With the ability today to run an OS, applications -- and even an entire PC desktop of applications -- in a virtual container using a hypervisor, the need to have the OS and applications installed natively on a PC is becoming less and less, said Brian Madden, an independent technology analyst.
Brian Madden is either talking about something else, or he's confused by references to hypervisors elsewhere. Midori will run under Hypervisors... but as one possible deployment of the OS, not as an essential part of the system. Singularity is more like ".NET" taken to the next level, with the entire OS running without hardware memory protection (let alone hypervisors), so it can run anywhere... even as a module inside another application... without any specific hardware support.
Hey, go back and look at the message that I was originally responding to. The OP explicitly recommended a Mac OSX server for small shops with little net expertise.
Hey, go back and look at who the OP was. I recommended OSX Server for small shops with little net expertise, and I'm explaining why your comment had nothing to do with that. Anyone who has the expertise to build a network using rsync for publishing and backups is NOT running "a small shop with little net expertise". Anyone who's hiring you to set up and run their network is also NOT running "a small shop with little net expertise".
I *never* considered OS X an appropriate server operating system for your kind of situation. Ever. Not now, not five years ago when I was investigating it (investigating it, by the way, means actually running OS X and understanding how it works... not getting confused about HFS and HFS+ or failing to recognize UTF-8 file names). Take away the desktop and all you have left is a rather anemic UNIX box. You can get a MUCH better UNIX box for MUCH less money from HP... I'd take a DL320 or DL360 over an XServe any day. But I wouldn't suggest that "a small shop with little net expertise" buy a DL320 and stick Debian or SUSE or FreeBSD on it and try and make a go of it.
I'm talking about shops with a few Mac and Windows desktops, no geeks, whose other option is Windows Server 200x.
And never fear, if you were interviewing with me and spent the interview slamming SNMP, I wouldn't make the mistake of hiring you.:)
The only part that needs to be converted to Objective C is the part that creates and manages the game's viewport. THe existing game logic and models can remain in C or C++ (or Fortran=Fortran+1, or "ADD ONE TO COBOL GIVING COBOL"), and the user interface would be rewritten from scratch for the iPhone regardless.
Niggle is a freeware Scrabble on the Palm that is fairly vanilla looking but is a far superior implementation of Scrabble to the official Hasbro version, but when they came out with official Scrabble on the Palm the authors of Niggle, of course, pulled it.
Apple's income primarily comes from sale of hardware - that's their business model. All their software sales put together are a small part of their income. So what it's a "loss leader" for, more than anything else, is sales of Macs. Other Apple software is lost in the noise.
And the cost of Mac OS is not just $129, it's actually most of their profit margin on the Mac. Apple's profit margins are 30-50% in a market where even boutique computers are lucky to get as much as 5%, and OS X is most of the reason they can make those margins... so what OS X on my Macbook Pro cost most of the difference in price between the Macbook Pro I bought and the Thinkpad I would have ended up with if OS X wasn't in the picture.
In my case, that's over $1000. If you average things out, including Mac minis and Macbooks where the margins are smaller in absolute terms, my guess is that the value of the average Mac includes almost $500 for OS X and iLife.
Paying only $400 for OS X to run on a Thinkpad? That's a bargain.
The 'demand bureaucracy' would have to live on a mere 1% of the 5%. So whatever a demand bureaucracy is, it would have to deliver a lot of social product analysis quite cheaply.
A demand bureaucracy is the mechanism by which a planned economy works. Demand is determined by the bureaucracy rather than the market. Someone in the center of the bureaucracy decides policy and creates long term plans, the layers of bureaucrats around it decide what will be made and what the state will pay for, and the factories make whatever they think they can get away with that satisfy the letter of the requirements from the bureaucracy.
Don't know what that has to do with bureaucrats who have to measure where and how the good of technology is flowing, and reward those that enable that flow.
That's precisely what a demand bureaucracy does. Control demand by taking advantage of the high level view of the economy and the long term goals of the central planners, and making payments based on those goals. It's how the old Soviet Union operated. It's how North Korea operates.
