Re:I remember that benchmark too.
on
Why FreeBSD
·
· Score: 1
In regards to the first point, the comments in the article that reveal that "PTHREAD_CONCURRENCY" was not set to 2 also reveal that at the time of publication, PTHREAD_CONCURRENCY was undocumented and even when building an SMP kernel PTHREAD_CONCURRENCY was set to 1. They mention that they were going to set it by default to the number of processors, but had to revert the change because of a regression test failure. Saying that the guy has "no clue what he is doing" is utter bullshit because the need to adjust PTHREAD_CONCURRENCY should be made loud and clear if it's that limiting of a factor. This was discussed specifically on the NetBSD mailing list:
------ It's worth mentioning that one needs to set $PTHREAD_CONCURRENCY environment to the number of CPUs to make it actually run on multiple CPUs on NetBSD 2.0. Especially because there is no existing common practice about this environment variable on other OS, and this may be a temporary measure. (If our implementation is mature enough, the pthread library
must set $PTHREAD_CONCURRENCY to the number of CPUs by default.) ------
Continuing along...
Of course SMP always has overhead. Saying that an operating system "scales linearly" means that to whatever line in the sand you draw, adding each extra processor results in a measurable and significant increase in performance. Which is why HP labs, that did the test, stated publically that Linux "scales linearly" to 64 processors on their Superdome.
Had they been running *bsd (do any of them even have NUMA support?), they would have stated that it does not "scale linearly", because in the case of BSD, the returns from adding additional processors would result in an exponential dropoff.
Please don't misinterpret the fact that I'm not a BSD fan to mean I'm a BSD hater. No doubt, many good things have come out of BSD. What I don't appreciate or like is BSD elitists who spend more time standing on razor-thin ice, throwing rocks at the Linux mansion, then they do working to improve one of the only other practical UNIX systems in existence today.
List of facts for clueless BSD users / users in denial:
1. Linux has a vastly greater development rate; 2. Linux has vastly more hardware support; 3. Linux is vastly more portable; 4. Linux is vastly more scalable; 5. All BSDs could have these things too if you could learn how to stop bickering amongst yourselves and others.
I remember my high school days when they had computers in the library. Suppose a student pops in one of these USB devices; they've just broken the system's security. Of course, they could have also done so by removing the chassis cover and resetting the BIOS, to get around the configuration password, allowing them to boot off their own media, mount and modify the hard drive. But chances are they might just get noticed.
Funny, I remember seeing some MySQL benchmarks a while back that touted three versions of FreeBSD (4.x with the normal threads, 4.x with LinuxThreads [hah], and 5.x) versus OpenBSD vs NetBSD vs Linux 2.4 and 2.6 vs Solaris. I believe that Linux did not come in first place on one single benchmark out of a batch, and when they went from 1 to 2 processors, *bsd fell on its fucking ass hardcore, with Solaris rocking and Linux rocking harder.
I'm seriously sick and tired of all the BSD whiners who drone on and on about how their OS is superior, and then when someone points out that they are in fact full of nothing but hot air, they whine and whine about how it's not their fault because Linux has better support. Perhaps if BSD politics didn't suck absolute balls, they wouldn't have lost that support and some of their key developers in the first place.
Then of course Linux gets bashed for things like RPM, as if RPM is a pill every Linux user must swallow. BSD has forked kernels, we have one kernel with occasional minor patch-level deviation and forks on the system level. It is therefore not surprising to me at all that Linux can do things like boot a 64-processor HP Superdome on a stock 2.6 kernel and scale linearly to the last processor.
Why fakeraid really sucks
on
Basics of RAID
·
· Score: 2, Insightful
My father has trusted his data (against my advice) to fakeraid chipsets on his various motherboards twice. He just got done *losing* all of his data for the second time.
Best we can tell, he had one drive go without his RAID controller warning him; then had a second drive go, killing the array. He spent weeks with a dead PC playing with all kinds of special Windows bootloaders and disk recovery tools trying to get his files back.
Fakeraid sucks because it's just a line item on the sale of a modern motherboard. The inclusion of the "RAID" functionality is borderline fradulent. REAL RAID controllers, of course, have a coprocessor and often battery backup and leave all of the storage details to themselves rather than some fly-by-night driver in the operating system.
It's no surprise that they come with virtually nothing in terms of recovery software.
