I agree with the original poster. According to this the K6 and PII were both released in 1997. I remember debating whether to buy a PII or K6 at the time, actually. While the K6 matched some early PIIs in integer performance, I remember reading many Boot magazine articles ridiculing the K6's phenomenally bad, not-fully-pipelined FPU. As the PII line progressed, you basically only got a K6 if you never executed an FPU operation (ie. played a game!)
That's what I don't get about those people lambasting Intel for the P4 design. The long pipeline of the P4 was a *design decision* It had certain negatives (huge pipeline flush penalties, insane interrupt and system call latencies, high power consumption, etc) and certain positives (extremely high clock speeds, massive bandwidth and throughput). At the end of the day, what matters is the end result. The P4 is simply faster than any other chip in its class, and is very competitive with a lot of chips (50-70% the FPU performance of a Power4) that are way out of its league. This is how engineering works, plain and simple.
After the columbine shooting, I saw a quote in a magazine that was along the lines of: "When I was a kid, I couldn't get a Playboy into my room without my mom finding out. How did these kids get away with bringing shotguns and pipe bombs into their rooms?"
When I was 16, I went to see Hollow Man at the theater. Since it was rated R, the attendent would not let me in without a parent. My mom came with me to the attendent, and told her that I had her permission to watch the movie, yet the attendent still wouldn't let me in. The conservatives, who think that a nipple in a movie automatically makes it inappropriate for those under 18, had won. They had effectively imposed their (in my opinion rather twisted) world view upon my mom and me. What was sad was that the government helped them do it.
An OpenGL driver isn't just a hardware bangger. It's an entire OpenGL implementation of all parts of the pipeline. Everything from the userspace OpenGL library to the kernel driver. Plus, in NVIDIA's, there are also replacements for DRI and XAA. There is a *whole* lot in there ATI could use.
It's a fine OS for certain tasks (web serving, render farms), but for certain others (desktop use) it will always be playing catch-up to commercial offerings (Windows, OSX, etc) >>>>>>>> That's rich. I use Linux exclusively as my desktop OS. Booting into Windows is painful these days. Using the brand new G4 flat-panel iMacs with OS-X in our computer centers is even more painful (nothing so slow since I ran Win95 on my 486!) Linux might not be easy, or appropriate for newbies, but it makes a damn good desktop/workstation OS.
A network render stresses an OS quite a bit. It impacts the VM subsystem, network subsystem, disk I/O subsystem, etc. It also takes a lot of tweeking of the OS to set something like this up. The fact that Linux was used shows that's it's more suitable in these catagories than other OSs.
Which is my point. Buying a gun isn't illegal (and shouldn't be, for reasons of freedom) just because it could potentially be used to commit a crime. The same should go for software!
Just looked this up. Apparently, the laws were changed not to long ago (60s-70s). But counseling someone to commit suicide (which your post does) is still illegal in Canada (don't know about the US) and other countries. So your post could very well be illegal:)
So let me get this straight. I can go buy a gun, with little or no background checking, and have the potential to kill dozens of children, and it isn't illegal. Or, I could make a program that could theoretically be used to pirate some stupid ebooks, and that's illegal. Wow. That's such a fucked up set of priorities my head hurts. I'm going to go drown myself now...oh fuck! That's illegal too!
The issue is far more complex than you understand. Bundling agreements is where a large percentage of the population gets Office. Because of Microsoft's dominence of the operating system, it becomes easy for them to give companies who bundle Office preferential licensing terms for Windows. Also, there is the ethical question. Should an OS monopoly be making other software at all? It's an undeniable fact that Microsoft has dominated all sorts of market segments with their products. In the segments they've conquered, the Microsoft product is basically the only choice. Are you going to tell me that their control of the underlying OS doesn't have something to do with this? If you look at other markets, even if a product is dominent, there are lots of other viable options. For example, lot's of programs share the 3D modeling applications market segment, even though a few key products dominate. Same thing for audio/video editing. How come the same situation doesn't exist for Word or Visual Studio?
