"You still haven't shown any evidence for that. A story on Apple's site about a Linux user switching isn't what you're talking about."
I made the claim that Apple is marketing against Linux. You said "I've never seen Apple pit its OS against Linux." I provided an example. Would you stop changing your argument midstream?
The patent contains no interesting, new methods. Instead, Apple is attempting to patent the general idea of transforming a program into a tamper resistant form.
The USPTO will probably grant this, or many of its claims, despite its lack of novelty. But this patent is a poster child for what is wrong with the patent system.
They may have one or two places on their site where they compare OS X to Linux or show some story of a Linux user switching but I have yet to see the kind of outright attack on Linux that they do on Windows and Microsoft.
Apple attacks Linux among Linux users.
Given some of their similarities (and I bet Apple realizes this) arguments against Linux can work both ways with OS X.
Apple's argument is simple: "Linux is technically OK, but its GUI is unusable for normal people". That's usually followed by bogus statements about X11 and a serious push to port applications to Cocoa.
You're actually trying to tell me that broad experience with a wide variety of operating systems and implementations of UNIX is a bad thing?
Whether your experience is "good" or "bad" or "broad" isn't at issue, at issue is your background, and the fact that you are a BSD guy is relevant to your judgement.
"BSD is still reasonably considered UNIX, but it's not canonical" And Linux is? Oh god, the irony. [...] Have since it wasn't much more than a rumor. What side of that debate were you on in 1993?
For Linux, the question is debatable depending on what criteria you apply, for Darwin, the answer is clear: Darwin implements many UNIX APIs, but it differs greatly from UNIX in philosophy, architecture, and functionality.
Big deal. I was at Berkeley when 2.8 and 4.1 were in development, and contributed code to some of those tapes. [...] and I can tell you with 30 years experience with UNIX, Darwin is pretty damn mainstream UNIX.
Ah, I see, your statements are not surprising given that your experience is mainly with bastardizations of UNIX and with BSD; you can look in the "Deprecated" section of the Research UNIX manuals about what the UNIX creators thought of a lot of the changes that Berkeley made to UNIX. Of course, BSD is still reasonably considered UNIX, but it's not canonical, and OS X's relation to UNIX through BSD is therefore even more tenuous.
In any case, can debate the meaning of "is" forever, but the fact remains that there are huge differences between OS X and traditional UNIX systems: large parts of the kernel are different in both code and architecture, IOKit is very different, administration is very different, and the file system fails to satisfy many of the fundamental properties that a UNIX file system ought to satisfy (even "cp" doesn't actually copy the entire contents of files).
OS X has big differences in many areas from traditional UNIX, as well as from the UNIX philosophy in general, and that those differences are getting bigger as Apple is hacking the system.
Quartz extreme uses 3D acceleration to render the desktop, something MS is just getting started with in Vista
Apple, too, is just getting started with this. Furthermore, Windows and X11 both support it, so OS X has no advantage. And Apple didn't invent this technology either.
many people don't even know how to get the 3D capabilities of their cards to work in Linux
The same way you get it to work on Macintosh: you buy supported hardware. It then works out of the box.
Cocoa is a VERY nice api. ObjC is a great language, FAR better than C++. [...] Combine ObjC with Python (Python is now a first class language on OS X) and you've got an excellent system that makes C#, Java etc. look awful.
Objective C's lack of runtime safety and the holes in its type system make it a dinosaur among programming languages.
C# (and Mono) at this point are supporting all the dynamic features of Objective C, plus the low-level features of C, plus runtime safety and garbage collection. They are clearly the future, and both Microsoft and Linux are ahead of Apple here. Apple need to either ditch Objective C or seriously overhaul it.
X11 is creaky -- pretty much everyone admits that.
I see nothing "creaky" about X11: its implementations have hardware accelerated drawing, full antialiasing, optional network transparency, and full control over compositing. Most importantly, it's a well-defined API and protocol, whereas Quartz is just whatever set of libraries Apple dreams up and ships with the current release.
