I think that's a little different. Netscape was already hopelessly behind when they open sourced their product. And there wasn't any hope of generating real revenue from the browser anyway. What they needed was to promote an open browser market so that Microsoft couldn't hijack HTML and HTTP. They succeeded posthumously in that respect, but their server business couldn't hold out that long.
The point of Java is to level the playing field. Sun needs people to write software for their systems. They won't do that if they have to write for Motif and UNIX, they'll write for Windows instead. They'll only write in Java if it will work well on their users' platforms. And that's the problem.
Sun doesn't know how to write a runtime that will work well for every platform. They need Apple to write an efficient JRE for Mac OS X, Windows experts for Windows, and Linux experts for Linux. And they need people to figure out how to compile Java straight to machine language (GCC's gcj) and make it fully compliant with the latest Java standards. The platform that Sun neglected the most, Linux, is the one that's most important to the success of Java now. Sun is adopting a Linux desktop (GNOME) so if they want Java software running on their own system they need to convince developers that Java is a viable solution for Linux desktops. That software can then be utilized on Sun's desktops. What does Sun gain if everyone just uses Java on Mac OS X or Windows? Their JVM and JDK are free.
The entire non-Solaris community cares. Sun simply does not know how to write efficient JVMs for platforms other than Solaris. The Linux JRE is slow. Microsoft Java runtime was tons faster than Sun's port. The whole point of Java is for it to be ubiquitous. It needs to work well on every platform, otherwise you would just use native code. The only way for this to happen is for Sun to let others maintain their own platform's ports. Sun is dependant on the GNOME project, which is primarily developed for Linux. So Sun needs the Linux community to buy into Java. If Java can't be distributed with major Linux distributions which can not happen. If the GNOME people don't buy into Java, then Sun's own Solaris platform will suffer. Sun can still control the standardization process of Java. They just need to open up the source for their own implementation so that others can quickly optimize it. If the GNU Java compiler (gcj) developers could interface with Sun's class library then it would help Java tremendously on UNIX.
I studied figure 3 in great detail. I even unfocused my eyes and tried to look *into* figure 3....No joy.
Really? Figure 3 has a black box that says "Compatibility Support Module". And then a blue box on top of that saying "Legacy OS Loader". And then the text above the figure says "In addition, the Framework provides support for legacy OS interfaces via a set of drivers."
I think that's you're looking for? That Compatibility Support Module must be a BIOS emulation layer.
All the mid to high end video cards from Nvidia and ATI have heat sinks and fans. For example, the Radeon 9600 and 9800 series. Radeon 9200 is fanless, but it's not a Direct X 9 card like the others. If you want something modern, you're not going to have a choice.
It's not a dumb statement at all. Even though Pentium64 will certainly be cheaper and faster than Itanium, it can't be used in large scale multiprocessors. The Xeons only scale 4 ways because the bus protocol sucks. That's not going to change with a Xeon64. SGI and HP simply won't be able to use it.
Consider that SGI was already having trouble making money with their MIPS line. The whole point of going with Intel was the hope that Intel's Itanium would be high volume and therefore cheaper. Obviously that didn't happen and probably never will. So SGI will just continue to go backrupt. And HP will have no competitive server chip to sell. IBM will kick their ass. HP and SGI won't sell enough Itanium systems for Intel to bother developing it. Intel needs more of a market than the old HP-UX and IRIX customers. They needed the traditional Intel server market to accept Itanium.
They can't make it cheap. It has a 6MB L2 cache on it. That's the only reason it can halfway decent SPEC scores. And it makes the chip damn expensive to produce.
Yes, G5's and Pentium 4's do run circles around Sun workstations. But Sun's money is in their servers. Most people writing software for Suns aren't even looking at their workstations now, only their servers. The only real reason to buy a Sun workstation these days is for developing and testing software that you run on a Sun server. If you want a 32+ processor system with fast I/O and a robust design, then you're not going to find a Mac or PC that's going to do that. And a multiprocessor is going to be more compact than a computing cluster.
I think you've gotten this confused with Xouvert. Xouvert's the one that's making the build changes. FreeDesktop is splitting Xlib into 2 pieces and writing a new server based on kdrive.
Most people on campuses (students) don't pay for shit anyway. They're broke. There's no point worrying about them not buying stuff. 5 hours wasted to save $25 doesn't make sense for working people. It does make sense for students and teens that have no source of income. But why worry about them?
Honestly I think people exaggerate how bad DVD piracy is. Most adults will pay the $15-$20 to buy a DVD rather than waste time trying to downloading it from the Internet. DVDs are selling like hotcakes. People want the special features, commentaries, all that good stuff.
