One of the most annoying problems with using alternative fuels for powering cars is that petroleum has pretty much the highest energy density of any car sized option. Ethanol is an interesting alternative, but a complete conversion would require a 10-20% increase in the amount of land farmed. (The 1 billion acres number hasn't changed much in 80+ years. Machines simply allow more farming by fewer people.)
A pie in the sky idea would be micro-fission reactors. The reality however, is that such a reactor would require more shielding than is reasonable for such a small vehicle.
Hydrogen fuel cells have an energy density of ~1/2 to 2/3 of gasoline. The upshot is that solar and fission power can be used to create more fuel at very cheap rates. i.e. You'll have to "gas up" more often, but you'll pay less. Such vehicles could even be designed to use solar panels to convert water into a little extra fuel for your car. This would make such a vehicle far more cost effective than today's vehicles.
All in all, there's no "better" solution, but there are a few "good enough" solutions.
Total crap regarding distro owners. SuSE, for example, includes installation support in the price of the their distro. Some of the community may give you that response and that refers to any community - this is not a Linux only community issue.
Do you people ever follow threads, or does it cause a buffer overflow in your brains? The whole thread is about the fact that the distro owners were of no help in resolving his issue. He was upset over this and the community skewered him for it.
Why shouldn't the consumer blame the vendor in this instance too if they didn't supply a working driver like HP or Nvidia etc do for Linux?
Because it's impossible to make a $%^#ing closed source driver for Linux? Just because NVidia has an UNSUPPORTED hack, doesn't mean that every Tom, Dick and Harry is going to follow suit. The hardware vendors put "Windows" (and sometimes "Mac") on the sides of their boxes, because that's what they support. They're not going to support Linux because Linux wants to do everything its own way.
Right here. I remember giving the salesman an incredulous look when he presented this monstrosity as an alternative to a Sun machine.
Him: "It's cheaper, and has twice as much I/O as a Sun machine!" Me: "Ok. And how about uptime?" Him: "I'm sure it's in the upper nines range!" Me: "You mean you haven't yet sold any of these yet?" Him: "No, it's brand new! You're getting a sneak peek!"
Gotta love Unisys. They're one of those companies that can never seem to quite "get it". Oh, and I'd like to strangle the guy that thought that combining NT and the MCP was a better solution than just adding new functionality to the MCP. No wonder IBM ate their breakfast time and time again.
Of course if you look at this as a "religious war" and then look that the history of religious wars, and then take into account the anonymous nature of message boards you can see how things might deteriorate quickly.
Believe it or not, the historical irony was intentional. As an example, WWI was sparked by the assassination of Archduke Ferdinand. What did his death have to do with anything? Absolutely nothing. The truth was that everyone had shiny new weapons thanks to technology, and were looking for a reason to go to war. WWII finally proved to everyone how absolutely stupid this thinking was.
What would Microsoft have done for him that his *nix vendor didn't?
First attempt standard troubleshooting. If that doesn't work, then forward him to the vendor of the hardware. If the vendor was unable to provide a driver that works, the consumer would blame the vendor. Contrast that to Linux where the distro owners (and community) have taken responsibility for supporting ALL hardware. If it doesn't work, then they stick their tongue out at you and tell you it's your own fault.
That was what caused me to try FreeBSD so many years ago. I got tired of the fact that "lp" wouldn't disable according to the instructions, that it wouldn't detect my zip drive, and that the community wouldn't even try to help. On FreeBSD everything just worked (they even had USB back then!), and the community was helpful when I couldn't figure something out. Oh, and they had the "FreeBSD Handbook". Linux had (usually out of date and difficult to understand) HOWTOs scattered across creation.
All I am saying is that the community operates on two different levels (community support and vendor support) and by not posting the specs of the card he basically insulted the community.
not a good excuse
Period, end of story. Perhaps we should also start world wars over the fact that an ambassador made a slight mistake in customs when he talked with a foreigner? Or maybe they could just accept that he's foreign, made a mistake (which he's probably more than willing to apologize for), has his own customs, and get on with it.
Please explain to me how a low slashdot user ID means that someone took their time signing up?
A higher id means that someone took their time. You understood me. Please don't be confusing the issue.
I won't disagree with you here on the foolishness part, but, particularly in light of the latter part of your comment, I would like to state that there are an infinite number of ways a person can be foolish...
