Actually, with a good implementation of the STL (SGI's for example), fast, standard, easy-to-use data structures and algorithms become a prime selling point for C++.
Hell, given gnome and kde, developers these days can't write acceptably fast software even in C/C++ (even on my 1.5GHz Athlon XP, KDE 3.x and GNOME 1.4 are pushing the limits of 'acceptable'), so maybe if they switched to languages that would allow them to add stupid features even more quickly without borking performance, software might even get faster! Seriously, though, it's all relative. You use whatever language you're comfortable with. Personally, I don't find C++ all that ardous, and with STL and whatnot, its as easy to write C++ applications as Java apps. When I need to interface at the shell level, however, I tend to use Perl, which can't be beat for its quickness and utility. But if you don't like C++, that's fine. Go use Java or something. I personally won't touch Java with a ten-foot pole, but then again, I grew up with C++. Of course, for some things, C/C++ is unavoidable. Kernel stuff, for example. Even C++ is almost too high-level for kernel development (lots of runtime support needed).
Actually, what deferencing the NULL pointer does depends on the platform. In plain DOS, it reads whatever value is in memory location 0 (part of the interrupt descriptor table on x86), since DOS doesn't use memory protection. In protected memory OSs, deferencing the NULL pointer essentially becomes an access to memory address 0, and this address is always unmapped by the kernel so referencing it generates a fault, which the OS catches, and results in the application being killed. It doesn't warp spacetime or anything like that, if that's what you were hoping for...
Ach. Damn blind Mac people. Font rendering under OS X is too damn blurry! Sorry, but BitStream's got the AA font renderer market all tied up! Take a look at QNX RtP sometime. FontFusion kicks ass!
Umm, you can always get source packages for important stuff. That's why binary distros are better. You can recompile glibc, X, and KDE for optimum performance, but can just download any application you need quickly (like I really need AbiWord ultra-optimized!) without waiting for compiles.
Re:So,We are not boycotting the evil MPAA this wee
on
Review: Spiderman
·
· Score: 2
They don't. Sure, it'll work under Linux as an optical device (SPDIF) but the OpenMG software, needed for NetMD mode (which allows LP mode and quick transfer of data) only runs on Windows.
Re:So,We are not boycotting the evil MPAA this wee
on
Review: Spiderman
·
· Score: 2
Sony is a big company, split into lots of piecies. Sony movies and sony games are great divisions. The funny thing is that one arm of Sony doesn't know what the other arm is doing. There's that new NetMD commercial, which tells you how you can use the Sony NetMD to rip music, meanwhile Sony Music is going freaky over the wave of MP3 "music stealing."
BTW> The NetMD rocks. 2.5 (at good quality) hours of music, 56 hours of battery life, cheap ($2) media. The players themselves are small, durable, and can take quite a beating without skipping. Now, if it only worked in Linux;)
Measuring it exactly is a little wonky. It's definately not 170 billion matrix multiplies. It could be 64+48, but most likely, it is they're counting 48 floating point multiply/accumulate (a mul and an add in one instruction) + 16 multiplies. For some reason, multiply-accumulate is easy enough that they can implement it as an individual instruction.
Actually, STLport is almost close to a standard implementation. It's extremely complient, high-quality, and free so you can use it wherever you need to.
It depends on your setup. For me at least, the only way to get my monitor to do 1152x864 at 85Hz (instead of the headache inducing 75Hz XFree 4.1 thought it should run at) was to manually add a modline. This involved not only editing XF86Config (simple edits like this shouldn't be a problem for most people) but learning a whole lot more CRT theory than I cared to. In the end, even I ran out of patience and wrote a BeOS program that grabbed the appropriate numbers. Until the X developers engineer some more "smarts" into XFree86, it won't approach the ease of configuration of other systems. Of course, I am happy with the config situation of most Linux components. KDE and GNOME are easy to configure, and the kernel is taking a lot of steps in the right direction (trying to make everything automagical). The real problem is in the user-level (ie. things that aren't part of the kernel or GUI) software, all of which have different configuration systems and none of which are designed by people skilled at UI (and yes, config files are a form of a user interface) design.
