic word that fits the naming scheme of the rest of the office suite it's part of. 'gAIM' is *just* the name of the product they are cloning, with a 'g' stuck on the front.
Not that I agree with the laws over patent and trademarks in your country, but with the laws as they are, 'gAIM' are in the wrong.
The research project I am currently engaged in involves finding 'averages' on spaces that aren't even manifolds. It's very easy to do on spaces that are uniquely geodesic, and I'd imagine you could get a plausible measure on spheres as well. All you need to do is to be able to measure distances between points -- and a sphere is certainly a metric space.
Just find the point (or points) which minimize the sum of the squares of distances to other points. Because the sphere isn't uniquely geodesic, you might get more than one point (if you had two points, one at the north, one at the south pole, then the points of minimum squared distance would be all the equator).
Does Konqueror have a simple button to do that, or do I have to drill down through several menus?
Well, if you install the kdeaddons package (this is with KDE 2.2, currently in beta), you get a drop down menu which lets you select the state of various features instantly. This includes cookies, Java, Javascript, image loading. That's generally good enough.
On a slightly different tack, you might be interested in
PyGame, a portable game programming framework for Python which wraps SDL. Of course, there are bindings for other languages as well (just look at the
SDL Homepage), but PyGame is very well implemented.
Here is a example of what a very simple demonstration app looks like in PyGame. One of the best applications/games using PyGame so far is
Solarwolf, a remake of an old Atari 2600 game (hmm... looks like the site is down at the moment, though).
One of the best ways to pick up game usage tips is to look at source code. One guy who's coded loads of SDL games (in C) can be found
here.
In particular - check out
Circus Linux - it's a lot of fun:).
To her mom, a computer is just a computer. It sends her e-mail, she can put picture from her digital camera on it, but she doesn't even think about things like licensing or freedom
And that's how it should be. You should use what works best for you. If you don't want to be bothered with anything technical, just send email, write recipies, etc., then use what is easiest, which is probably Windows. I'm quite happy for Linux to stay away from the mass market - it means it will stay oriented toward the areas which are best for me.
Don't assume that the operating system space should be a monoculture. For all the wierdos harping about Windows, it's a very good product, particularly given the (mostly achieved) objective of maintaining compatibility with code going all the way back to Windows 2.
Smiley noted:). That said - I'm not rabid. I've defended GNOME before from idiot 'KDE RULEZ GNOM SUKS' posters. I just love the fact that *twice* (with GNOME and KDE) a volunteer network of part-time coders has managed to produce someting better than CDE...
But anyway: this discussion was supposed to be about Galeon, so we're both off-topic. Sorry for shouting at you:)
No, we're talking about Microsoft, not Lateinos and Romiith.
(ah, who am I kidding... none of you are going to know what I'm talking about. sigh).
Just to get you a little less spooked:
'16. And he causeth all, both small and great, rich and poor, free and bond to receive a mark in their right hand or in their foreheads.
17. And that no man might buy or sell save that he had the mark or the name of the beast or the number of his name.
18. Here is wisdom. Let he that hath understanding count the number of the beast; for it is the number of a man, and his number is six hundred, three score and six.'
So don't get worried until you need to get that new identity barcode tattooed to your forehead:)
It's too late to sound clear and reasonable now: you should have thought of that before you trolled.
now:
'I wasn't implying that KDE sucked or should be tossed, far from it.'
earlier:
earlier:
'KDE is plain and simple, the most disgusting user interface experience one could ever have'
Have you ever used CDE? Better, how about Xaw applications (Xfig, Xmag, etc)? You may not have meant it, but you really pissed me off. Don't insult an entire desktop environment because you don't think you can make it look k3wl enough.
Re:KDE vs Gnome vs The rest of the Window Managers
on
Galeon At A Glance
·
· Score: 1
... you do know that blackbox is a window manager, and KDE and GNOME *aren't*, don't you?
