If OSNews is reporting this (I can't get to OSNews right now), then they appear to be exaggerating a comment by Prabhakar. The only posting I can find from him on the entire list in the last two months is this one.
In it, he simply makes it clear that no decision has been made public yet. This is what he writes:
Hi all,
Just to be clear, Tom Yager was *speculating* about why we have -- so far -- not released the source code of the kernel for Intel-based Macintoshes. We continue to release *all* the Darwin sources for our PowerPC systems, and so far has released all the non-kernel Darwin sources for Intel.
Nothing has been announced, so he (and everyone else) certainly has the right to speculate. But please don't confuse "speculation" with "fact."
Thanks,
-- Ernie P.
If the intent is to release the source to XNU for Intel, then Prabhakar went a long way to avoid saying explicitly that it's going to happen. He most certainly did NOT say this was a matter of them not having released it "yet".
There are multiple reasons why XNU for Intel might not have been released, from over-paranoid and not terribly bright PHB types thinking that somehow it would help hackers (because hackers never hack the copy-prevention measures of binary-only programs...) to support for an unannounced feature, like Virtualization, being present in all new 10.4.1+ kernels. But the way I read Prabhakar, it sounds like more of the former type of issue, with key figures in Apple being against the release of the source now and in the future, and with the final decision not being made. If it was as simple as the "unannounced new feature", I think he'd explicitly say Apple plans to release the source, wouldn't he?
That's standard Intel branding. The names tend to go with specific markets, price points, and relative power (to other contemporary CPUs) rather than architecture.
The Pentium and Celeron would spring to mind. There's little in common between the Pentium 1, the Pentium 3, and the Pentium 4. The Pentium Pro, 2, 3, and M, are to some extent architecturally related, albeit loosely. The Celeron has been the "cut down" version of whatever the current Pentium was. Meanwhile, the Xeon exists in so many forms it's hard to tell what it is these days.
Intel's in the process of a rebranding, but they're not doing away with the stategy, just getting away from the Pentium/Celeron brands. The Core 2 will replace the current Core. That's why it's still a Core, even when it's been re-engineered to be percentage% faster per whatever-the-measure-is-today and contain a significant amount of 64-bit stuff. Allegedly, it'll be a drop in replacement, OEMs will not even have to redesign their motherboards.
But that's because I'm interested in laptops as mobile desktops that play games occasionally. The more serious you are with a laptop, the less interested you are in games and the more you're interested in battery life.
I criticised Apple for bunging a GMA950 in the Mac mini. I still think it's the wrong choice. A populist computer that's poor at modern games? That doesn't make a lot of sense. But a GMA950 is perfect for a laptop for most people. It performs very well when it comes to accellerating Mac OS X's graphics, video, and a handful of other applications, and it's exceptionally low power. The only thing I can think of that would be "better" (but it'd take some work on Apple's side to make it work cleanly) would be to have a decent graphics card in there but only enable it when it needs to be used, having it powered off otherwise (the GMA950 is actually present in all Intel Apple computers, it's built-in to the chipset, it's just not used in those with Radeons)
I think the criticism of Apple for using the GMA950 in the Macbook is overblown. It's a good choice for a laptop GPU. The criticisms of the Mac mini need to be kept up, but they shouldn't translate into a "GMA950 always bad, X300 always good" mentality, because that just isn't the case.
I would. I mean its pretty much between RMS and Linus, and I would give it to Linus.
You would choose between two people who aren't even in the running?
RMS doesn't consider himself a supporter of Open Source. And Linus Torvalds has made it clear on several occasions, most infamously with the BitKeeper debacle, that he's NOT an Open Source, or Free Software, advocate, he's quite happy with proprietary software.
I'm not sure who qualifies as an Open Source leader. There's the official one, Michael Tiemann, the current head of the Open Source Initiative. There's the most vocal one, Eric S. Raymond, who arguably was the person who made the single biggest contribution, for better or worse, to popularizing the Open Source concept. And then there's Bruce Perens. In terms of people heading specific projects, it's hard to really come up with a name. Torvalds simply is not in the running, he's popular, but he's not someone anyone who considers FS or OS important takes seriously when it comes to pronouncements on the future of Open Source because he's not an Open Source advocate. Other "Open Source" projects, be they Apache, Mozilla, or whatever, have less well known characters in charge, and generally the few that do have fairly controvertial characters (OpenBSD's Theo springs to mind) rather than natural leaders.