But for every power user tapping his fingers waiting for the compiler, there are a hundred Grandmas typing things into MS Word or reading web pages. Which mean a CPU utilization of about one per cent, since the computer spends most of its time waiting for human input. >>>>>>>> This is very misleading. While the CPU utilization over a period of 10 minutes might be 1% for someone browsing the web, it doesn't mean that the user doesn't wait an agonizing several seconds waiting for a complex page to load. With my DSL connection on my 1.5GHz machine, using Konqueror is a much more pleasent experience than using Mozilla because Konq is so much faster. Of course, neither program taxes my hardware much overall, but what counts for the user experience is maximum latency, not total throughput (so to speak). With more and more complex content coming out (in particular SVG, which is pretty slow to render, and is even slower if complex animations are used) CPU's will need to keep getting faster just to keep up with the internet.
I don't know what kind of trade-offs they are making. KDE 3.0 has so many features that I'd guess they're trading speed for features. The difference probably is that Microsoft can implement lots of features and keep them fast while KDE (with less resources) can implement lots of features and make them slow. In that case, is that the kind of trade-off they should be making? It's not just an efficiency thing. There is NO machine that will run KDE 3.0 at an acceptable speed for me. I'm not asking them to make it usable on a Pentium 90, but at least make it usable on my 1.5 GHz Athlon. It is probably *your* setup. Win98 is a good deal slower than Win2k/XP. BTW, I'm on a debian box as well.
The problem is that binaries on Gentoo aren't released nearly as often as on other systems, and aren't as comprehensive. Gentoo is a source-based distro at heart.
I never said it was easy to fix. I simply pointed out that it has been done before many times, so it can't be all that hard. If OSS software is so great, why can't they beat MS in performance (one of their weak areas!)?
Making something complex doesn't automatically make it bad design. The EE has a huge amount of resources available for vector processing. The PS2 needs all those resources if not more. If the only way to achieve that level of power is a complex chip, so be it. It's not just Sony, however. NVIDIA does it all the time. They introduce gigantic complex GPUs every six months at very high prices. Eventually, those chips stabilize and trickle down into the mainstream, but the high-end gets the complex stuff, always. There's nothing wrong with that, as long as you can make money on it (and oh yes, Sony is making a ton of money on PS2 licenses).
Wow! You can browse and watch AVI's at the same time on a 500Mhz processor! Windows can do that on an old P90. On anything above a 500Mhz machine, NO desktop should have ANY lag. On my 1.5Ghz machine, KDE 2.x is still agonizingly slow (so much so that I switched to GNOME, which is as slow, but at least apps start quickly) to the point where using it is irritating.
In that time, I could install and configure a dozen debian setups... Gentoo is a great distro, but there is no way in hell I'd ever survive the compile times. I'm a "install tools as I need them" kind of guy.
I doubt the problem with Sony is a lack of careful chip design. The problem Sony is having with the EE is that it is freak-ass complicated. Read This ArsTechnica article for details about why the EE is so complex. Also read This article for information about why the GS is so complex. In total, the EE + GS consist of about 55 million transistors, which is comparable to the 63 million in the GeForce4. Unlike NVIDIA, however, Sony did not have the luxury of an established, evolving architecture (GeForce1 -> GeForce 4) that allowed NVIDIA to implement it's complex chips with relatively few problems.
While it may or may not be true that the Arab press has a bad reputation, is this kind of statement any better than "jews can't be trusted with money" or "christians are all prosletyzers" crap?
The clone now has no rights, since she was purchased, created and programmed (raised) by the company. >>>>>> Umm, this statement doesn't make sense logically. The situation of a clone is no different from any adoption or surrogate mother case. A women would have to carry the baby. In order to get the baby, the corporation would have to make formal procedures to transfer guardianship to the corporation. Once that guardianship is transferred, the corporation has to follow all the rules any guardian does. Once the child passes the age of 18, the corporation no longer has any legal hold on the child, just as with any other guardianship situation. Even if you consider that the corporation rasies a test-tube baby, you still have to remember that people are people. This is simple biology. An organism that is produced by cloning human DNA must also be a human. Legal rights are granted to people, and thus as long as the clone is a person, they have the same legal rights as anyone else.
I'm just waiting for the KDE 3.0 to get into sid, so I have a few questions for people already using it:
1) Is it really faster? I keep hearing this, but apparently, most people think fast actually means slow. I've been trying to find a decently fast desktop for my 1.5GHz/256MB Athlon XP, and so far, only Fluxbox+GTK fits the bill. It runs Galeon, Evolution, etc at a pace only slightly slower than Win2K on my 750MHz Duron... Still, I like KDE better (prettier, certain apps like KDevelop are nicer than comparative GNOME ones) but so far, 2.2.2 is unusably slow. I don't want to hear anything about packages and optimizing and whatnot, been there done that. I'm running Debian sid with 2.4.18 + xfs + preempt + lock-break + O(1) sched, with X at -11 (per Debian defaults) and Fluxbox at -10. Doubt I can optimize much more. I just want to know: Does KDE 3.0 finally make KDE even remotely comparable to (well-optimized installs of) Win2k/XP?