You can also run Linux natively on PowerPC hardware. Some people do quite successfully, but most just use OS X.
I know: I have had a Linux/PPC machine running for several years. The reason to run Linux on PPC is for the available software for Linux, and because Linux runs much more efficiently on the same hardware than OS X (OS X is rather bloated).
X is about as well integrated as it can be. I'm not sure how you think it should work differently.
-- It should be preinstalled and start up automatically when needed; right now, it requires a separate installation step.
-- It should be integrated with the desktop in a way similar to what CodeTek VirtualDesktop does it; this should work out of the box.
-- It needs better support for drag-and-drop, conversions, etc.
-- Apple should add support to the X11 version of Gtk+ to support Mac-specific features (like menu bars at the top, etc.)
And Linux already runs just fine on Apple hardware, switching to x86 isn't going to make much of a difference there.
A lot of packages haven't been ported or tested sufficient on PPC hardware. That doesn't matter for simple desktop applications, but compilers, numerical libraries, and video are kind of weak under Linux/PPC.
And anyone that tells you that Linux is a great desktop OS in a thread about Mac OS X simply hasn't had enough experience with OSX.
It's nice that your needs are modest and OS X fulfills them. But, in fact, I do own several Macs and use them daily, and it is not a replacement for a Linux or UNIX desktop. Sorry.
Keynote stands head and shoulders above any other presentation software package I've seen - and to bring a mac to a staff meeting for a presentation is a joy.
Let's just take this as an example. I'm sure Keynote appeals to you: it looks clean, it's easy to use, and it makes slides look nice. I use it for quick presentations. But it is primitive and limited compared to PowerPoint or OpenOffice. Check the discussion forums; you'll find plenty of examples.
But I'm tired of going over the same territory with kids who have absolutely no clue of what UNIX is or where it came from
I started using, developing for, and managing, UNIX with Research V6, 2.8BSD, and 4.1BSD, and have been involved in it ever since on every major workstation and server, and I can tell you from nearly 30 years experience with UNIX: OS X is not UNIX, not even close.
Apart from simple issues like codebase and heritage, more importantly, Apple just doesn't follow the UNIX philosophy; they are busy adding a lot of the mess into Darwin that the original UNIX developers deliberately left out.
and go on to explain how X11 isn't actually a UNIX window system,
I'm glad you realize that. X11 is, in fact, a window system that runs on many platforms, both as client and as a server. That is one of its strengths. And that makes it particularly unfortunate that the support for it on OS X is so poor.
Linux and Darwin are both first class implementations of UNIX
Darwin is not an implementation of UNIX--it has a very different structure.
Furthermore, there are big differences, such as the file system and Netinfo.
OS X and Quartz/Aqua/Cocoa is a much more complete environment than Gnome or KDE and X11/Gtk/Qt/...
I suggest you actually look at the software that is available for Gnome and KDE, and you'll find that statement to be false. OS X is particularly weak when it comes to scientific applications. Cocoa also has fewer language bindings, there are few toolkits available for Quartz, there is less ability to customize (and none out of the box).
OS X (including the GUI) is also a resource hog and quite slow compared to a UNIX workstation setup in my experience.
But, more importantly, Quartz/Aqua/Cocoa are no the UNIX standard, and X11 support on OS X is not particularly good.
And, as other people have noted, Apple hardly mentions Linux.
They most certainly do: when they identify Linux or UNIX users, they try to push OS X aggressively as "a better implementation of Linux" and things like that.
And Apple has been going around universities in the region actually misleading people about OS X, telling them that OS X "is Linux with Apple enhancements".
Of course, Apple isn't going to do that on television, where they mostly reach Windows users. But they are making a strong push against Linux among Linux users they can actually reach. It is not pretty.
Open source allows people to fork software if they want to. That means that if I don't like RedHat, I can start my own distribution. And, guess what, that's exactly what people have done. That's what matters. It guarantees that software is written for users' needs, not for the needs of Bill Gates's bank account. And it guarantees that the software will be around long after the original developers are gone.