The happy median, I've discovered, is Linux md. md supports many RAID levels, and according to some benchmarks will certainly outrun fakeraid in performance (which doesn't particularly surprise me). The administration tools let you simulate drive failure, monitor array health, create degraded arrays, and the documentation tells you what to do when something goes wrong.
Fortunately, you're totally wrong. I'm posting from KDE Konqueror right nwo, and I seem to be able to:
1. Read my fonts, which are aligned perfectly inside my widgets; 2. Witness consistent spacing between elements of the interface; 3. Use contrast to clearly see differences in the elements of my interface; 4. Witness excellent consistency not only in appearance but actual *functionality* between my KDE apps.
There is still the matter that a lot of X11 apps disagree about their desktop toolkit. Thankfully, the majority of the apps I use have powerful KDE versions, and gtk-qt-engine cleans up around the edges (though is certainly far from perfect).
I think it would be interesting for the KDE devs to implement some kind of preset system for the level of user experience... if you're a 'NOVICE' user, it hides most of the toolbar buttons and simplifies the interface.
But personally, I hate OS X and can't get any work done in it. I am a software developer, and so I maintain a very intimate relationship with the computer and lean a lot on all of the advanced features the modern GNU/Linux desktop offers me. Moving from Windows 2000 to Linux won me a 2x increase in productivity; moving from either to OS X is comically tragic.
Everything you have said is true. Of course, the problem is most prevalent in fucking Java software, which the KDE developers thankfully seem to be avoiding. (I remember a remark someone made about Eclipse... 512MB RAM = slow, 1GB RAM = fast... which is ASSINIE for a text editor).
I run KDE both at work and at home, on decent machines, and I find the performance to be excellent (certainly faster than Windows on either machine). I've grown to love the so called "bloat" - that is, it seems that any time I think to myself "I wonder if KDE will let me do Y" I find out that I'm not the first to have thought of it; it's actually implemented!
You cannot compare versions of Windows to Linux distributions -- even classes of Windows.
The point is that Microsoft is the single vendor of Windows. The "Microsoft Certified" solution is a Microsoft Windows flavor and Microsoft enterprise / datacenter software.
Linux does not certify anything - Apache, Sendmail, BIND, whatever. Linux is a kernel.
Thus, by comparing the two operating systems in this capacity, your only valid comparison is between two kernels, which this comparison is not of.
You're crossing into entirely new territory with this one.
1. As I stated in my last post, I'm fully aware of user stupidity. That does NOT invalidate the value of having good security systems.
2. I accept that you don't blame UNIX, but you go on to say that Windows is not responsible for its poor security. That's assinine. Windows, given that they cater to a particularly retarded genre of computer user, should not have its users dancing over razor blades and hot coals.
I could draft an e-mail to thousands that would tell them that switching from 110v to 220v on their power supply would result in a speed boost. I'm sure there would be plenty of gullible people who would flip that switch, but does that mean that it's just fine to make that switch a huge toggle switch, rather than a small switch behind a steel plate that requires utility and intention to move it?
I never asserted that Windows is impossible to secure, but if you want to start getting into this debate, might I remind you that the OSS model has a demonstrated advantage over closed source solutions. The reasons are vast and numerous, and I don't feel compelled to repeat the lecture here.
Microsoft sells a car that has cheap windows and shitty locks. If you're willing to spend the time (and often in their world, the money), you can upgrade the shitty locks and weld steel over the windows. Me, I'd feel better off with my free tank.
The only thing I hate more than Microsoft's mediocrity is its apologists who believe not only that it is in our nature as computer scientists to be imperfect, but that given our nature as humans we might as well not care at all.
You (and parents) are somewhat right about idiot users... You can't hand them a car they can't crash unless you take away their driving privileges. But if you put on skid control, good windsheild, ABS, etc, you lower the odds they'll fuck up.
Likewise, someone is always a handful of clicks and keystrokes from fucking themselves - all someone has to say is "start the terminal and run rm -rf/" in a convincing manner.
My point, however, is that UNIX mechanisms should not be discarded as useless defenses in the hands of clueless users.
My grandmother is one of the world's most clueless computer users - she even calls all the different programs 'My Microsoft'. But she knows better than to follow strangely complicated instructions in some e-mail she receives. That only helps so much in Windows because Windows will happily execute anything, and has no way other than extensions (which mailers honor) to know any better.
In any case, where my contribution began on comments regarding this subject, the parent post was retarded. Newsflash, people are gullible, and kids can be convinced to kill themselves by telling them to play with knives. Let's see how well the theory of PUTTING THE KNIVES AWAY AND NOT LEAVING THEM ON THE FLOOR holds up, and you're smoking weed if you think it's a worthwhile defense. (?)