Indeed. I can actually answer why I still hang around here. I'm trolling for minds. Not trolling as in the immature "look what I got them to argue about" sense, but in the "looking for intelligent people to have conversations with". Despite all the noise and BS many slashdotters throw out, I've managed to have intelligent conversions with Nuclear Engineers, Rocket Scientists (Aerospace Engineers), former and present Navy personnel, airship experts, history buffs, etc, etc, etc. Slashdot manages to attract so many intelligent people that sparking a gem of a conversation is worth 10 times any abuse I take.
You know the most interesting thing about many of these intelligent people? They signed up just to join the conversation. Thus my point about the stupidity of using UIDs.
Looking at the Java Desktop System, I think that it is GNOME with a custom theme. Nothing more.
That's not my point. I stated that JDS is currently just another distro. My point is that Sun's long term goal is to migrate the desktop toward a mix of Open Source and their own (Java-based) proprietary stuff.
If you're 16 years old, and you've only had access to a computer for 4 years, and you see someone with a low ID on/., this means that you're probably looking at someone whose own personal experience with computing in general is older than yours./. is what, 8 years old now (I can't remember)... to a fairly large demographic of web-browsing public (teenagers), thats a farkin' long time ago... half of some peoples' lives.
Or it just means that they took their time signing up. I lurked on Slashdot (and actually largely ignored it) for a year or two before I actually created an account. I've now been here for about 4 years. I was already a longtime professional in the field when I signed up, and I am still a professional in the field.
I'll say it again: Using a slashdot UID to judge someone's opinion is foolishness. Sometimes I wonder why I still hang around this place...
So if you have a problem you have to turn to other users and give them the information they need to help you. Of course since he paid "near windows prices" for a distro, he expected that help from the distro vendor.
Is there anything wrong with him expecting support from the vendor? He did exchange money for a promise of support. Receiving a "well it should work, but I guess you're SOL" for that money is not a particularly good value proposition. Also, many of the distros are directly responsible for employing the programmers who did a lot of the work on the open source software in their product. This *should* give them a support advantage, but it seems that many of the distros don't make effective use of it.
I think he's talking about the fragmentation caused by multiple non-interoperative desktop APIs.
That's the thing though. From everything I've seen of Looking Glass, it supports GNOME and KDE apps just fine. It's actually less of a Desktop Environment (like GNOME), and more of a Window Manager (like Metacity). I don't know where Sun is going with this, but they may decide to integrate GNOME and Nautilus into Looking Glass. They've certainly hinted at it by saying that "Looking Glass is a technology demo. Expect to see parts of it slowly migrate into the Java Desktop System."
what makes you think they aren't independent OSes already?
There's a lot of pressure on Linux vendors to develop true "Linux" (or GNU/Linux depending on who you are) distros instead of branching off in incompatible ways. This is compounded by the fact that many of the distros are based on one another. It's like horse breeding. e.g. Mandrake is based on Man O'War... err... RedHat, while Knoppix is based on Debian, etc.
Why do they have to be proprietary GUI's to make them different OSes?
They don't. My point is that Sun is attempting to reinvent the desktop. To meet that ever precious "time-to-market", they've started with a Linux and GNOME desktop.
What is it about being all locked up like sun that makes it distinct?
Nothing at the moment. In the future they obviously want to be a distinct entity with as much software as possible written in Java.
Is it the incompatibility caused by being locked up, or is it that people consider everything with a free license to be part of the system?
Neither. It's a marketing scheme referred to as "Value Add". i.e. Sun develops a desktop based on existing and compatible components, but adds in proprietary stuff that people want. The theory is that people are willing to choose their product over a similar one because Sun has added value to the product proposition.
Relax. What I meant was that other dude seemed to be around at an earlier time than yourself and thus had a different view because s/he saw what went down.
Fair enough. But an UID is a poor way of determining that. Believe it or not, I do remember the Unix wars. However, I have my doubts that it was really the Unix wars that allowed Microsoft to supplant Unix. The Unix wars certainly did prevent them from seeing the threat that Microsoft posed, but Microsoft took hold because a new generation of hobby programmers grew up with PCs. Since PCs were really the only machine they knew, they wanted to use it for everything.
This put pressure on Intel for ever faster processors and chipsets, eventually allowing PCs to be reasonably comparable to Unix machines. Unix machines were still better in nearly all respects, but since none of the former hobbyists had used features like redundant power supplies, 64 processor scaling, 99.999% uptime abilities, etc., they simply didn't know they needed it. By the time they did realize this, PC manufacturers were getting smart any trying to backport Unix innovation into the PC architecture. (Who remembers the Unisys 32 Intel processor machine with a gigaplex backplane? Just me? Do they still sell those things?)