The primary alternative to the planned is the market economy, which replaces the bureaucracy by the market. To make the market work, you have the entire population of the economy working full time making buying decisions... it takes advantage of the fact that they're doing that anyway to get the rewards for free.
But the key idea is to reward the total delivery, not the creation.
That's what the market does. That's why it costs more to buy a copy of Photoshop from Best Buy than Best Buy paid Adobe for it. That's also why it's possible for an inventor to sell patents to companies to develop them... yes, that has led to the creation of patent holding companies, but there are less radical solutions to them (such as requiring a patent to be developed within some small number of years) than eliminating patents altogether.
The goal of the patent system is to harness the market to promote research and development. It may not be necessary any more, though I'm by far from convinced of that. But if it were abolished it would be better to replace it with nothing than replace it with a 5% tax doled out at the whims of an underfunded bureaucracy. An underfunded bureaucracy is how we got INTO this mess.
One of the issues was the "Internet Sharing" buzz phrase. If you google that now, you'll find lots of warnings that if you enable this in OSX, it silently starts up a DHCP server.
Well, yes. It's emulating a home/office firewall/router. Every SOHO firewall/router I've ever used has done the same thing.
This works if everything was developed on OSX.
Or Windows. Which also has the same file naming convention.
But if you're running a web server, you're probably going to be including things from other machines in the vicinity.
If you're running a network and you already have UNIX systems on the network, why on earth would you even consider OS X? You already have better UNIX server boxes (just about any other UNIX system is a better UNIX server than OS X). You're not the target market. You already have more technical expertise than the people I'm talking about, AND you're not looking for a system designed for an all-Mac-and-Windows environment.
OS X doesn't run HFS. It only runs HFS+. You can build an HFS+ file system with case sensitive or case insensitive behavior. That's not "turning HFS into HFS+", it's "running HFS+ case insensitive". No wonder the poor folks at Apple were confused. And, as you found out, it's probably a bad idea, because Mac software is written with the assumption that the file system is case-insensitive, and it is likely to misbehave. The same thing would happen if you ran Windows on a non-translating file system (I've tried it... you can allegedly do it with NTFS but I've found that it pretends it's done it while keeping the case-insensitive file name matching, the 8+3 file name translation, and so on).
The solution to your problem, by the way, would be to create a separate *UFS* (not "HFSanything" partition and keep your work on there.
There's been a LOT of work done on rsync on Windows to deal with the NTFS oddness. There hasn't been that much done on HFS+... because people are less likely to be stuck using HFS+ on OS X because their company is "Mac Only", the way people get stuck using NTFS and Windows because so many companies are Windows-only.
But on OSX, we'd see non-ASCII chars simply garbaged with no obvious pattern.
All file names in OS X are stored in UTF-8: you were probably just reading them as raw 8-bit characters and trying to interpret them as ISO-8859.1. That trick never works.
If you were starting with Mac-and-Windows you would be using UTF-8 and Unicode file names to begin with, you'd be using case-insensitive file names, you'd never run into this stuff. That's the market Apple's looking at, not people who already have FOSS UNIX boxes around the place. People who have FOSS UNIX have much better servers already.
Well, setting aside the fact that we're talking about "Apple vs Microsoft" rather than "Apple vs Novell" or "Apple vs IBM" (where you might have a point) the big problem with OpenDarwin wasn't Apple... it was the fact that OpenDarwin was a particularly pointless project.
Apple provided a lot more support for OpenDarwin than Jolitz did for 386BSD, and yet we still managed to turn 386BSD into a pretty damn good OS. How did this happen? Because there wasn't any alternative. This was before Linux. The closest things to free UNIX were Minix and the Software Tools VOS. Linus Torvalds started with even less, and produced the free UNIX of choice. There was a purpose to the 386BSD patchkits, there was a purpose to FreeBSD and Linux.
There's no open niche for OpenDarwin to scratch. That niche had long been filled, by Linux and other BSD derivitives. If you really wanted Mach (despite it being the kiss of death for open source operating systems), you had Lites and Hurd to build from.