DivX and XviD at 1080i aren't going to be small either. Going from 640x480 standard def. to 1900x1080 is more than quadruple the size. Also keep in mind that HD-DVD is not going to be using MPEG-2. They can't fit a decent sized movie in MPEG-2 on a measly 15GB. They'll be using MPEG-4 or Windows Media. In which case, you're not getting any size reduction whatsoever. It's already as compressed as you can get it. You need to transfer the whole damn 15 GB. The BluRay people are the ones still promoting MPEG-2. BluRay has quite a bit more capacity than HD-DVD so they can pull it off.
I think you should give Babylon 5 another shot. How much have you seen? You have to be patient with it. A lot of other sci fi series are designed so that each episode can stand by itself. Star Trek comes to mind.
Babylon 5 doesn't work that way. There are several episodes which I think are superb, but to appreciate those you'd have to watch most of the episodes prior.
The entire series is essentially a giant novel. Can you accurately judge the Lord of the Rings trilogy by the first 30 pages? At that point, they're not even out the Shire. It's kind of the same thing.
It's weaker on a per-episode basis, but as a series it really stands out. Consider after 7 seasons of Star Trek: The Next Generation, how much have the characters and the Star Trek universe changed? Not a whole lot. You start out with the Federation flagship, with a seasoned captain and a slightly inexperienced crew. In the end you have a less prickly captain and a crew that's a bit more interesting.
In contrast, the Babylon 5 station starts out as a joke. Becoming an ambassador to Babylon 5 was considered a demotion. Earth politicians had trouble justifying its existence. It was about as productive as the United Nations General Assembly. Nothing resolved, just political infighting. Hardly the place you'd expect anything interesting to happen.
The commander of Babylon 5 even has to justify space travel. When asked, he concludes that we have to be out in space because someday our sun will become cold. If we want to survive more than a few billion years we have to get ourselves somewhere else.
In the last episode, you actually get to see the death of the Sun billions of years in the future. Along the way, worlds and entire civilizations are destroyed. Governments collapse. But greater things are created along way.
Sorry to all the nerds who think Star Trek is real science. We are not getting off of the Earth.
The laws of physics as we know them aren't carved in stone. They're simply approximations. The General Theory of Relativity does not fit with Quantum Mechanics. The unified theories being developed do offer possibilities (higher dimensions, wormholes). No one knows that faster than light travel is impossible. There wouldn't have been a Breakthrough Propulsion Laboratory at NASA if respected physicists weren't willing to entertain the possibility.
Even if you could determine a place to go you don't have the right equipment. I mean you. Humans are not designed for long term exposure to space. Sorry. I don't know of one respected scientist who thinks our physiology and psychology permit long term exposure to radiation, low gravity, and isolation.
And you think these problems can't be handled? We can't build radiation shields? We can't produce artificial gravity? We can't build communities in space?
Human beings are willing to do some crazy things. We have people living in the Sahara desert, living in high altitudes, living in the Arctic. We could have just stayed in the trees. Our physiology is certainly more suited to it. Maybe there was a grand debate a long time ago about whether to leave the trees and whether to live in any of the godforsaken places human beings are willing to. No respected "scientist" then would have said we'd survive okay but people did it anyway. And the people living in these terrible places end up living a lot longer than we do, sitting on our asses watching TV and popping cheesy poofs.
I sure as hell don't want to pack up and move to Mars. But when the colony ships launch, there are going to be plenty of volunteers.
A bunch quit Compaq and went to AMD several years ago and helped develop the Athlon. Probably a lot of best jumped ship. But the developers that were left got transferred off to Intel when HP acquired Compaq.
The desktop Suns are an exception to the rule, I think. The Ultra 5's and 10's have shipped (especially in the beginning) with poor quality unreliable parts. I've seen their floppy drives and CD-ROM drives fail, and of course there was that fiasco with L2 cache on the UltraSparc 2 processors. But the Ultra 5's and 10's were the first attempt by Sun to make cheap PC hardware compatible systems. It was the first time Sun used IDE and PCI on a system. (The replacements (Blade 100 and 150) are much improved.) They aren't intended to be used for mission critical applications. They're just cheap computers that can run the same operating system as the servers. They're useful for developers who want to compile and test their software before deploying onto a Sun server. That's probably all that Sun workstations are useful for nowadays. Linux/x86 workstations would be cheaper and faster but the hardware and operating system incompatibility makes them tough to use for that purpose. Except in the case of Java applications, of course. Solaris/x86 took the operating system differences out of the picture, but the poor hardware support makes them infeasible for desktops.
So yes, comparing two RISC CPUs that both execute one instruction every two cycles on a MHz basis will give you a pretty good comparison of their relative performance.
That's not true. That would have been true back before there were pipelines, multiple functional units (superscalar), and branch prediction. It's no longer as simple as an Integer Addition taking 1 clock cycle. Now, 1 clock cycle is just the time it takes to complete a stage. And the number of actual stages varies greatly. There are stages for fetching the instruction, decoding the instruction, reading the registers, writing back the registers, and many stages for the actual execution in the functional unit. And because we're starting the execution of the next instruction before a current instruction completes, the pipeline may have to stall if there is a dependency between the instructions.