Look who's talking. Does your exceedingly high user id mean that we should not take your post seriously? Or does it automatically mark you as a troll? Or perhaps your just despondent over the fact that you're missing a +1 modifier?
Evaluate peoples opinions for their content, not how soon they signed up for an account on Slashdot.
It is exactly this sort of shit that nearly killed UNIX in the 1980s and allowed Microsoft the opportunity to supplant technically superior systems with their shoddy software and then leverage that toehold into a desktop monopoly.
I believe you mean the 1990's. Unix was not having all that many troubles in the 1980's. It's biggest competitor was mainframes and "smart" terminals that allowed things like an independent Word Processor to interface to the mainframe.
What you are describing for Sun et. al. is a narrowing of their (Linux) platform, and undermining one of the great values of Linux... that it is a defacto standard system that runs the same basic flavor of *NIX on multiple hardware platforms, irrespective of distribution, CPU type, 32-bit vs. 64-bit, 1-way vs. N-way processors, etc.
Hmm... Not sure I follow you here. Sun has release JDS for x86. They've promised to also release JDS on Solaris for x86, AMD, and Sparc. In this way, you can choose what features meet your needs. If you need a kernel that handles all the exotic x86 hardware, then a Linux kernel is best. If you want to run JDS for 200 people using SunRay stations, then JDS Solaris is for you.
Lose that and your right back to the state of UNIX circa 1990, and that wasn't a pretty picture (or a viable state of affairs, with every hardware manufacturer's proprietary system incompatible with everyone elses).
True. But I'd like to point out that Sun isn't bucking any specs here. Their JDS system still interops with all Unix and Windows machines. In the process, they're attempting to blaze a trail for what Linux can be used for. Assuming they don't open source Looking Glass, what stops the Linux community from creating a better engineered knockoff? And how will Sun's desktop fragment the market? It's still X11. It just has 3D features not present in today's GNOME and KDE desktops.
Agree or disagree with the author, there is one thing he shows quite clearly: Many Linux users would rather attack than help. Regardless of whether it's an EBKAC problem or not, don't you people think that you should be using polite language to discuss the issue? A little bit of "Oh, it's all right. You merely did this wrong. Now you're up and going and you know for next time.:-)" would go a long way toward getting Linux a positive review. Instead users are assailed as "stoopid" and "the real problem is that you don't know what you're doing". This is extremely frustrating!
I myself have years of experience with Linux, *BSD, Solaris, and several other Unixes. When I try to point out a deficiency that I think should be fixed (binary compatibly, PLEASE) I merely get the "you're stupid and don't know anything about Linux", or the "You're using the wrong distro. MY distro doesn't have this problem!" Of course, you can switch, run into some other problem, then be told, "Well this OTHER distro (which you were previously using) doesn't have this issue! You should switch!"
In all fairness, many people have managed to be polite, as evidenced by many of the replies I received in my Linux reviews. Unfortunately, one bad apple tends to spoil the bunch. Stop the fighting and name calling! Work together! So much more will be accomplished that way.
...that future Linux distros will move away from being "Linux" and toward being independent OSes. They'd still retain the Linux kernel, and perhaps some of the CLI userland, but the GUI and standard programs will be proprietary.
Apple has already accomplished this with BSD and OS X. Looking at the Java Desktop System, I think that this is Sun's endgame as well. For now they'll leverage everything Linux, then slowly replace all programs with Java ones, and the Desktop with Java Looking Glass. It's hard to say how it will work out, but I wish them the best.
Arguably, making orbit or Earth escape trajectory is a much less arbitrary definition than simply going up to where the atmosphere is really thin. After all, "space" is a nebulous concept. We're always in space. It just so happens that this little planet has air for us to breath. Achieving orbit means that the craft is actually a craft capable of flight outside the influence of our gravity. (sort of)
Fair enough. But that means that I'm correct in assuming that it's really a problem with Windows as opposed to a flaw in the x86 architecture. Obviously, it was very easy for Windows to simply memory map the executable instead of parsing and loading it. But that same decision is what has gotten us into the mess.
If Microsoft needs a new executable format, they should develop it. (Or just use ELF like everyone else.) However, I think they're a little too focused on the CLR right now to pay much attention to memory segmentation issues.