Speaking of Lites and Hurd... where are they now? Lites last release was in the '90s, and Hurd hasn't had a production-quality release yet. Why didn't they get anywhere? Because the Free Software Foundation didn't have any commitment to Open Source? No, it's because they didn't have a market. They had no killer application. Same problem as OpenDarwin. What would you actually use OpenDarwin for if it was still actively being developed?
I do think that inventors should be rewarded, as should those who support them.
The purpose of the patent system is not "rewarding the inventors".
The purpose of the patent system is to encourage research and development. Rewarding the inventors is the method, not the goal.
Increasing the cost of goods and services 5% doesn't encourage R&D, it reduces the potential profit from inventions by 5%, or reduces the potential market by... anything from 0% to 100%, depending on the marginal benefit and potential market.
And creating a demand bureaucracy to administer it, well, that worked SO well for the Soviet Union.
Apple is a corporation. You don't call corporations (whether they be Google, Microsoft, Apple, or IBM) good or evil, you just look at what they're actually doing.
So, let's look at what Apple and Microsoft are actually doing:
iPod/iPhone vs Zune/XBox: iTunes has weak DRM, Windows Media Player has strong DRM with kernel support that's getting stronger in Vista (trusted media path and tilt switches). None of their consumer entertainment products are open, but it's a lot cheaper and easier to get into software development for the iPhone than the XBox.
Open source: Microsoft uses GPLed software in Interix, and used OpenBSD extensively. They haven't released their versions of any of the BSD-licensed components they used in Interix (or Windows), but they do have a copy of their GCC source in the Interix tree. They have *recently* decided to accept the LGPL, but there's no GPLed software in Windows proper. Apple uses GPLed software in Darwin, and used FreeBSD extensively. They have released the open source code in Darwin and kept releasing it with every new release of OS X. They have added their own open source components, to the point where the majority of the traditional userland in OS X, as well as many major new components like launchd, are open source. Microsoft's open source poster boys are things like Windows installers.
This is like claiming that Gandalf's "turning evil" because he's wearing a grey robe, while cheering on Sauron for having the Nazgul stop off to pick up litter on the way back from scouring the Shire.
Mac OS X isn't my first choice for a server OS, I'd rather run FreeBSD straight without spiking it with Mach. But it's probably a better choice for small sites without much technical expertise.
Let's say someone finds a new way to cut logs that let you build log cabins almost as easily but many times more sturdily and with better isolation. This would surely be patentable.
I assume you mean "better insulation". How much do you think it would cost to research such a process to the point where it was patentable? You have to actually cut logs and fit them together... even if you prototype the design in Second Life you're going to have to build it to tell if it actually works.
Let's try another example: Let's say someone patents an idea (putting adhesive strips on a piece of soft plastic to hold it against a curved monitor screen on a fish-finder to protect it), then later sees someone using van-der-waal forces to hold a flat piece of soft plastic against a flat touch screen, and files a modified patent that eliminates all the claims of his original patent (because the first few claims made a big point of the difficulty of attaching he plastic to the curved screen, and the use of adhesive strips) and replaces them with a patent for this handheld screen protector, then starts going after people who had been making these things long before he filed his amended patent?
The thing about software patents is that most of them are more like the second case than the first. They're overly broad, are based on simple and obvious ideas, and are frequently re-interpreted to apply to problems that nobody had even thought of when the original patent was filed.
If software patents were restricted to completed systems with comparable levels of research and development costs as your examples, I suspect there would be a good deal fewer objections to software patents. Some people might have the kind of hidden agenda you're talking about, but most people wouldn't care. The problem is that they're not, and most of them seem to be about things that are routinely implemented hundreds of times a day by people who have no idea that they're creating a "patentable invention".
I don't know how to come up with rules that would allow your "hard problems" patents without leaving the floodgates open. I'm not a lawyer. I don't play one on TV.
$90-$180 FOB Shanghai, QTY 500. Runs Linux or Windows CE.
Looks like they have variants of this from 7" to 12.1", which is why the range of prices.
Re:Once the company founders are gone...
on
Apple After Jobs
·
· Score: 1
Commodore's end was nigh when Jack Tramiel defected and bought Atari.
So was Atari's.