You can use KDE with blackbox (and sawmill/sawfish, and icewm, and TWM...) very easily. If they don't implement the new window manager specification, then you lose some features, but the programs still work.
Well, if you're going for the standard Konqueror install, then you need the 'kdelibs' and 'kdebase' package. You don't have to install everything in 'kdebase', but it's easier:).
If you want something smaller, look for
'Konqueror embedded', which doesn't need KDE or X.
(you're trolling, but you have a decent point somewhere in there)
There was no enforced copyright in Shakespeare's time. Hence, by abolishing copyright we will immediately find many more Shakespeares. Silly? Yes - just as your argument is. Oh, and the whole anticommunist thing doesn't play so well outside the USA.
People have been scanning in and posting and trading copywrited material (books, software, music) for long before Napster and Gnutella, and they'll still be doing it long after the easy-to-use corporate front ends have all been hounded out of business. Take a look at
the alt.binaries.e-book* newsgroups, or
certain IRC channels, or Hotline, and so on and so on.
Increasing the major version number normally (i.e., in the part that isn't Microsoft:) indicates major incompatible changes to interfaces (both user and programmatic).
In this case, the GTK people are making binary incompatible changes to GTK+ (to include things like anti-aliasing, for example), so they will be increasing the major version number of their library. Because the GNOME people wish to use this library, they'll be using the opportunity to make binary incompatible changes to their system, and bump the release number up to 2 as well.
The same thing will happen with KDE when they port it to QT 3 (and QT 3 is QT 3 because it's a major binary incompatible change over QT 2).
Having two 'rival' desktop environments is a very good thing for UNIX. Each has areas where they excel, and which point the way for the other. To take just a small example, Nautilus has a much higher 'prettyness' factor than Konqueror in KDE 2.1 -- so, what did the KDE developers do? They improved some features on Konqueror (the detailed list view and the file previews) in the direction of the trail blazed by Nautilus. Look for KDE 2.2 to see what I mean:). Similarly, KDE shows to the GNOME guys things to do and not do in getting a proper component and object model: it's always easier to implement something if someone else has done it first:). GNOME's panel, it's default window manager, many major applications (Pan, Evolution,...), are all superior to KDE's.
With two seperate environments with distinct but similar goals around, they will *both* advance faster than they would do if only one of them was around. Hell, they're both infinitely better than CDE was (and I hate to think about how much was spent on developing that).
If I wasn't being forced to use Windows 98 at the moment, which means I can't moderate when I have points because the boxes get randomly mixed around the screen whenever I scroll, I'd mod you up:)
A very interesting idea, your colour beacons. I'm not sure how well it would work in practise though -- different applications would map different colours to different ideas. Perhaps a better idea is to generalise the idea slightly, into icons which reflect state. You see this, for instance, when using GetRight on windows - you can minimize each download to an icon in the task bar, which visually represents the progress of the download. You also get icon feedback in KDE: when you load and application, an icon appears on the task bar with a spinning gear, which disappears once the application is loaded (it's amazing how simple feedback like this reduces frustation, and stops people loads 10 copies of a program). Possibly, when you wrote the article, animation of icons would have been too much of a drain on the system.
From the looks of your article, you were using something like WindowMaker at the time. One of the real disappointments of the last couple of years, for me, is how little progress there has been on GnuStep. Okay, it's a ripoff of an existing system, but one with a much more interesting and coherent architecture than Windows.
Sadly, coders work on what they want to work on -- and it appears that, despite all the flaming and vitriol aimed at Windows, people really do want to work on a system like it. After all, if there was a real ground swell of opposition to the direction Gnome is headed in, then people would be free to fork the code and move in their own direction. The lack of forks (and the lack of major dissent I see on the Gnome mailing lists), seems to show that you are in the minority.
I'm a reasonable person, though, and I gave Python a look. I don't doubt that its a great language, but I wouldn't want to use a language that relies on code formating for actual program structure.