It isn't as easy to use as it looks, though, and there can only be one code segment at a time.
I assume you mean that there can only be one segment the CS register points to at a time, right? The processor still shouldn't let code write to a segment marked as execute only. And since the x86 processor uses segment selectors, it should be an easy rule to enforce.
JIT compiled code can be handled by creating both a data segment and code segment that point to the same address, writing the compiled code, then deleting the data segment. The remaining code segment should be well protected then. For more dynamic code, you can leave the data segment, but then you have to be careful that the segment selectors don't end up where they shouldn't be. One wrong pop off the stack and POW!, you're executing an overflowed buffer.
Don't get me wrong. The x86 is not that great of an architecture. When their own manual states that use of the task switching instructions is not recommended, you know that something is wrong. I just hate to see people claiming things about it that simply aren't true.
I'm actually more concerned with the lack of prosecution. These scammers often register domain names and leave a long paper trail. When I send a notification to a company of a fraudulant attempt, what do they do to attempt to prosecute the scammer? Do they contact the police? The FBI? Bring in their crack team of technology investigators? No. They tell you not to trust anyone. THAT'S NOT HELPFUL. What would be helpful is to pursue criminals who misuse your name in an attempt to steal money.
Hacked?! Oh, please. You're one of an innumerable number of fools who responded to a "please verify your account" email or a "security update" email that asked you to enter your eBay or PayPal userid and password.
At first glance, it LOOKS legit. And when you click on the link, it takes you to a page that LOOKS like the real thing. Many even have links to the real site. The only way to know that it's fake is to look in the URL bar, and keep your head straight about them asking for passwords and credit cards.
The sad part is that companies seems to be doing very little to prosecute these scammers. I've received emails for both eBay and Citibank. Both times I've gotten no response on my fraud report.
Slashdot Mirror
One of the most annoying problems with using alternative fuels for powering cars is that petroleum has pretty much the highest energy density of any car sized option. Ethanol is an interesting alternative, but a complete conversion would require a 10-20% increase in the amount of land farmed. (The 1 billion acres number hasn't changed much in 80+ years. Machines simply allow more farming by fewer people.)
A pie in the sky idea would be micro-fission reactors. The reality however, is that such a reactor would require more shielding than is reasonable for such a small vehicle.
Hydrogen fuel cells have an energy density of ~1/2 to 2/3 of gasoline. The upshot is that solar and fission power can be used to create more fuel at very cheap rates. i.e. You'll have to "gas up" more often, but you'll pay less. Such vehicles could even be designed to use solar panels to convert water into a little extra fuel for your car. This would make such a vehicle far more cost effective than today's vehicles.
All in all, there's no "better" solution, but there are a few "good enough" solutions.
Total crap regarding distro owners. SuSE, for example, includes installation support in the price of the their distro.
Some of the community may give you that response and that refers to any community - this is not a Linux only community issue.
Do you people ever follow threads, or does it cause a buffer overflow in your brains? The whole thread is about the fact that the distro owners were of no help in resolving his issue. He was upset over this and the community skewered him for it.
Why shouldn't the consumer blame the vendor in this instance too if they didn't supply a working driver like HP or Nvidia etc do for Linux?
Because it's impossible to make a $%^#ing closed source driver for Linux? Just because NVidia has an UNSUPPORTED hack, doesn't mean that every Tom, Dick and Harry is going to follow suit. The hardware vendors put "Windows" (and sometimes "Mac") on the sides of their boxes, because that's what they support. They're not going to support Linux because Linux wants to do everything its own way.
There are (probably) no 'Windows Mainframes'.
Right here. I remember giving the salesman an incredulous look when he presented this monstrosity as an alternative to a Sun machine.
Him: "It's cheaper, and has twice as much I/O as a Sun machine!"
Me: "Ok. And how about uptime?"
Him: "I'm sure it's in the upper nines range!"
Me: "You mean you haven't yet sold any of these yet?"
Him: "No, it's brand new! You're getting a sneak peek!"
Gotta love Unisys. They're one of those companies that can never seem to quite "get it". Oh, and I'd like to strangle the guy that thought that combining NT and the MCP was a better solution than just adding new functionality to the MCP. No wonder IBM ate their breakfast time and time again.
Good analogy.
:-)
Thank you.
Of course if you look at this as a "religious war" and then look that the history of religious wars, and then take into account the anonymous nature of message boards you can see how things might deteriorate quickly.