Eh, my enterprise is running qmail, qmail-scanner, clam antivirus, spam assassin, pyzor and razor2 without much difficulty.
Admittedly, milter does seem to be a sane approach now that I read up on it. Given that Red Hat often funds OSS in the form of developer hours, they could have integrated it with a more secure mailer. Sendmail (and BIND) are just bad, bad, bad.
----------------
Please save the file to your hard drive and execute it from the command line. If you have problems executing it, please type "chmod +x filename.sh" and then execute it.
----------------
You're saying that this hypothetical gullible dingbat is going to be educated enough to bring up a terminal and execute it on the command line? Especially considering that running something on the command line involves adding./ unless you're braindead enough to have . in $PATH?
I would now like to take credit, on behalf of my open source breatheren, as the first "side" to have thought of the concept of the Keyboardsaver. KEuphoria on the keyboard would kick ass!
I'm talking about an article that discusses Microsoft Windows security, not your OEM computer security. My point stands - until the articles / studies specify exactly what they're comparing, they're totally meaningless.
Yes, but the reason these studies are absolute horseshit is because they call it 'LINUX' and fail to mention any of the choices. I could run apache, or I could run publicfile. You could choose apache or IIS for Windows.
Windows Server is a solution ENDORSED by Microsoft - they recommend IIS, SQL Server, and Exchange for your enterprise needs.
Apache isn't ENDORSED by Linus Torvalds, or lkml. It's ENDORSED by Red Hat Linux. So if you're going to make the comparison, compare a Linux server distribution (and specify WHICH ONE) to the Microsoft server product.
Yeah, as a seasoned Linux user, I patently refuse to take responsibility for the fucking disaster that is Sendmail, that STILL COMES INSTALLED BY DEFAULT ON RED HAT "ENTERPRISE LINUX".
Average user is too dumb to add execute permission to something. If all they do is use the software that is installed on the machine (or nice user interfaces to install more), then they rely on that software to do it for them.
If someone gets a message that tells them save this, type this command, type this command... the odds they actually take the time to do it are tremendously low unless it comes from a friend, and with good explanation.
The reason viruses and worms are so deadly on Windows is because the extension assigns something execute permission... all the user must do is click (which they are trained to do) as a result of their curiosity as to the nature of an attachment (which they are born with).
I hate these studies. Saying Linux isn't secure is like saying that fruit isn't red... it depends on what you're looking at. Are we talking about kernels? GNU tools? Common server software?
More importantly, which distribution? Windows comes with f*cking notepad and Solitaire. Linux distributions typically come with an order of magnitude more applications.
I'm on the Gentoo Security Mailing List. I get a few messages each day about vulnerabilities in software. Is each of these a ding on Linux? No, certainly not... it's a piece of software that happens to be available via portage.
If they want to be fair, then every ding on every Windows application counts against Windows.
More importantly, why the hell does every one of these boneheaded articles make it on the front page of Slashdot? Just helps spread the FUD.
Today, I sigh in pleasure as I type this message in KDE Konqueror. Glad my browser isn't vulerable to a kitchen full of exotic security holes; taste of the week style.
I, for one, was pleased to be a user of Gentoo Linux when this bug hit. I got notification via e-mail before the CERT report was even live, and after an emerge --sync and emerge zlib (the compiling of zlib took hardly any time at all), I ran lsof | grep libz and restarted the affected packages. Fixing my cluster at work took a few minutes, and I fixed my machines at home over SSH.
Perhaps you've used a particularly lucky combination of drivers, because I used Win2000 professional for quite some time and suffered from numerous BSODs.
Wonderful! So all the proprietary vendors will continue to port to OS X, which will now be on x86. They're *that* much closer to being able to run natively on Linux.
If anything, success OS X has will raise awareness of the possibility of non-Microsoft. And once they start to gain ground in the corporate world, Microsoft might just be forced to interoperate.
This issue reminds me of the Webcore / KHTML argument, and why I laughed at the idiocy of the Firefox guy who came down on the KDE developers for not swallowing the Webcore pill.
Open source does have some security advantages universally (think transparency), but transparency doesn't make something secure by default. Much of open source security comes from good design - a divorce from business deadlines and the concept of 'good enough'. I will always trust Konqueror over Internet Explorer, Safari, or Firefox for this reason.