It would be great if Torvalds took the concept more seriously and stepped into the shoes so many people want him to wear. He clearly has the charisma, but he's no more the leader of Open Source than John McCain is leader of the Republican Party.
Apple loses out on the thing they do BEST, which just so happens to be their biggest profit margin.
Not using the figures you quoted. They gain on the thing they do BEST (selling 2,750,000 copies of OS X vs 1,000,000), though this isn't their profit margin either. They lose 25% of hardware sales, but I don't think anyone can seriously argue that hardware is something Apple does "best" and I seriously doubt you're trying to argue that!
GPRS (and EDGE) has relatively high latency, it's that, not bandwidth, that's the issue.
I'm not sure how the CDMA and WCDMA based alternatives are with latency. It may be that 802.16 works out better than any "grafted onto a cellphone network" solution anyway.
Piracy is a synonym for copyright infringement. It has been since before I was born, and copyright infringement appears as a valid definition in every dictionary I own.
So give it a rest. You're the umpteenth slashbot to "correct" someone for using the term "pirate" correctly. It's pretty hard to believe that deep down you're unaware that the term does, actually, mean "copyright infringment" pretty much everywhere.
25 hours? I suspect most pirated copies are people downloading the latest cracked version, and then installing it on their PCs. At that point, it either works or it doesn't.
The people who are doing the cracks themselves may be taking 25 hours, but I doubt the majority of people using the cracks are.
A better way to word this is: Why give legitimate users a version of Darwin for x86 when pirates are perfectly capable of creating illegal copies of Mac OS X?
To which, I guess the answer is: Flubajugalubums.
I honestly doubt closing the sources will make much of a difference. There's an awful lot of pirated binary-only programs out there. Meanwhile, those with legitimate uses are locked out. Doesn't really make much sense to me.
Perhaps a better solution would be to deal with the piracy issue by, well, catering to a clear demand. What's better, 1,000,000 MacBook sales, and 10,000,000 people with illegal copies of Mac OS X on their Dells, or 1,000,000 MacBook sales, 2,000,000 people with paid-for, unsupported, $100 copies of Mac OS X on their Dells, and 8,000,000 illegal copies? If people are going to do it anyway, and they are, you might as well make some money from them.
Adding the extensions to Kaffe wouldn't disrupt Java, as the extensions would be open, documented, and thus possible to incorporate into a standard distribution. It might disrupt Sun's control over Java, but that would be over a period of years. The fundamental concept of "Write once, run everywhere" wouldn't be affected, and if the extensions were useful, Java itself would become more viable at the end of it, not less.
It's possible Microsoft were trying "to help their case", I'll give you that. It would be nice to think of Microsoft as occasionally good, once in a while. Maybe they'll inject a little funding into Mono one day, with no court case to encourage them.
I wasn't aware of that. That kind of undermines the argument that Microsoft was trying the "extend and embrace" route to killing Java with proprietary incompatibilities, and suggests their motive in creating extensions really was to improve it. They just went about it the wrong way.
Or maybe they didn't, maybe they presented their improvements to Sun, which at the time didn't really have a community development process, and Sun rejected them.
Either way, supporting an Free Software project and documenting their extensions and asking the FS project to include these features doesn't mesh with the E&E strategy they've always been painted as using.
If you had any idea about the state of Java in the IT industry...
Who are you addressing? Because you quote a few words from my comment, and then talk about "open sourcers" "complaining" no matter what Sun does, something that has nothing to do with anything I wrote. Indeed, depicting me as yet another "open sourcer" is utterly ridiculous given my comments.
The full comment was, of course:
The real problem with Sun's strategy hasn't been forsaking the development model advantages of the OSI's "Open Source", it's been that it's harder to integrate the official Sun Java, the reference implementation, with the non-Java world, because of licensing issues.
That's absolutely right and I dare you to find holes in that, rather than dubious and irrelevent complaints about "open sourcers".
No, GPL's fine. It will help prevent a certain type of "embrace and extend" forking.