Believe it or not, the historical irony was intentional. As an example, WWI was sparked by the assassination of Archduke Ferdinand. What did his death have to do with anything? Absolutely nothing. The truth was that everyone had shiny new weapons thanks to technology, and were looking for a reason to go to war. WWII finally proved to everyone how absolutely stupid this thinking was.
What would Microsoft have done for him that his *nix vendor didn't?
First attempt standard troubleshooting. If that doesn't work, then forward him to the vendor of the hardware. If the vendor was unable to provide a driver that works, the consumer would blame the vendor. Contrast that to Linux where the distro owners (and community) have taken responsibility for supporting ALL hardware. If it doesn't work, then they stick their tongue out at you and tell you it's your own fault.
That was what caused me to try FreeBSD so many years ago. I got tired of the fact that "lp" wouldn't disable according to the instructions, that it wouldn't detect my zip drive, and that the community wouldn't even try to help. On FreeBSD everything just worked (they even had USB back then!), and the community was helpful when I couldn't figure something out. Oh, and they had the "FreeBSD Handbook". Linux had (usually out of date and difficult to understand) HOWTOs scattered across creation.
All I am saying is that the community operates on two different levels (community support and vendor support) and by not posting the specs of the card he basically insulted the community.
not
a
good
excuse
Period, end of story. Perhaps we should also start world wars over the fact that an ambassador made a slight mistake in customs when he talked with a foreigner? Or maybe they could just accept that he's foreign, made a mistake (which he's probably more than willing to apologize for), has his own customs, and get on with it.
Did I say there was anything wrong with it?
Please explain to me how a low slashdot user ID means that someone took their time signing up?
...
A higher id means that someone took their time. You understood me. Please don't be confusing the issue.
I won't disagree with you here on the foolishness part, but, particularly in light of the latter part of your comment, I would like to state that there are an infinite number of ways a person can be foolish
Indeed. I can actually answer why I still hang around here. I'm trolling for minds. Not trolling as in the immature "look what I got them to argue about" sense, but in the "looking for intelligent people to have conversations with". Despite all the noise and BS many slashdotters throw out, I've managed to have intelligent conversions with Nuclear Engineers, Rocket Scientists (Aerospace Engineers), former and present Navy personnel, airship experts, history buffs, etc, etc, etc. Slashdot manages to attract so many intelligent people that sparking a gem of a conversation is worth 10 times any abuse I take.
You know the most interesting thing about many of these intelligent people? They signed up just to join the conversation. Thus my point about the stupidity of using UIDs.
Looking at the Java Desktop System, I think that it is GNOME with a custom theme. Nothing more.
That's not my point. I stated that JDS is currently just another distro. My point is that Sun's long term goal is to migrate the desktop toward a mix of Open Source and their own (Java-based) proprietary stuff.
If you're 16 years old, and you've only had access to a computer for 4 years, and you see someone with a low ID on /., this means that you're probably looking at someone whose own personal experience with computing in general is older than yours. /. is what, 8 years old now (I can't remember) ... to a fairly large demographic of web-browsing public (teenagers), thats a farkin' long time ago ... half of some peoples' lives.
Or it just means that they took their time signing up. I lurked on Slashdot (and actually largely ignored it) for a year or two before I actually created an account. I've now been here for about 4 years. I was already a longtime professional in the field when I signed up, and I am still a professional in the field.
I'll say it again: Using a slashdot UID to judge someone's opinion is foolishness. Sometimes I wonder why I still hang around this place...
So if you have a problem you have to turn to other users and give them the information they need to help you. Of course since he paid "near windows prices" for a distro, he expected that help from the distro vendor.
Is there anything wrong with him expecting support from the vendor? He did exchange money for a promise of support. Receiving a "well it should work, but I guess you're SOL" for that money is not a particularly good value proposition. Also, many of the distros are directly responsible for employing the programmers who did a lot of the work on the open source software in their product. This *should* give them a support advantage, but it seems that many of the distros don't make effective use of it.
I think he's talking about the fragmentation caused by multiple non-interoperative desktop APIs.
That's the thing though. From everything I've seen of Looking Glass, it supports GNOME and KDE apps just fine. It's actually less of a Desktop Environment (like GNOME), and more of a Window Manager (like Metacity). I don't know where Sun is going with this, but they may decide to integrate GNOME and Nautilus into Looking Glass. They've certainly hinted at it by saying that "Looking Glass is a technology demo. Expect to see parts of it slowly migrate into the Java Desktop System."
what makes you think they aren't independent OSes already?