Commodore's problem was the whole "Tramiel vs Gould" thing, and that kept going even after Tramiel left. Buying Atari simply roped Atari into the same doomed feud.
I thought about building a Hackintosh, but by the time I had my parts list down, I was nearing $2000 in parts. And even that wasn't at the quality of the Mac Pro.
If you're trying to build something comparable to a Mac Pro, you're not even in the same volume (let alone the same page) as most of the people building their own white boxes.
I mean, I've built a bunch of computers and the only time I went over $1000 in parts was in the early '90s...
Think "Mac mini Pro", not "Mac Pro".
Even though I'd be the first to agree that the Mini is overpriced for the specs, remember that it is a tiny computer. It's smaller than a laptop.
It's bigger than a lot of laptops (it's 3* as high as most!), and more expensive than a few. If you're looking for a small, quiet computer and don't care about the OS are you going to buy a Mac mini or are you going to buy a laptop? Let's see... for $600 you can get "Bring your own display, keyboard, and mouse" or you can get something with better specs than a Mini plus a keyboard, WXGA screen, trackpad, battery... or if you want to really cut the size done the EeePC 901 is pretty nice.
So, to return to the GP, yes, people do pay for the whole package.
Some people do. That's the point. People buy Macs for all kinds of reasons. But how many people really want the whole package, and how many are just putting up with the package because it's the only way to get the part they want?
Well, how's the *rest* of the package sell when OS X isn't in the picture?
Dismally.
If there really was significant demand for the Mac mini hardware, then you'd see more people buying Wintel PCs with that kind of form factor. People have sure tried to *sell* tiny PCs, from things like the Multia in the '90s to "Mac mini clones" today. The demand for that kind of machine just isn't there, or they'd have sold enough that you'd be used to seeing them along side the minitowers in every Office Depot in the US. The small size and accompanying lack of expandability are actually negatives for most people... at least most of the people I've tried to talk into getting a Mac.
SOME people buy Macs because they want that kind of hardware. But I don't believe MOST people do, because it just doesn't sell.
Name one open source project that Apple created that is useful?
launchd is the obvious one.
Yes, they were totally on the level and fulfilling the letter of their responsibilities, but it was hardly "bent over backwards".
What they were doing before the uproar was totally on the level and fulfilling their responsibilities, how they responded was far more than anyone had a right to expect. More importantly they were and have continued to make releases of their updated versions of numerous open source components that they are not obligated to.
I think the problem is that you're comparing Apple with the handful of "open source superstars", and I'm comparing them to the average company that uses FOSS. There are a number of companies that have done more than Apple, yes, but by far the majority of companies using open source projects do far less... most actively avoid using GPLed software at all, even LGPLed software, and many have policies in place to prevent employees contributing to open source projects. Many have to be sued by the FSF before they will even come near to meeting the letter of their responsibilities.
Yes, Apple has a history of being controlling. Every time there's a fuss about Apple's Open Source work, or about people using Apple's Open Source releases, I honestly expect them to dump all but the GPLed core of the Darwin open source repository. That would be completely consistent with Apple's past, and would still leave them being better than average in their open source policies among large US corporations. And yet they continue to release more software, and maintain their releases. Maybe, eventually, they won't, but right now while they are not a superstar like IBM... they're still way ahead of the pack.
What they make that people want to buy is software.
Their business model is to sell hardware at a high margin, but people *buy* the hardware because of the software. They wouldn't spend 40% more than a comparable pc or handheld if it was running Vista or Windows CE.
This is not a rare model, either. Cisco hardware is not exceptional, people buy it because it's running IOS. Cisco is a software company with a hardware business model. Network Appliance hardware is nothing special, but you can't run the Filer software without it. They're another company with a hardware business model.
Slashdot Mirror
Brian Madden is either talking about something else, or he's confused by references to hypervisors elsewhere. Midori will run under Hypervisors... but as one possible deployment of the OS, not as an essential part of the system. Singularity is more like ".NET" taken to the next level, with the entire OS running without hardware memory protection (let alone hypervisors), so it can run anywhere... even as a module inside another application... without any specific hardware support.
Hmmmm?
