there was some argument that Microsoft should have stayed with a windowing manager on top of core OS paradigm as they previously had and beefed up DOS to be something like Unix. Fortunately, saner minds prevailed.
No, they didn't. That's exactly what Windows is; you can boot NT into a command line environment and run it completely without a GUI. It would have been insane if they hadn't done that.
Use of a text interface and system fiddling is inevitable. Not so with Windows.
That's total bullshit. There are millions of Linux systems that don't even have a command line and do everything graphically (e.g., Linksys routers). And desktop Linux installations usually come with a full complement of graphical administration tools.
On the other hand, a lot of Windows system management involves going to the command line. It's just that many people give up at that point and just reinstall Windows.
The differences between Linux and BSD are minor; anybody who thinks they matter needs to have their head examined, and the kind of (implicit) Linux bashing represented by the article is pointless.
Yes linux is powerful but it can by no means replace the average desktop user *YET*.
Linux is an excellent drop-in replacement for a Windows desktop: existing Windows users feel comfortable with it and become productive right away. That's what counts in a corporate environment.
In contrast, OS X differs significantly from the de-facto standard desktop environment, making it harder for users to become productive on it.
Don't forget which chipset is in each, which VERSION of which chipset is in each, and as for wireless forget an easy install of WPA when you want (more?) secure wireless networking......
You OS X users must feel really threatened by Linux to keep making up such lies.
In real life, Linux "just works" when you buy it preinstalled, just like OS X (in fact, better than OS X).
In addition, Linux actually autodetects and autoconfigures most hardware. That's a bonus: OS X doesn't even boot on hardware not specifically designed for it.
The install process is considerably easier than a linux install because I don't need to know a damn thing about who made my graphics card, who made my NIC or who made my wireless system and I don't have to know anything about partitioning.
If you buy your Linux machines like you do your OS X machines, namely preinstalled, that's all you have to do: turn it on. The fact that Linux also installs on other PCs is an added bonus.
And the last Linux install I did required me trying 20 different drivers for my NIC since the ones that were supposed to work with my card was failing with some obscure error that a web search turned no hits on.
Well, then stop being stupid and buy your Linux machines preinstalled, like you do your OS X machines.
I'm tired of Linux fanboys like you spreading uninformed bullshit about OS X
We Linux fanboys don't generally give a damn about OS X. We don't tell OS X users to switch to Linux. We think you guys are a lost cause. We'd just like you to return the favor. OK?
If a Linux system "just worked" then yeah, maybe they'd choose it over OS X or Windows.
When you buy Linux machines from a Linux vendor, they do "just work". In fact, even when you don't, something like Knoppix or Ubuntu will "just work" on most Windows PCs--just boot it--it's simpler than OS X.
The mere existence of multiple windowing systems means that applications will be targeted at one of them.
You mean like Carbon, Cocoa, Classic, Mozilla, MS Office, and Java on OS X? Yes, that sure wears on me. When will Apple finally manage to produce a consistent system? They haven't so far.
They just want to go to Edit -> Paste (the more savvy ones may know Control+V) and have it work.
Maybe they'd also like a consistent view of keybindings and the file system. Too bad OS X doesn't provide either of those (Carbon and Cocoa handle those differently from one another). That's something that bothers me a great deal on OS X.
I already pay for it, through my workplace. Now I get to pay for it twice.
No, you don't. Your workplace retirement benefits are not guaranteed (even if they say they are). Millions of people have paid into insurance and workplace-related benefits and were left with nothing when they retired.
I've seen no valid argument for wealth redistribution yet.
Social security and similar government programs have nothing to do with "wealth redistribution"; they are there to guarantee you a minimum income in case all other precautions fail.
(Some wealth redistribution is essential for a free society, and everybody benefits from it, but that's a separate argument.)
Stop focusing on the structure of the operating system and START focusing on user productivity. There is a STEEEEEEEEP learning curve for the average user to comprehend much less USE linux on a day today basis.
I challenge you to cite studies that actually demonstrate that OS X is any easier to learn or use than Gnome, KDE, or Windows.
In our own experience with the platform, we find that users have about the same number of problems with OS X as with other platforms.
In fact, the idiosyncratic design of OS X (e.g., menu bar, single button mouse, admin tools, drag-and-drop installation, different visual appearance, etc.) makes for high retraining costs of existing users in our experience, and makes the platform difficult to use for people who need to use multiple platforms. In contrast, Linux desktops can be configured to look and behave nearly identically to their Windows counterparts, lowering retraining and support costs.