Perhaps but i can't go to the mall and buy a linux laptop. The ONLY unix like OS i can buy is a mac at the mall.
Oh, I fully agree with that. And it's one of the reasons why I actually have several Macintoshes (they also look nice). But that's not a problem with Linux software, it's a problem with business development. Linux is ready for the desktop, and its use in that area will expand.
While I like KDE better, I want GNOME to win because of the licensing issues with Qt. Commercial vendors shouldn't have to pay more than visual studio pricing to develop on any platform
I feel the same way.
and mac os has FREE developer tools with every mac. Think about that.
Well, I have thought about it, and it worries me. In the end, one of the problems I see with Macintosh is that its underlying technologies are a bit creaky. While people manage to create nice apps with them, Objective C, Cocoa, and Quartz are not the future IMO. I think Linux actually, at this point, has better technologies with Mono, Gtk#, and X11+Render.
And I find the business and marketing games Apple is playing with technologies annoying; for example, rather than making it easy for users to use X11-based apps, they go out of their way to keep them un-integrated with the Apple desktop. I think in the long term, Apple is headed for trouble with that kind of approach. I hope they'll come around and embrace Linux a bit more.
In any case, after Apple's switch to x86, it will be a lot easier to have the advantages of Apple's hardware and still be able to run Linux: we'll be able to run Linux both natively on Apple hardware, and virtualize it much more easily as well.
I would just like an Apple marketshare at around 35%-45% again
Macintosh never had a 35-45% market share.
So before you Mac-hating Linux kids start flaming another Apple article, most of us just want less Windows domination.
Nice words, except that people like you don't live by them. Apple and Apple proponents are constantly claiming that OS X is a superior replacement for Linux and UNIX. The antipathy between Linux and Macintosh originates with Apple zealots and Apple marketing.
Linux is a great desktop OS: KDE and Gnome have all the functionality you would want in a modern desktop OS, they are easy to use, attractive, and reliable.
I think WiFi on my Dell notebook is about the best example that I can come up with off the top of my head. it's a dual band 802.11a/b/g card for which Linux drivers just don't exist.
So, you bought a laptop that isn't well supported by Linux; that's your fault, not Linux's. If you buy your laptop from a Linux laptop vendor, Linux just works out of the box.
The main advantage on Macintosh is that OpenGL, wireless, and other built-in devices generally work on Apple hardware. But the same is true for Linux if you actually buy Linux from Linux hardware vendor.
For third party hardware, I don't see much of a difference. And don't talk to me about Epson scanners on OS X; I think they have gotten better by now, but they used to be a huge pain.
Have you seen the customer list of Trolltech? Commercial support can be a big plus for companies.. They have one company to contact for support questions, instead of developers somewhere on the net.
Yes, and that's a nice thing and I'm not saying that Troll Tech should disappear from the face of the earth. Indeed, all the companies I have worked at have had Qt licenses for some of their developers.
But there is a difference between being a vendor a cross-platform toolkit for custom application development, and being at the core of the desktop platform of one of the top three desktop operating systems in the world. The former doesn't make you a good choice for the latter.
The whole community thing means free support, but also also scares companies away. Companies who feel commercial partners are more reliable.
Some companies indeed do want commercial support, and they can get it for both Gtk+ and Qt.
Nope. Not in an open-source world. The point is to take EVERYTHING on an equal footing, get the best out of both and ditch the cruft. It's like software evolution
I like the concept of software evolution and have repeatedly argued for it.
But at the level of individual distributions, distributions need to make a choice where to put their resources. That's particularly true for distributions that are trying to give their users a nice desktop experience.
Whoever wins out of KDE and GNOME will, by definition, be the better system.
I disagree. I think Gnome will win this one no matter which one is "the better system", simply because the Qt dual-licensing scheme is too much of a problem. And, in the unlikely event that KDE does win, the Linux desktop is in serious trouble, because it simply ain't going to happen that a large number of commercial Linux developers are going to send their money to Norway.