;-)
There's a lot of pressure on Linux vendors to develop true "Linux" (or GNU/Linux depending on who you are) distros instead of branching off in incompatible ways. This is compounded by the fact that many of the distros are based on one another. It's like horse breeding. e.g. Mandrake is based on Man O'War... err... RedHat, while Knoppix is based on Debian, etc.
Why do they have to be proprietary GUI's to make them different OSes?
They don't. My point is that Sun is attempting to reinvent the desktop. To meet that ever precious "time-to-market", they've started with a Linux and GNOME desktop.
What is it about being all locked up like sun that makes it distinct?
Nothing at the moment. In the future they obviously want to be a distinct entity with as much software as possible written in Java.
Is it the incompatibility caused by being locked up, or is it that people consider everything with a free license to be part of the system?
Neither. It's a marketing scheme referred to as "Value Add". i.e. Sun develops a desktop based on existing and compatible components, but adds in proprietary stuff that people want. The theory is that people are willing to choose their product over a similar one because Sun has added value to the product proposition.
(why did i phrase everything as questions?)
Me: You've got questions, I've got answers.
Radioshack: You've got questions, we've got blank stares.
Relax. What I meant was that other dude seemed to be around at an earlier time than yourself and thus had a different view because s/he saw what went down.
Fair enough. But an UID is a poor way of determining that. Believe it or not, I do remember the Unix wars. However, I have my doubts that it was really the Unix wars that allowed Microsoft to supplant Unix. The Unix wars certainly did prevent them from seeing the threat that Microsoft posed, but Microsoft took hold because a new generation of hobby programmers grew up with PCs. Since PCs were really the only machine they knew, they wanted to use it for everything.
This put pressure on Intel for ever faster processors and chipsets, eventually allowing PCs to be reasonably comparable to Unix machines. Unix machines were still better in nearly all respects, but since none of the former hobbyists had used features like redundant power supplies, 64 processor scaling, 99.999% uptime abilities, etc., they simply didn't know they needed it. By the time they did realize this, PC manufacturers were getting smart any trying to backport Unix innovation into the PC architecture. (Who remembers the Unisys 32 Intel processor machine with a gigaplex backplane? Just me? Do they still sell those things?)
Look who's talking. Does your exceedingly high user id mean that we should not take your post seriously? Or does it automatically mark you as a troll? Or perhaps your just despondent over the fact that you're missing a +1 modifier?
Evaluate peoples opinions for their content, not how soon they signed up for an account on Slashdot.
It is exactly this sort of shit that nearly killed UNIX in the 1980s and allowed Microsoft the opportunity to supplant technically superior systems with their shoddy software and then leverage that toehold into a desktop monopoly.
... that it is a defacto standard system that runs the same basic flavor of *NIX on multiple hardware platforms, irrespective of distribution, CPU type, 32-bit vs. 64-bit, 1-way vs. N-way processors, etc.
I believe you mean the 1990's. Unix was not having all that many troubles in the 1980's. It's biggest competitor was mainframes and "smart" terminals that allowed things like an independent Word Processor to interface to the mainframe.
What you are describing for Sun et. al. is a narrowing of their (Linux) platform, and undermining one of the great values of Linux
Hmm... Not sure I follow you here. Sun has release JDS for x86. They've promised to also release JDS on Solaris for x86, AMD, and Sparc. In this way, you can choose what features meet your needs. If you need a kernel that handles all the exotic x86 hardware, then a Linux kernel is best. If you want to run JDS for 200 people using SunRay stations, then JDS Solaris is for you.
Lose that and your right back to the state of UNIX circa 1990, and that wasn't a pretty picture (or a viable state of affairs, with every hardware manufacturer's proprietary system incompatible with everyone elses).
True. But I'd like to point out that Sun isn't bucking any specs here. Their JDS system still interops with all Unix and Windows machines. In the process, they're attempting to blaze a trail for what Linux can be used for. Assuming they don't open source Looking Glass, what stops the Linux community from creating a better engineered knockoff? And how will Sun's desktop fragment the market? It's still X11. It just has 3D features not present in today's GNOME and KDE desktops.