I dont think its so much understanding what Linux is...as it is that companies want to obtain technologies that get them where they need or want to be
And your point is what? That you think that OS X is that technology? I don't thinks so. Even leaving aside the question of which one is actually the better platform, the fact that Linux is nearly fully compatible with existing UNIX workstations and that it runs on existing PC hardware saves companies a ton of money. Their IT staff will already know how to run Linux machines, while OS X administration is rather different (Netinfo and all that). On OS X, X11 is a separate install, and its native window system and GUI is completely incompatible with anything else. Linux also has a much wider choice of mature software for remote administration, network administration, and other functions.
Lee built up a research lab for Microsoft in China and he is supposed to do the same for Google. That doesn't pose a "conflict of interest", nor should it be difficult for him to avoid disclosure of confidential information in his new position.
but it does show that most people get their work done more efficiently with them (I know I do).
That shows preference, not efficacy.
Could you give a few examples?
Remote logins, remote applications, preferences, and file system access, for example. Basically, Gnome and KDE are Windows-like desktops that happen to use X11 for graphics (and inefficiently at that).
I'm typing this on a laptop with Linux installed. What did I have to do to get it running? Stick in an Ubuntu Linux Install CD, boot, and confirm that I want to install it. After the install, it automatically connected to the nearest wireless network.
The install process was easier than a fresh install of OS X (which asks lots more annoying questions). And the last OS X upgrade I did on my iBook required me doing a web search for why it was failing with an obscure message, going into the command line, and patching up the HFS+ file system manually.
I'm tired of OS X fanboys like you spreading uninformed bullshit about Linux.
I think that story and those statistics are implausible, both given how big businesses operate and given Apple's actual sales figures.
Furthermore, the suggestion that OS X is an alternative to Linux indicates a lack of understanding of what Linux is all about and why it's being adopted.
"When you add things, it breaks?" That's one of the main problems with Windows: everytime you do anything (add h/w, s/w,...) with Windows, you risk having to reinstall. Maybe Taylor got confused about which platform is which.
What practical use is KDE? Has anybody actually ever demonstrated that KDE is a better tool for getting work done than emacs on a vt100? Or xedit under twm?
At this point, assertions about the utility of GUI features are based on faith, for any GUI. But assertions of non-innovativeness are not: neither KDE nor Gnome are innovative in any significant way.
And if it's good, why is no one using it?
Xt-based toolkits were very widely used and probably are still far more widely used commercially than either Gtk+ or Qt. They were by no means perfect, but they got a lot of things right.
That is indeed interesting, I had no idea this existed.
And that illustrates the core of the problem: KDE and Gnome have been developed with ignorance and disregard of prior work. It's hard to improve on something you don't know. And KDE and Gnome both actually break a lot of things that X11 users used to take for granted.
Explain welfare and medicaid, and that guy FICA who steals part of every one of my paychecks then.
Simple: it's so that you pay for the risk of you becoming indigent. If you weren't forced to pay for insurance against that, you wouldn't bother (as your comment indicates), and other people would have to assume the risk that you so foolishly take.
while simultaneously believing YOU know what's best for everyone else.
See, the difference is that when liberals think they know what's best for someone, they don't force them to do it by law, even if they tell you it's bad for you.
When conservatives think they know what's best for someone (and they do as much as liberals), they'll try to pass laws to force them to do it.
What makes KDE 'slow' (note that I've had no problems running it on old hardware) is that it tries to do too much. Changing the language it's implemented in is not going to speed it up.
Of course it is. Implemented in a decent language, it wouldn't take a 10MByte process to put up a bunch of hotkeys, it would take a small function that takes up a few hundred bytes. And it would take more than a dozen processes just to get a simple desktop. Implemented in a decent language, I wouldn't have to download dozens of megabytes of source code and libraries just to change the behavior of a little applet, I'd click on whatever I want to change, get an editor, make the change, and go right along.
99.999% of the 'problems' that programming languages are claimed to have are the result of inexperienced, incompetent and down right crap programmers.
Yes, like the ones that think that "C++ is an excellent language for OO coding".
Thankfully increased competition in the job market has removed a large number of them from the industry.
Many competent programmers have left active programming in disgust because--what's the point? The tide of crap written in C/C++/Java just can't be stemmed. The best thing one can do is move up the food chain and hire people like you who apparently enjoy shoveling shit for a living.
First, he basically ripped this idea off from a previous chart built in 1951, modernized, gave it a better "UI" and is now shipping it out to the masses. Sound familiar?
