Microsoft Corp. is hard at work to make “Longhorn,” the next iteration of its Windows Server product, do more for less by integrating various server software systems. To do it, the company plans to extend .Net’s capabilities even further with common execution environments and complete .Net Common Language Runtime support, said Valerie Olague, a Windows Server System director, in an interview.
Microsoft Building Integration, Management into Longhorn Server
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I thought that Longhorn was only going to be a client release and that Blackcomb was to be the server. Anyway… This whole push towards having all apps run as .net managed code is great, I guess, but since I cannot even access a COM port without creating a .net wrapper around win32, how well is it going to work using .net to replace win32?
I thought that Longhorn was only going to be a client release and that Blackcomb was to be the server. Anyway… This whole push towards having all apps run as .net managed code is great, I guess, but since I cannot even access a COM port without creating a .net wrapper around win32, how well is it going to work using .net to replace win32?
Aye? what are you going on about? it will use a method similar to how Java interacts with a USB port, via a USB interface class.
“extend .Net’s capabilities even further with … complete .Net Common Language Runtime support”
What is meant by this? Does not .Net already by definition have complete .Net CLR support?
If the topic is beyond your intellect just don’t reply!
Microsoft promises the world with it’s operating systems but only delievers a few continents. Many of the things that you hear are vaporware, especially with a product that isn’t due out in another 2 – 3 years, a lot can happen in that time. Don’t get your hopes up.
Sorry, you are wrong. This isn’t some open source project where the developers only work when they run out of cheetohs. Microsoft pays actual employees to work, over 5000 of them in fact.
>Sorry, you are wrong. This isn’t some open source project >where the developers only work when they run out of cheetohs. >Microsoft pays actual employees to work, over 5000 of them in >fact.
No wonder IE and windoze is such a solid, robust, no BSOD product.
The main advantage of windows is that most people already have all their data in Microsoft formats, your employees know how to use it, it have bin relatively inexpensive.
If Microsoft makes too much changes, people will start evaluate other solutions. Tying into the server side too heavily may increase the cost of upgrades. Major changes in GUI gives more training costs. The new security technologies based on crypto technology is good, but that too increases the cost. It also have the problem of introducing a single point of failure. And as the article spells out higher integration between desktop and server makes it harder and more expensive to integrate legacy systems.
One of the alterantives IT managers will look at is opensource systems like Linux. The “Get the facts” ads are a sign that Microsoft takes this competition very seriously. Right now it is relatively easy to beat Linux on speed and usability. But that is going to change.
Today they compare brand new windows technology with Linux based on an old soon to be phased out Linux kernel. And my guess is that if we make the “right” tests the “facts” could be very different even if you compare this old Linux with brand new windows technology. And no doubt, the Linux community will provide such tests.
The Linux community is no longer just Linus Torvalds and his friends doing hacks on their spare time, but often employees of large companies like IBM, Novell,… Large companies that both have the power to make Linux a truly better system, and to sponsor studies of their own, showing the facts just as “independently” as the Microsoft campaign.
Another factor that could be problematic for Microsoft is that if you look at the increase of user value of a specific development effort you will find that this value is lower if you are near the end of life of a product (as they are with their winNT/win2k/XP series). In fact this is why you start something new (like Longhorn/WinFS)
But if you start something new, you will have to establish new types of infrastructure that gives little immediate increase in user value. This is where .net is today.
That leaves Microsoft in the wrong end of the value/effort curve for both new and old technology. They have bin so certain that there would be no competing technology to take their place so they have postponed development of new things to the last possible moment.
The Linux community on the other hand have a 30 year old infrastructure that oddly enough still works well, and they have not reached the state where some more polishing costs you arms and legs. (E.g. you could quite easily make a much more useable system out of KDE, merely by changing
the defaults.)
This battle will be fun to watch!
What’s your point? Were you expecting rock solid software? Show me any rock solid software that is as complex as Internet Explorer and windows then I will appreciate your comment, unless of course you are not from the idiots camp. So far these two software are used by more users than any other software.
Proving that piece of software is not good is not as simple as stating that there are problems with the software. You have to come up with an example which is as complex as the software you are questioning and then you have to show that your example is much better than the software you are critizing. Ohterwise, making simple stupid statements is usually a way to prove your idiocy, not the idiocy of the software. I don’t see any other software which is as complex as these two software and yet significantly better. In fact overall I think windows is the best so far. Internet Explorer is also the best when considering all the issues, but Firebird as a “browser only” is better.