As far as ensuring the GPL license doesn't leak into applications that run over the Java run-time system, you simply supply an additional, optional, license, that allows for linking code with a pre-built binary distribution of Sun's Java using the published APIs. Developers can choose which they use on a project-by-project basis.
It depends on whether they prohibit or merely discourage forking. Indeed, Sun could even go the trademark route with some success, with only the official Sun Java, and specific licensees (such as creators of alternative Java implementations that conform to the spec) being allowed to use the trademark. This is compatible with the GPL. The fact you can't call your fork "Java" doesn't mean your freedom to change and distribute it has been affected.
There's a more interesting issue here. Sun Java is an embarassment to the OSI. Over the last few years, by using a community driven development process, Java has improved leaps and bounds. Essentially, Sun said "What the Open Source movement says is right, except for the freedom part". And given the OSI keeps being at pains to argue that it's merely a front for software freedom, trying to encourage the development of free software by promoting community-driven development processes which, supposedly, rely upon the software being developed to be Free, this really doesn't hasn't helped it much.
Essentially, the OSI says "We must have free software, because free software means a community of interested parties can develop a program to a much higher standard than would otherwise be the case if it was proprietary. We describe this whole thing as "Open Source"."
Sun responds with: "Aha! But Java isn't free, and it too is developed by a community of interested parties, and they've generated a much higher standard of product than would otherwise have been the case if it wasn't developed using a community process. So your argument fails because you don't need software to be free to use your "open source" development model!"
ESR responds with: "You all suck. Set Java free!!!1!"
So why's Sun "open sourcing" Java? I think they're just looking at ensuring the official Sun implementation has wider adoption, by removing licensing barriers. Free software licenses happen to be a great way to get there. Sun wants to get Java "out there", especially with.NET nipping at its heels. The real problem with Sun's strategy hasn't been forsaking the development model advantages of the OSI's "Open Source", it's been that it's harder to integrate the official Sun Java, the reference implementation, with the non-Java world, because of licensing issues.
And as such, I don't think Sun gives a rats arse what the OSI thinks.
Macs don't BSOD. They GSOD. The screen gets dimmed and a large gray box appears telling you in three different languages to hold down the power button until the machine turns off.
I should know. It happens to me once every few days. Most likely explanation? Hardware failure. I play games a lot, causing the fans to go full blast, and crashes generally follow that.
Most likely explanation for the BSODs on XP? A combination of two: hardware failures, as with OS X, and (to a much lesser extent) viruses. Viruses don't exist for the Mac for the most part because it's difficult to get a virus to spread when 95% of the computers it would "hit" are incapable of passing it on.
Can we cut out the BSOD BS? Both Macs and PCs are, these days, more or less equally vulnerable to them. 99% of BSODs are almost certainly caused by hardware issues, such as corrupted RAM.
Neither company has "innovated" yet. Neither the PS3 nor the Wii are actually on sale. Innovation doesn't occur until the seemingly-new technology is in people's hands.
It *might* be a Firefox bug. It might equally be a GMail bug. You have no way of knowing if it's Firefox that's not freeing unused memory, or GMail not clearing references to data it no longer needs.
GMail is a Javascript application, and it does allocate memory. Without someone pointing at specific code and saying "This is where the leak is occuring", it's simply not possible to point at either Firefox or GMail and say "This is where the bug's occurring".
Garbage collectors are not magic. They don't release memory where there's a reference to it. If an extension is storing references to objects on a window that's subsequently closed, and the extension doesn't then remove those references, the GC isn't going to discard the objects, because as far as it's concerned, they're still in use.
Because most Windows users run as admin and thereby have full access to every spot on the HD, also means that any code they may pick up somewhere, either purposely or by someone's stealth, will have free reign of the ENTIRE computer. On Macs that is not the case and that alone makes Macs more secure.
Most Mac users run as admin too. While it's slightly harder to access the entire hard disk in the Mac's version of admin, key areas such as/Applications can be accessed and modified as an admin user without further authentication.
There are some ridiculously stupid aspects to certain subsystems of Windows that create big holes (ActiveX springs to mind), but while I don't like ordinary users being "admin" either, the fact is 99% of Mac OS X users are admin just as 99% of Windows users are, so this is an area OS X most certainly isn't more secure than Windows. Most OS X users are admin because:
- The initial account you set up during the installation process is an admin account
- You're encouraged to treat it as your default account, given it's based upon the name you enter when registering Mac OS X
- Mac OS X has no manual, and no messages are presented, or warnings given, suggesting you should be using a separate account for your day to day use.