So, in other words, you didn't actually try and write or alter anything written in Python, but just went 'ooh, yuk, you have to indent code'?
Amazingly enough, relying on code formatting
for actual program structure is one of the real
strong points of Python - it both reduces clutter, and ensures readability.
That said, Perl is an amusing language as well, it's just neither readable nor maintainable, so using it for large projects involving more than one coder is an absolute nightmare.
Don't go away thinking that C++ is some paragon of Object Orientated virtue - it's a horrible mixture of low level and high level, with bits of generic programming (templates) and functional programming (STL) thrown in for good measure; all mixed and blended to produce a very unintuitive language, and one of the largest specifications since PL/1.
There are advantages to doing OO code in C -- it shows you exactly what is going on, and doesn't hide anything from you. Most people would use lots of boiler-plate templates, and just fill in the blanks.
If you want simplicity and ease of use, try some real high-level languages. Python, for example, shows you that mixed metaphor languages don't have to be a cluttered, unreadable mess. Lisp (and its modern relatives like Scheme) is beautifully pure, with a simple specification (ignoring the abomination of CL), and is also very readable once you've hit your head against the monitor 5 or 6 times (although I wouldn't want to code it in an editor without a decent Lisp-mode).
Getting back to the main point: KDE may be more polished, but at the moment Gnome has better apps in many areas. Hopefully, Gnome will gain polish, KDE will gain applications, and in 2 years time we will have two excellent free UNIX desktops.
Come on, this is an article about Evolution (which looks like it's getting very interesting) -- what's the point of sneering at theKompany?
One of the reasons mindless trolls have left to refuse to use QT and KDE is that QT is GPLed, not LGPLed, which makes it impossible to write closed source programs using QT without paying TrollTech a licence fee. In contrast, you are perfectly free to write a closed source application using GTK and Gnome, because they are LGPLed.
So, you are using a system which makes it easier to produce closed source programs, and you are sneering at a company which produces closed source programs. Are you feeling ok?
I'd be interested to know what restrictions ECMA place on the licensing of the standards they endorse. ECMA do say in the article that "There are no known rights owned by Microsoft that would require a licensing agreement," but who would bet on that in advance? What's the point of having an open standard if no-one is allowed to actually implement it?
:)
Interesting move on the part of Microsoft if they really are going to go through with standardisation. I suppose it's a similar issue to MPEG layer 3 and Fraunhofer... and look at how successful Fraunhofer have been at stopping open source MP3 encoders.
Anyhow, I wish Ximian luck with their reimplementation of Microsoft standards. I just hope they don't get burned like all the other companies in the past which have cuddled up to Microsoft. I would love it if, just this once, Microsoft did something nice
I've tried out Vorbis on a couple of albums (started looking when it got to beta 2). Beta 4 is on a par with Lame (except I find 128Kb Vorbis nicer to listen to than 160Kb Lame MP3). There are a couple of builds of the alpha version 1 release candidate around, and they're even more impressive - they have a couple of bugs with amplifying noise, but have encoded a couple of my albums to approx. 85Kb/s streams that sound incredible. It's still not the best codec to choose for very low bit rates (i.e. IP telephony), but it's definately the best cross platform format you can get for music and mid-to-high-quality reproduction.
So, yes, I'll be reencoding my 150 albums to Vorbis when RC1 is out. If only for the 40% space saving on my hard drive.
Bollocks.
We have quite strong laws in the UK about libel and slander - if they posted anything knowing it to be untrue they could be sued for a lot of money.
A lot of the whining about The Register comes from wierdo Americans who don't understand humour.
You're half correct. The release candidate for the version 1 *decoder* has been released, but the *encoder* is still at beta4. The release candidate for the encoder will be released at the end of the month, and that will be the first version supporting channel coupling. There are also several bugs in the psycho-acoustic model in beta4 that will be fixed in the release candidate.