Longhorn is quite dangerous for the Linux. These guys are building things that we can only dream of on Linux. They are changing lots of stuff now. You can not spend one year programming a software and then claim that you are competing with Windows. It is not as simple as that anymore. Well, certainly we will still see some idiots claiming that Microsoft sucks, but nobody cares about them, as the market share clearly shows.
First of all, Microsoft is moving to .net. This means higher productivity for Microsoft developers. To achieve a similar productivity on Linux we need something like .net, maybe jvm. But we need a much better integrated solution. You should safely expect to have that platform in all linux desktops out there. Then you need to change kde codebase etc… so that they are written in managed code. Many services such as search has to be built into the os. More importantly, we need APIs that we can use for our applications running on Linux. Even today, commercial companies do not produce much software for Linux. Linux not only has to solve the existing API problems, but also has to quickly add new APIs that match the ones found in the windows. There are some technologies that can only be donated by a company, such as IBM. Speech technology is an example.
Clearly these problems do not worry many idiots claiming that Microsoft sucks. They have no idea, what it takes to write software or develop APIs. Linux is already behind Windows, I don’t know how many years Longhorn will put Linux behind. On the other hand, I am sure Linux on server will be a big competitive for the Longhorn.
A secure integrated environment? A contradiction in terms?
No, you might not be familar with it if you have used Linux for too long, though. Think BeOS, Mac OS X, basically every other commercial operating system.
Wow, some people really hate things they haven’t tried here. Quite a lot of flame bait too, so I thought I’d even up the windows vs the world debate.
“Linux is already behind Windows, I don’t know how many years Longhorn will put Linux behind”
Let’s see, Linux has been working on 64 bit processors for a while now, ahem, Windows is already behind Linux here. Plus Linux works on many other architectures, like Macs for instance, Windows is limited to the intel x86 architecture and probably didn’t even bother working on a 64-bit version till Intel decided it was a good idea to play catch up with AMD.
“They have no idea, what it takes to write software or develop APIs”
Ever had to program in a DLL environment? Check out the Java API and the tool javadoc if you want to see how it’s done properly.
To quote a good point: “Windows’ API is only partially documented. Much of the operating system and the API remain essentially a black box to third-party developers. This causes problems during application development that often cannot be traced. Developers are forced to use workarounds, and may even be tempted to circumvent or ignore the API altogether. This, and the fact that nobody outside Microsoft really knows what goes on in the underlying code, leads to application software that won’t run correctly (if at all) on updated or new versions of Windows.”
.NET is just a way to tie developers to Windows, Java works on most OSes and is a far better developing environment due to it’s portability and well developed class libraries.
“You can not spend one year programming a software and then claim that you are competing with Windows”
Err… yeah sure, Linux kernel was started in 1991 and the GNU project started in the 80s.
“In fact overall I think windows is the best so far.”
uhuh, check out http://www.sophos.com to see how stable it is, btw is your mouse wheel still working properly? An update was released recently that crippled it I heard.
Also, if you want to see something really enlightening check out:
http://news.netcraft.com/archives/2004/01/01/january_2004_web_serve…
and see how many people use Windows online…you may notice the the 67% market share by Linux/BSD boxes, I think MS is around 21%.
“Internet Explorer is also the best when considering all the issues, but Firebird as a “browser only” is better. ”
Ok, this one’s too easy. Let’s compare IE to a browser that’s had *real* improvements since 1998. Opera has mouse gestures which are very useful though I feel Mozilla has teh crown atm. Mozilla has “tabbed browsing” so windows open inside the application (as opposed to opening new windows which require around 30mb or RAM per window) and also has a fantastic pop-up blocker (btw, check the links at the bottom of this page if you want to see what the profesionals use to browse the web). IE needs to be patched to hell to get anywhere near the features of Mozilla etc.
