Does anyone know of an easy way to reduce CPU/disks power consumption?
Isn't the power supply the main consumer of power in a PC? I mean an old PCs 400 watt power supply is running full time no matter how much power the CPU or disks are actually using. Is there some way to reduce the amount of energy being eaten by the power supply?
When I read the article earlier today, the pledge link was down (probably slashdotted) so I couldn't read the details. But the upshot is that there is at least a proposal to make the laptops available for $300.
After seeing the details, I agree there is no mechanism in place to sell them for that price at this time. But if OS X were in the picture, even making such a proposal would not be an option. At least Fedora gives OLPC the flexibility to make such offers in the future.
We believe health is the cornerstone of human development.
Funny how, when there is a healthcare crisis in other countries, the U.S. solution is to pour billions of dollars into direct healthcare services for those countries. But when there is a healthcare crisis in the U.S. ( U.S. has second worst newborn death rate in modern world), the U.S. solution is to privatize healthcare, maximize profits for the healthcare industry, and essentially tell its people to "go get an education and a job".
If education and employment is the solution for the healthcare crisis in the U.S., why is not not a plausible solution for other countries as well?
Possibly the biggest problem working on this laptop is its small 12' screen.
The classic Macintosh shipped with a 9" monochrome screen, sold millions of units, and revolutionized personal computing. If a 12" screen is the biggest problem of this unit, I'd say we may be looking at a new personal computing revolution.
1) You can't run OSX on a 400 MHz AMD processor with 128 Meg of RAM. (If you know how, please let me know!)
2) Apple would never allow an OSX laptop to retail at $300 in the U.S., which is what OLPC is doing.
3) One of the design goals of OLPC was to be totally open source, to allow third parties to tinker with it and improve the entire system at will. I don't see how OSX could be part of a purely open-source project.
The author says he needs to share complex Excel spreadsheets with his office. The vast majority of "ordinary" PC users don't even know what Excel is. And for those who do know what it is, the most complex spreadsheet they ever see is their Christmas card list. Besides, his inability to share files with his office has more to do with Excel's proprietary file format than anything Linux can be blamed for.
As for his multimedia woes, that is the result of not chosing the right Linux distribution for the job. Free Linux distros will never be abled to play encrypted DVDs. That is a decision of the movie industry that choses to use proprietary encryption formats, that is not the fault Linux. The author even admits there is a $40 version of Linux that chould do the job that he wants, but instead of using that distro he wastes his time on six other distros that clearly will not provide the functions he wants in their standard configuration.
Rather than blaming Linux for not being able to use proprietary file formats for free, he should be questioning the computer industry's over-reliance on proprietary formats. And when trying out Linux he should choose the distro that fits his needs, not complain about the half-dozen other distros don't do what he wants them to do.
Whatever solution you come up with, it is going to be pain in the neck and is only going to be the beginning of your problems. Once you get the OS running everywhere, you get the fun task of making sure all the apps perform adequately everywhere on your various hardware and OS configurations. Then you get to do it all over again when the OS has a patch or you need to tweek an application, which in my experience will be about once a day.
The common solution to this is to throw all your computers away and move all of your desktops to identical hardware. But why should we fill our landfills with perfectly working computers that will only leak toxic waste once discarded, just because Windows is a pain in the arse to administer?
Here's what I would do: Move all of your critical enterprise applications to some web application environment. I don't care which one:.NET, J2EE, Zope, Zend, Ruby on Rails, whatever. Then all you have to worry about is having a functioning web browser on your desktops. Problem is solved. And when you want to do an upgrade or modify an app, just update the web server once and you are done.
Not only does that solve your current problem but it makes your future plan to move to Linux on the desktop a breeze. Your web apps won't care what OS your browser is running in. (Well, at least as long as you don't choose.NET!)
If you want to deliver applications to the desktop, why waste your time playing the MS/Intel perpetual upgrade gauntlet? Just deliver your apps inside a web application environment and be done with it.