Even with 32 registers, you end up running out of registers. If you run out of registers, you would normally have to stall a pipeline until the previous instructions have finished using them, So processors have "renaming registers", and use them to direct results between instructions without having to compete for the same registers. Even though x86 has only 8 registers, every architecture really has around 128 registers underneath. More renaming registers would mean fewer pipeline stalls. But bigger registers sets mean slower register access. Then there's the issue of how many read ports can register sets have. Superscalar processors can have multiple instructions competing to read the same register.
Conditional branches induce long stalls in the pipeline because you won't know which instructions to pipeline until you've actually determined where you branch to. So the quality of the branch prediction can make a huge impact on performance.
Even memory access is complicated to compare because there are multiple levels of cache with different performance and sizes and load/store buffers in the processor so that instructions don't have to wait for memory operations to finish.
The fact of the matter is that MHz is not a reliable metric for either CISC or RISC. It's irrelevant whether it's CISC or RISC. What actually happens in a single clock cycle in independant of the complexity of the instruction set.
For example, the G3 processor had to run at a higher clock rate to match the performance on the 604. It's same situation between the Pentium 4 and Pentium 3. There are more pipeline stages, so the processor is doing less per clock cycle.
What a lot of people don't realize is that everything Intel does to squeeze more performance out of a stagnant instruction set architecture is valid to use on any architecture. Even RISC instructions are too complex to execute in a single clock cycle.
5-7% may not seem huge, but consider that if you bought a 3GHz processor and added 7% that it's now running at 3.2GHz.
With the 3GHz priced at $417 and the 3.2GHz priced at $637, you are essentially saving $220 with this change, not to mention the $100 savings over the 875 board.
Actually, the 875 board gives you that 5-7% increase by itself. This Asus hack just adds the same 5-7% speed increase to their 865 board. So you're only saving on the price difference between the 875 and 865, which is more like $70.
I agree that there is a potential for history to be rewritten, but I think the "Yesterday's Enterprise" episode would be a bad example. In the time travel scenarios in the original Star Trek and the The Next Generation, we've been seeing self reenforcing loops in time.
Take "Yesterday's Enterprise". If the Enterprise-D had never helped the Enterprise-C in the "original" timeline, wouldn't the C have been destroyed? They were on the verge of destruction before they went forward in time. My understanding was that history had never really been rewritten, except for the short span of time where the "C" had gone into the future. When the "C" went back, they eventually caused the Klingon-Federation peace treaty and the normal timeline was made possible. Nothing the "C" did has caused the normal timeline to be any different from our perspective. Any evidence of Tasha Yar existing in the past was erased as far as the Federation knew until Sela Yar made her appearance later.
Also, in the "Voyage Home", the Enterprise crew didn't make any changes to the past that we'd be aware of. A couple of whales and a whale researcher disappeared. Kirk sold off his pair of glasses, which he later re-acquired as antiques. Scotty gave away transparent aluminum technology, but then points out that maybe the guy he gave it to was recorded as the inventor anyway.
And in "First Contact", we don't really know that history turned out to be any different than the way the Next Generation crew had originally perceived it. There hasn't (yet) been any evidence of it. Cochrane did everything they remembered him doing. And in all the cases where they divulged the future to Cochrane, it might have made him do all the things they remembered him to have done.
The only case I remember of a non-sustaining loop in time was that TNG episode where the Enterprise-D kept getting destroyed until Data figured out what message to send back to get them out of the loop. That might establish the possibility in Trekdom that history can be rewritten. All that Star Trek series might just be one iteration of that non-sustaining loop.
That would be really lame of B&B to do this. But I wouldn't put it past them. They've already demonstrated that they have little regard for Roddenberry's work. Hasn't Braga been quoted as saying that he didn't like the original Star Trek series? Honestly, if those guys want to rejuvinate Star Trek they don't need to rewrite Star Trek history. They could try sticking to the spirit of Star Trek instead of just recycling old plots and techonobabble. How often did the original Star Trek revisit planets or mention the events of previous episodes? Star Trek 2 might have been the first time that's happened.
But I'm not saying that every episode should be independant. Babylon 5's strength was that hardly anything happened in an episode that wasn't important later on. But even Babylon 5 had a introduction, middle, and conclusion. It was like a novel told over 5 years. But there was a real ending, and the ending had signficance. But Star Trek just drags on now. If B&B want creative freedom, maybe they shouldn't have made a show that occurs between First Contact and the original Star Trek. Haven't we explored Vulcans to death by now? It was interesting the first time around with Spock, and probably because he was only half Vulcan.
The 4x SCSI's were great drives. But a lot of people had problems with their 16x and faster drives. I have 2 friends that each have had their 16x Yamaha drives die on them.