2) Did they fix the annoying font display problems? There is this peculier issue with all KDE applications that causes the Microsoft Courier New font to be displayed very strangely. The text itself is fine, but it looks like all the lines are double spaced. When using Courier new in KDevelop or KWrite, one gets about 70% as many lines on screen at once as one can in any GTK+ or regular Xlib program. I've seen this problem reported in a few places (such as here ) but I've never seen a resolution. I doubt its on my end, since I've had this problem in every single install of KDE 2.x I've ever used, including Mandrake and Debian.
3) Does anyone besides me think that the whole OS-X UI on Linux is an asthetic nightmare? First, I don't much like the OS-X look. Second, KDE isn't OS-X. It's KDE. It should have it's own personality. It's getting hard these days to find a theme that doesn't look like some other OS. Even then, the themes are never as well put-together and consistant as the originals they copy.
4) Is it hard to understand that you can have good looks without all the glitzy, performance robbing features? I have yet to see a Linux UI that matches the elegance and polish of BeOS or MacOS. In the end, the UI does little more than move some bitmaps and text around the screen. XF86 4.x on my RivaTNT can blit 3000 100x100 bitmaps to the screen every second. It is trivial to make a UI that looks good (which is more a function of the quality of the drawing in the bitmap) and is also fast. Additionally, stuff like transparency and animation, while nifty, doesn't make up for the fact that there is not a single Linux UI that looks polished and elegant. GTK+ looks dated, and 3rd party themes are of depressing quality (and not really developed much anymore, apparently). Plus, GTK+'s container mechanism makes for some rather annoying display quality. Resizing a Gaim window, for example, causes the 4 icons along the bottom to spread farther apart, to the point a comical distance seperates them. KDE is a bit better, but there are still many places where it just doesn't look professionally put together. It's like American cars vs. European cars. Sure American cars have all the things that European cars do, but they're just not as well put together.
Slashdot Mirror
Actually, with a good implementation of the STL (SGI's for example), fast, standard, easy-to-use data structures and algorithms become a prime selling point for C++.
Hell, given gnome and kde, developers these days can't write acceptably fast software even in C/C++ (even on my 1.5GHz Athlon XP, KDE 3.x and GNOME 1.4 are pushing the limits of 'acceptable'), so maybe if they switched to languages that would allow them to add stupid features even more quickly without borking performance, software might even get faster! Seriously, though, it's all relative. You use whatever language you're comfortable with. Personally, I don't find C++ all that ardous, and with STL and whatnot, its as easy to write C++ applications as Java apps. When I need to interface at the shell level, however, I tend to use Perl, which can't be beat for its quickness and utility. But if you don't like C++, that's fine. Go use Java or something. I personally won't touch Java with a ten-foot pole, but then again, I grew up with C++. Of course, for some things, C/C++ is unavoidable. Kernel stuff, for example. Even C++ is almost too high-level for kernel development (lots of runtime support needed).
Actually, what deferencing the NULL pointer does depends on the platform. In plain DOS, it reads whatever value is in memory location 0 (part of the interrupt descriptor table on x86), since DOS doesn't use memory protection. In protected memory OSs, deferencing the NULL pointer essentially becomes an access to memory address 0, and this address is always unmapped by the kernel so referencing it generates a fault, which the OS catches, and results in the application being killed. It doesn't warp spacetime or anything like that, if that's what you were hoping for...
Ach. Damn blind Mac people. Font rendering under OS X is too damn blurry! Sorry, but BitStream's got the AA font renderer market all tied up! Take a look at QNX RtP sometime. FontFusion kicks ass!
Umm, you can always get source packages for important stuff. That's why binary distros are better. You can recompile glibc, X, and KDE for optimum performance, but can just download any application you need quickly (like I really need AbiWord ultra-optimized!) without waiting for compiles.
They don't. Sure, it'll work under Linux as an optical device (SPDIF) but the OpenMG software, needed for NetMD mode (which allows LP mode and quick transfer of data) only runs on Windows.
Sony is a big company, split into lots of piecies. Sony movies and sony games are great divisions. The funny thing is that one arm of Sony doesn't know what the other arm is doing. There's that new NetMD commercial, which tells you how you can use the Sony NetMD to rip music, meanwhile Sony Music is going freaky over the wave of MP3 "music stealing."