Have you ever noticed that when soldiers fight they hardly ever do so in a firing range and never police up the brass?
(note for the humor-challenged, this is sarcasm)
Hey, go back and look at the message that I was originally responding to. The OP explicitly recommended a Mac OSX server for small shops with little net expertise.
Hey, go back and look at who the OP was. I recommended OSX Server for small shops with little net expertise, and I'm explaining why your comment had nothing to do with that. Anyone who has the expertise to build a network using rsync for publishing and backups is NOT running "a small shop with little net expertise". Anyone who's hiring you to set up and run their network is also NOT running "a small shop with little net expertise".
I *never* considered OS X an appropriate server operating system for your kind of situation. Ever. Not now, not five years ago when I was investigating it (investigating it, by the way, means actually running OS X and understanding how it works... not getting confused about HFS and HFS+ or failing to recognize UTF-8 file names). Take away the desktop and all you have left is a rather anemic UNIX box. You can get a MUCH better UNIX box for MUCH less money from HP... I'd take a DL320 or DL360 over an XServe any day. But I wouldn't suggest that "a small shop with little net expertise" buy a DL320 and stick Debian or SUSE or FreeBSD on it and try and make a go of it.
I'm talking about shops with a few Mac and Windows desktops, no geeks, whose other option is Windows Server 200x.
And never fear, if you were interviewing with me and spent the interview slamming SNMP, I wouldn't make the mistake of hiring you. :)
You do realize you get more of a workout doing 50 or so jumping jacks (takes about a minute) than you do playing Wii Sports for 5 or 10 minutes.
Probably so, but 50 jumping jacks I don't do isn't much of a workout compared to an hour of playing Wii Tennis I do.
Just don't blow out your knees trying to do your kata with your feet in the wrong position because your sensei doesn't understand anatomy.
Not that I'm bitter. Well, not... hell yes, I am.
If you play it right, it's a workout. You won't get as good scores at the weenies who sit on the couch and twitch the controllers, but who cares?
Where's the defamation?
The only part that needs to be converted to Objective C is the part that creates and manages the game's viewport. THe existing game logic and models can remain in C or C++ (or Fortran=Fortran+1, or "ADD ONE TO COBOL GIVING COBOL"), and the user interface would be rewritten from scratch for the iPhone regardless.
So if they called it "X-Words" they'd have been OK?
If it was simply a matter of Trademark then Scrabulous and Niggle would have no more problem than MAD Magazine.
Niggle is a freeware Scrabble on the Palm that is fairly vanilla looking but is a far superior implementation of Scrabble to the official Hasbro version, but when they came out with official Scrabble on the Palm the authors of Niggle, of course, pulled it.
Apple's income primarily comes from sale of hardware - that's their business model. All their software sales put together are a small part of their income. So what it's a "loss leader" for, more than anything else, is sales of Macs. Other Apple software is lost in the noise.
And the cost of Mac OS is not just $129, it's actually most of their profit margin on the Mac. Apple's profit margins are 30-50% in a market where even boutique computers are lucky to get as much as 5%, and OS X is most of the reason they can make those margins... so what OS X on my Macbook Pro cost most of the difference in price between the Macbook Pro I bought and the Thinkpad I would have ended up with if OS X wasn't in the picture.
In my case, that's over $1000. If you average things out, including Mac minis and Macbooks where the margins are smaller in absolute terms, my guess is that the value of the average Mac includes almost $500 for OS X and iLife.
Paying only $400 for OS X to run on a Thinkpad? That's a bargain.
The 'demand bureaucracy' would have to live on a mere 1% of the 5%. So whatever a demand bureaucracy is, it would have to deliver a lot of social product analysis quite cheaply.
A demand bureaucracy is the mechanism by which a planned economy works. Demand is determined by the bureaucracy rather than the market. Someone in the center of the bureaucracy decides policy and creates long term plans, the layers of bureaucrats around it decide what will be made and what the state will pay for, and the factories make whatever they think they can get away with that satisfy the letter of the requirements from the bureaucracy.
Don't know what that has to do with bureaucrats who have to measure where and how the good of technology is flowing, and reward those that enable that flow.