It's pretty simply really, it turns out that all installations of VNC will accept the following password:
' or 'x' = 'x
* That's a joke. If you don't get it, don't worry. People who've had to fix crappy password authenticators in software based upon SQL will know what I'm refering to.
Here's a fact: 50% of working-age adults are NOT employed in full-time, salaried jobs.
Why on Earth would you expect them to be?
Half of working age adults are women, who, as a group, are more likely to be homemakers, and if employed, working part time to supplement, rather than provide, the household income. Even today. And I don't know where you work, but most of my office is comprised of hourly, not salaried, workers.
I hear this kind of crap all the time. The truth is that every household needs one breadwinner. That we may be at a stage where people are working two jobs or both adults feel obliged to gain employment, salaried or hourly, is the real problem, not the opposite.
(And before anyone accuses me of anything, I believe in equal opportunities, and I don't like the imbalances we see, however, I think it's good for one parent to be at home bringing up the children. That's a full time job.)
Who says the winner doesn't ALSO get the patent? You think the winner has to surrender his patents to the public domain to collect the prize?
I never said anything of the sort. All I said was that I much prefer the concept of using prizes to granting patents. This particular scheme tries to reward inventors by offering a prize for the first inventor to come up with something. That's a great idea. We should replace patents with prize money.
It's not a bad amount. Small inventors are likely to have budgets well below that, meaning the 10 million prize is more than enough to recoup the costs, and leave the inventor able to spend the rest of their lives inventing things if they choose.
Should it be higher? If it doesn't need to be higher, then no, even if there are supposedly less important things that require more cash.
On a wider issue, I much prefer the idea of prizes and grants, from government and private industry, than patents. Arguably, many patents, by granting the inventor a monopoly on something that there's no reason to believe wouldn't have been created otherwise (and often whose victims are independent developers who created the technologies without knowing about the patented version's existance) cause far more than $10 million to be arbitrarily moved from "the public" to a consortium of lawyers and an inventor.
Slashdot Mirror
In it, he simply makes it clear that no decision has been made public yet. This is what he writes:
If the intent is to release the source to XNU for Intel, then Prabhakar went a long way to avoid saying explicitly that it's going to happen. He most certainly did NOT say this was a matter of them not having released it "yet".There are multiple reasons why XNU for Intel might not have been released, from over-paranoid and not terribly bright PHB types thinking that somehow it would help hackers (because hackers never hack the copy-prevention measures of binary-only programs...) to support for an unannounced feature, like Virtualization, being present in all new 10.4.1+ kernels. But the way I read Prabhakar, it sounds like more of the former type of issue, with key figures in Apple being against the release of the source now and in the future, and with the final decision not being made. If it was as simple as the "unannounced new feature", I think he'd explicitly say Apple plans to release the source, wouldn't he?
The Pentium and Celeron would spring to mind. There's little in common between the Pentium 1, the Pentium 3, and the Pentium 4. The Pentium Pro, 2, 3, and M, are to some extent architecturally related, albeit loosely. The Celeron has been the "cut down" version of whatever the current Pentium was. Meanwhile, the Xeon exists in so many forms it's hard to tell what it is these days.
Intel's in the process of a rebranding, but they're not doing away with the stategy, just getting away from the Pentium/Celeron brands. The Core 2 will replace the current Core. That's why it's still a Core, even when it's been re-engineered to be percentage% faster per whatever-the-measure-is-today and contain a significant amount of 64-bit stuff. Allegedly, it'll be a drop in replacement, OEMs will not even have to redesign their motherboards.
But that's because I'm interested in laptops as mobile desktops that play games occasionally. The more serious you are with a laptop, the less interested you are in games and the more you're interested in battery life.