;)
BTW> The NetMD rocks. 2.5 (at good quality) hours of music, 56 hours of battery life, cheap ($2) media. The players themselves are small, durable, and can take quite a beating without skipping. Now, if it only worked in Linux
Measuring it exactly is a little wonky. It's definately not 170 billion matrix multiplies. It could be 64+48, but most likely, it is they're counting 48 floating point multiply/accumulate (a mul and an add in one instruction) + 16 multiplies. For some reason, multiply-accumulate is easy enough that they can implement it as an individual instruction.
Actually, STLport is almost close to a standard implementation. It's extremely complient, high-quality, and free so you can use it wherever you need to.
It depends on your setup. For me at least, the only way to get my monitor to do 1152x864 at 85Hz (instead of the headache inducing 75Hz XFree 4.1 thought it should run at) was to manually add a modline. This involved not only editing XF86Config (simple edits like this shouldn't be a problem for most people) but learning a whole lot more CRT theory than I cared to. In the end, even I ran out of patience and wrote a BeOS program that grabbed the appropriate numbers. Until the X developers engineer some more "smarts" into XFree86, it won't approach the ease of configuration of other systems. Of course, I am happy with the config situation of most Linux components. KDE and GNOME are easy to configure, and the kernel is taking a lot of steps in the right direction (trying to make everything automagical). The real problem is in the user-level (ie. things that aren't part of the kernel or GUI) software, all of which have different configuration systems and none of which are designed by people skilled at UI (and yes, config files are a form of a user interface) design.
But for every power user tapping his fingers waiting for the compiler, there are a hundred Grandmas typing things into MS Word or reading web pages. Which mean a CPU utilization of about one per cent, since the computer spends most of its time waiting for human input.
>>>>>>>>
This is very misleading. While the CPU utilization over a period of 10 minutes might be 1% for someone browsing the web, it doesn't mean that the user doesn't wait an agonizing several seconds waiting for a complex page to load. With my DSL connection on my 1.5GHz machine, using Konqueror is a much more pleasent experience than using Mozilla because Konq is so much faster. Of course, neither program taxes my hardware much overall, but what counts for the user experience is maximum latency, not total throughput (so to speak). With more and more complex content coming out (in particular SVG, which is pretty slow to render, and is even slower if complex animations are used) CPU's will need to keep getting faster just to keep up with the internet.
It's amazing how computing technology has advanced so. It used to be that we'd see fake HTML in postings. Now, its fake XML. Ah, progress...
I don't know what kind of trade-offs they are making. KDE 3.0 has so many features that I'd guess they're trading speed for features. The difference probably is that Microsoft can implement lots of features and keep them fast while KDE (with less resources) can implement lots of features and make them slow. In that case, is that the kind of trade-off they should be making? It's not just an efficiency thing. There is NO machine that will run KDE 3.0 at an acceptable speed for me. I'm not asking them to make it usable on a Pentium 90, but at least make it usable on my 1.5 GHz Athlon. It is probably *your* setup. Win98 is a good deal slower than Win2k/XP. BTW, I'm on a debian box as well.
The problem is that binaries on Gentoo aren't released nearly as often as on other systems, and aren't as comprehensive. Gentoo is a source-based distro at heart.
I never said it was easy to fix. I simply pointed out that it has been done before many times, so it can't be all that hard. If OSS software is so great, why can't they beat MS in performance (one of their weak areas!)?
Making something complex doesn't automatically make it bad design. The EE has a huge amount of resources available for vector processing. The PS2 needs all those resources if not more. If the only way to achieve that level of power is a complex chip, so be it. It's not just Sony, however. NVIDIA does it all the time. They introduce gigantic complex GPUs every six months at very high prices. Eventually, those chips stabilize and trickle down into the mainstream, but the high-end gets the complex stuff, always. There's nothing wrong with that, as long as you can make money on it (and oh yes, Sony is making a ton of money on PS2 licenses).
Wow! You can browse and watch AVI's at the same time on a 500Mhz processor! Windows can do that on an old P90. On anything above a 500Mhz machine, NO desktop should have ANY lag. On my 1.5Ghz machine, KDE 2.x is still agonizingly slow (so much so that I switched to GNOME, which is as slow, but at least apps start quickly) to the point where using it is irritating.
In that time, I could install and configure a dozen debian setups... Gentoo is a great distro, but there is no way in hell I'd ever survive the compile times. I'm a "install tools as I need them" kind of guy.