That's precisely what a demand bureaucracy does. Control demand by taking advantage of the high level view of the economy and the long term goals of the central planners, and making payments based on those goals. It's how the old Soviet Union operated. It's how North Korea operates.
The primary alternative to the planned is the market economy, which replaces the bureaucracy by the market. To make the market work, you have the entire population of the economy working full time making buying decisions... it takes advantage of the fact that they're doing that anyway to get the rewards for free.
But the key idea is to reward the total delivery, not the creation.
That's what the market does. That's why it costs more to buy a copy of Photoshop from Best Buy than Best Buy paid Adobe for it. That's also why it's possible for an inventor to sell patents to companies to develop them... yes, that has led to the creation of patent holding companies, but there are less radical solutions to them (such as requiring a patent to be developed within some small number of years) than eliminating patents altogether.
The goal of the patent system is to harness the market to promote research and development. It may not be necessary any more, though I'm by far from convinced of that. But if it were abolished it would be better to replace it with nothing than replace it with a 5% tax doled out at the whims of an underfunded bureaucracy. An underfunded bureaucracy is how we got INTO this mess.
One of the issues was the "Internet Sharing" buzz phrase. If you google that now, you'll find lots of warnings that if you enable this in OSX, it silently starts up a DHCP server.
Well, yes. It's emulating a home/office firewall/router. Every SOHO firewall/router I've ever used has done the same thing.
This works if everything was developed on OSX.
Or Windows. Which also has the same file naming convention.
But if you're running a web server, you're probably going to be including things from other machines in the vicinity.
If you're running a network and you already have UNIX systems on the network, why on earth would you even consider OS X? You already have better UNIX server boxes (just about any other UNIX system is a better UNIX server than OS X). You're not the target market. You already have more technical expertise than the people I'm talking about, AND you're not looking for a system designed for an all-Mac-and-Windows environment.
OS X doesn't run HFS. It only runs HFS+. You can build an HFS+ file system with case sensitive or case insensitive behavior. That's not "turning HFS into HFS+", it's "running HFS+ case insensitive". No wonder the poor folks at Apple were confused. And, as you found out, it's probably a bad idea, because Mac software is written with the assumption that the file system is case-insensitive, and it is likely to misbehave. The same thing would happen if you ran Windows on a non-translating file system (I've tried it... you can allegedly do it with NTFS but I've found that it pretends it's done it while keeping the case-insensitive file name matching, the 8+3 file name translation, and so on).
The solution to your problem, by the way, would be to create a separate *UFS* (not "HFSanything" partition and keep your work on there.
There's been a LOT of work done on rsync on Windows to deal with the NTFS oddness. There hasn't been that much done on HFS+... because people are less likely to be stuck using HFS+ on OS X because their company is "Mac Only", the way people get stuck using NTFS and Windows because so many companies are Windows-only.
But on OSX, we'd see non-ASCII chars simply garbaged with no obvious pattern.
All file names in OS X are stored in UTF-8: you were probably just reading them as raw 8-bit characters and trying to interpret them as ISO-8859.1. That trick never works.
If you were starting with Mac-and-Windows you would be using UTF-8 and Unicode file names to begin with, you'd be using case-insensitive file names, you'd never run into this stuff. That's the market Apple's looking at, not people who already have FOSS UNIX boxes around the place. People who have FOSS UNIX have much better servers already.
Well, setting aside the fact that we're talking about "Apple vs Microsoft" rather than "Apple vs Novell" or "Apple vs IBM" (where you might have a point) the big problem with OpenDarwin wasn't Apple... it was the fact that OpenDarwin was a particularly pointless project.
Apple provided a lot more support for OpenDarwin than Jolitz did for 386BSD, and yet we still managed to turn 386BSD into a pretty damn good OS. How did this happen? Because there wasn't any alternative. This was before Linux. The closest things to free UNIX were Minix and the Software Tools VOS. Linus Torvalds started with even less, and produced the free UNIX of choice. There was a purpose to the 386BSD patchkits, there was a purpose to FreeBSD and Linux.
