Linux already runs about 40% to 50% of servers, so you are clearly underestimating it.
Windows will lose some people in every upgrade cycle, while I don't see anybody going from Linux to Windows...
I think 2004 will be the year when Linux runs so many more servers than Windows that it can no longer be denied (read: It will be reported in the news).
The question is, will they choose Linux or Windows?
Yes, this will be moderated as flamebait, so what. But it *is* a fact that those using older versions of Windows (read: those not doing everything Microsoft sais) are the first to switch to Linux which is a much more friendly environment if you don't want to do needless upgrades.
In worst-case, Windows-Palladium apps don't run on Linux, which means pretty much the status quo (if you ignore Wine) so nothing really changes.
So, I don't see Palladium hindering Linux acceptance.
On the other hand, nobody likes Palladium, especially Windows pirates (= 90% of Windows users). If Joe Average won't be able to play his mp3s, he will install Linux, no questions asked.
If you ask me, Palladium is the best thing that can happen to Linux acceptance.
(bear in mind that end of life for most software is 5 years)
In late 1999 and January 2000, NT4 was Microsoft's newest server OS.
I don't know about you, but for me 2000 + 5 is not 2003.
Re:They don't know anything about a DDOS attack...
on
4l-j4z333ra 0wn3d
·
· Score: 1
Actually, the USA tries to hold back information - DDOSing al-jazeera and having the incident on slashdot's front page is just accomplishing the contrary - a lot more people will get interested.
First of all, open source is part of the global IT marketplace, even if you don't like that fact.
Second, only a tiny fraction of the IT marketplace is about selling and sending shiny shrinkwrapped boxes around the world. Most programmers are employed to write in-house software or special software for one customer or one need. - If the customer or your boss wants the end product to be open source, it is open source.
Third, the shrinkwrap-part of the IT marketplace tends to form dominating products because of network effects and almost zero variable (= per unit) costs: MS for operating systems and office suites, Adobe for graphics software (Photoshop), Macromedia for animation software (flash), Intuit for money management software (Quicken), AutoCAD for CAD, etc. etc.
These dominating products are virtually impossible to attack in the CSS world, because you have more costs than the market leader (you would have to waste a lot of time being compatible) and a lot less revenues (because of much fewer sales). The only way to come over this hurdles is to use another dominating product to push it (like Microsoft did with Windows for MS Office). An upstart company has absolutely no chance of getting a profitable software product in the shrinkwrap market (except games).
So yes, Andalucia does have indeed no chance to compete in the global shrinkwrap market - no upstart company in the world has a chance to compete in the global shrinkwrap software market (unless they invent a whole new market).
OSS changes the picture, it's just a more efficient way to develop software, that's why Linux is there and going strong. The OSS community created a whole operating system with thousands of drivers and applications - something Microsoft despite their huge ressources just couldn't do. Microsoft depends very much on hardware vendors to write drivers and software developers (which are very often in-house developers) to write software for it.
Just take any non-RedHat distribution (SuSE, Mandrake if you want it easy or debian, Gentoo if you don't mind a little harder installation) and try KDE 3.1.1
Instant networking: No problem, just type "fish://server" into Konq and you got secure networking without installation.
You also can handle windows faster and more efficient, for example Alt+Right Mousebutton will resize a window by dragging near the border, Alt+Left MB will let you grab a window everywhere.
On top you get remote desktop abilities, loads of nice helper programs and in the case of SuSE a fully consistent configuration center for everything from IP-number to font color.
The reason why I don't recommend RedHat for desktop use is because their BlueCurve obsession causes KDE packages to appear much later than for all other distros, causes features to be taken out and introduces even more inconsistency in the Linux world.
You hear all the time about how bad X is, but when you ask what exactly is wrong with X, you get nothing but - silence.
But maybe you are different and can actually name something specific that is wrong about the X protocol.
(Note that we are both talking about the protocol, not the implementation - although XFree4.3 also took care of much of the implementation problems - so I don't really see any problems with the implementation either.)
Yes, I know I'm not agreeing to the FUD-fest here, but confusion and chaos is synonymous with a successful platform.
The PC platform is successful because there are many different "confusing" vendors. A very primitive PC running DOS was more successful than what Apple and Amiga had to offer.
The Windows platform is successful because of the above reason and because there are many different "confusing" software packages.