But what to start from? Where to get those ideas? Where to find those approaches? How to determine which of the new approaches works and which was better off the old way? By putting them all together, fighting it out (by a vote of user popularity) and, as if by magic, a victor will appear. They can take bits of each other, they can "steal" each other's ideas but they shouldn't be written off just because you don't like them. Many, many people do.
That kind of exchange between Gnome and KDE works for incremental tweaks. But, in the grand scheme of things, the two desktops are very similar and have very similar faults. I think right now is the opportunity to do something substantially different and better.
Why exactly do you care about proprietary, closed source, non-free software anyway?
Because we need vendors of proprietary software to port to the Linux desktop for Linux to succeed. And they aren't going to do that if we make it difficult for their employees to choose porting to Linux.
And that argument is no different from what Troll Tech proponents make for Qt's commercial license: it's a marriage of convenience. By giving people the opportunity to develop commercial software, we are helping free software succeed. It's just that among marriages of convenience, the Gnome/Gtk+ one is better.
Proprietary software is immoral, and if that's what you're developing, then you SHOULD be penalized by having to pay licensing fees.
I think that's called "cutting off your nose to spite your face".
Seriously, you should think about how to make open source succeed, not about penalizing people. As a platform with very little desktop market share and and perception of being hard to use, we need all the help we can get.
I thought the point was to make Linux more poular rather than less popular. Restricting the choice of window mamager
Who is talking about "restricting"? I just think that the logical and sensible thing for Ubuntu and other commercial distros is to focus on Gnome support.
Personally I prefer KDE
Purely as an end user, so do I. As a programmer, I think KDE and Gnome suck about equally. But as the basis for a distribution, Gnome is the only viable choice IMO.
Slashdot Mirror
"You still haven't shown any evidence for that. A story on Apple's site about a Linux user switching isn't what you're talking about."
I made the claim that Apple is marketing against Linux. You said "I've never seen Apple pit its OS against Linux." I provided an example. Would you stop changing your argument midstream?
The patent contains no interesting, new methods. Instead, Apple is attempting to patent the general idea of transforming a program into a tamper resistant form.
The USPTO will probably grant this, or many of its claims, despite its lack of novelty. But this patent is a poster child for what is wrong with the patent system.
I don't think you should place a price upon the value of saving civilization.
Well, a good start would be creating a civilization to save.
They may have one or two places on their site where they compare OS X to Linux or show some story of a Linux user switching but I have yet to see the kind of outright attack on Linux that they do on Windows and Microsoft.
Apple attacks Linux among Linux users.
Given some of their similarities (and I bet Apple realizes this) arguments against Linux can work both ways with OS X.
Apple's argument is simple: "Linux is technically OK, but its GUI is unusable for normal people". That's usually followed by bogus statements about X11 and a serious push to port applications to Cocoa.
No, the problem is that you hate Linux and X11 for some reason.
You're actually trying to tell me that broad experience with a wide variety of operating systems and implementations of UNIX is a bad thing?
Whether your experience is "good" or "bad" or "broad" isn't at issue, at issue is your background, and the fact that you are a BSD guy is relevant to your judgement.
"BSD is still reasonably considered UNIX, but it's not canonical" And Linux is? Oh god, the irony. [...] Have since it wasn't much more than a rumor. What side of that debate were you on in 1993?
For Linux, the question is debatable depending on what criteria you apply, for Darwin, the answer is clear: Darwin implements many UNIX APIs, but it differs greatly from UNIX in philosophy, architecture, and functionality.
Big deal. I was at Berkeley when 2.8 and 4.1 were in development, and contributed code to some of those tapes. [...] and I can tell you with 30 years experience with UNIX, Darwin is pretty damn mainstream UNIX.
Ah, I see, your statements are not surprising given that your experience is mainly with bastardizations of UNIX and with BSD; you can look in the "Deprecated" section of the Research UNIX manuals about what the UNIX creators thought of a lot of the changes that Berkeley made to UNIX. Of course, BSD is still reasonably considered UNIX, but it's not canonical, and OS X's relation to UNIX through BSD is therefore even more tenuous.