There's no open niche for OpenDarwin to scratch. That niche had long been filled, by Linux and other BSD derivitives. If you really wanted Mach (despite it being the kiss of death for open source operating systems), you had Lites and Hurd to build from.
Speaking of Lites and Hurd... where are they now? Lites last release was in the '90s, and Hurd hasn't had a production-quality release yet. Why didn't they get anywhere? Because the Free Software Foundation didn't have any commitment to Open Source? No, it's because they didn't have a market. They had no killer application. Same problem as OpenDarwin. What would you actually use OpenDarwin for if it was still actively being developed?
I do think that inventors should be rewarded, as should those who support them.
The purpose of the patent system is not "rewarding the inventors".
The purpose of the patent system is to encourage research and development. Rewarding the inventors is the method, not the goal.
Increasing the cost of goods and services 5% doesn't encourage R&D, it reduces the potential profit from inventions by 5%, or reduces the potential market by... anything from 0% to 100%, depending on the marginal benefit and potential market.
And creating a demand bureaucracy to administer it, well, that worked SO well for the Soviet Union.
Apple is a corporation. You don't call corporations (whether they be Google, Microsoft, Apple, or IBM) good or evil, you just look at what they're actually doing.
So, let's look at what Apple and Microsoft are actually doing:
iPod/iPhone vs Zune/XBox: iTunes has weak DRM, Windows Media Player has strong DRM with kernel support that's getting stronger in Vista (trusted media path and tilt switches). None of their consumer entertainment products are open, but it's a lot cheaper and easier to get into software development for the iPhone than the XBox.
Open source: Microsoft uses GPLed software in Interix, and used OpenBSD extensively. They haven't released their versions of any of the BSD-licensed components they used in Interix (or Windows), but they do have a copy of their GCC source in the Interix tree. They have *recently* decided to accept the LGPL, but there's no GPLed software in Windows proper. Apple uses GPLed software in Darwin, and used FreeBSD extensively. They have released the open source code in Darwin and kept releasing it with every new release of OS X. They have added their own open source components, to the point where the majority of the traditional userland in OS X, as well as many major new components like launchd, are open source. Microsoft's open source poster boys are things like Windows installers.
This is like claiming that Gandalf's "turning evil" because he's wearing a grey robe, while cheering on Sauron for having the Nazgul stop off to pick up litter on the way back from scouring the Shire.
Mac OS X isn't my first choice for a server OS, I'd rather run FreeBSD straight without spiking it with Mach. But it's probably a better choice for small sites without much technical expertise.
I think a bigger issue may be Internet Sharing.
Let's say someone finds a new way to cut logs that let you build log cabins almost as easily but many times more sturdily and with better isolation. This would surely be patentable.
I assume you mean "better insulation". How much do you think it would cost to research such a process to the point where it was patentable? You have to actually cut logs and fit them together... even if you prototype the design in Second Life you're going to have to build it to tell if it actually works.
Let's try another example: Let's say someone patents an idea (putting adhesive strips on a piece of soft plastic to hold it against a curved monitor screen on a fish-finder to protect it), then later sees someone using van-der-waal forces to hold a flat piece of soft plastic against a flat touch screen, and files a modified patent that eliminates all the claims of his original patent (because the first few claims made a big point of the difficulty of attaching he plastic to the curved screen, and the use of adhesive strips) and replaces them with a patent for this handheld screen protector, then starts going after people who had been making these things long before he filed his amended patent?
The thing about software patents is that most of them are more like the second case than the first. They're overly broad, are based on simple and obvious ideas, and are frequently re-interpreted to apply to problems that nobody had even thought of when the original patent was filed.
If software patents were restricted to completed systems with comparable levels of research and development costs as your examples, I suspect there would be a good deal fewer objections to software patents. Some people might have the kind of hidden agenda you're talking about, but most people wouldn't care. The problem is that they're not, and most of them seem to be about things that are routinely implemented hundreds of times a day by people who have no idea that they're creating a "patentable invention".
I don't know how to come up with rules that would allow your "hard problems" patents without leaving the floodgates open. I'm not a lawyer. I don't play one on TV.