My own pet peeve is when headlines talk about "disability access" or "disabled people" when the article is all about blind people. Certainly linux is no better or worse than Windows for physically impaired, deaf, or learning disabled users. It is blind users who are running into a bit of trouble. I wish people would be more clear that it is blind users whose needs are unmet, not clump the issue together with all other disabilities.
That said, the issue with blind access is not an OS problem. Microsoft has gone to no more effort to make Windows accessible than the FOSS community has gone to make linux accessible. The issue with Linux access for the blind is twofold: third-party vendors and the way blind people are taught computers.
Windows is usable by the blind not becuase Microsoft makes it that way, but becuase Microsoft shares their code with third-party developers who write and sell the software to make Windows work for the blind. Guess what? FOSS community shares it's code too -- much more than Microsoft does! The problem isn't that FOSS people haven't done enough, the problem is that no third-party developer has stepped forward to write the software that blind people need. Maybe they haven't found a business model that will work with FOSS. More likely -- the blind market is so small and the Linux market is so small and Linux distros are so splintered that there is no way to justify the amount of work it would take for such a small potential market.
In addition to the difficulty in mounting an effort for software for the blind, you have to look at how blind people are taught computers. There are a few that learn from their friends. But the vast majority of blind people I know (I've provided computer support to blind users for 15 years) are introduced to computers at school through Special Education programs, then receive further training after school through Vocational Rehab services. There are a growing number of linux software tools for the blind, but I know of no school system or voc rehab service that is prepaired to introduce these tools to their blind clients. They are all Microsoft based.
Linux may not be accessible to the blind, but the problem is not with Linux itself. The problem is that no big third-party provider has stepped up to write linux software, and no schools or training centers are prepared to teach linux to blind users.
Sure, it starts as "one laptop per child" but I'm sure these laptops are going to be picked up and fiddled with by moms and uncles once they get out there. And many of these moms and uncles may be more literate than you give credit. I'm thinking specifically about Latin America where most communities have some adults who've had access to basic education, even though the community itself remains desperately poor and any opportunities that might pop up are usually quashed by a corrupt local government.
For people that are mildly literate and fairly bright, all you have to do is give them a networked computer, show them the internet, and point out that HTML, JS, and PHP exist. It is amazing what people can do with it when they take it from there -- all the documentation and tutorials you need can be found for free right on the internet. Here in the U.S. we have 14-year-olds writing slick, decent web sites with just that set of skills. And I dare say that is exactly the way the majority of "professional" web developers today, including myself, got started. I don't see why adults in other parts of the world couldn't do the same thing.
Regarding your main point: "I've seen poor people climb out of poverty with no help from anyone, just because a simple opportunity opened up near them." -- Isn't it possible that access to a networked laptop would actually open up opportunity to thousands of people in poverty?
Most of these people are pretty bright, just very poor. Why not set up their laptops with Nvu, teach them a little HTML, CSS and Javascript -- maybe even PHP, Python or Ruby, and set them loose as website developers. They could sell their services for pennies on the dollar to developed nations -- all through the internet, making it easier to avoid the corruption and control of their local government.
Their prices for web development would be so low they could be in very high demand in industrial nations, yet still they would make more money than they have any hope for now in their current situation.
That's just one idea. I think the link between networked computers and opportunity may be greater than you think.
ALL retail sales are an auction with the list price being the "buy it now" price. That is called the "free market". I walked into Best Buy last year, told them I wanted to buy a washer, dryer, fridge and stove all at once, and asked if would they take 10% off of the total list price for me. The salesrep looked through his books, and agreed. In some countries, negotiating over listed price for consumer products is common practice. In the U.S. it isn't -- except, oddly, for cars.
Certainly the sticker price for a car is the "buy it now" price, even though the vast magority of people will make a counter-offer that the car dealer will accept.
HTML is certainly not very standard and some implementations of it aren't even very open. HTML is basically a bastardization of SGML, while XML is a well-defined, strict subset of SGML. You can write HTML in XML if you want to make it open and standard, but you'd still be writing in XML.