Linux is successful (say what you want: Being number 1 on web/news/mail servers and number 2 on file/print servers in a saturated market is a great success. The desktop market is moving slowly but moving towards Linux, too - at least outside the US, especially in Europe and Japan.) because of many "confusing" distributions and choices.
No, I'm saying that their business heavily relies upon being as backward compatible as possible so that corporate types buy truckloads of their desktops. That means not doing anything radical like omitting ISA (heh) and making damned sure to have every other port created since time began. Apple doesn't do things this way.
OK, maybe I didn't make that clear, so I will now:
Apple is wrong.
Including every port ever created is one of the most important features. My digital camera works damn fine and uses the serial port and I will not buy a new one just because Apple wants to save 0.3$ per unit. My HP49 calculator also has a serial interface.
The reason why I use PS2 instead of USB for my mouse is that I have a PS2-extension cord and no USB-extension cord. PS2 works fine, I have no reason to jump to USB just because - well, I just don't know why. And I hate cordless mice, BTW (I avoid anything battery-driven as much as possible).
In short, the "make the whole widget" vertical integration Apple has in the Mac platform is both the worst and the best thing about them. It keeps prices high and marketshare low, but without it they'd dissapear in a sea of other boxmakers and software makers, with no means of distinguishing themselves.
Parallel and serial ports are maybe 1$ per unit and would make their products interesting for a whole slew of people currently needing these interfaces. Same goes for PCI slots in their low-end hardware.
You are lying to yourself. There is nothing preventing Apple from testing the market by releasing 2 low-end models, one with, another without PCI-slots.
As Apple has always been forward thinking to gain market share and attention
As long as you can't get reasonably priced spare parts from Apple, their marketshare will go down and no 64-Bit CPU will change that. If I can't get spare parts I won't even consider Apple.
What are the advantages though? Seriously? The end user won't understand them.
If the end user would understand even the simplest things about computing performance, nobody would by 2.5GHz CPUs and everybody would buy 15000 RPM harddisks.
But he doesn't understand them, that's why CPUs have seen a tenfold increase in clockspeed in the last 5 years (IIRC) while harddrives haven't become a lot faster in the same period.
Opteron won't be a lot more expensive than a 32-Bit Athlon (it will be more expensive because of more cache, but the 64-Bit part is not making the chip a lot more expensive, contrary to popular belief), so I don't see why they shouldn't at least outsell the 32-Bit Athlons.
If you take into account that
People might not need over 4GB RAM right now, but they might want to upgrade their RAM in 3-4 years.
64-Bit is a big marketing instrument, actually needed or not.
Quite some people need/want 64-Bits already.
Opteron is a guaranteed success if priced similarily as the Athlon.
Yes, they've got the monopolist's desktop advantage now, but they got it because they've put out products that, while far from perfect, did the job well enough to serve as the focal point for PC development for the last couple decades.
Wrong. Wrong. Wrong.
The only reason they got their OS-domination on x86 is because IBM gave it to them and Asian hardware makers made their platform popular.
People would still use DOS if there were no alternative on cheap hardware - oh well, people did use DOS for half a decade when everybody else was already using graphical OSes (Apple, Amiga, Unix-GUIs).
But that changed, Linux is available on cheap hardware, too and is itself cheaper, more secure, stable and with Opteron clearly more advanced. That's why Microsoft is so scared of Linux and not of Unix or MacOS. Microsoft has already lost their dominating position on x86-servers.
Opteron is just another nail into the coffin of Windows as a server and will also have a big effect on the desktop.
Essentially you say that all PC-motherboard makers are ignoring the market needs by including PCI-slots and you have the one true insight that nobody needs them and Apple with it's 2% and shrinking marketshare is the only computer maker which really, truely adresses market needs.
Sorry, but your hypothesis just doesn't match with reality.
I would really like Apple to succeed, I even have a Powerbook myself. But if Apple's don't start listening to the market, their own arrogance and greed will be their downfall.
Another example is their spare parts policy. I don't consider Apple for a desktop because I can't get a spare motherboard/CPU. Even if Apple were cheaper than comparable PCs, I wouldn't even consider them for desktops because of that. Now you can lie to yourself and claim that "the enduser market doesn't need spare parts", but you know as well as me that's it's only Apple's greed preventing PPC-CPUs and motherboards from appearing in stores at reasonable prices.