I criticised Apple for bunging a GMA950 in the Mac mini. I still think it's the wrong choice. A populist computer that's poor at modern games? That doesn't make a lot of sense. But a GMA950 is perfect for a laptop for most people. It performs very well when it comes to accellerating Mac OS X's graphics, video, and a handful of other applications, and it's exceptionally low power. The only thing I can think of that would be "better" (but it'd take some work on Apple's side to make it work cleanly) would be to have a decent graphics card in there but only enable it when it needs to be used, having it powered off otherwise (the GMA950 is actually present in all Intel Apple computers, it's built-in to the chipset, it's just not used in those with Radeons)
I think the criticism of Apple for using the GMA950 in the Macbook is overblown. It's a good choice for a laptop GPU. The criticisms of the Mac mini need to be kept up, but they shouldn't translate into a "GMA950 always bad, X300 always good" mentality, because that just isn't the case.
RMS doesn't consider himself a supporter of Open Source. And Linus Torvalds has made it clear on several occasions, most infamously with the BitKeeper debacle, that he's NOT an Open Source, or Free Software, advocate, he's quite happy with proprietary software.
I'm not sure who qualifies as an Open Source leader. There's the official one, Michael Tiemann, the current head of the Open Source Initiative. There's the most vocal one, Eric S. Raymond, who arguably was the person who made the single biggest contribution, for better or worse, to popularizing the Open Source concept. And then there's Bruce Perens. In terms of people heading specific projects, it's hard to really come up with a name. Torvalds simply is not in the running, he's popular, but he's not someone anyone who considers FS or OS important takes seriously when it comes to pronouncements on the future of Open Source because he's not an Open Source advocate. Other "Open Source" projects, be they Apache, Mozilla, or whatever, have less well known characters in charge, and generally the few that do have fairly controvertial characters (OpenBSD's Theo springs to mind) rather than natural leaders.
It would be great if Torvalds took the concept more seriously and stepped into the shoes so many people want him to wear. He clearly has the charisma, but he's no more the leader of Open Source than John McCain is leader of the Republican Party.
I'm not sure how the CDMA and WCDMA based alternatives are with latency. It may be that 802.16 works out better than any "grafted onto a cellphone network" solution anyway.
So give it a rest. You're the umpteenth slashbot to "correct" someone for using the term "pirate" correctly. It's pretty hard to believe that deep down you're unaware that the term does, actually, mean "copyright infringment" pretty much everywhere.
The people who are doing the cracks themselves may be taking 25 hours, but I doubt the majority of people using the cracks are.
To which, I guess the answer is: Flubajugalubums.
I honestly doubt closing the sources will make much of a difference. There's an awful lot of pirated binary-only programs out there. Meanwhile, those with legitimate uses are locked out. Doesn't really make much sense to me.
Perhaps a better solution would be to deal with the piracy issue by, well, catering to a clear demand. What's better, 1,000,000 MacBook sales, and 10,000,000 people with illegal copies of Mac OS X on their Dells, or 1,000,000 MacBook sales, 2,000,000 people with paid-for, unsupported, $100 copies of Mac OS X on their Dells, and 8,000,000 illegal copies? If people are going to do it anyway, and they are, you might as well make some money from them.
Try compiling those XNU sources. My understanding is that its only possible for PPC. XNU for Intel is dead.
It's possible Microsoft were trying "to help their case", I'll give you that. It would be nice to think of Microsoft as occasionally good, once in a while. Maybe they'll inject a little funding into Mono one day, with no court case to encourage them.
Or maybe they didn't, maybe they presented their improvements to Sun, which at the time didn't really have a community development process, and Sun rejected them.
Either way, supporting an Free Software project and documenting their extensions and asking the FS project to include these features doesn't mesh with the E&E strategy they've always been painted as using.
The full comment was, of course:
That's absolutely right and I dare you to find holes in that, rather than dubious and irrelevent complaints about "open sourcers".No, that wouldn't help at all.
As far as ensuring the GPL license doesn't leak into applications that run over the Java run-time system, you simply supply an additional, optional, license, that allows for linking code with a pre-built binary distribution of Sun's Java using the published APIs. Developers can choose which they use on a project-by-project basis.
There's a more interesting issue here. Sun Java is an embarassment to the OSI. Over the last few years, by using a community driven development process, Java has improved leaps and bounds. Essentially, Sun said "What the Open Source movement says is right, except for the freedom part". And given the OSI keeps being at pains to argue that it's merely a front for software freedom, trying to encourage the development of free software by promoting community-driven development processes which, supposedly, rely upon the software being developed to be Free, this really doesn't hasn't helped it much.