Slashdot Mirror
In regards to the first point, the comments in the article that reveal that "PTHREAD_CONCURRENCY" was not set to 2 also reveal that at the time of publication, PTHREAD_CONCURRENCY was undocumented and even when building an SMP kernel PTHREAD_CONCURRENCY was set to 1. They mention that they were going to set it by default to the number of processors, but had to revert the change because of a regression test failure. Saying that the guy has "no clue what he is doing" is utter bullshit because the need to adjust PTHREAD_CONCURRENCY should be made loud and clear if it's that limiting of a factor. This was discussed specifically on the NetBSD mailing list:
------
It's worth mentioning that one needs to set $PTHREAD_CONCURRENCY
environment to the number of CPUs to make it actually run on multiple
CPUs on NetBSD 2.0. Especially because there is no existing common
practice about this environment variable on other OS, and this may be
a temporary measure.
(If our implementation is mature enough, the pthread library
must set $PTHREAD_CONCURRENCY to the number of CPUs by default.)
------
Continuing along...
Of course SMP always has overhead. Saying that an operating system "scales linearly" means that to whatever line in the sand you draw, adding each extra processor results in a measurable and significant increase in performance. Which is why HP labs, that did the test, stated publically that Linux "scales linearly" to 64 processors on their Superdome.
Had they been running *bsd (do any of them even have NUMA support?), they would have stated that it does not "scale linearly", because in the case of BSD, the returns from adding additional processors would result in an exponential dropoff.
Please don't misinterpret the fact that I'm not a BSD fan to mean I'm a BSD hater. No doubt, many good things have come out of BSD. What I don't appreciate or like is BSD elitists who spend more time standing on razor-thin ice, throwing rocks at the Linux mansion, then they do working to improve one of the only other practical UNIX systems in existence today.
List of facts for clueless BSD users / users in denial:
1. Linux has a vastly greater development rate;
2. Linux has vastly more hardware support;
3. Linux is vastly more portable;
4. Linux is vastly more scalable;
5. All BSDs could have these things too if you could learn how to stop bickering amongst yourselves and others.
I remember my high school days when they had computers in the library. Suppose a student pops in one of these USB devices; they've just broken the system's security. Of course, they could have also done so by removing the chassis cover and resetting the BIOS, to get around the configuration password, allowing them to boot off their own media, mount and modify the hard drive. But chances are they might just get noticed.
Funny, I remember seeing some MySQL benchmarks a while back that touted three versions of FreeBSD (4.x with the normal threads, 4.x with LinuxThreads [hah], and 5.x) versus OpenBSD vs NetBSD vs Linux 2.4 and 2.6 vs Solaris. I believe that Linux did not come in first place on one single benchmark out of a batch, and when they went from 1 to 2 processors, *bsd fell on its fucking ass hardcore, with Solaris rocking and Linux rocking harder.
I'm seriously sick and tired of all the BSD whiners who drone on and on about how their OS is superior, and then when someone points out that they are in fact full of nothing but hot air, they whine and whine about how it's not their fault because Linux has better support. Perhaps if BSD politics didn't suck absolute balls, they wouldn't have lost that support and some of their key developers in the first place.
Then of course Linux gets bashed for things like RPM, as if RPM is a pill every Linux user must swallow. BSD has forked kernels, we have one kernel with occasional minor patch-level deviation and forks on the system level. It is therefore not surprising to me at all that Linux can do things like boot a 64-processor HP Superdome on a stock 2.6 kernel and scale linearly to the last processor.
My father has trusted his data (against my advice) to fakeraid chipsets on his various motherboards twice. He just got done *losing* all of his data for the second time.
Best we can tell, he had one drive go without his RAID controller warning him; then had a second drive go, killing the array. He spent weeks with a dead PC playing with all kinds of special Windows bootloaders and disk recovery tools trying to get his files back.
Fakeraid sucks because it's just a line item on the sale of a modern motherboard. The inclusion of the "RAID" functionality is borderline fradulent. REAL RAID controllers, of course, have a coprocessor and often battery backup and leave all of the storage details to themselves rather than some fly-by-night driver in the operating system.
It's no surprise that they come with virtually nothing in terms of recovery software.
The happy median, I've discovered, is Linux md. md supports many RAID levels, and according to some benchmarks will certainly outrun fakeraid in performance (which doesn't particularly surprise me). The administration tools let you simulate drive failure, monitor array health, create degraded arrays, and the documentation tells you what to do when something goes wrong.