(With notebooks, the situation is equally bad for all vendors, that's why I have a Powerbook)
I have always said that most 95% of people who buy computers don't know what a PCI slot is, and 95% of the people who do know never use the ones they have anyway.
Where do you have these numbers from? Did you did a survey?
Well, I agree that few people probably know what a PCI slot is, however most know very well that they can upgrade their computers and put cards into them. While it's perfectly OK to offer a PCI-less device it would be better for Apple if they would offer a low-cost model with PCI slots, too.
Oh my god, I can't believe how somebody can be so naive and paranoid at the same time.
Policies? Courtcases? Judges legalizing piracy because of "they didn't prosecute in the past"? What a bunch of nonsense.
The only reasons for closing the betas is:
So that Steve gets a good feeling showing off new features at the next keynote
Integrating it in a new version of MacOSX (and possibly making sure it doesn't run on an older version) to create upgrade incentive and buying incentive. Apple has a long history of crippling their own products (just like the lack of PCI-slots in their customer machines), they just don't get that they hurt themselves in the long run.
Company stupidity/paranoia. Essentially that's the same point as the one above.
The xbox makes up for it with the brute force of a 733Mhz CPU
Exactly that's my point. XBox only has a chance because it's 2 years newer.
If the PS2 would have come out at the same time as the XBox, it would also run at about 600-800Mhz and completely wipe the floor with XBox.
Similarily, if the XBox would have come out at the same time as the PS2, it would get it's ass kicked because at that time only 200-300MHz P2s and early nVidia cards (TNT?) were available.
That's my whole point. Without 2 years of technological advantage, the x86 architecture doesn't stand a remote chance against the PS2 architecture.
And again, the PS2 isn't cheaper then the Xbox.
It is a lot cheaper to make, the royalties are also cheaper.
I'll asked once, I'll ask again: Who will pay the billion/year that Microsoft loses now on XBox? Does MS plan to never break even or even make a profit? So who will pay it? You, the gamers?
Of course you won't. XBox is a product that is unable to support itself and will die the second MS stops pouring money in it.
what if the manufacturers wanted to save money on hardware
They would waste >6 months and lots of money on development just to find out that the hardware has become cheap enough for Linux in the meantime.
linux is certainly going to be larger than a homegrown o/s
Not true.
Linux is opensource in case you didn't notice, which means you can do anything with it. So in theory, you could optimize it so much to be as small as an in-house system - and still save development time. Of course in real life nobody does this because ROM and RAM is cheap enough and optimizing everything down to the last byte takes by far too much time.
With Linux, you also get a very reliable codebase which is well tested on millions of different architectures and configurations. With an in-house OS, you need years until you get to the same quality of Linux.
Contrary to what you may want to believe, linux is not always the best tool for the job.
Well if design a new system with a CPU, it is - if you don't have any special requests like PalmOS (or some other OS) compatibility or some very narrow and specialized requirements. Even on most devices that don't necessarily need Linux, the faster time-to-market will make it worth it.
There are a lot of devices that use DSPs and no CPU - on those you of course can't and won't use Linux. But a DSP is no classical CPU.
Of course I am only talking about new developments, it might always make more sense to refine an older in-house software if available.
Slashdot Mirror
Windows will lose some people in every upgrade cycle, while I don't see anybody going from Linux to Windows...
I think 2004 will be the year when Linux runs so many more servers than Windows that it can no longer be denied (read: It will be reported in the news).
Yes, this will be moderated as flamebait, so what. But it *is* a fact that those using older versions of Windows (read: those not doing everything Microsoft sais) are the first to switch to Linux which is a much more friendly environment if you don't want to do needless upgrades.
So, I don't see Palladium hindering Linux acceptance.
On the other hand, nobody likes Palladium, especially Windows pirates (= 90% of Windows users). If Joe Average won't be able to play his mp3s, he will install Linux, no questions asked.
If you ask me, Palladium is the best thing that can happen to Linux acceptance.
In late 1999 and January 2000, NT4 was Microsoft's newest server OS.
I don't know about you, but for me 2000 + 5 is not 2003.
Actually, the USA tries to hold back information - DDOSing al-jazeera and having the incident on slashdot's front page is just accomplishing the contrary - a lot more people will get interested.
First of all, open source is part of the global IT marketplace, even if you don't like that fact.