Open Document Format, on other hand, is a strict XML format so it is both open and standard.
Sure, plain text is open and standard too, but most applications require a more structured document than you can get with plain text.
Are the staffing issues associated with OSS enough to outweigh the benefits?
Simple answer: NO. There are staffing issues with any computer infrastructure. Those issues have more to do with how well the thing is managed than why software or platform it is based on. There is no reason why F/OSS can't be just as efficient to manage as proprietary software.
To really see what this report is about, just look at the funders: http://www.imsglobal.org/members.html. It is a who's who of commercial educational software vendors, including Blackboard, WebCT, and, yes, Microsoft. That is all you need to know about this report.
I will agree that the weight of an emac makes it a low theft risk. But I don't know any personal computer product that survives by using extreme weight as an asset. And remember, the teachers in the classroom need to move the computer around for overhead presentations. Can't do that with an emac. Other than weight, I don't see much difference between an emac and a mini.
I think the ideal student lab would be a bunch of minis locked down to the desktop. Or better yet, mounted under the desktop out of sight.
Interestingly, Apple is now featuring the iMac on its education site. But I can't imagine selling $1200 Macs to the administration when $500 windows boxes are easy to come by.
I don't think apple will be bringing the emac back, and schools are going to have to learn to deal with the theft issue -- just like they already do with that laptops that all the administrators carry around.
The new imacs now start at $1,199.00, which, if I recall, is as cheap as emacs ever were. And for $800 a school could set up a mini with an LCD display. So there doesn't seem to be much need for emacs any more.
It's not just engineers. It is the worker bees in any organization.
I happen to work for a school district. Our main (only?) job is supposed to be teaching students. Yet our administration spends almost no time asking teachers what they need to be more productive. They come out with these great initiatives that they've heard or read about somewhere and the teachers are supposed to fall in line.
Then they wonder why teachers aren't more "productive". Its quite simple to figure out. The teachers spend half of the time jumping through hoops that the administration has set up, rather than the administration providing what the teachers need to do their job.
Teachers know how to teach, engineers know how to engineer, what the suits are supposed to do is listen to what those people need and provide it for them.
Instead, and especially with IT, the administration thinks they know it all and can tell the teachers what to do. It should be the other way around!
Since it has already been widely reported that the Chinese censorship of Google can be easily circumvented with common typos, having Google enforce such ineffective censorship will eventually weaken Chinese censorship efforts and make them irrelevant.
The Chinese government is acting like the parent who, because they run NetNanny or Cyberpatrol while their kids roam the web, they can magically prevent their kids from ever seeing pornography. All the kid has to do is look at Google cache and they can find all the porn they want, no matter what filter is in place. I don't care what filter you user -- once you have access to Google, there is no way to stop you from eventually finding what you are looking for.
Likewise the Chinese government. Having a filtered Google that can easily be circumvented will result in the propogation of the truth to the Chinese people, not the hiding of truth that the Chinese government expects. Once the people have access to the truth through Google, censorship becomes irrelevant.
The Chinese people benefit more from a Google that is ineffectively censored than they do from no Google at all.
Screen readers for Office uniformly suck, partly because they have to hack Office and reverse-engineer the Word format to work (just like OpenOffice).
I agree that current screen readers suck, but it is not because they have to reverse-engineer a proprietary format. Freedom Scientific, the largest maker of Screen Readers in the world, is a close Microsoft partner and has full access to the Microsoft Windows and Word specs to develop their product. Screen readers suck because turning a 2-dimensional graphic interface into 1-dimensional speech is practically impossible, not because the document structure is proprietary.
ODF's free spec will allow people to build screen readers that actually work properly.
I believe it may be possible to build a better screen reader with ODF, but where is it? Where is the robust screen reader for ODF? Saying that something may be better than the Microsoft solution some day is no different than the vapor-ware claims that Microsoft makes all the time. Bad as it is, at least Word has a screen reader right now, ODF doesn't.