Competition in this area is a great thing. There is no competition in the Office Suite space in Windows, and thus the products there are sub-par. MS Office might be standard, but's it's not a great product. It's large, slow, the menus are confusing, and it's about as innovative as dirt. Word Perfect is an example of a great piece of software (that had a nifty text and graphics layout capabilities Word still can't match) that isn't here anymore because there is no competition in the Office Suite space. The current Linux office suits all offer something different. Open Office is big, bloated, and full of features. AbiWord is lightweight and very cross-platform. KWord is lightweight, well integrated with KDE, and has a nifty frames-based design that neither of the above have. The problem with choosing only one product is that a lot of these features are contridictory. You can be well integrated with KDE or GNOME, but then you can't be cross-platform. You can have tons of features, but then you can't be lightweight. You can use a frames-based layout system, but then you can't use a more conventional one.
I find that Windows users have the same problem when moving to Linux. They see so many features and scream "I need structure!" If you're finding the same problem with languages, just do what I tell those Windows people to do. Pretend no other languages exist and just use one. I'd even be happy to help you. You *must* use BrainFuck. No more options. Now, bend over...
Seriously, though, just because all these languages exist, there's no need to *learn* all of them. Just survey a variety of them and pick the ones that suit you. Personally, I use C++ for systems level stuff and Python for everything else. I'm looking into learning Prolog for AI type stuff. The important thing is to learn one or two in each catagory, and if you ever need the others, you can pick it up quite quickly.
People who know nothing shouldn't speak. The card was tested against a bunch of budget/midrange cards. On the low-end, the Radeon 9000 Pro starts at $80, while the "high-end" Radeon 9500 Pro goes all the way to $150. As for drivers, it could very well be that these drivers suck royally. But if they're this bad this late in the game, then what's Trident doing sending them out for review? Also, it must be noted that ATI's drivers aren't that great either, and that lots of beta drivers get tested (they are 'beta' not 'internal release' after all) and almost always they perform 50-70% as well as the final ones. So this card will get (at best) maybe 20 fps at 1024x768 with final drivers.
A lot of laptops have 1600x1200 resolution. But what do laptops have to do with anything? These are desktop chips. And they do take into account price: they mention that even if these are significantly cheaper than the Radeon 9000 Pro, they suck so bad it still wouldn't be worth it. A Radeon 9000 Pro runs $80 these days. This thing would have to sell around $30 for it to be any good, and projected retail prices are a whole lot higher than that ($100). As for power consumption, who cares? AGP only provides 25 watts of power, and none of the tested cards used an extra power connector. Even if the Radeon 9000 used the full power, and this card use 1/5 the power, the difference of 20 watts is worth jack shit.
I second that sentiment, though I'm too young to have have a degree in either. There is something about literature that attracts programmer types. It's probably the parallel relationship they have with using lanuage to express abstract concepts.
This was in Tannenbaum's book. Designing a huge software program is nowhere near as hard as (the example he used) designing an aircraft carrier. The latter involves a far larger total investment in skilled man hours. However, the former is far more complex, because the seperate parts of a program are tightly coupled and can interact in unforseen ways. The seperate parts of the aircraft carrier, however, are loosly coupled, and changes in (again, his example) the toilets don't effect the rader system.
It was Jan 2000 they dropped support for PPC, so that makes it Release 4. I think BeBoxes were supported as long as PPC was.
Re:Similar problem with Adaptec
on
Sun vs. OpenBSD?
·
· Score: 2
This is a far better performance than, for example, NVidia, whose drivers are well-known for breaking every few kernel releases because of their binary-only nature. >>>>>>>>> Bullshit. The binary nature has nothing to do with it. They don't touch the kernel at all. The only thing that could be breaking is the kernel glue code, which is provided in source form with the drivers. I've never had the drivers break on me in a stable kernel release (in their default form) and (once patched) I've been able to track kernel 2.5 since about 2.5.38 (now up to 2.5.50) without them breaking a single time.
Slashdot Mirror
I agree with the original poster. According to this the K6 and PII were both released in 1997. I remember debating whether to buy a PII or K6 at the time, actually. While the K6 matched some early PIIs in integer performance, I remember reading many Boot magazine articles ridiculing the K6's phenomenally bad, not-fully-pipelined FPU. As the PII line progressed, you basically only got a K6 if you never executed an FPU operation (ie. played a game!)