Re:A lesson the Linux worlds needs to learn
on
Why VHS Was Better
·
· Score: 1
If you want help, Slashdot is not the proper place to find it. Your coumment will very quickly be marked Offtopic. You'll have much better luck on Linux IRC channels for newbie questions and on Usenet for more advanced ones. Most often a question has been asked before so you'll want to use Google Groups to search Usenet.
That said, it's still not going to be easy. The learning curve is steep and there's very little help to be found. For most people, it's a waste of time.
I think a major reason people are willing to put up with it is because the easy to use proprietary operating systems can be impossible to troubleshoot when they fail. Everything is hidden. Given enough knowledge, you can always troubleshoot a problem in an open source UNIX. Every layer can be debugged because each layer is visible and accessible.
I'm sorry, but you are the idiot. No code was "stolen" from SCO at all.
USL (then owned by AT&T GAVE the code away to Berkeley and other universities almost 30 years ago. In 1995, USL (then owned by Novell) settled its lawsuit with Berkeley. They expressly gave permission to use all the code that is in BSD as long as the USL copyright was in the code. This is all BSD licensed code. And USL agreed never to sue anyone for using that code. So they did more then allow people do use their patented ideas. They allowed people to use the actual code. Think about that.
Linux doesn't even use that code. It uses the ideas. And many many other operating systems (probably all) use the same ideas. They were given the right to do so.
And USL got plenty back in return. Berkeley implemented and contributed all the networking code in UNIX. The UNIX Filesystem was designed and implemented by BSD. USL never paid for any of that stuff.
SCO uses Linux under a license (GPL) which says they cannot sue Linux for violating their patents. By enforcing patents on UNIX, they are violating the GPL for Linux and the GNU tools.
It's all a matter of perspective. In our everyday human existence, we deal only with humans. So differences that are tiny on the biological scale are easily noticed by us.
To a microbiologist, humans and chimpanzees could be considered indistinguishable. The part of the genome which governs sentience is very very small. When you're working at such a low level, it's not meaningful to make that distinction.
Similarly, for a zoologist the differences between humans is too tiny to even classify us into difference subspecies. Subspecies itself doesn't have any concrete biological defintion. And below the level of subspecies there doesn't exist any other classification scheme. There's no reason for them to consider human races.
Significance is a relative word. What may be significant in human society has nothing to do with what is significant in science. When a scientist says something is not significant, you must always associate that with the context.
The situation is more complicated when it comes to races. Our notion of race categories has more to do with local society. Living in the Western World, we get the impression that race is more clear cut than it really is. We don't think about all the people of in-between racial categories which have made biological classification impossible. In India, for example, you'll see a skin color variation from near white to pure black over a span of a mere 2 thousand miles. Most Americans just encounter white Europeans and black Africans and so their concept of race is defined by the extremes. In reality, there is only a smooth variation. Biology has to deal with reality. In this case, the common incorrect perception of reality shapes what is significant in society.
You're a little out of date. Solaris ships with the Sun Freeware CD, which has pretty much everything you need. You can also download all the packages from sun.com.
I know about that. But my point is that defects in 3rd party applications/servers, even ones that one provides in the Freeware CD, aren't counted as defects in Solaris. They're not officially part of Solaris. So when people compare the security of Linux and Solaris, they're comparing a Redhat system with tons of packages (and tons of their security holes) against a system that just comes with a basic UNIX installation. That's not a fair comparison. To be fair, you need to compare the base Solaris installation against only the Linux kernel, C library, the GNU *utils stuff, X11, and the basic network servers. And you can find a Linux distribution that is as barebones as that if you want that security.
Seems unlikely that's why your computers are crashing, since the defect only affected 400 and 450 MHz UltraSparc II CPUs.
The Ultra 10 never shipped at either of those speeds, and it used the UltraSparc IIi.
It affected the 440Mhz Ultra 10 also. The crash listed something about "Ecache". It's been over a year and it was a previous job so I don't remember the exact message. The only thing that worked was replacing the CPU. The floppy disk crash is a different issue I'm sure. It's probably an operating system bug rather than a hardware defect.
Solaris notebooks will satisfy their own market niche - users who need a stable, secure Unix with good development tools.
What development tools? Solaris doesn't even come with a compiler. You're just going to end up using gcc, gnu make, and emacs/vi like the Linux users. And there's an easy way to make Linux secure. Stop installing all that crap that you don't need!
As far as stability, I've seen Ultra 10's crash a hell of a lot more than Linux PCs. Damn Ultra 10 kernel panicked today when I tried writing to a floppy disk. That's the second Ultra 10 that's happened on. And at my previous job, at least 5 percent of the Ultra 10's we received had that L2 cache defect:
For at least 2 years, Sun produced CPUs with that defect. They weren't able to identify the faulty chips in their labs. The time it would be take to crash was random, anywhere between hours to a week. Sun offered help to customers (i.e. replaced the chips) if they signed agreements to keep silent about the problem. That worked the first year.