Essentially, the OSI says "We must have free software, because free software means a community of interested parties can develop a program to a much higher standard than would otherwise be the case if it was proprietary. We describe this whole thing as "Open Source"."
Sun responds with: "Aha! But Java isn't free, and it too is developed by a community of interested parties, and they've generated a much higher standard of product than would otherwise have been the case if it wasn't developed using a community process. So your argument fails because you don't need software to be free to use your "open source" development model!"
ESR responds with: "You all suck. Set Java free!!!1!"
So why's Sun "open sourcing" Java? I think they're just looking at ensuring the official Sun implementation has wider adoption, by removing licensing barriers. Free software licenses happen to be a great way to get there. Sun wants to get Java "out there", especially with .NET nipping at its heels. The real problem with Sun's strategy hasn't been forsaking the development model advantages of the OSI's "Open Source", it's been that it's harder to integrate the official Sun Java, the reference implementation, with the non-Java world, because of licensing issues.
And as such, I don't think Sun gives a rats arse what the OSI thinks.
FWIW, I wrote about this in my journal.
I should know. It happens to me once every few days. Most likely explanation? Hardware failure. I play games a lot, causing the fans to go full blast, and crashes generally follow that.
Most likely explanation for the BSODs on XP? A combination of two: hardware failures, as with OS X, and (to a much lesser extent) viruses. Viruses don't exist for the Mac for the most part because it's difficult to get a virus to spread when 95% of the computers it would "hit" are incapable of passing it on.
Can we cut out the BSOD BS? Both Macs and PCs are, these days, more or less equally vulnerable to them. 99% of BSODs are almost certainly caused by hardware issues, such as corrupted RAM.
Neither company has "innovated" yet. Neither the PS3 nor the Wii are actually on sale. Innovation doesn't occur until the seemingly-new technology is in people's hands.
GMail is a Javascript application, and it does allocate memory. Without someone pointing at specific code and saying "This is where the leak is occuring", it's simply not possible to point at either Firefox or GMail and say "This is where the bug's occurring".
Garbage collectors are not magic. They don't release memory where there's a reference to it. If an extension is storing references to objects on a window that's subsequently closed, and the extension doesn't then remove those references, the GC isn't going to discard the objects, because as far as it's concerned, they're still in use.
There are some ridiculously stupid aspects to certain subsystems of Windows that create big holes (ActiveX springs to mind), but while I don't like ordinary users being "admin" either, the fact is 99% of Mac OS X users are admin just as 99% of Windows users are, so this is an area OS X most certainly isn't more secure than Windows. Most OS X users are admin because:
- The initial account you set up during the installation process is an admin account
- You're encouraged to treat it as your default account, given it's based upon the name you enter when registering Mac OS X
- Mac OS X has no manual, and no messages are presented, or warnings given, suggesting you should be using a separate account for your day to day use.
' or 'x' = 'x
* That's a joke. If you don't get it, don't worry. People who've had to fix crappy password authenticators in software based upon SQL will know what I'm refering to.
Half of working age adults are women, who, as a group, are more likely to be homemakers, and if employed, working part time to supplement, rather than provide, the household income. Even today. And I don't know where you work, but most of my office is comprised of hourly, not salaried, workers.
I hear this kind of crap all the time. The truth is that every household needs one breadwinner. That we may be at a stage where people are working two jobs or both adults feel obliged to gain employment, salaried or hourly, is the real problem, not the opposite.
(And before anyone accuses me of anything, I believe in equal opportunities, and I don't like the imbalances we see, however, I think it's good for one parent to be at home bringing up the children. That's a full time job.)
Should it be higher? If it doesn't need to be higher, then no, even if there are supposedly less important things that require more cash.
On a wider issue, I much prefer the idea of prizes and grants, from government and private industry, than patents. Arguably, many patents, by granting the inventor a monopoly on something that there's no reason to believe wouldn't have been created otherwise (and often whose victims are independent developers who created the technologies without knowing about the patented version's existance) cause far more than $10 million to be arbitrarily moved from "the public" to a consortium of lawyers and an inventor.