As for K6, it beat Intel in MHz, but K6 existed during the early PII era, and was regularly ridiculed for its inner city high-school FPU capabilities.
That's what I don't get about those people lambasting Intel for the P4 design. The long pipeline of the P4 was a *design decision* It had certain negatives (huge pipeline flush penalties, insane interrupt and system call latencies, high power consumption, etc) and certain positives (extremely high clock speeds, massive bandwidth and throughput). At the end of the day, what matters is the end result. The P4 is simply faster than any other chip in its class, and is very competitive with a lot of chips (50-70% the FPU performance of a Power4) that are way out of its league. This is how engineering works, plain and simple.
After the columbine shooting, I saw a quote in a magazine that was along the lines of: "When I was a kid, I couldn't get a Playboy into my room without my mom finding out. How did these kids get away with bringing shotguns and pipe bombs into their rooms?"
When I was 16, I went to see Hollow Man at the theater. Since it was rated R, the attendent would not let me in without a parent. My mom came with me to the attendent, and told her that I had her permission to watch the movie, yet the attendent still wouldn't let me in. The conservatives, who think that a nipple in a movie automatically makes it inappropriate for those under 18, had won. They had effectively imposed their (in my opinion rather twisted) world view upon my mom and me. What was sad was that the government helped them do it.
Wow. Animal porn and graph theory in the same post. Only on Slashdot...
An OpenGL driver isn't just a hardware bangger. It's an entire OpenGL implementation of all parts of the pipeline. Everything from the userspace OpenGL library to the kernel driver. Plus, in NVIDIA's, there are also replacements for DRI and XAA. There is a *whole* lot in there ATI could use.
That's probably a function of hardware more than anything else. Alpha's have always been floating point screamers in a way MIPS's never have.
It's a fine OS for certain tasks (web serving, render farms), but for certain others (desktop use) it will always be playing catch-up to commercial offerings (Windows, OSX, etc)
>>>>>>>>
That's rich. I use Linux exclusively as my desktop OS. Booting into Windows is painful these days. Using the brand new G4 flat-panel iMacs with OS-X in our computer centers is even more painful (nothing so slow since I ran Win95 on my 486!) Linux might not be easy, or appropriate for newbies, but it makes a damn good desktop/workstation OS.
A network render stresses an OS quite a bit. It impacts the VM subsystem, network subsystem, disk I/O subsystem, etc. It also takes a lot of tweeking of the OS to set something like this up. The fact that Linux was used shows that's it's more suitable in these catagories than other OSs.
What's the purpose of handguns other than to kill people? Do you go hunting with your handguns?
Which is my point. Buying a gun isn't illegal (and shouldn't be, for reasons of freedom) just because it could potentially be used to commit a crime. The same should go for software!
Just looked this up. Apparently, the laws were changed not to long ago (60s-70s). But counseling someone to commit suicide (which your post does) is still illegal in Canada (don't know about the US) and other countries. So your post could very well be illegal :)
This is actually more important than you'd think. Most current PDAs just don't handle PDF fast enough. Converting it to text is very useful.
So let me get this straight. I can go buy a gun, with little or no background checking, and have the potential to kill dozens of children, and it isn't illegal. Or, I could make a program that could theoretically be used to pirate some stupid ebooks, and that's illegal. Wow. That's such a fucked up set of priorities my head hurts. I'm going to go drown myself now...oh fuck! That's illegal too!
The issue is far more complex than you understand. Bundling agreements is where a large percentage of the population gets Office. Because of Microsoft's dominence of the operating system, it becomes easy for them to give companies who bundle Office preferential licensing terms for Windows. Also, there is the ethical question. Should an OS monopoly be making other software at all? It's an undeniable fact that Microsoft has dominated all sorts of market segments with their products. In the segments they've conquered, the Microsoft product is basically the only choice. Are you going to tell me that their control of the underlying OS doesn't have something to do with this? If you look at other markets, even if a product is dominent, there are lots of other viable options. For example, lot's of programs share the 3D modeling applications market segment, even though a few key products dominate. Same thing for audio/video editing. How come the same situation doesn't exist for Word or Visual Studio?