Slashdot Mirror
I think that's a little different. Netscape was already hopelessly behind when they open sourced their product. And there wasn't any hope of generating real revenue from the browser anyway. What they needed was to promote an open browser market so that Microsoft couldn't hijack HTML and HTTP. They succeeded posthumously in that respect, but their server business couldn't hold out that long.
The point of Java is to level the playing field. Sun needs people to write software for their systems. They won't do that if they have to write for Motif and UNIX, they'll write for Windows instead. They'll only write in Java if it will work well on their users' platforms. And that's the problem.
Sun doesn't know how to write a runtime that will work well for every platform. They need Apple to write an efficient JRE for Mac OS X, Windows experts for Windows, and Linux experts for Linux. And they need people to figure out how to compile Java straight to machine language (GCC's gcj) and make it fully compliant with the latest Java standards. The platform that Sun neglected the most, Linux, is the one that's most important to the success of Java now. Sun is adopting a Linux desktop (GNOME) so if they want Java software running on their own system they need to convince developers that Java is a viable solution for Linux desktops. That software can then be utilized on Sun's desktops. What does Sun gain if everyone just uses Java on Mac OS X or Windows? Their JVM and JDK are free.
The entire non-Solaris community cares. Sun simply does not know how to write efficient JVMs for platforms other than Solaris. The Linux JRE is slow. Microsoft Java runtime was tons faster than Sun's port. The whole point of Java is for it to be ubiquitous. It needs to work well on every platform, otherwise you would just use native code. The only way for this to happen is for Sun to let others maintain their own platform's ports. Sun is dependant on the GNOME project, which is primarily developed for Linux. So Sun needs the Linux community to buy into Java. If Java can't be distributed with major Linux distributions which can not happen. If the GNOME people don't buy into Java, then Sun's own Solaris platform will suffer. Sun can still control the standardization process of Java. They just need to open up the source for their own implementation so that others can quickly optimize it. If the GNU Java compiler (gcj) developers could interface with Sun's class library then it would help Java tremendously on UNIX.
Really? Figure 3 has a black box that says "Compatibility Support Module". And then a blue box on top of that saying "Legacy OS Loader". And then the text above the figure says "In addition, the Framework provides support for legacy OS interfaces via a set of drivers."
I think that's you're looking for? That Compatibility Support Module must be a BIOS emulation layer.
All the mid to high end video cards from Nvidia and ATI have heat sinks and fans. For example, the Radeon 9600 and 9800 series. Radeon 9200 is fanless, but it's not a Direct X 9 card like the others. If you want something modern, you're not going to have a choice.
It's not a dumb statement at all. Even though Pentium64 will certainly be cheaper and faster than Itanium, it can't be used in large scale multiprocessors. The Xeons only scale 4 ways because the bus protocol sucks. That's not going to change with a Xeon64. SGI and HP simply won't be able to use it.
Consider that SGI was already having trouble making money with their MIPS line. The whole point of going with Intel was the hope that Intel's Itanium would be high volume and therefore cheaper. Obviously that didn't happen and probably never will. So SGI will just continue to go backrupt. And HP will have no competitive server chip to sell. IBM will kick their ass. HP and SGI won't sell enough Itanium systems for Intel to bother developing it. Intel needs more of a market than the old HP-UX and IRIX customers. They needed the traditional Intel server market to accept Itanium.
They can't make it cheap. It has a 6MB L2 cache on it. That's the only reason it can halfway decent SPEC scores. And it makes the chip damn expensive to produce.
Yes, G5's and Pentium 4's do run circles around Sun workstations. But Sun's money is in their servers. Most people writing software for Suns aren't even looking at their workstations now, only their servers. The only real reason to buy a Sun workstation these days is for developing and testing software that you run on a Sun server. If you want a 32+ processor system with fast I/O and a robust design, then you're not going to find a Mac or PC that's going to do that. And a multiprocessor is going to be more compact than a computing cluster.
I think you've gotten this confused with Xouvert. Xouvert's the one that's making the build changes. FreeDesktop is splitting Xlib into 2 pieces and writing a new server based on kdrive.
Most people on campuses (students) don't pay for shit anyway. They're broke. There's no point worrying about them not buying stuff. 5 hours wasted to save $25 doesn't make sense for working people. It does make sense for students and teens that have no source of income. But why worry about them?
Honestly I think people exaggerate how bad DVD piracy is. Most adults will pay the $15-$20 to buy a DVD rather than waste time trying to downloading it from the Internet. DVDs are selling like hotcakes. People want the special features, commentaries, all that good stuff.