Nissan (Datsun) is famous for their Z series cars, hence z.com.
No, they weren't. Take a look at This diagram
I doubt the problem with Sony is a lack of careful chip design. The problem Sony is having with the EE is that it is freak-ass complicated. Read This ArsTechnica article for details about why the EE is so complex. Also read This article for information about why the GS is so complex. In total, the EE + GS consist of about 55 million transistors, which is comparable to the 63 million in the GeForce4. Unlike NVIDIA, however, Sony did not have the luxury of an established, evolving architecture (GeForce1 -> GeForce 4) that allowed NVIDIA to implement it's complex chips with relatively few problems.
While it may or may not be true that the Arab press has a bad reputation, is this kind of statement any better than "jews can't be trusted with money" or "christians are all prosletyzers" crap?
The clone now has no rights, since she was purchased, created and programmed (raised) by the company.
>>>>>>
Umm, this statement doesn't make sense logically. The situation of a clone is no different from any adoption or surrogate mother case. A women would have to carry the baby. In order to get the baby, the corporation would have to make formal procedures to transfer guardianship to the corporation. Once that guardianship is transferred, the corporation has to follow all the rules any guardian does. Once the child passes the age of 18, the corporation no longer has any legal hold on the child, just as with any other guardianship situation. Even if you consider that the corporation rasies a test-tube baby, you still have to remember that people are people. This is simple biology. An organism that is produced by cloning human DNA must also be a human. Legal rights are granted to people, and thus as long as the clone is a person, they have the same legal rights as anyone else.
But, can't you just change the "viewed" attribute on the file from false to true? Oh, wait, this isn't BeOS... ;)
I'm just waiting for the KDE 3.0 to get into sid, so I have a few questions for people already using it:
1) Is it really faster? I keep hearing this, but apparently, most people think fast actually means slow. I've been trying to find a decently fast desktop for my 1.5GHz/256MB Athlon XP, and so far, only Fluxbox+GTK fits the bill. It runs Galeon, Evolution, etc at a pace only slightly slower than Win2K on my 750MHz Duron... Still, I like KDE better (prettier, certain apps like KDevelop are nicer than comparative GNOME ones) but so far, 2.2.2 is unusably slow. I don't want to hear anything about packages and optimizing and whatnot, been there done that. I'm running Debian sid with 2.4.18 + xfs + preempt + lock-break + O(1) sched, with X at -11 (per Debian defaults) and Fluxbox at -10. Doubt I can optimize much more. I just want to know: Does KDE 3.0 finally make KDE even remotely comparable to (well-optimized installs of) Win2k/XP?
2) Did they fix the annoying font display problems? There is this peculier issue with all KDE applications that causes the Microsoft Courier New font to be displayed very strangely. The text itself is fine, but it looks like all the lines are double spaced. When using Courier new in KDevelop or KWrite, one gets about 70% as many lines on screen at once as one can in any GTK+ or regular Xlib program. I've seen this problem reported in a few places (such as here ) but I've never seen a resolution. I doubt its on my end, since I've had this problem in every single install of KDE 2.x I've ever used, including Mandrake and Debian.
3) Does anyone besides me think that the whole OS-X UI on Linux is an asthetic nightmare? First, I don't much like the OS-X look. Second, KDE isn't OS-X. It's KDE. It should have it's own personality. It's getting hard these days to find a theme that doesn't look like some other OS. Even then, the themes are never as well put-together and consistant as the originals they copy.
4) Is it hard to understand that you can have good looks without all the glitzy, performance robbing features? I have yet to see a Linux UI that matches the elegance and polish of BeOS or MacOS. In the end, the UI does little more than move some bitmaps and text around the screen. XF86 4.x on my RivaTNT can blit 3000 100x100 bitmaps to the screen every second. It is trivial to make a UI that looks good (which is more a function of the quality of the drawing in the bitmap) and is also fast. Additionally, stuff like transparency and animation, while nifty, doesn't make up for the fact that there is not a single Linux UI that looks polished and elegant. GTK+ looks dated, and 3rd party themes are of depressing quality (and not really developed much anymore, apparently). Plus, GTK+'s container mechanism makes for some rather annoying display quality. Resizing a Gaim window, for example, causes the 4 icons along the bottom to spread farther apart, to the point a comical distance seperates them. KDE is a bit better, but there are still many places where it just doesn't look professionally put together. It's like American cars vs. European cars. Sure American cars have all the things that European cars do, but they're just not as well put together.