No: companies universally give things they rip off from other companies worse UIs. I think it's because the people who do the ripping off don't spend much time thinking about the problem and prefer a simplistic solution to a good solution.
Would you like an OS/DE that lets everything integrate with everything? Then you might as well make that KDE (or GNOME, or whichever you prefer), and consider only apps that stick to KDE's rules. What am I missing?
The fact that mechanisms like KParts and DCOP are so cumbersome, opaque, and user-hostile that even experienced programmers don't bother using them.
Maybe I'm just not imaginative enough, but what kind of system are you envisioning? How could a desktop environment, no matter how innovative, let the user create buttons in a program that provide a functionality that the programmer hasn't provided for?
Have a look at Kay's Squeak: you can click on any UI element, move it around, replicate it, edit the corresponding code, script it, connect it to other elements, etc.
Heck, even editres for Xt was more flexible than what KDE is doing (you can still try it--it ships with X11, but toolkits like Qt and Gtk+ are too primitive and too non-standard to talk to it). The current crop of UIs have enormous amount of time invested in them in terms of theming, placement of buttons, etc., but they are technically worse than what we had 20 years ago.
Is that really innovative, given that you can do the same thing with "Setting -> configure toolbars"?
That only lets you add buttons that the programmer decided to let you have.
I don't know, but it's certainly not too far-fetched, thanks to KParts.
KParts only lets you embed those things that have explicit code for being embedded.
I assume you mean things like http://www.croquetproject.org/ ? That's certainly a more radical change from what we have now, but it's just not ready yet for everyday use.
It's a "radical change" in the same sense that eating healthy is a "radical change" from eating junk food. If it weren't for 20 years of junk GUI programming in C/C++ by Microsoft, Apple, Gnome, KDE, DEC, and HP, it wouldn't be a "radical change"--Alan Kay's current work is merely the evolution of ideas that go back to the 1970's.
The guy used to be at Apple's Advanced Technology Lab (back when Apple still had an Advanced Technology Lab) and at SGI. This is a bad move by Microsoft and something that's gonna make people of Lee's caliber think twice about working for them. They should be grateful about what the guy did for them and congratulate him on his new job.
Slashdot Mirror
there was some argument that Microsoft should have stayed with a windowing manager on top of core OS paradigm as they previously had and beefed up DOS to be something like Unix. Fortunately, saner minds prevailed.
No, they didn't. That's exactly what Windows is; you can boot NT into a command line environment and run it completely without a GUI. It would have been insane if they hadn't done that.
Use of a text interface and system fiddling is inevitable. Not so with Windows.
That's total bullshit. There are millions of Linux systems that don't even have a command line and do everything graphically (e.g., Linksys routers). And desktop Linux installations usually come with a full complement of graphical administration tools.
On the other hand, a lot of Windows system management involves going to the command line. It's just that many people give up at that point and just reinstall Windows.
The differences between Linux and BSD are minor; anybody who thinks they matter needs to have their head examined, and the kind of (implicit) Linux bashing represented by the article is pointless.
Well, if anybody asserts that "Windows is the operating system that Linux should have been", they clearly deserve a bashing.
Yes linux is powerful but it can by no means replace the average desktop user *YET*.
Linux is an excellent drop-in replacement for a Windows desktop: existing Windows users feel comfortable with it and become productive right away. That's what counts in a corporate environment.
In contrast, OS X differs significantly from the de-facto standard desktop environment, making it harder for users to become productive on it.
Don't forget which chipset is in each, which VERSION of which chipset is in each, and as for wireless forget an easy install of WPA when you want (more?) secure wireless networking......
You OS X users must feel really threatened by Linux to keep making up such lies.
In real life, Linux "just works" when you buy it preinstalled, just like OS X (in fact, better than OS X).
In addition, Linux actually autodetects and autoconfigures most hardware. That's a bonus: OS X doesn't even boot on hardware not specifically designed for it.
The install process is considerably easier than a linux install because I don't need to know a damn thing about who made my graphics card, who made my NIC or who made my wireless system and I don't have to know anything about partitioning.
If you buy your Linux machines like you do your OS X machines, namely preinstalled, that's all you have to do: turn it on. The fact that Linux also installs on other PCs is an added bonus.
And the last Linux install I did required me trying 20 different drivers for my NIC since the ones that were supposed to work with my card was failing with some obscure error that a web search turned no hits on.
Well, then stop being stupid and buy your Linux machines preinstalled, like you do your OS X machines.
I'm tired of Linux fanboys like you spreading uninformed bullshit about OS X
We Linux fanboys don't generally give a damn about OS X. We don't tell OS X users to switch to Linux. We think you guys are a lost cause. We'd just like you to return the favor. OK?