DivX and XviD at 1080i aren't going to be small either. Going from 640x480 standard def. to 1900x1080 is more than quadruple the size. Also keep in mind that HD-DVD is not going to be using MPEG-2. They can't fit a decent sized movie in MPEG-2 on a measly 15GB. They'll be using MPEG-4 or Windows Media. In which case, you're not getting any size reduction whatsoever. It's already as compressed as you can get it. You need to transfer the whole damn 15 GB. The BluRay people are the ones still promoting MPEG-2. BluRay has quite a bit more capacity than HD-DVD so they can pull it off.
Firebird uses XUL, Gecko, and PKI which were developed for Mozilla.
I think you should give Babylon 5 another shot. How much have you seen? You have to be patient with it. A lot of other sci fi series are designed so that each episode can stand by itself. Star Trek comes to mind.
Babylon 5 doesn't work that way. There are several episodes which I think are superb, but to appreciate those you'd have to watch most of the episodes prior.
The entire series is essentially a giant novel. Can you accurately judge the Lord of the Rings trilogy by the first 30 pages? At that point, they're not even out the Shire. It's kind of the same thing.
It's weaker on a per-episode basis, but as a series it really stands out. Consider after 7 seasons of Star Trek: The Next Generation, how much have the characters and the Star Trek universe changed? Not a whole lot. You start out with the Federation flagship, with a seasoned captain and a slightly inexperienced crew. In the end you have a less prickly captain and a crew that's a bit more interesting.
In contrast, the Babylon 5 station starts out as a joke. Becoming an ambassador to Babylon 5 was considered a demotion. Earth politicians had trouble justifying its existence. It was about as productive as the United Nations General Assembly. Nothing resolved, just political infighting. Hardly the place you'd expect anything interesting to happen.
The commander of Babylon 5 even has to justify space travel. When asked, he concludes that we have to be out in space because someday our sun will become cold. If we want to survive more than a few billion years we have to get ourselves somewhere else.
In the last episode, you actually get to see the death of the Sun billions of years in the future. Along the way, worlds and entire civilizations are destroyed. Governments collapse. But greater things are created along way.
Sorry to all the nerds who think Star Trek is real science. We are not getting off of the Earth.
The laws of physics as we know them aren't carved in stone. They're simply approximations. The General Theory of Relativity does not fit with Quantum Mechanics. The unified theories being developed do offer possibilities (higher dimensions, wormholes). No one knows that faster than light travel is impossible. There wouldn't have been a Breakthrough Propulsion Laboratory at NASA if respected physicists weren't willing to entertain the possibility.
Even if you could determine a place to go you don't have the right equipment. I mean you. Humans are not designed for long term exposure to space. Sorry. I don't know of one respected scientist who thinks our physiology and psychology permit long term exposure to radiation, low gravity, and isolation.
And you think these problems can't be handled? We can't build radiation shields? We can't produce artificial gravity? We can't build communities in space?
Human beings are willing to do some crazy things. We have people living in the Sahara desert, living in high altitudes, living in the Arctic. We could have just stayed in the trees. Our physiology is certainly more suited to it. Maybe there was a grand debate a long time ago about whether to leave the trees and whether to live in any of the godforsaken places human beings are willing to. No respected "scientist" then would have said we'd survive okay but people did it anyway. And the people living in these terrible places end up living a lot longer than we do, sitting on our asses watching TV and popping cheesy poofs.
I sure as hell don't want to pack up and move to Mars. But when the colony ships launch, there are going to be plenty of volunteers.
A bunch quit Compaq and went to AMD several years ago and helped develop the Athlon. Probably a lot of best jumped ship. But the developers that were left got transferred off to Intel when HP acquired Compaq.
The desktop Suns are an exception to the rule, I think. The Ultra 5's and 10's have shipped (especially in the beginning) with poor quality unreliable parts. I've seen their floppy drives and CD-ROM drives fail, and of course there was that fiasco with L2 cache on the UltraSparc 2 processors. But the Ultra 5's and 10's were the first attempt by Sun to make cheap PC hardware compatible systems. It was the first time Sun used IDE and PCI on a system. (The replacements (Blade 100 and 150) are much improved.) They aren't intended to be used for mission critical applications. They're just cheap computers that can run the same operating system as the servers. They're useful for developers who want to compile and test their software before deploying onto a Sun server. That's probably all that Sun workstations are useful for nowadays. Linux/x86 workstations would be cheaper and faster but the hardware and operating system incompatibility makes them tough to use for that purpose. Except in the case of Java applications, of course. Solaris/x86 took the operating system differences out of the picture, but the poor hardware support makes them infeasible for desktops.
That's not true. That would have been true back before there were pipelines, multiple functional units (superscalar), and branch prediction. It's no longer as simple as an Integer Addition taking 1 clock cycle. Now, 1 clock cycle is just the time it takes to complete a stage. And the number of actual stages varies greatly. There are stages for fetching the instruction, decoding the instruction, reading the registers, writing back the registers, and many stages for the actual execution in the functional unit. And because we're starting the execution of the next instruction before a current instruction completes, the pipeline may have to stall if there is a dependency between the instructions.