Though I’m not sure what you mean by the “best all round” comment, best all round for being messed over maybe? By the way watch out for stuff like:
http://ibank.barclays.co.uk%01%01%01%01%01%…
%01%01%01%01%01%01%01%01%01%01 %01%01%01%01%01%01%01%01%01%01 %01
%01%01%01%01%01%01%01%01%01%01 %01%01%01%01%01%01%01%01%01%01 %01
%01%01%01%01%01%01%01%01%01%01 %01%01%01%01%01%01%01%01%01%01 %01
%01%01%01%01%01%01%01%01%01%01 @%77%77%77%2E%6E%65%77%79%65%7 2
%73%6D%2E%63%6F%6D:%38%30/%31% 2C%2C%6C%6F%67%6F%6E%2C%30%30% 2E
%70%68%70
As IE can be exploited to fake URLs, a fix will be available in your monthly “IE’s broken” update (though hopefully a fix is out already).
“Right now it is relatively easy to beat Linux on speed and usability.”
Really? Maybe so on usability due to many users already being adjusted to the Windows interface, though MacOS has easily got the most “user-friendly” interface as it’s error-checking is top notch, KDE/Gnome still have a way to go before they are totally “User-Friendly” to the Mac standards, though it’s about par with Windows atm. Though doing a test on this is quite difficult. For speed we could do tests on my 64 bit processor and your 32bit one yeah?
“Sorry, you are wrong. This isn’t some open source project where the developers only work when they run out of cheetohs. Microsoft pays actual employees to work, over 5000 of them in fact.”
Ignoring the fact that you pulled that statistic out of your ass, I’m sure Microsoft have many more employees that number. Major developers like Sun and IBM are creating many programs that are made by coders also (see OpenOffice which runs on java so works on many OSes as oppsoed to 1 like .NET). KDE is also a German company making the Desktop Environment, not some ad-hoc developer circle.
“(E.g. you could quite easily make a much more useable system out of KDE, merely by changing
the defaults.) ”
See Red Hat.
“Today they compare brand new windows technology with Linux based on an old soon to be phased out Linux kernel. And my guess is that if we make the “right” tests the “facts” could be very different even if you compare this old Linux with brand new windows technology. And no doubt, the Linux community will provide such tests. ”
Perhaps I’ll test it on my 64bit processor I’m using now? Also the 2.6 kernel rocks and you can check out the source code at http://www.kernel.org if you want. I don’t see how we can really compare .NET to anything else though, it’s not like we can test it’s performance on any other operating system is it? Though I have heard of a project called “mono” which is attempting to port .NET apps to linux (I hear they’ve managed to get ASP.NET over quite well).
So yeah, .NET is a good idea, though it’s being setup by the wrong people and is destined to go the same way as any other MS project and you can bet your left testicle it’s applications will be ripe to exploits. Personally I’ll stick with Java if I want a language which works by converting to byte-code. As said before, this is just management speak atm, they will still be in the design stages for this product and we will not be able to judge it till it’s released on the beta-testing public.
Nice post Uno. One of the most level-headed, non-partisan things I’ve read about MS vs. Linux in a long time.
Yes, clue (get one).
First, Windows is available for 64 bit processors.
http://www.microsoft.com/WindowsXP/64bit/default.asp
Second, just because Linux can be compiled and run on 64 bit systems, does not mean it is production ready. Nor does support for two architectures mean that it is more advanced than any other operating system. If we play by that book, I guess NetBSD is the most advanced operating system in the world, and we should all be using it.
Also, your link does not work. Both because of osnews’ parsing, and the fact that the patch was released two weeks ago. I, for one, am glad that Microsoft takes the time to test their patches. I always hear that open source software has patches released within hours of vulnerabilities being found. It’s doubtful that any regression tests are done beyond “yup it compiles, grab the cheetos”, let alone QA or a distributed testing. No wonder development kernels fry hardware, and production kernels have remote root exploits.
Fair enough,
I actually use FreeBSD mainly at home, I was just a bit worried by the comments made without people actually trying things out first. I don’t think that if one kernel can compile on more machines (I wouldn’t be surprised if NetBSD compiled on a toaster soon) then it’s more superior I just believe in open standards which .NET and Windows do not promote, and only working on 1 architecture doesn’t really highlight Windows good points. When considering every factor to decide what is “best, fastest, most secure” most OSes turn out to be even and they all have their advantages and disadvantages. Mainly what’s best depends on the experience of the user.
“Trial Software Now Available
If you have a 64-bit ready PC, you now have the option to receive trial software for Windows XP 64-Bit Edition for 64-Bit Extended Systems via CD or download.”
My 64 kernel works fine and the linux stable kernel development is going well. That windows 64 bit is actually for testers (has anyone here tried it? I’d be interested to hear how it performs) as far as I can see, though personally I think there is something suspicious about Windows 64 being released coinciding with the official announcement of Intel releasing their 64 bit offering next month.