If a Linux system "just worked" then yeah, maybe they'd choose it over OS X or Windows.
When you buy Linux machines from a Linux vendor, they do "just work". In fact, even when you don't, something like Knoppix or Ubuntu will "just work" on most Windows PCs--just boot it--it's simpler than OS X.
The mere existence of multiple windowing systems means that applications will be targeted at one of them.
You mean like Carbon, Cocoa, Classic, Mozilla, MS Office, and Java on OS X? Yes, that sure wears on me. When will Apple finally manage to produce a consistent system? They haven't so far.
They just want to go to Edit -> Paste (the more savvy ones may know Control+V) and have it work.
Maybe they'd also like a consistent view of keybindings and the file system. Too bad OS X doesn't provide either of those (Carbon and Cocoa handle those differently from one another). That's something that bothers me a great deal on OS X.
Thanks for demonstrating typical behaviors and attitudes of many Macintosh users. Your posting speaks for itself.
I already pay for it, through my workplace. Now I get to pay for it twice.
No, you don't. Your workplace retirement benefits are not guaranteed (even if they say they are). Millions of people have paid into insurance and workplace-related benefits and were left with nothing when they retired.
I've seen no valid argument for wealth redistribution yet.
Social security and similar government programs have nothing to do with "wealth redistribution"; they are there to guarantee you a minimum income in case all other precautions fail.
(Some wealth redistribution is essential for a free society, and everybody benefits from it, but that's a separate argument.)
Stop focusing on the structure of the operating system and START focusing on user productivity. There is a STEEEEEEEEP learning curve for the average user to comprehend much less USE linux on a day today basis.
I challenge you to cite studies that actually demonstrate that OS X is any easier to learn or use than Gnome, KDE, or Windows.
In our own experience with the platform, we find that users have about the same number of problems with OS X as with other platforms.
In fact, the idiosyncratic design of OS X (e.g., menu bar, single button mouse, admin tools, drag-and-drop installation, different visual appearance, etc.) makes for high retraining costs of existing users in our experience, and makes the platform difficult to use for people who need to use multiple platforms. In contrast, Linux desktops can be configured to look and behave nearly identically to their Windows counterparts, lowering retraining and support costs.
I dont think its so much understanding what Linux is...as it is that companies want to obtain technologies that get them where they need or want to be
And your point is what? That you think that OS X is that technology? I don't thinks so. Even leaving aside the question of which one is actually the better platform, the fact that Linux is nearly fully compatible with existing UNIX workstations and that it runs on existing PC hardware saves companies a ton of money. Their IT staff will already know how to run Linux machines, while OS X administration is rather different (Netinfo and all that). On OS X, X11 is a separate install, and its native window system and GUI is completely incompatible with anything else. Linux also has a much wider choice of mature software for remote administration, network administration, and other functions.
Lee built up a research lab for Microsoft in China and he is supposed to do the same for Google. That doesn't pose a "conflict of interest", nor should it be difficult for him to avoid disclosure of confidential information in his new position.
but it does show that most people get their work done more efficiently with them (I know I do).
That shows preference, not efficacy.
Could you give a few examples?
Remote logins, remote applications, preferences, and file system access, for example. Basically, Gnome and KDE are Windows-like desktops that happen to use X11 for graphics (and inefficiently at that).
I'm typing this on a laptop with Linux installed. What did I have to do to get it running? Stick in an Ubuntu Linux Install CD, boot, and confirm that I want to install it. After the install, it automatically connected to the nearest wireless network.
The install process was easier than a fresh install of OS X (which asks lots more annoying questions). And the last OS X upgrade I did on my iBook required me doing a web search for why it was failing with an obscure message, going into the command line, and patching up the HFS+ file system manually.
I'm tired of OS X fanboys like you spreading uninformed bullshit about Linux.
I think that story and those statistics are implausible, both given how big businesses operate and given Apple's actual sales figures.
Furthermore, the suggestion that OS X is an alternative to Linux indicates a lack of understanding of what Linux is all about and why it's being adopted.
"When you add things, it breaks?" That's one of the main problems with Windows: everytime you do anything (add h/w, s/w, ...) with Windows, you risk having to reinstall. Maybe Taylor got confused about which platform is which.
But what practical use is it good for?
What practical use is KDE? Has anybody actually ever demonstrated that KDE is a better tool for getting work done than emacs on a vt100? Or xedit under twm?