Even with 32 registers, you end up running out of registers. If you run out of registers, you would normally have to stall a pipeline until the previous instructions have finished using them, So processors have "renaming registers", and use them to direct results between instructions without having to compete for the same registers. Even though x86 has only 8 registers, every architecture really has around 128 registers underneath. More renaming registers would mean fewer pipeline stalls. But bigger registers sets mean slower register access. Then there's the issue of how many read ports can register sets have. Superscalar processors can have multiple instructions competing to read the same register.
Conditional branches induce long stalls in the pipeline because you won't know which instructions to pipeline until you've actually determined where you branch to. So the quality of the branch prediction can make a huge impact on performance.
Even memory access is complicated to compare because there are multiple levels of cache with different performance and sizes and load/store buffers in the processor so that instructions don't have to wait for memory operations to finish.
The fact of the matter is that MHz is not a reliable metric for either CISC or RISC. It's irrelevant whether it's CISC or RISC. What actually happens in a single clock cycle in independant of the complexity of the instruction set.
For example, the G3 processor had to run at a higher clock rate to match the performance on the 604. It's same situation between the Pentium 4 and Pentium 3. There are more pipeline stages, so the processor is doing less per clock cycle. What a lot of people don't realize is that everything Intel does to squeeze more performance out of a stagnant instruction set architecture is valid to use on any architecture. Even RISC instructions are too complex to execute in a single clock cycle.
With the 3GHz priced at $417 and the 3.2GHz priced at $637, you are essentially saving $220 with this change, not to mention the $100 savings over the 875 board.
Actually, the 875 board gives you that 5-7% increase by itself. This Asus hack just adds the same 5-7% speed increase to their 865 board. So you're only saving on the price difference between the 875 and 865, which is more like $70.
I agree that there is a potential for history to be rewritten, but I think the "Yesterday's Enterprise" episode would be a bad example. In the time travel scenarios in the original Star Trek and the The Next Generation, we've been seeing self reenforcing loops in time.
Take "Yesterday's Enterprise". If the Enterprise-D had never helped the Enterprise-C in the "original" timeline, wouldn't the C have been destroyed? They were on the verge of destruction before they went forward in time. My understanding was that history had never really been rewritten, except for the short span of time where the "C" had gone into the future. When the "C" went back, they eventually caused the Klingon-Federation peace treaty and the normal timeline was made possible. Nothing the "C" did has caused the normal timeline to be any different from our perspective. Any evidence of Tasha Yar existing in the past was erased as far as the Federation knew until Sela Yar made her appearance later.
Also, in the "Voyage Home", the Enterprise crew didn't make any changes to the past that we'd be aware of. A couple of whales and a whale researcher disappeared. Kirk sold off his pair of glasses, which he later re-acquired as antiques. Scotty gave away transparent aluminum technology, but then points out that maybe the guy he gave it to was recorded as the inventor anyway.
And in "First Contact", we don't really know that history turned out to be any different than the way the Next Generation crew had originally perceived it. There hasn't (yet) been any evidence of it. Cochrane did everything they remembered him doing. And in all the cases where they divulged the future to Cochrane, it might have made him do all the things they remembered him to have done.
The only case I remember of a non-sustaining loop in time was that TNG episode where the Enterprise-D kept getting destroyed until Data figured out what message to send back to get them out of the loop. That might establish the possibility in Trekdom that history can be rewritten. All that Star Trek series might just be one iteration of that non-sustaining loop.
That would be really lame of B&B to do this. But I wouldn't put it past them. They've already demonstrated that they have little regard for Roddenberry's work. Hasn't Braga been quoted as saying that he didn't like the original Star Trek series? Honestly, if those guys want to rejuvinate Star Trek they don't need to rewrite Star Trek history. They could try sticking to the spirit of Star Trek instead of just recycling old plots and techonobabble. How often did the original Star Trek revisit planets or mention the events of previous episodes? Star Trek 2 might have been the first time that's happened.
But I'm not saying that every episode should be independant. Babylon 5's strength was that hardly anything happened in an episode that wasn't important later on. But even Babylon 5 had a introduction, middle, and conclusion. It was like a novel told over 5 years. But there was a real ending, and the ending had signficance. But Star Trek just drags on now. If B&B want creative freedom, maybe they shouldn't have made a show that occurs between First Contact and the original Star Trek. Haven't we explored Vulcans to death by now? It was interesting the first time around with Spock, and probably because he was only half Vulcan.
The 4x SCSI's were great drives. But a lot of people had problems with their 16x and faster drives. I have 2 friends that each have had their 16x Yamaha drives die on them.