If you think you can build a better screen reader, then do it. The ODF specs are fully available, and Blind people the world over will thank you. But the problem with screen readers is not the availability of specs. The problem is it is impossible to squeeze two dimensions into one.
An increase in spending is not a cut in spending.
But the article never said there was a cut in spending. The article said the government is cutting back on college loans. Which is true -- the government is cutting back the rate of increase in college loans.
All the fuzzy-warm-happy programs handing money out to people are going to get more next year
The fact that education and food are considered "fuzzy-warm-happy programs" speaks volumes about the current state of our country.
This political nonsense has nothing to do with the switch to digital tv.
The switch to digital tv is one of several current technology proposals that needs more political scrutiny, not less. Government funding priorities has everything to do with this issue. Switching to digital TV would not even be an issue if government funding priorities were not in play.
Let's not play games with semantics. Cutting back an increase certainly is cutting back. It is taking action to reduce something to less than what would have occured without the action. There are many ways to cut back. You can cut back a number or you can cut back the rate of increase in a number -- they are both cut backs. The author wasn't specific about what type of cut back is occuring, but a cut back in a rate of increase is still a cut back.
By not positioning it as a replacement for a Windows PC, but as an additional accessory, it doesn't have to replace every esoteric piece of software available for Windows.
That is exactly what Apple figured out at least five years ago. That is why Apple is staking its future on media-centered products like the iPod, Front Row, and the soon-to-be-announced Media Center Mac Mini
And it is what I've already figured out in my household. I keep one old Windows machine around to run the CDs that my kids get on the back of cereal boxes. Anyone can pick up an adequate machine for this purpose out of garbage bins. Then I have Mac and Linux machines around when I want to do real work -- organizing family movies and photo albums on the Mac and doing web development in Linux.
Packaging windows alternatives as media appliances rather than Windows replacements is the proven way to get non-MS computers into people's homes. Apple is already getting rich doing it. There is plenty of room for Linux and others to play a role on the lower-price side of the same market.
Universities have been doing enterprise deployment of Macintosh desktops since before MS Windows existed, even since before the term "enterprise deployment" existed. The early days were pretty rough, of course. But now OSX and its Unix core have a full set of enterprise desktop management tools available. Check out MacEnterprise and tools like Radmind. Many Universities are now doing large-scale deployments of Mac desktops. And since those tools are all Unix-based, perhaps many of the same techniques will work for Linux desktops?
When are they going to fix the way Outlook Web Access runs on non-IE browsers? The current non-IE OWA is a 1990's-era frame-based piece of crap that lacks any ability to spell-check emails or search mailboxes. And they try to pass it off as a state-of-the-art web-based email client. What a joke.
If your offended by the price, you'd be doubly offended to know that that is almost always paid for by your tax dollars. Most screen readers are paid for either by schools or by public vocational service programs. Very few blind people, especially students, are able to shell out that kind of money to make a $400 computer accessible.
There is a big opportunity for some open-source solution to step in here. And the government sector would gobble it up once it became workable -- they're the ones shelling out big bucks on current screen readers.
Macintosh Tiger has a rich screen reader built-in. There are several open source projects underway for screen readers in KDE. But none of these new alternatives will have any success until the blind users are aware of them, teachers of the blind begin incorporateing these new products into their lesson plans, and employers of the blind figure out how to use these new products to make their blind employees productive.
That is why the JAWS/Windows combination is so pervasive right now. Not because it is the only solution, and certainly not because it is affordable. Only because no one else has ben able to piece together a decent screen reader on another platform and integrate it into real-world school and work environments.
Slashdot Mirror
Isn't the power supply the main consumer of power in a PC? I mean an old PCs 400 watt power supply is running full time no matter how much power the CPU or disks are actually using. Is there some way to reduce the amount of energy being eaten by the power supply?
After seeing the details, I agree there is no mechanism in place to sell them for that price at this time. But if OS X were in the picture, even making such a proposal would not be an option. At least Fedora gives OLPC the flexibility to make such offers in the future.
Besides, points 1) and 3) still stand.