In any case, can debate the meaning of "is" forever, but the fact remains that there are huge differences between OS X and traditional UNIX systems: large parts of the kernel are different in both code and architecture, IOKit is very different, administration is very different, and the file system fails to satisfy many of the fundamental properties that a UNIX file system ought to satisfy (even "cp" doesn't actually copy the entire contents of files).
OS X has big differences in many areas from traditional UNIX, as well as from the UNIX philosophy in general, and that those differences are getting bigger as Apple is hacking the system.
Quartz extreme uses 3D acceleration to render the desktop, something MS is just getting started with in Vista
Apple, too, is just getting started with this. Furthermore, Windows and X11 both support it, so OS X has no advantage. And Apple didn't invent this technology either.
many people don't even know how to get the 3D capabilities of their cards to work in Linux
The same way you get it to work on Macintosh: you buy supported hardware. It then works out of the box.
Cocoa is a VERY nice api. ObjC is a great language, FAR better than C++. [...] Combine ObjC with Python (Python is now a first class language on OS X) and you've got an excellent system that makes C#, Java etc. look awful.
Objective C's lack of runtime safety and the holes in its type system make it a dinosaur among programming languages.
C# (and Mono) at this point are supporting all the dynamic features of Objective C, plus the low-level features of C, plus runtime safety and garbage collection. They are clearly the future, and both Microsoft and Linux are ahead of Apple here. Apple need to either ditch Objective C or seriously overhaul it.
X11 is creaky -- pretty much everyone admits that.
I see nothing "creaky" about X11: its implementations have hardware accelerated drawing, full antialiasing, optional network transparency, and full control over compositing. Most importantly, it's a well-defined API and protocol, whereas Quartz is just whatever set of libraries Apple dreams up and ships with the current release.
You can also run Linux natively on PowerPC hardware. Some people do quite successfully, but most just use OS X.
I know: I have had a Linux/PPC machine running for several years. The reason to run Linux on PPC is for the available software for Linux, and because Linux runs much more efficiently on the same hardware than OS X (OS X is rather bloated).
X is about as well integrated as it can be. I'm not sure how you think it should work differently.
-- It should be preinstalled and start up automatically when needed; right now, it requires a separate installation step.
-- It should be integrated with the desktop in a way similar to what CodeTek VirtualDesktop does it; this should work out of the box.
-- It needs better support for drag-and-drop, conversions, etc.
-- Apple should add support to the X11 version of Gtk+ to support Mac-specific features (like menu bars at the top, etc.)
And Linux already runs just fine on Apple hardware, switching to x86 isn't going to make much of a difference there.
A lot of packages haven't been ported or tested sufficient on PPC hardware. That doesn't matter for simple desktop applications, but compilers, numerical libraries, and video are kind of weak under Linux/PPC.
And anyone that tells you that Linux is a great desktop OS in a thread about Mac OS X simply hasn't had enough experience with OSX.
It's nice that your needs are modest and OS X fulfills them. But, in fact, I do own several Macs and use them daily, and it is not a replacement for a Linux or UNIX desktop. Sorry.
Keynote stands head and shoulders above any other presentation software package I've seen - and to bring a mac to a staff meeting for a presentation is a joy.
Let's just take this as an example. I'm sure Keynote appeals to you: it looks clean, it's easy to use, and it makes slides look nice. I use it for quick presentations. But it is primitive and limited compared to PowerPoint or OpenOffice. Check the discussion forums; you'll find plenty of examples.
But I'm tired of going over the same territory with kids who have absolutely no clue of what UNIX is or where it came from
I started using, developing for, and managing, UNIX with Research V6, 2.8BSD, and 4.1BSD, and have been involved in it ever since on every major workstation and server, and I can tell you from nearly 30 years experience with UNIX: OS X is not UNIX, not even close.