Basically, they released the decoder as soon as they could, so other people can include a decoder in their applications that will play all version 1 Vorbis files.
So -- version 1 will have channel coupling, but it's not quite there yet.
This seems to be the way that British hardware companies develop. There is a close parallel between Psion moving from hardware (the 5mx, Revo, etc) to intellectual property (via Symbian), and the fate of Acorn.
To those of you that haven't been in a British school in the last 15 years - Acorn used to be the main supplier of computers to educational establishments, with the BBCs in the early 80s, and the Archimedes in the late 80s, early 90s. Being British, the Archimedes was an incredible ground-breaking mass-market system which absolutely no-one bought: it was the system the original ARM chip was designed for (indeed, ARM used to mean 'Acorn RISC Machines'). The Archimedes, which came out in 1987, had a 32-bit, graphical, multi-tasking operating system with the best version of Basic I've used used.
Luckily, they were better at marketing the ARM chip than they were in marketing the actual computer - ARM was spun off at a seperate company, and is now worth much much more than Acorn ever was. Much of the money ARM makes comes from licensing it's designs to other companies.
Similarly, Psion designed achingly wonderful handheld machines (I bought a Psion 5 when it came out. Recently, it ran for 44 hours on a pair of ordinary AA batteries). Incredible battery life, wonderful keyboard, very well designed OS (Epoc), integrated programming language. Now it looks like Psion the hardware company will fade away, and Symbian the software company will grow.
If you're using ICQ in Windows, there are a couple of alternative clients you might be interested in. The first is Miranda, which looks like the ICQ client did back before it got big and bloated. The second is Jabber, which I've not really looked at, but seems to be aiming to glob support for all the other main instant messaging systems into one client, together with it's own system.
Konqueror is very configurable about the browser ID string it sends (and the upcoming 2.2 release will be even more configurable. I believe you can make Mozilla change it's ID string, but I'm not sure how easy it is to configure.
Interesting to compare and contrast the reaction over 'gAIM' and 'kIllustator'. At least 'kIllustrator' has some claim to just be using a generic word that fits the naming scheme of the rest of the office suite it's part of. 'gAIM' is *just* the name of the product they are cloning, with a 'g' stuck on the front.
Not that I agree with the laws over patent and trademarks in your country, but with the laws as they are, 'gAIM' are in the wrong.
The research project I am currently engaged in involves finding 'averages' on spaces that aren't even manifolds. It's very easy to do on spaces that are uniquely geodesic, and I'd imagine you could get a plausible measure on spheres as well. All you need to do is to be able to measure distances between points -- and a sphere is certainly a metric space.
Just find the point (or points) which minimize the sum of the squares of distances to other points. Because the sphere isn't uniquely geodesic, you might get more than one point (if you had two points, one at the north, one at the south pole, then the points of minimum squared distance would be all the equator).
Well, if you install the kdeaddons package (this is with KDE 2.2, currently in beta), you get a drop down menu which lets you select the state of various features instantly. This includes cookies, Java, Javascript, image loading. That's generally good enough.
One of the best ways to pick up game usage tips is to look at source code. One guy who's coded loads of SDL games (in C) can be found here. In particular - check out Circus Linux - it's a lot of fun :).
To her mom, a computer is just a computer. It sends her e-mail, she can put picture from her digital camera on it, but she doesn't even think about things like licensing or freedom
And that's how it should be. You should use what works best for you. If you don't want to be bothered with anything technical, just send email, write recipies, etc., then use what is easiest, which is probably Windows. I'm quite happy for Linux to stay away from the mass market - it means it will stay oriented toward the areas which are best for me.
Don't assume that the operating system space should be a monoculture. For all the wierdos harping about Windows, it's a very good product, particularly given the (mostly achieved) objective of maintaining compatibility with code going all the way back to Windows 2.