“Microsoft takes the time to test their patches”
What about that bug that affected the mouse wheel?
If you want an example of response to bugs check out GAIM which recently had a remote overflow vun bug due to a rushed release to comply with the new AIM protocol, a fix was released for the cvs version on the same day the exploit went public:
http://www.securityfocus.com/archive/1/351235
Joejag, I know java more than I know .net. I know it inside out. Not just programming in java, I mean all the details, class structure, bytecodes, jvm, garbage collection. I had the chance to study the jvm source code of several jvms. But look at your statements, they are all political.
.net is produced to tie developers to windows.
a fix will be available in your monthly “IE’s broken” update (though hopefully a fix is out already).
IE needs to be patched to hell to get anywhere near the features of Mozilla etc
check the links at the bottom of this page if you want to see what the profesionals use to browse the web
Again and again, you get everything all wrong. You are extremely simple minded even to discuss with some intuition. You know how to use Linux or some few open source stuff and you think you are a genius. Believe me you are no where near being a genius.
Let’s first state something you got it all wrong. I use Firebird. Now just one sentence puts you into a very awkard position. You assume that I was using IE, however I don’t. So let’s think for a moment. You can’t get even this right, your whole arguments are full of false assumptions, madeup stories, stupid attacks. You don’t have any technical knowledge to offer us. You are simply saying Microsoft sucks and repeat that to us. You attack other Linux and Mozilla users, not their arguments. Obviously you come from what I call as “idiots camp”. In this camp, people can favor Microsoft or Linux, doesn’t matter, but they don’t make sense. Unfortunately looks like we have too many people from Linux in this camp, but I am hoping that we are going to change it and make Linux a more serious OS. Right now, people scream, rant, but at the end Linux is still small and many of those people in this camp do not care about Linux or Mozilla. They are ranting because it makes them feel good about themselves.
Another stupid comment was that I didn’t try using Linux. I have been using Linux more than 6 years now. You think by using Linux, I will automatically start bashing Microsoft and say stupid things like .net is for tying developers to Microsoft. Just because I use Linux or Firebird doesn’t mean that I think you are free to bs here. I don’t think that you have any right to lie in public so openly. If someone from MS bs’es about Linux, I am not going to accept that. Similarly I am not going to accept so many lies from you guys. You are going to lie about Linux and Windows anyway, whether you are pro-Linux or pro-MS or whatever you are. It is not that you are pro-Linux you are lying. You are saying these stupid things, because of your personality.
DLL hell, yes I heard about it. But only a moron will compare the dll hell with java. Why? Java is not an OS platform where millions of developers are developing client programs for. Linux is, but java is not. On Linux there is something similar to dll hell and it is called dependency hell. Have you ever tried to compile things which do not exist in the distribution yet. When KDE comes up with a new version, I have to wait debian guys to use it reliably. You try to install KDE, you have to dig internet to figure out how to install it on your debian without breaking it. Installing a new program shouldn’t be tricky. Some idiots will claim that you just don’t know how to do it, but we all know that these people are simply idiots. There is no way you can claim that people should waste their time to learn how to install a new program and learn more about the libraries, how to compile things and so on.
Anyway, the arguments presented by the people in the idiots camp unfortunately just points out that Linux is quite unlikely to replace Windows on the desktop. I am hoping to see people with more knowledge to show us exactly what is the plan and how are we going to address the current problems. I have read Miguel Icaza’s comments and it seems that he is begging to developers to make APIs more stable and unite behind some strict standards so that third party developers can write code easily. Unlike people from the idiots camp he is quite direct and he talks about the issues. You can never see him bashing Microsoft like many idiots do. Serious people do not do that. We need to force idiots to realize that, they shouldn’t have the freedom to abuse Linux to cure their own personal psychological problems. If they really love Linux, they should talk about the problems themselves. Microsoft is wayyyyyyyy ahead of us. Many serious guys on the Linux admit that. People without any responsibility will rant on the public forums, will say all sorts of stupid things, but remember these people have nothing to lose. Check out what HP’s linux business head says about Linux on desktop. Also check out why Redhat dropped linux on desktop.
Jimbo –
“…but since I cannot even access a COM port without creating a .net wrapper around win32, how well is it going to work using .net to replace win32?”