Fortunately, you're totally wrong. I'm posting from KDE Konqueror right nwo, and I seem to be able to:
1. Read my fonts, which are aligned perfectly inside my widgets;
2. Witness consistent spacing between elements of the interface;
3. Use contrast to clearly see differences in the elements of my interface;
4. Witness excellent consistency not only in appearance but actual *functionality* between my KDE apps.
There is still the matter that a lot of X11 apps disagree about their desktop toolkit. Thankfully, the majority of the apps I use have powerful KDE versions, and gtk-qt-engine cleans up around the edges (though is certainly far from perfect).
I think it would be interesting for the KDE devs to implement some kind of preset system for the level of user experience... if you're a 'NOVICE' user, it hides most of the toolbar buttons and simplifies the interface. But personally, I hate OS X and can't get any work done in it. I am a software developer, and so I maintain a very intimate relationship with the computer and lean a lot on all of the advanced features the modern GNU/Linux desktop offers me. Moving from Windows 2000 to Linux won me a 2x increase in productivity; moving from either to OS X is comically tragic.
Everything you have said is true. Of course, the problem is most prevalent in fucking Java software, which the KDE developers thankfully seem to be avoiding. (I remember a remark someone made about Eclipse... 512MB RAM = slow, 1GB RAM = fast... which is ASSINIE for a text editor).
I run KDE both at work and at home, on decent machines, and I find the performance to be excellent (certainly faster than Windows on either machine). I've grown to love the so called "bloat" - that is, it seems that any time I think to myself "I wonder if KDE will let me do Y" I find out that I'm not the first to have thought of it; it's actually implemented!
You cannot compare versions of Windows to Linux distributions -- even classes of Windows.
The point is that Microsoft is the single vendor of Windows. The "Microsoft Certified" solution is a Microsoft Windows flavor and Microsoft enterprise / datacenter software.
Linux does not certify anything - Apache, Sendmail, BIND, whatever. Linux is a kernel.
Thus, by comparing the two operating systems in this capacity, your only valid comparison is between two kernels, which this comparison is not of.
You're crossing into entirely new territory with this one.
1. As I stated in my last post, I'm fully aware of user stupidity. That does NOT invalidate the value of having good security systems.
2. I accept that you don't blame UNIX, but you go on to say that Windows is not responsible for its poor security. That's assinine. Windows, given that they cater to a particularly retarded genre of computer user, should not have its users dancing over razor blades and hot coals.
I could draft an e-mail to thousands that would tell them that switching from 110v to 220v on their power supply would result in a speed boost. I'm sure there would be plenty of gullible people who would flip that switch, but does that mean that it's just fine to make that switch a huge toggle switch, rather than a small switch behind a steel plate that requires utility and intention to move it?
I never asserted that Windows is impossible to secure, but if you want to start getting into this debate, might I remind you that the OSS model has a demonstrated advantage over closed source solutions. The reasons are vast and numerous, and I don't feel compelled to repeat the lecture here.
Microsoft sells a car that has cheap windows and shitty locks. If you're willing to spend the time (and often in their world, the money), you can upgrade the shitty locks and weld steel over the windows. Me, I'd feel better off with my free tank.
The only thing I hate more than Microsoft's mediocrity is its apologists who believe not only that it is in our nature as computer scientists to be imperfect, but that given our nature as humans we might as well not care at all.
It's spelled libel, and slander and libel are civil matters, not criminal ones.
But I would like to see him tried for fraud.
You (and parents) are somewhat right about idiot users... You can't hand them a car they can't crash unless you take away their driving privileges. But if you put on skid control, good windsheild, ABS, etc, you lower the odds they'll fuck up.
/" in a convincing manner.
Likewise, someone is always a handful of clicks and keystrokes from fucking themselves - all someone has to say is "start the terminal and run rm -rf
My point, however, is that UNIX mechanisms should not be discarded as useless defenses in the hands of clueless users.
My grandmother is one of the world's most clueless computer users - she even calls all the different programs 'My Microsoft'. But she knows better than to follow strangely complicated instructions in some e-mail she receives. That only helps so much in Windows because Windows will happily execute anything, and has no way other than extensions (which mailers honor) to know any better.
In any case, where my contribution began on comments regarding this subject, the parent post was retarded. Newsflash, people are gullible, and kids can be convinced to kill themselves by telling them to play with knives. Let's see how well the theory of PUTTING THE KNIVES AWAY AND NOT LEAVING THEM ON THE FLOOR holds up, and you're smoking weed if you think it's a worthwhile defense. (?)