Funny how, when there is a healthcare crisis in other countries, the U.S. solution is to pour billions of dollars into direct healthcare services for those countries. But when there is a healthcare crisis in the U.S. ( U.S. has second worst newborn death rate in modern world), the U.S. solution is to privatize healthcare, maximize profits for the healthcare industry, and essentially tell its people to "go get an education and a job".
If education and employment is the solution for the healthcare crisis in the U.S., why is not not a plausible solution for other countries as well?
Possibly the biggest problem working on this laptop is its small 12' screen. The classic Macintosh shipped with a 9" monochrome screen, sold millions of units, and revolutionized personal computing. If a 12" screen is the biggest problem of this unit, I'd say we may be looking at a new personal computing revolution.
2) Apple would never allow an OSX laptop to retail at $300 in the U.S., which is what OLPC is doing.
3) One of the design goals of OLPC was to be totally open source, to allow third parties to tinker with it and improve the entire system at will. I don't see how OSX could be part of a purely open-source project.
As for his multimedia woes, that is the result of not chosing the right Linux distribution for the job. Free Linux distros will never be abled to play encrypted DVDs. That is a decision of the movie industry that choses to use proprietary encryption formats, that is not the fault Linux. The author even admits there is a $40 version of Linux that chould do the job that he wants, but instead of using that distro he wastes his time on six other distros that clearly will not provide the functions he wants in their standard configuration.
Rather than blaming Linux for not being able to use proprietary file formats for free, he should be questioning the computer industry's over-reliance on proprietary formats. And when trying out Linux he should choose the distro that fits his needs, not complain about the half-dozen other distros don't do what he wants them to do.
Whatever solution you come up with, it is going to be pain in the neck and is only going to be the beginning of your problems. Once you get the OS running everywhere, you get the fun task of making sure all the apps perform adequately everywhere on your various hardware and OS configurations. Then you get to do it all over again when the OS has a patch or you need to tweek an application, which in my experience will be about once a day. The common solution to this is to throw all your computers away and move all of your desktops to identical hardware. But why should we fill our landfills with perfectly working computers that will only leak toxic waste once discarded, just because Windows is a pain in the arse to administer? Here's what I would do: Move all of your critical enterprise applications to some web application environment. I don't care which one: .NET, J2EE, Zope, Zend, Ruby on Rails, whatever. Then all you have to worry about is having a functioning web browser on your desktops. Problem is solved. And when you want to do an upgrade or modify an app, just update the web server once and you are done.
Not only does that solve your current problem but it makes your future plan to move to Linux on the desktop a breeze. Your web apps won't care what OS your browser is running in. (Well, at least as long as you don't choose .NET!)
If you want to deliver applications to the desktop, why waste your time playing the MS/Intel perpetual upgrade gauntlet? Just deliver your apps inside a web application environment and be done with it.
That said, the issue with blind access is not an OS problem. Microsoft has gone to no more effort to make Windows accessible than the FOSS community has gone to make linux accessible. The issue with Linux access for the blind is twofold: third-party vendors and the way blind people are taught computers.
Windows is usable by the blind not becuase Microsoft makes it that way, but becuase Microsoft shares their code with third-party developers who write and sell the software to make Windows work for the blind. Guess what? FOSS community shares it's code too -- much more than Microsoft does! The problem isn't that FOSS people haven't done enough, the problem is that no third-party developer has stepped forward to write the software that blind people need. Maybe they haven't found a business model that will work with FOSS. More likely -- the blind market is so small and the Linux market is so small and Linux distros are so splintered that there is no way to justify the amount of work it would take for such a small potential market.
In addition to the difficulty in mounting an effort for software for the blind, you have to look at how blind people are taught computers. There are a few that learn from their friends. But the vast majority of blind people I know (I've provided computer support to blind users for 15 years) are introduced to computers at school through Special Education programs, then receive further training after school through Vocational Rehab services. There are a growing number of linux software tools for the blind, but I know of no school system or voc rehab service that is prepaired to introduce these tools to their blind clients. They are all Microsoft based.