Microsoft is actually working to reduce the tens of thousands of Win32 calls into about 8,000 .NET ones.
Hopefully, you’ll be able to get at quite a bit of the guts through managed code by the time Longhorn comes out.
At the moment, though, it’s true – doing something as simple as accessing a serial port is a pain in the butt. That *will* change.
One of the most idiotic complains the Linux community uses against Windows is the amount of virus designed for MS Technologies. Let’s say that Microsoft explodes and suddenly all computers in the world uses some version of Linux. You know what will happen then? All the people making viruses for Windows will begin coding them for Linux. Better security you say? Even if that’s true, my answer would be ‘Trojans’. Most computers get infected because the unsuspected user (the vast majority of the users) download some program and it is full of spyware and Trojans.
On one post I read something about MS not giving enough information about the WinAPI. Only an idiot would say that MS is shy about giving information to its developers. No other development platform has the amount of information like the MSDN. The only perk is that the information has to stop somewhere. It’s not open source after all. They have to draw the line somewhere and say “this is as far as we can let you know. Here are the inputs and outputs. We take the inputs and do this and that but we can’t tell you HOW we do it.” That kind of information is only useful if you’re trying to modify the OS itself, not to make applications.
My 2cp
“Also, your link does not work. Both because of osnews’ parsing, and the fact that the patch was released two weeks ago. I, for one, am glad that Microsoft takes the time to test their patches. I always hear that open source software has patches released within hours of vulnerabilities being found.”
Well, they actually do. But as you say at that stage they are seldom extensively tested, how could they be. Some users test themselves, others leave the testing to Linux distributers that do the same job as Microsoft does. As a result those checked patches are somewhat delayed. Still they are usually faster to market than Microsoft. The question is what is best. To do like Microsoft, and not provide any solution at all and leave the bug open for quite some time, or make it possible for advanced users or even consultants to get access to the patches early.
“It’s doubtful that any regression tests are done beyond “yup it compiles, grab the cheetos”, let alone QA or a distributed testing. No wonder development kernels fry hardware, and production kernels have remote root exploits.”
I doubt that Microsoft tells us when any of their development kernels fries hardware and I do think that there are remote exploits to windows as well.
If you want high security you could always turn on mandatory access control (Standard feature of Linux 2.6). That way you could limit the damages of an exploit to an application.
Your 2nd post seemed a lot more level headed then your first that just seemed like a tirade against anything Linux, and I agree my first post did have a dose of anti-MS venom (people were troll baiting c’mon).
I found it quite surprising that you think IE was the best browser when you’ve discovered FireBird (or any other browser) already, I usually hear this argument from people who tried something new once and decided it wasn’t exactly the same as before so they went back. I apologise for making this assumption.
I used to use Debian and I know of the Depedency hell you speak of, it was because of that I checked out the Gentoo project, which has a lovely system based on BSD ports which sorts out dependencies much cleaner then apt-get did when I was using it. By the way, installing KDE was simply a matter of typing “emerge kde” and that was it (and a 9 hour wait since I was doing it from source!).
I don’t believe that my “whole arguments are full of false assumptions, madeup stories, stupid attacks” though if I said anything like this please discuss (so I could also learn why I am wrong) that rather then labelling me in the “idiot’s camp”.
“You attack other Linux and Mozilla users, not their arguments.”
I don’t remember doing that. I just was trying to promote the idea that it’s not quite that easy to say “this is the best”. MacOS has clearly done what Linux hasn’t done on the desktop by taking FreeBSD and putting a user friendly GUI on top of it, and kudos for them for showing everyone how it’s done. It would be very interesting if they released a x86 port.
.NET is a great idea and it will make programming for Windows a lot easier, c# is very similar to Java in structure and I managed to pick it up quite easily for my Uni course due to it’s OO nature. The sentiments I have about the .NET philosophy were summed up by IBM when the .NET ball starting rolling which you can read at: http://www.zdnet.com.au/news/business/0,39023166,20221630,00.htm
Ah yeah, and I had an IQ test, I’m not a genius, just a “normal” person, I have no delusions to this. I’ve met a Debain developer too, some people can easily put you in your place.