At this point, assertions about the utility of GUI features are based on faith, for any GUI. But assertions of non-innovativeness are not: neither KDE nor Gnome are innovative in any significant way.
And if it's good, why is no one using it?
Xt-based toolkits were very widely used and probably are still far more widely used commercially than either Gtk+ or Qt. They were by no means perfect, but they got a lot of things right.
That is indeed interesting, I had no idea this existed.
And that illustrates the core of the problem: KDE and Gnome have been developed with ignorance and disregard of prior work. It's hard to improve on something you don't know. And KDE and Gnome both actually break a lot of things that X11 users used to take for granted.
Explain welfare and medicaid, and that guy FICA who steals part of every one of my paychecks then.
Simple: it's so that you pay for the risk of you becoming indigent. If you weren't forced to pay for insurance against that, you wouldn't bother (as your comment indicates), and other people would have to assume the risk that you so foolishly take.
Sarah Brady is trying to force you to stop doing something that is bad for someone else.
while simultaneously believing YOU know what's best for everyone else.
See, the difference is that when liberals think they know what's best for someone, they don't force them to do it by law, even if they tell you it's bad for you.
When conservatives think they know what's best for someone (and they do as much as liberals), they'll try to pass laws to force them to do it.
What makes KDE 'slow' (note that I've had no problems running it on old hardware) is that it tries to do too much. Changing the language it's implemented in is not going to speed it up.
Of course it is. Implemented in a decent language, it wouldn't take a 10MByte process to put up a bunch of hotkeys, it would take a small function that takes up a few hundred bytes. And it would take more than a dozen processes just to get a simple desktop. Implemented in a decent language, I wouldn't have to download dozens of megabytes of source code and libraries just to change the behavior of a little applet, I'd click on whatever I want to change, get an editor, make the change, and go right along.
99.999% of the 'problems' that programming languages are claimed to have are the result of inexperienced, incompetent and down right crap programmers.
Yes, like the ones that think that "C++ is an excellent language for OO coding".
Thankfully increased competition in the job market has removed a large number of them from the industry.
Many competent programmers have left active programming in disgust because--what's the point? The tide of crap written in C/C++/Java just can't be stemmed. The best thing one can do is move up the food chain and hire people like you who apparently enjoy shoveling shit for a living.
First, he basically ripped this idea off from a previous chart built in 1951, modernized, gave it a better "UI" and is now shipping it out to the masses. Sound familiar?
No: companies universally give things they rip off from other companies worse UIs. I think it's because the people who do the ripping off don't spend much time thinking about the problem and prefer a simplistic solution to a good solution.
Would you like an OS/DE that lets everything integrate with everything? Then you might as well make that KDE (or GNOME, or whichever you prefer), and consider only apps that stick to KDE's rules. What am I missing?
The fact that mechanisms like KParts and DCOP are so cumbersome, opaque, and user-hostile that even experienced programmers don't bother using them.
Maybe I'm just not imaginative enough, but what kind of system are you envisioning? How could a desktop environment, no matter how innovative, let the user create buttons in a program that provide a functionality that the programmer hasn't provided for?
Have a look at Kay's Squeak: you can click on any UI element, move it around, replicate it, edit the corresponding code, script it, connect it to other elements, etc.
Heck, even editres for Xt was more flexible than what KDE is doing (you can still try it--it ships with X11, but toolkits like Qt and Gtk+ are too primitive and too non-standard to talk to it). The current crop of UIs have enormous amount of time invested in them in terms of theming, placement of buttons, etc., but they are technically worse than what we had 20 years ago.
Is that really innovative, given that you can do the same thing with "Setting -> configure toolbars"?
That only lets you add buttons that the programmer decided to let you have.
I don't know, but it's certainly not too far-fetched, thanks to KParts.
KParts only lets you embed those things that have explicit code for being embedded.
I assume you mean things like http://www.croquetproject.org/ ? That's certainly a more radical change from what we have now, but it's just not ready yet for everyday use.
It's a "radical change" in the same sense that eating healthy is a "radical change" from eating junk food. If it weren't for 20 years of junk GUI programming in C/C++ by Microsoft, Apple, Gnome, KDE, DEC, and HP, it wouldn't be a "radical change"--Alan Kay's current work is merely the evolution of ideas that go back to the 1970's.
The guy used to be at Apple's Advanced Technology Lab (back when Apple still had an Advanced Technology Lab) and at SGI. This is a bad move by Microsoft and something that's gonna make people of Lee's caliber think twice about working for them. They should be grateful about what the guy did for them and congratulate him on his new job.