Actually, can you even buy a new PC these days without Word (or MS Works) being bundled in?
Competition in this area is a great thing. There is no competition in the Office Suite space in Windows, and thus the products there are sub-par. MS Office might be standard, but's it's not a great product. It's large, slow, the menus are confusing, and it's about as innovative as dirt. Word Perfect is an example of a great piece of software (that had a nifty text and graphics layout capabilities Word still can't match) that isn't here anymore because there is no competition in the Office Suite space. The current Linux office suits all offer something different. Open Office is big, bloated, and full of features. AbiWord is lightweight and very cross-platform. KWord is lightweight, well integrated with KDE, and has a nifty frames-based design that neither of the above have. The problem with choosing only one product is that a lot of these features are contridictory. You can be well integrated with KDE or GNOME, but then you can't be cross-platform. You can have tons of features, but then you can't be lightweight. You can use a frames-based layout system, but then you can't use a more conventional one.
I find that Windows users have the same problem when moving to Linux. They see so many features and scream "I need structure!" If you're finding the same problem with languages, just do what I tell those Windows people to do. Pretend no other languages exist and just use one. I'd even be happy to help you. You *must* use BrainFuck. No more options. Now, bend over...
Seriously, though, just because all these languages exist, there's no need to *learn* all of them. Just survey a variety of them and pick the ones that suit you. Personally, I use C++ for systems level stuff and Python for everything else. I'm looking into learning Prolog for AI type stuff. The important thing is to learn one or two in each catagory, and if you ever need the others, you can pick it up quite quickly.
People who know nothing shouldn't speak. The card was tested against a bunch of budget/midrange cards. On the low-end, the Radeon 9000 Pro starts at $80, while the "high-end" Radeon 9500 Pro goes all the way to $150. As for drivers, it could very well be that these drivers suck royally. But if they're this bad this late in the game, then what's Trident doing sending them out for review? Also, it must be noted that ATI's drivers aren't that great either, and that lots of beta drivers get tested (they are 'beta' not 'internal release' after all) and almost always they perform 50-70% as well as the final ones. So this card will get (at best) maybe 20 fps at 1024x768 with final drivers.
A lot of laptops have 1600x1200 resolution. But what do laptops have to do with anything? These are desktop chips. And they do take into account price: they mention that even if these are significantly cheaper than the Radeon 9000 Pro, they suck so bad it still wouldn't be worth it. A Radeon 9000 Pro runs $80 these days. This thing would have to sell around $30 for it to be any good, and projected retail prices are a whole lot higher than that ($100). As for power consumption, who cares? AGP only provides 25 watts of power, and none of the tested cards used an extra power connector. Even if the Radeon 9000 used the full power, and this card use 1/5 the power, the difference of 20 watts is worth jack shit.
I second that sentiment, though I'm too young to have have a degree in either. There is something about literature that attracts programmer types. It's probably the parallel relationship they have with using lanuage to express abstract concepts.
This was in Tannenbaum's book. Designing a huge software program is nowhere near as hard as (the example he used) designing an aircraft carrier. The latter involves a far larger total investment in skilled man hours. However, the former is far more complex, because the seperate parts of a program are tightly coupled and can interact in unforseen ways. The seperate parts of the aircraft carrier, however, are loosly coupled, and changes in (again, his example) the toilets don't effect the rader system.
It was Jan 2000 they dropped support for PPC, so that makes it Release 4. I think BeBoxes were supported as long as PPC was.
This is a far better performance than, for example, NVidia, whose drivers are well-known for breaking every few kernel releases because of their binary-only nature.
>>>>>>>>>
Bullshit. The binary nature has nothing to do with it. They don't touch the kernel at all. The only thing that could be breaking is the kernel glue code, which is provided in source form with the drivers. I've never had the drivers break on me in a stable kernel release (in their default form) and (once patched) I've been able to track kernel 2.5 since about 2.5.38 (now up to 2.5.50) without them breaking a single time.