Smiley noted :). That said - I'm not rabid. I've defended GNOME before from idiot 'KDE RULEZ GNOM SUKS' posters. I just love the fact that *twice* (with GNOME and KDE) a volunteer network of part-time coders has managed to produce someting better than CDE...
But anyway: this discussion was supposed to be about Galeon, so we're both off-topic. Sorry for shouting at you :)
No, we're talking about Microsoft, not Lateinos and Romiith.
:)
(ah, who am I kidding... none of you are going to know what I'm talking about. sigh).
Just to get you a little less spooked:
'16. And he causeth all, both small and great, rich and poor, free and bond to receive a mark in their right hand or in their foreheads.
17. And that no man might buy or sell save that he had the mark or the name of the beast or the number of his name.
18. Here is wisdom. Let he that hath understanding count the number of the beast; for it is the number of a man, and his number is six hundred, three score and six.'
So don't get worried until you need to get that new identity barcode tattooed to your forehead
It's too late to sound clear and reasonable now: you should have thought of that before you trolled.
now:
'I wasn't implying that KDE sucked or should be tossed, far from it.'
earlier:
earlier:
'KDE is plain and simple, the most disgusting user interface experience one could ever have'
Have you ever used CDE? Better, how about Xaw applications (Xfig, Xmag, etc)? You may not have meant it, but you really pissed me off. Don't insult an entire desktop environment because you don't think you can make it look k3wl enough.
... you do know that blackbox is a window manager, and KDE and GNOME *aren't*, don't you?
You can use KDE with blackbox (and sawmill/sawfish, and icewm, and TWM...) very easily. If they don't implement the new window manager specification, then you lose some features, but the programs still work.
Well, if you're going for the standard Konqueror install, then you need the 'kdelibs' and 'kdebase' package. You don't have to install everything in 'kdebase', but it's easier :).
If you want something smaller, look for
'Konqueror embedded', which doesn't need KDE or X.
There was no enforced copyright in Shakespeare's time. Hence, by abolishing copyright we will immediately find many more Shakespeares. Silly? Yes - just as your argument is. Oh, and the whole anticommunist thing doesn't play so well outside the USA.
People have been scanning in and posting and trading copywrited material (books, software, music) for long before Napster and Gnutella, and they'll still be doing it long after the easy-to-use corporate front ends have all been hounded out of business. Take a look at the alt.binaries.e-book* newsgroups, or certain IRC channels, or Hotline, and so on and so on.
Increasing the major version number normally (i.e., in the part that isn't Microsoft :) indicates major incompatible changes to interfaces (both user and programmatic).
In this case, the GTK people are making binary incompatible changes to GTK+ (to include things like anti-aliasing, for example), so they will be increasing the major version number of their library. Because the GNOME people wish to use this library, they'll be using the opportunity to make binary incompatible changes to their system, and bump the release number up to 2 as well.
The same thing will happen with KDE when they port it to QT 3 (and QT 3 is QT 3 because it's a major binary incompatible change over QT 2).
Having two 'rival' desktop environments is a very good thing for UNIX. Each has areas where they excel, and which point the way for the other. To take just a small example, Nautilus has a much higher 'prettyness' factor than Konqueror in KDE 2.1 -- so, what did the KDE developers do? They improved some features on Konqueror (the detailed list view and the file previews) in the direction of the trail blazed by Nautilus. Look for KDE 2.2 to see what I mean :). Similarly, KDE shows to the GNOME guys things to do and not do in getting a proper component and object model: it's always easier to implement something if someone else has done it first :). GNOME's panel, it's default window manager, many major applications (Pan, Evolution, ...), are all superior to KDE's.
With two seperate environments with distinct but similar goals around, they will *both* advance faster than they would do if only one of them was around. Hell, they're both infinitely better than CDE was (and I hate to think about how much was spent on developing that).
If I wasn't being forced to use Windows 98 at the moment, which means I can't moderate when I have points because the boxes get randomly mixed around the screen whenever I scroll, I'd mod you up :)