If you want help, Slashdot is not the proper place to find it. Your coumment will very quickly be marked Offtopic. You'll have much better luck on Linux IRC channels for newbie questions and on Usenet for more advanced ones. Most often a question has been asked before so you'll want to use Google Groups to search Usenet.
That said, it's still not going to be easy. The learning curve is steep and there's very little help to be found. For most people, it's a waste of time.
I think a major reason people are willing to put up with it is because the easy to use proprietary operating systems can be impossible to troubleshoot when they fail. Everything is hidden. Given enough knowledge, you can always troubleshoot a problem in an open source UNIX. Every layer can be debugged because each layer is visible and accessible.
I'm sorry, but you are the idiot. No code was "stolen" from SCO at all.
USL (then owned by AT&T GAVE the code away to Berkeley and other universities almost 30 years ago. In 1995, USL (then owned by Novell) settled its lawsuit with Berkeley. They expressly gave permission to use all the code that is in BSD as long as the USL copyright was in the code. This is all BSD licensed code. And USL agreed never to sue anyone for using that code. So they did more then allow people do use their patented ideas. They allowed people to use the actual code. Think about that.
Linux doesn't even use that code. It uses the ideas. And many many other operating systems (probably all) use the same ideas. They were given the right to do so.
And USL got plenty back in return. Berkeley implemented and contributed all the networking code in UNIX. The UNIX Filesystem was designed and implemented by BSD. USL never paid for any of that stuff.
SCO uses Linux under a license (GPL) which says they cannot sue Linux for violating their patents. By enforcing patents on UNIX, they are violating the GPL for Linux and the GNU tools.
It's all a matter of perspective. In our everyday human existence, we deal only with humans. So differences that are tiny on the biological scale are easily noticed by us.
To a microbiologist, humans and chimpanzees could be considered indistinguishable. The part of the genome which governs sentience is very very small. When you're working at such a low level, it's not meaningful to make that distinction.
Similarly, for a zoologist the differences between humans is too tiny to even classify us into difference subspecies. Subspecies itself doesn't have any concrete biological defintion. And below the level of subspecies there doesn't exist any other classification scheme. There's no reason for them to consider human races.
Significance is a relative word. What may be significant in human society has nothing to do with what is significant in science. When a scientist says something is not significant, you must always associate that with the context.
The situation is more complicated when it comes to races. Our notion of race categories has more to do with local society. Living in the Western World, we get the impression that race is more clear cut than it really is. We don't think about all the people of in-between racial categories which have made biological classification impossible. In India, for example, you'll see a skin color variation from near white to pure black over a span of a mere 2 thousand miles. Most Americans just encounter white Europeans and black Africans and so their concept of race is defined by the extremes. In reality, there is only a smooth variation. Biology has to deal with reality. In this case, the common incorrect perception of reality shapes what is significant in society.
You're a little out of date. Solaris ships with the Sun Freeware CD, which has pretty much everything you need. You can also download all the packages from sun.com.
I know about that. But my point is that defects in 3rd party applications/servers, even ones that one provides in the Freeware CD, aren't counted as defects in Solaris. They're not officially part of Solaris. So when people compare the security of Linux and Solaris, they're comparing a Redhat system with tons of packages (and tons of their security holes) against a system that just comes with a basic UNIX installation. That's not a fair comparison. To be fair, you need to compare the base Solaris installation against only the Linux kernel, C library, the GNU *utils stuff, X11, and the basic network servers. And you can find a Linux distribution that is as barebones as that if you want that security.
Seems unlikely that's why your computers are crashing, since the defect only affected 400 and 450 MHz UltraSparc II CPUs.
The Ultra 10 never shipped at either of those speeds, and it used the UltraSparc IIi.
It affected the 440Mhz Ultra 10 also. The crash listed something about "Ecache". It's been over a year and it was a previous job so I don't remember the exact message. The only thing that worked was replacing the CPU. The floppy disk crash is a different issue I'm sure. It's probably an operating system bug rather than a hardware defect.
Solaris notebooks will satisfy their own market niche - users who need a stable, secure Unix with good development tools.
What development tools? Solaris doesn't even come with a compiler. You're just going to end up using gcc, gnu make, and emacs/vi like the Linux users. And there's an easy way to make Linux secure. Stop installing all that crap that you don't need!
As far as stability, I've seen Ultra 10's crash a hell of a lot more than Linux PCs. Damn Ultra 10 kernel panicked today when I tried writing to a floppy disk. That's the second Ultra 10 that's happened on. And at my previous job, at least 5 percent of the Ultra 10's we received had that L2 cache defect:
http://www.theregister.co.uk/content/2/17417.html
For at least 2 years, Sun produced CPUs with that defect. They weren't able to identify the faulty chips in their labs. The time it would be take to crash was random, anywhere between hours to a week. Sun offered help to customers (i.e. replaced the chips) if they signed agreements to keep silent about the problem. That worked the first year.