Apart from simple issues like codebase and heritage, more importantly, Apple just doesn't follow the UNIX philosophy; they are busy adding a lot of the mess into Darwin that the original UNIX developers deliberately left out.
and go on to explain how X11 isn't actually a UNIX window system,
I'm glad you realize that. X11 is, in fact, a window system that runs on many platforms, both as client and as a server. That is one of its strengths. And that makes it particularly unfortunate that the support for it on OS X is so poor.
Linux and Darwin are both first class implementations of UNIX
e ns/
Darwin is not an implementation of UNIX--it has a very different structure.
Furthermore, there are big differences, such as the file system and Netinfo.
OS X and Quartz/Aqua/Cocoa is a much more complete environment than Gnome or KDE and X11/Gtk/Qt/...
I suggest you actually look at the software that is available for Gnome and KDE, and you'll find that statement to be false. OS X is particularly weak when it comes to scientific applications. Cocoa also has fewer language bindings, there are few toolkits available for Quartz, there is less ability to customize (and none out of the box).
OS X (including the GUI) is also a resource hog and quite slow compared to a UNIX workstation setup in my experience.
But, more importantly, Quartz/Aqua/Cocoa are no the UNIX standard, and X11 support on OS X is not particularly good.
And, as other people have noted, Apple hardly mentions Linux.
They most certainly do: when they identify Linux or UNIX users, they try to push OS X aggressively as "a better implementation of Linux" and things like that.
Here is an example from their web site:
http://www.apple.com/uk/hotnews/articles/simoncoz
See, for example, http://www.apple.com/uk/hotnews/articles/simoncoze ns/
And Apple has been going around universities in the region actually misleading people about OS X, telling them that OS X "is Linux with Apple enhancements".
Of course, Apple isn't going to do that on television, where they mostly reach Windows users. But they are making a strong push against Linux among Linux users they can actually reach. It is not pretty.
It had to happen: a compiler with a JavaScript backend...
what's the news in that?
Seriously, there has been very little science accomplished on the ISS, at least relative to the enormous amounts of money it has consumed.
Their business model is obviously to be acquired by Yahoo!, Google, or Microsoft.
Open source allows people to fork software if they want to. That means that if I don't like RedHat, I can start my own distribution. And, guess what, that's exactly what people have done. That's what matters. It guarantees that software is written for users' needs, not for the needs of Bill Gates's bank account. And it guarantees that the software will be around long after the original developers are gone.
Perhaps but i can't go to the mall and buy a linux laptop. The ONLY unix like OS i can buy is a mac at the mall.
Oh, I fully agree with that. And it's one of the reasons why I actually have several Macintoshes (they also look nice). But that's not a problem with Linux software, it's a problem with business development. Linux is ready for the desktop, and its use in that area will expand.
While I like KDE better, I want GNOME to win because of the licensing issues with Qt. Commercial vendors shouldn't have to pay more than visual studio pricing to develop on any platform
I feel the same way.
and mac os has FREE developer tools with every mac. Think about that.
Well, I have thought about it, and it worries me. In the end, one of the problems I see with Macintosh is that its underlying technologies are a bit creaky. While people manage to create nice apps with them, Objective C, Cocoa, and Quartz are not the future IMO. I think Linux actually, at this point, has better technologies with Mono, Gtk#, and X11+Render.
And I find the business and marketing games Apple is playing with technologies annoying; for example, rather than making it easy for users to use X11-based apps, they go out of their way to keep them un-integrated with the Apple desktop. I think in the long term, Apple is headed for trouble with that kind of approach. I hope they'll come around and embrace Linux a bit more.
In any case, after Apple's switch to x86, it will be a lot easier to have the advantages of Apple's hardware and still be able to run Linux: we'll be able to run Linux both natively on Apple hardware, and virtualize it much more easily as well.
I would just like an Apple marketshare at around 35%-45% again
Macintosh never had a 35-45% market share.
So before you Mac-hating Linux kids start flaming another Apple article, most of us just want less Windows domination.
Nice words, except that people like you don't live by them. Apple and Apple proponents are constantly claiming that OS X is a superior replacement for Linux and UNIX. The antipathy between Linux and Macintosh originates with Apple zealots and Apple marketing.