Second, only a tiny fraction of the IT marketplace is about selling and sending shiny shrinkwrapped boxes around the world. Most programmers are employed to write in-house software or special software for one customer or one need. - If the customer or your boss wants the end product to be open source, it is open source.
Third, the shrinkwrap-part of the IT marketplace tends to form dominating products because of network effects and almost zero variable (= per unit) costs: MS for operating systems and office suites, Adobe for graphics software (Photoshop), Macromedia for animation software (flash), Intuit for money management software (Quicken), AutoCAD for CAD, etc. etc.
These dominating products are virtually impossible to attack in the CSS world, because you have more costs than the market leader (you would have to waste a lot of time being compatible) and a lot less revenues (because of much fewer sales). The only way to come over this hurdles is to use another dominating product to push it (like Microsoft did with Windows for MS Office). An upstart company has absolutely no chance of getting a profitable software product in the shrinkwrap market (except games).
So yes, Andalucia does have indeed no chance to compete in the global shrinkwrap market - no upstart company in the world has a chance to compete in the global shrinkwrap software market (unless they invent a whole new market).
OSS changes the picture, it's just a more efficient way to develop software, that's why Linux is there and going strong. The OSS community created a whole operating system with thousands of drivers and applications - something Microsoft despite their huge ressources just couldn't do. Microsoft depends very much on hardware vendors to write drivers and software developers (which are very often in-house developers) to write software for it.
Ironically, double-clicking is the worst usability nightmare - introduced by Apple because of lack of mouse-buttons, then later copied by Microsoft.
I agree that Linux is not a very good gaming platform - because there are not many games, not because of any usability issues.
If you are not a gamer, KDE/Linux is much better and more efficient to work with than the primitive Windows GUI.
Instant networking: No problem, just type "fish://server" into Konq and you got secure networking without installation.
You also can handle windows faster and more efficient, for example Alt+Right Mousebutton will resize a window by dragging near the border, Alt+Left MB will let you grab a window everywhere.
On top you get remote desktop abilities, loads of nice helper programs and in the case of SuSE a fully consistent configuration center for everything from IP-number to font color.
The reason why I don't recommend RedHat for desktop use is because their BlueCurve obsession causes KDE packages to appear much later than for all other distros, causes features to be taken out and introduces even more inconsistency in the Linux world.
But maybe you are different and can actually name something specific that is wrong about the X protocol.
(Note that we are both talking about the protocol, not the implementation - although XFree4.3 also took care of much of the implementation problems - so I don't really see any problems with the implementation either.)
The PC platform is successful because there are many different "confusing" vendors. A very primitive PC running DOS was more successful than what Apple and Amiga had to offer.
The Windows platform is successful because of the above reason and because there are many different "confusing" software packages.
Linux is successful (say what you want: Being number 1 on web/news/mail servers and number 2 on file/print servers in a saturated market is a great success. The desktop market is moving slowly but moving towards Linux, too - at least outside the US, especially in Europe and Japan.) because of many "confusing" distributions and choices.
Serious question - will RedHat:
Does anybody know?
I'll get flamed for saying that, but BlueCurve was a very stupid idea, IMO.
Thanks.
OK, maybe I didn't make that clear, so I will now:
Apple is wrong.
Including every port ever created is one of the most important features. My digital camera works damn fine and uses the serial port and I will not buy a new one just because Apple wants to save 0.3$ per unit. My HP49 calculator also has a serial interface.
The reason why I use PS2 instead of USB for my mouse is that I have a PS2-extension cord and no USB-extension cord. PS2 works fine, I have no reason to jump to USB just because - well, I just don't know why. And I hate cordless mice, BTW (I avoid anything battery-driven as much as possible).
In short, the "make the whole widget" vertical integration Apple has in the Mac platform is both the worst and the best thing about them. It keeps prices high and marketshare low, but without it they'd dissapear in a sea of other boxmakers and software makers, with no means of distinguishing themselves.
Parallel and serial ports are maybe 1$ per unit and would make their products interesting for a whole slew of people currently needing these interfaces. Same goes for PCI slots in their low-end hardware.
You are lying to yourself. There is nothing preventing Apple from testing the market by releasing 2 low-end models, one with, another without PCI-slots.
As long as you can't get reasonably priced spare parts from Apple, their marketshare will go down and no 64-Bit CPU will change that. If I can't get spare parts I won't even consider Apple.