You got any ideas?
http://www.alibaba.com/product-gs/206720976/7_mini_laptop.html
$90-$180 FOB Shanghai, QTY 500. Runs Linux or Windows CE.
Looks like they have variants of this from 7" to 12.1", which is why the range of prices.
Commodore's end was nigh when Jack Tramiel defected and bought Atari.
So was Atari's.
Commodore's problem was the whole "Tramiel vs Gould" thing, and that kept going even after Tramiel left. Buying Atari simply roped Atari into the same doomed feud.
The only winners were the lawyers.
I thought about building a Hackintosh, but by the time I had my parts list down, I was nearing $2000 in parts. And even that wasn't at the quality of the Mac Pro.
If you're trying to build something comparable to a Mac Pro, you're not even in the same volume (let alone the same page) as most of the people building their own white boxes.
I mean, I've built a bunch of computers and the only time I went over $1000 in parts was in the early '90s...
Think "Mac mini Pro", not "Mac Pro".
Even though I'd be the first to agree that the Mini is overpriced for the specs, remember that it is a tiny computer. It's smaller than a laptop.
It's bigger than a lot of laptops (it's 3* as high as most!), and more expensive than a few. If you're looking for a small, quiet computer and don't care about the OS are you going to buy a Mac mini or are you going to buy a laptop? Let's see... for $600 you can get "Bring your own display, keyboard, and mouse" or you can get something with better specs than a Mini plus a keyboard, WXGA screen, trackpad, battery... or if you want to really cut the size done the EeePC 901 is pretty nice.
So, to return to the GP, yes, people do pay for the whole package.
Some people do. That's the point. People buy Macs for all kinds of reasons. But how many people really want the whole package, and how many are just putting up with the package because it's the only way to get the part they want?
Well, how's the *rest* of the package sell when OS X isn't in the picture?
Dismally.
If there really was significant demand for the Mac mini hardware, then you'd see more people buying Wintel PCs with that kind of form factor. People have sure tried to *sell* tiny PCs, from things like the Multia in the '90s to "Mac mini clones" today. The demand for that kind of machine just isn't there, or they'd have sold enough that you'd be used to seeing them along side the minitowers in every Office Depot in the US. The small size and accompanying lack of expandability are actually negatives for most people... at least most of the people I've tried to talk into getting a Mac.
SOME people buy Macs because they want that kind of hardware. But I don't believe MOST people do, because it just doesn't sell.
Name one open source project that Apple created that is useful?
launchd is the obvious one.
Yes, they were totally on the level and fulfilling the letter of their responsibilities, but it was hardly "bent over backwards".
What they were doing before the uproar was totally on the level and fulfilling their responsibilities, how they responded was far more than anyone had a right to expect. More importantly they were and have continued to make releases of their updated versions of numerous open source components that they are not obligated to.
I think the problem is that you're comparing Apple with the handful of "open source superstars", and I'm comparing them to the average company that uses FOSS. There are a number of companies that have done more than Apple, yes, but by far the majority of companies using open source projects do far less... most actively avoid using GPLed software at all, even LGPLed software, and many have policies in place to prevent employees contributing to open source projects. Many have to be sued by the FSF before they will even come near to meeting the letter of their responsibilities.
Yes, Apple has a history of being controlling. Every time there's a fuss about Apple's Open Source work, or about people using Apple's Open Source releases, I honestly expect them to dump all but the GPLed core of the Darwin open source repository. That would be completely consistent with Apple's past, and would still leave them being better than average in their open source policies among large US corporations. And yet they continue to release more software, and maintain their releases. Maybe, eventually, they won't, but right now while they are not a superstar like IBM... they're still way ahead of the pack.
What they make that people want to buy is software.
Their business model is to sell hardware at a high margin, but people *buy* the hardware because of the software. They wouldn't spend 40% more than a comparable pc or handheld if it was running Vista or Windows CE.
This is not a rare model, either. Cisco hardware is not exceptional, people buy it because it's running IOS. Cisco is a software company with a hardware business model. Network Appliance hardware is nothing special, but you can't run the Filer software without it. They're another company with a hardware business model.