Another factor that could be problematic for Microsoft is that if you look at the increase of user value of a specific development effort you will find that this value is lower if you are near the end of life of a product (as they are with their winNT/win2k/XP series). In fact this is why you start something new (like Longhorn/WinFS)
Longhorn is the continuing development of NT. It has some significant changes, to be sure, but it’s not a ground-up from-scratch new product (like NT was). Similarly, WinFS is not replacing existing technology, it’s building on top of it. It’s basically just a pervasive, integrated “File Search” tool.
Let’s see, Linux has been working on 64 bit processors for a while now, ahem, Windows is already behind Linux here.
For a sufficiently loose definition of “working”. How many distros that are feature comparable with their 32 bit brethren (and Windows) are available _and supported_ on 64 bit hardware ? How long have they been available ?
A 64 bit version of Windows has been around for a while. It’s not a finished product but, then again, 64 bit hardware is a fairly recent entrant to the market Windows is aimed at, so that’s not surprising.
Plus Linux works on many other architectures, like Macs for instance, Windows is limited to the intel x86 architecture and probably didn’t even bother working on a 64-bit version till Intel decided it was a good idea to play catch up with AMD.
Once again, for a sufficiently loose definition of “works”.
Also, not “limited”, “available on”. NT is very portable. NT4 (and earlier versions) was publically available for four platforms until ~1998ish, Win2k was available on x86 and Alpha until well into the beta cycle and was been ported internally to several other platforms as well. You’d be very naive to think NT development isn’t parallelled on at _least_ two (very) different platforms (certainly, if it’s still like the early days, x86 is the platform NT is ported to, not developed on). Similarly, 64 bit development for NT has probably been going on for several years.
I also find it ironic you say intel are playing “catch up” when they are the ones who made a *new* 64 bit architecture while AMD just hacked some extensions onto intel’s x86 design.
Microsoft are a *company*. They have to justify spending money on things like alternative hardware platform support by *making money*. Now, what alternative platforms are you thinking Microsoft can target where they’ll actually have a reasonable chance at recovering their costs ? Most of them, for all practical purposes, are proprietry and/or single sourced by a seller who also produces the Operating System and/or offer SFA advantages in terms of price/performance over intel.
Cross-platform availability and support is nice for appealing to the 0.000001% of consumers who either require it, or who are swayed by its “coolness”, but from a business perspective, it’s a disaster.
[…]
My 64 kernel works fine and the linux stable kernel development is going well. That windows 64 bit is actually for testers (has anyone here tried it? I’d be interested to hear how it performs) as far as I can see, though personally I think there is something suspicious about Windows 64 being released coinciding with the official announcement of Intel releasing their 64 bit offering next month.
Yes, particularly since that 64bit version of Windows is specifically for AMD’s 64 bit CPU, Microsoft have stated several times they will only target *one* x86-64 platform (AMD’s), and it’s been a well known about program for quite some time. Very suspicious indeed. Undoubtedly the whole thing is being manipulated behind the scenes by intel to garner some sort of sympathy vote.
If you want an example of response to bugs check out GAIM which recently had a remote overflow vun bug due to a rushed release to comply with the new AIM protocol, a fix was released for the cvs version on the same day the exploit went public:
Wow, a fix in the CVS. That would be *so* helpful to the consumers out there who have trouble with the concept of double clicking. Or the ones who need to run software on certified platforms.
The speed that some fixes are produced by some (if not most) OSS developers is admirable, but of little interest to the vast majority of customers who are looking for patches that are tested, stable and easy to apply. It’s a bit like people who think individual machine uptimes are the holy grail of stability.
[…]
I don’t remember doing that. I just was trying to promote the idea that it’s not quite that easy to say “this is the best”. MacOS has clearly done what Linux hasn’t done on the desktop by taking FreeBSD and putting a user friendly GUI on top of it, and kudos for them for showing everyone how it’s done. It would be very interesting if they released a x86 port.
Unfortunately it would also be a) financially disastrous to Apple and b) a step backwards with regards to some of OS X’s best aspects.
The thing with OS X is, while it’s certainly unixy underneath (note that it’s not based on FreeBSD, it just has a BSDish user space), this is neither apparent nor relevant to most of the people who use it. The OS X GUI and most applications do not really resemble existing unixes, or their philosophies – it’s still a Mac more than it’s a unix OS with an above-average GUI. It could just as easily be sitting on top of a BeOS or NT core and be just as useful, just as cool, just as popular and just as good. Indeed, plonk Cygwin or Microsoft’s Services for Unix onto NT and for all intents and purposes you’ve got the same “unix with a good GUI”.