If the end user would understand even the simplest things about computing performance, nobody would by 2.5GHz CPUs and everybody would buy 15000 RPM harddisks.
But he doesn't understand them, that's why CPUs have seen a tenfold increase in clockspeed in the last 5 years (IIRC) while harddrives haven't become a lot faster in the same period.
If you take into account that
Opteron is a guaranteed success if priced similarily as the Athlon.
Now you tell me if you want that machine to support more than 4 Gigs of RAM or not.
Samba easily outperforms Windows' native SMB implementation: http://www.itweek.co.uk/News/1131114
So there is nothing preventing Wine from beating Windows at it's own game.
Wrong. Wrong. Wrong.
The only reason they got their OS-domination on x86 is because IBM gave it to them and Asian hardware makers made their platform popular.
People would still use DOS if there were no alternative on cheap hardware - oh well, people did use DOS for half a decade when everybody else was already using graphical OSes (Apple, Amiga, Unix-GUIs).
But that changed, Linux is available on cheap hardware, too and is itself cheaper, more secure, stable and with Opteron clearly more advanced. That's why Microsoft is so scared of Linux and not of Unix or MacOS. Microsoft has already lost their dominating position on x86-servers.
Opteron is just another nail into the coffin of Windows as a server and will also have a big effect on the desktop.
Sorry, but your hypothesis just doesn't match with reality.
I would really like Apple to succeed, I even have a Powerbook myself. But if Apple's don't start listening to the market, their own arrogance and greed will be their downfall.
Another example is their spare parts policy. I don't consider Apple for a desktop because I can't get a spare motherboard/CPU. Even if Apple were cheaper than comparable PCs, I wouldn't even consider them for desktops because of that. Now you can lie to yourself and claim that "the enduser market doesn't need spare parts", but you know as well as me that's it's only Apple's greed preventing PPC-CPUs and motherboards from appearing in stores at reasonable prices.
(With notebooks, the situation is equally bad for all vendors, that's why I have a Powerbook)
Where do you have these numbers from? Did you did a survey?
Well, I agree that few people probably know what a PCI slot is, however most know very well that they can upgrade their computers and put cards into them. While it's perfectly OK to offer a PCI-less device it would be better for Apple if they would offer a low-cost model with PCI slots, too.
Policies? Courtcases? Judges legalizing piracy because of "they didn't prosecute in the past"? What a bunch of nonsense.
The only reasons for closing the betas is:
Imaginary court cases have nothing to do with it.
Exactly that's my point. XBox only has a chance because it's 2 years newer.
If the PS2 would have come out at the same time as the XBox, it would also run at about 600-800Mhz and completely wipe the floor with XBox.
Similarily, if the XBox would have come out at the same time as the PS2, it would get it's ass kicked because at that time only 200-300MHz P2s and early nVidia cards (TNT?) were available.
That's my whole point. Without 2 years of technological advantage, the x86 architecture doesn't stand a remote chance against the PS2 architecture.
And again, the PS2 isn't cheaper then the Xbox.
It is a lot cheaper to make, the royalties are also cheaper.
I'll asked once, I'll ask again: Who will pay the billion/year that Microsoft loses now on XBox? Does MS plan to never break even or even make a profit? So who will pay it? You, the gamers?
Of course you won't. XBox is a product that is unable to support itself and will die the second MS stops pouring money in it.
They would waste >6 months and lots of money on development just to find out that the hardware has become cheap enough for Linux in the meantime.
linux is certainly going to be larger than a homegrown o/s
Not true.
Linux is opensource in case you didn't notice, which means you can do anything with it. So in theory, you could optimize it so much to be as small as an in-house system - and still save development time. Of course in real life nobody does this because ROM and RAM is cheap enough and optimizing everything down to the last byte takes by far too much time.
With Linux, you also get a very reliable codebase which is well tested on millions of different architectures and configurations. With an in-house OS, you need years until you get to the same quality of Linux.
Contrary to what you may want to believe, linux is not always the best tool for the job.
Well if design a new system with a CPU, it is - if you don't have any special requests like PalmOS (or some other OS) compatibility or some very narrow and specialized requirements. Even on most devices that don't necessarily need Linux, the faster time-to-market will make it worth it.
There are a lot of devices that use DSPs and no CPU - on those you of course can't and won't use Linux. But a DSP is no classical CPU.
Of course I am only talking about new developments, it might always make more sense to refine an older in-house software if available.