Eh, my enterprise is running qmail, qmail-scanner, clam antivirus, spam assassin, pyzor and razor2 without much difficulty. Admittedly, milter does seem to be a sane approach now that I read up on it. Given that Red Hat often funds OSS in the form of developer hours, they could have integrated it with a more secure mailer. Sendmail (and BIND) are just bad, bad, bad.
---------------- Please save the file to your hard drive and execute it from the command line. If you have problems executing it, please type "chmod +x filename.sh" and then execute it. ---------------- You're saying that this hypothetical gullible dingbat is going to be educated enough to bring up a terminal and execute it on the command line? Especially considering that running something on the command line involves adding ./ unless you're braindead enough to have . in $PATH?
I would now like to take credit, on behalf of my open source breatheren, as the first "side" to have thought of the concept of the Keyboardsaver. KEuphoria on the keyboard would kick ass!
I'm talking about an article that discusses Microsoft Windows security, not your OEM computer security. My point stands - until the articles / studies specify exactly what they're comparing, they're totally meaningless.
Yes, but the reason these studies are absolute horseshit is because they call it 'LINUX' and fail to mention any of the choices. I could run apache, or I could run publicfile. You could choose apache or IIS for Windows.
Windows Server is a solution ENDORSED by Microsoft - they recommend IIS, SQL Server, and Exchange for your enterprise needs.
Apache isn't ENDORSED by Linus Torvalds, or lkml. It's ENDORSED by Red Hat Linux. So if you're going to make the comparison, compare a Linux server distribution (and specify WHICH ONE) to the Microsoft server product.
Yeah, as a seasoned Linux user, I patently refuse to take responsibility for the fucking disaster that is Sendmail, that STILL COMES INSTALLED BY DEFAULT ON RED HAT "ENTERPRISE LINUX".
Average user is too dumb to add execute permission to something. If all they do is use the software that is installed on the machine (or nice user interfaces to install more), then they rely on that software to do it for them. If someone gets a message that tells them save this, type this command, type this command... the odds they actually take the time to do it are tremendously low unless it comes from a friend, and with good explanation. The reason viruses and worms are so deadly on Windows is because the extension assigns something execute permission... all the user must do is click (which they are trained to do) as a result of their curiosity as to the nature of an attachment (which they are born with).
You're right. I keep my family's Windows machines safe, by putting them behind my iptables Linux firewall.
I hate these studies. Saying Linux isn't secure is like saying that fruit isn't red... it depends on what you're looking at. Are we talking about kernels? GNU tools? Common server software?
More importantly, which distribution? Windows comes with f*cking notepad and Solitaire. Linux distributions typically come with an order of magnitude more applications.
I'm on the Gentoo Security Mailing List. I get a few messages each day about vulnerabilities in software. Is each of these a ding on Linux? No, certainly not... it's a piece of software that happens to be available via portage.
If they want to be fair, then every ding on every Windows application counts against Windows.
More importantly, why the hell does every one of these boneheaded articles make it on the front page of Slashdot? Just helps spread the FUD.
Today, I sigh in pleasure as I type this message in KDE Konqueror. Glad my browser isn't vulerable to a kitchen full of exotic security holes; taste of the week style.
I, for one, was pleased to be a user of Gentoo Linux when this bug hit. I got notification via e-mail before the CERT report was even live, and after an emerge --sync and emerge zlib (the compiling of zlib took hardly any time at all), I ran lsof | grep libz and restarted the affected packages. Fixing my cluster at work took a few minutes, and I fixed my machines at home over SSH.
Perhaps you've used a particularly lucky combination of drivers, because I used Win2000 professional for quite some time and suffered from numerous BSODs.
Wonderful! So all the proprietary vendors will continue to port to OS X, which will now be on x86. They're *that* much closer to being able to run natively on Linux.
If anything, success OS X has will raise awareness of the possibility of non-Microsoft. And once they start to gain ground in the corporate world, Microsoft might just be forced to interoperate.
This issue reminds me of the Webcore / KHTML argument, and why I laughed at the idiocy of the Firefox guy who came down on the KDE developers for not swallowing the Webcore pill. Open source does have some security advantages universally (think transparency), but transparency doesn't make something secure by default. Much of open source security comes from good design - a divorce from business deadlines and the concept of 'good enough'. I will always trust Konqueror over Internet Explorer, Safari, or Firefox for this reason.