As a desktop OS, Linux sucks.
Linux is a great desktop OS: KDE and Gnome have all the functionality you would want in a modern desktop OS, they are easy to use, attractive, and reliable.
I think WiFi on my Dell notebook is about the best example that I can come up with off the top of my head. it's a dual band 802.11a/b/g card for which Linux drivers just don't exist.
So, you bought a laptop that isn't well supported by Linux; that's your fault, not Linux's. If you buy your laptop from a Linux laptop vendor, Linux just works out of the box.
The main advantage on Macintosh is that OpenGL, wireless, and other built-in devices generally work on Apple hardware. But the same is true for Linux if you actually buy Linux from Linux hardware vendor.
For third party hardware, I don't see much of a difference. And don't talk to me about Epson scanners on OS X; I think they have gotten better by now, but they used to be a huge pain.
Have you seen the customer list of Trolltech? Commercial support can be a big plus for companies.. They have one company to contact for support questions, instead of developers somewhere on the net.
Yes, and that's a nice thing and I'm not saying that Troll Tech should disappear from the face of the earth. Indeed, all the companies I have worked at have had Qt licenses for some of their developers.
But there is a difference between being a vendor a cross-platform toolkit for custom application development, and being at the core of the desktop platform of one of the top three desktop operating systems in the world. The former doesn't make you a good choice for the latter.
The whole community thing means free support, but also also scares companies away. Companies who feel commercial partners are more reliable.
Some companies indeed do want commercial support, and they can get it for both Gtk+ and Qt.
Nope. Not in an open-source world. The point is to take EVERYTHING on an equal footing, get the best out of both and ditch the cruft. It's like software evolution
I like the concept of software evolution and have repeatedly argued for it.
But at the level of individual distributions, distributions need to make a choice where to put their resources. That's particularly true for distributions that are trying to give their users a nice desktop experience.
Whoever wins out of KDE and GNOME will, by definition, be the better system.
I disagree. I think Gnome will win this one no matter which one is "the better system", simply because the Qt dual-licensing scheme is too much of a problem. And, in the unlikely event that KDE does win, the Linux desktop is in serious trouble, because it simply ain't going to happen that a large number of commercial Linux developers are going to send their money to Norway.
But what to start from? Where to get those ideas? Where to find those approaches? How to determine which of the new approaches works and which was better off the old way? By putting them all together, fighting it out (by a vote of user popularity) and, as if by magic, a victor will appear. They can take bits of each other, they can "steal" each other's ideas but they shouldn't be written off just because you don't like them. Many, many people do.
That kind of exchange between Gnome and KDE works for incremental tweaks. But, in the grand scheme of things, the two desktops are very similar and have very similar faults. I think right now is the opportunity to do something substantially different and better.
Why exactly do you care about proprietary, closed source, non-free software anyway?
Because we need vendors of proprietary software to port to the Linux desktop for Linux to succeed. And they aren't going to do that if we make it difficult for their employees to choose porting to Linux.
And that argument is no different from what Troll Tech proponents make for Qt's commercial license: it's a marriage of convenience. By giving people the opportunity to develop commercial software, we are helping free software succeed. It's just that among marriages of convenience, the Gnome/Gtk+ one is better.
Proprietary software is immoral, and if that's what you're developing, then you SHOULD be penalized by having to pay licensing fees.
I think that's called "cutting off your nose to spite your face".
Seriously, you should think about how to make open source succeed, not about penalizing people. As a platform with very little desktop market share and and perception of being hard to use, we need all the help we can get.
I thought the point was to make Linux more poular rather than less popular. Restricting the choice of window mamager
Who is talking about "restricting"? I just think that the logical and sensible thing for Ubuntu and other commercial distros is to focus on Gnome support.
Personally I prefer KDE
Purely as an end user, so do I. As a programmer, I think KDE and Gnome suck about equally. But as the basis for a distribution, Gnome is the only viable choice IMO.