Linux may not be accessible to the blind, but the problem is not with Linux itself. The problem is that no big third-party provider has stepped up to write linux software, and no schools or training centers are prepared to teach linux to blind users.
For people that are mildly literate and fairly bright, all you have to do is give them a networked computer, show them the internet, and point out that HTML, JS, and PHP exist. It is amazing what people can do with it when they take it from there -- all the documentation and tutorials you need can be found for free right on the internet. Here in the U.S. we have 14-year-olds writing slick, decent web sites with just that set of skills. And I dare say that is exactly the way the majority of "professional" web developers today, including myself, got started. I don't see why adults in other parts of the world couldn't do the same thing.
Regarding your main point: "I've seen poor people climb out of poverty with no help from anyone, just because a simple opportunity opened up near them." -- Isn't it possible that access to a networked laptop would actually open up opportunity to thousands of people in poverty?
Most of these people are pretty bright, just very poor. Why not set up their laptops with Nvu, teach them a little HTML, CSS and Javascript -- maybe even PHP, Python or Ruby, and set them loose as website developers. They could sell their services for pennies on the dollar to developed nations -- all through the internet, making it easier to avoid the corruption and control of their local government.
Their prices for web development would be so low they could be in very high demand in industrial nations, yet still they would make more money than they have any hope for now in their current situation.
That's just one idea. I think the link between networked computers and opportunity may be greater than you think.
Certainly the sticker price for a car is the "buy it now" price, even though the vast magority of people will make a counter-offer that the car dealer will accept.
Open Document Format, on other hand, is a strict XML format so it is both open and standard.
Sure, plain text is open and standard too, but most applications require a more structured document than you can get with plain text.
Simple answer: NO. There are staffing issues with any computer infrastructure. Those issues have more to do with how well the thing is managed than why software or platform it is based on. There is no reason why F/OSS can't be just as efficient to manage as proprietary software.
To really see what this report is about, just look at the funders: http://www.imsglobal.org/members.html. It is a who's who of commercial educational software vendors, including Blackboard, WebCT, and, yes, Microsoft. That is all you need to know about this report.
I think the ideal student lab would be a bunch of minis locked down to the desktop. Or better yet, mounted under the desktop out of sight.
Interestingly, Apple is now featuring the iMac on its education site. But I can't imagine selling $1200 Macs to the administration when $500 windows boxes are easy to come by.
I don't think apple will be bringing the emac back, and schools are going to have to learn to deal with the theft issue -- just like they already do with that laptops that all the administrators carry around.
The new imacs now start at $1,199.00, which, if I recall, is as cheap as emacs ever were. And for $800 a school could set up a mini with an LCD display. So there doesn't seem to be much need for emacs any more.
I happen to work for a school district. Our main (only?) job is supposed to be teaching students. Yet our administration spends almost no time asking teachers what they need to be more productive. They come out with these great initiatives that they've heard or read about somewhere and the teachers are supposed to fall in line.
Then they wonder why teachers aren't more "productive". Its quite simple to figure out. The teachers spend half of the time jumping through hoops that the administration has set up, rather than the administration providing what the teachers need to do their job.
Teachers know how to teach, engineers know how to engineer, what the suits are supposed to do is listen to what those people need and provide it for them.Instead, and especially with IT, the administration thinks they know it all and can tell the teachers what to do. It should be the other way around!
The Chinese government is acting like the parent who, because they run NetNanny or Cyberpatrol while their kids roam the web, they can magically prevent their kids from ever seeing pornography. All the kid has to do is look at Google cache and they can find all the porn they want, no matter what filter is in place. I don't care what filter you user -- once you have access to Google, there is no way to stop you from eventually finding what you are looking for.
Likewise the Chinese government. Having a filtered Google that can easily be circumvented will result in the propogation of the truth to the Chinese people, not the hiding of truth that the Chinese government expects. Once the people have access to the truth through Google, censorship becomes irrelevant.
The Chinese people benefit more from a Google that is ineffectively censored than they do from no Google at all.