Agree or disagree with the author, there is one thing he shows quite clearly: Many Linux users would rather attack than help. Regardless of whether it's an EBKAC problem or not, don't you people think that you should be using polite language to discuss the issue? A little bit of "Oh, it's all right. You merely did this wrong. Now you're up and going and you know for next time. :-)" would go a long way toward getting Linux a positive review. Instead users are assailed as "stoopid" and "the real problem is that you don't know what you're doing". This is extremely frustrating!
I myself have years of experience with Linux, *BSD, Solaris, and several other Unixes. When I try to point out a deficiency that I think should be fixed (binary compatibly, PLEASE) I merely get the "you're stupid and don't know anything about Linux", or the "You're using the wrong distro. MY distro doesn't have this problem!" Of course, you can switch, run into some other problem, then be told, "Well this OTHER distro (which you were previously using) doesn't have this issue! You should switch!"
In all fairness, many people have managed to be polite, as evidenced by many of the replies I received in my Linux reviews. Unfortunately, one bad apple tends to spoil the bunch. Stop the fighting and name calling! Work together! So much more will be accomplished that way.
...that future Linux distros will move away from being "Linux" and toward being independent OSes. They'd still retain the Linux kernel, and perhaps some of the CLI userland, but the GUI and standard programs will be proprietary.
Apple has already accomplished this with BSD and OS X. Looking at the Java Desktop System, I think that this is Sun's endgame as well. For now they'll leverage everything Linux, then slowly replace all programs with Java ones, and the Desktop with Java Looking Glass. It's hard to say how it will work out, but I wish them the best.
Arguably, making orbit or Earth escape trajectory is a much less arbitrary definition than simply going up to where the atmosphere is really thin. After all, "space" is a nebulous concept. We're always in space. It just so happens that this little planet has air for us to breath. Achieving orbit means that the craft is actually a craft capable of flight outside the influence of our gravity. (sort of)
Fair enough. But that means that I'm correct in assuming that it's really a problem with Windows as opposed to a flaw in the x86 architecture. Obviously, it was very easy for Windows to simply memory map the executable instead of parsing and loading it. But that same decision is what has gotten us into the mess.
If Microsoft needs a new executable format, they should develop it. (Or just use ELF like everyone else.) However, I think they're a little too focused on the CLR right now to pay much attention to memory segmentation issues.
It isn't as easy to use as it looks, though, and there can only be one code segment at a time.
I assume you mean that there can only be one segment the CS register points to at a time, right? The processor still shouldn't let code write to a segment marked as execute only. And since the x86 processor uses segment selectors, it should be an easy rule to enforce.
JIT compiled code can be handled by creating both a data segment and code segment that point to the same address, writing the compiled code, then deleting the data segment. The remaining code segment should be well protected then. For more dynamic code, you can leave the data segment, but then you have to be careful that the segment selectors don't end up where they shouldn't be. One wrong pop off the stack and POW!, you're executing an overflowed buffer.
Don't get me wrong. The x86 is not that great of an architecture. When their own manual states that use of the task switching instructions is not recommended, you know that something is wrong. I just hate to see people claiming things about it that simply aren't true.
I'm actually more concerned with the lack of prosecution. These scammers often register domain names and leave a long paper trail. When I send a notification to a company of a fraudulant attempt, what do they do to attempt to prosecute the scammer? Do they contact the police? The FBI? Bring in their crack team of technology investigators? No. They tell you not to trust anyone. THAT'S NOT HELPFUL. What would be helpful is to pursue criminals who misuse your name in an attempt to steal money.
Hacked?! Oh, please. You're one of an innumerable number of fools who responded to a "please verify your account" email or a "security update" email that asked you to enter your eBay or PayPal userid and password.
Arguably, scammers are getting very clever at their email attacks. They'll send you an email with a link like this: http://www.ebay.com/cgi-bin/verify.dll?Acct=1234
At first glance, it LOOKS legit. And when you click on the link, it takes you to a page that LOOKS like the real thing. Many even have links to the real site. The only way to know that it's fake is to look in the URL bar, and keep your head straight about them asking for passwords and credit cards.
The sad part is that companies seems to be doing very little to prosecute these scammers. I've received emails for both eBay and Citibank. Both times I've gotten no response on my fraud report.
Thank God! Another person who has actually READ the Intel Manuals!
Sorry, the actual table of flags is Section 3.4.3.1.
BTW, here's the manual if you don't believe me. Section 3.2.3, Multi-segment Model. Scroll down to Section 3.4.3.1 to see the table of types.