Any school that is not managed as an enterprise is wasting its students' money.
I agree that current screen readers suck, but it is not because they have to reverse-engineer a proprietary format. Freedom Scientific, the largest maker of Screen Readers in the world, is a close Microsoft partner and has full access to the Microsoft Windows and Word specs to develop their product. Screen readers suck because turning a 2-dimensional graphic interface into 1-dimensional speech is practically impossible, not because the document structure is proprietary.
ODF's free spec will allow people to build screen readers that actually work properly.
I believe it may be possible to build a better screen reader with ODF, but where is it? Where is the robust screen reader for ODF? Saying that something may be better than the Microsoft solution some day is no different than the vapor-ware claims that Microsoft makes all the time. Bad as it is, at least Word has a screen reader right now, ODF doesn't.
If you think you can build a better screen reader, then do it. The ODF specs are fully available, and Blind people the world over will thank you. But the problem with screen readers is not the availability of specs. The problem is it is impossible to squeeze two dimensions into one.
An increase in spending is not a cut in spending. But the article never said there was a cut in spending. The article said the government is cutting back on college loans. Which is true -- the government is cutting back the rate of increase in college loans. All the fuzzy-warm-happy programs handing money out to people are going to get more next year The fact that education and food are considered "fuzzy-warm-happy programs" speaks volumes about the current state of our country. This political nonsense has nothing to do with the switch to digital tv. The switch to digital tv is one of several current technology proposals that needs more political scrutiny, not less. Government funding priorities has everything to do with this issue. Switching to digital TV would not even be an issue if government funding priorities were not in play.
Let's not play games with semantics. Cutting back an increase certainly is cutting back. It is taking action to reduce something to less than what would have occured without the action. There are many ways to cut back. You can cut back a number or you can cut back the rate of increase in a number -- they are both cut backs. The author wasn't specific about what type of cut back is occuring, but a cut back in a rate of increase is still a cut back.
That is exactly what Apple figured out at least five years ago. That is why Apple is staking its future on media-centered products like the iPod, Front Row, and the soon-to-be-announced Media Center Mac Mini
And it is what I've already figured out in my household. I keep one old Windows machine around to run the CDs that my kids get on the back of cereal boxes. Anyone can pick up an adequate machine for this purpose out of garbage bins. Then I have Mac and Linux machines around when I want to do real work -- organizing family movies and photo albums on the Mac and doing web development in Linux.
Packaging windows alternatives as media appliances rather than Windows replacements is the proven way to get non-MS computers into people's homes. Apple is already getting rich doing it. There is plenty of room for Linux and others to play a role on the lower-price side of the same market.
Universities have been doing enterprise deployment of Macintosh desktops since before MS Windows existed, even since before the term "enterprise deployment" existed. The early days were pretty rough, of course. But now OSX and its Unix core have a full set of enterprise desktop management tools available. Check out MacEnterprise and tools like Radmind. Many Universities are now doing large-scale deployments of Mac desktops. And since those tools are all Unix-based, perhaps many of the same techniques will work for Linux desktops?
When are they going to fix the way Outlook Web Access runs on non-IE browsers? The current non-IE OWA is a 1990's-era frame-based piece of crap that lacks any ability to spell-check emails or search mailboxes. And they try to pass it off as a state-of-the-art web-based email client. What a joke.
There is a big opportunity for some open-source solution to step in here. And the government sector would gobble it up once it became workable -- they're the ones shelling out big bucks on current screen readers.
Macintosh Tiger has a rich screen reader built-in. There are several open source projects underway for screen readers in KDE. But none of these new alternatives will have any success until the blind users are aware of them, teachers of the blind begin incorporateing these new products into their lesson plans, and employers of the blind figure out how to use these new products to make their blind employees productive.
That is why the JAWS/Windows combination is so pervasive right now. Not because it is the only solution, and certainly not because it is affordable. Only because no one else has ben able to piece together a decent screen reader on another platform and integrate it into real-world school and work environments.