You're kidding right? Not only is it commercial "payware" but it is "software that'll cost you an arm and a leg and you have to buy a new computer to use." Hell, this thing costs more than a lot of CARS! If you can find a GNU animation tool that comes anywhere within the same solar system as this thing, then you've found a really dedicated OSS guy because he just gave away a few million dollars. Geez, Linux users and the whole "perspective" problem.
Well, it just makes their life harder. Maybe if companies started requiring a certain distribution, we'd put an end to all the incompatibility between distributions. I mean, I don't seriously think they're going to allow you to use any "Bob's custom distro" here. This is a company who certifies the hardware you're running the thing on, and for something as quirky as the various Linux distributions, I'm sure they'll narrow it down to one or two that they'll support.
A) You're six month figure is ridiculous. Maybe in time, but nowhere near six months. B) What does a render farm have anything to do with it? And you MIGHT choose BSD, since it supposedly does better under that kind of load. Either way, the 3D modeler part won't actually ever run on the render farm, just the backend renderer. C) If Intergraph supports the WildCat on Linux within 6 months, I'll print out an entire/. story and eat it. GeForce support is nice, but without support from the big guys (Evans and Sutherland, 3Dlabs, etc) Linux GL will have a hard time. D) I'll argue the better platform part. For me, NT tends to be extremely stable, and I for one prefer all the great tools available on NT over their less than spectacular Linux counterparts.
Its a port of a program from a major company to Linux. Why WOULDN'T LinuxDot report on this? (Not that there's anything wrong with being Linux oriented) Of course, if this results in a port of SoftImage, it will be a very funny thing. SoftImage, Blender, and Moonlight renderer are all programs from UI hell. After taking a look at the Maya demo, you'll find that SoftImage's UI is less functional. (Not to mention butt ugly)
Now if MS said that,/. would be up in arms! The truth is, that comment (write your own interface) is just an excuse for a crappy product. That's fine if the guys are OSS guys who do this for free, but if I pay for a product, I want the interface done correctly dammit! I mean you can replace the GUI in Windows even though its not OSS (its called modularity) but hell, its MS's product, I paid for it, and I want THEM to do it.
What exactly does the "Welcome Crackers" setting do? I mean what does it turn of?... Oh, wait. Never mind, I didn't say that. My box isn't on that setting. Really, its not... You can't get an IP from a/. posting, can you?
Oh come on. You know if they reviewed Debian/. would be up in arms because they forgot to upgrade some "critical package that just recieved bug fixes this morning, and it REALLY works better and would have changed the outcome entierly!"
Now you can go back. Win2K doesn't require reboots anymore. Seriously, though, BeOS (and other microkernels like the HURD) is the king of non-rebooting. You can replace drivers on the fly. Hell, I reinstalled the GUI (Tracker and Deskbar) while I was still browsing/.! Just pops up a message telling you that you've been upgraded.
People don't have time to check with the maker of every program in the distro. If there's a problem in Solaris, I know I can just easily to go Sun and ask about it. Or I know that if a security hole is found, it will be on Sun's website soon as its discovered. People expect the same of a Linux vendor. (Hell, even MS takes responsibility once they want to admit a security flaw.) They don't expect to root through all the different vendors, and they really shouldn't be expected too. After all, the vendor IS selling a seperate OS as far as most people are concerned. It certainly is a seperate product, and the vendor is responsible for the product in its entirity. Some vendors do that job better than other. To put it into perspective, what would you say if your computer's RAM went south and Dell told you to go ask Micron because they're the ones who made the RAM? I can hear to obsceneties now.
A) You're point is utterly irrelevant. I'm saying that KDE and GNOME should use fully compatible object libraries so I don't have to load both at the same time. That makes the system LESS bloated, not more so. I don't care if its in X, or there is a seperate library entierly, they should be together. Also, objects are very usefull for desktop applications. The problem with the current X model is that it allows too much diversity. For those people who don't only run one application, all this diversity leads to a lot of bloat and redundency. B) Windows seems to have gotten the handle of objects pretty well. My point is that I want to be able to use GIMP object from within KDE apps. Its a bad idea to build up two libraries of objects, because it not only is duplicated effort, but almost always a person will need an object not availabe on their native platform, and thus have to run both GNOME AND KDE. C)The "old fashioned way" is what made Linux so bloated to begin with. Here is a clean kernel with top notch components, competent libraries, and some great desktop environemnts, but simply because of the traditional way of doing things and all the numerous libraries they result in, you've got something that approaches the size of Win2K. And its truly a waste. And linking the old fashioned way leads to bugs as well. Take DirectX. It has been nearly rewritten and has 8 versions, each of which introduce new features, and still it manages to stay compatible with all the different applications that use it. And not only that, it does so without resorting to kludges like compatibility libraries and wrappers. Until I see something like that on Linux, the object approach seems to have an edge. As for BeOS, it doesn't HAVE an object model. I'm not advocating it at all. If anything, I'm advocating COM. It might be messy to code for, but its fast, fairly easy to use once you understand it, and really redefines ways of making libraries.
Its stupid to use BeOS in its current state as a webserver. It can be done (especially since BeTips is running the experimental BONE networking) but why? BeTips simply runs the site on BeOS, because A) That's his personal machine and he uses BeOS on it, and B) He wants to show support for the platform. The GScube is not designed for web serving. Its got small caches and unspectacular integer performance. Additionally, you're essentially wasting the 512MB or graphics RAM and the 16 graphics processors. Its like buying an SGI Infinate Reality to do webserving. Sure its doable (maybe it won't be, the GScube might use a custom OS) but you'd be wasting your money on the additional graphics hardware, and you'd be getting less performance than you would have from a much cheaper machine that was designed to handle webserving.
That's always bugged me. Why in hell are all daemons running automatically? If I want to set up a Sendmail server, I've got a bigger job ahead than installing the RPM and setting it to run. However, those who aren't running sendmail, have to go and disable it. I'm guessing less than 50% of Linux machines run Sendmail, so why inconvenience the majority, while making the job insignificantly easier for the minority?
I'd just like to nitpick on their definition of who's an early adopter. Recently, Suse and Mandrake have been much more agressive about introducing new features than has RedHat. Currently, Suse 6.4 comes with both ResierFS and XFree 4.0. Mandrake 7.1 does as well. However, RedHat has neither.
First things first. They probably won't run Linux on this thing. Even if they do, it will be either A) Heavily tweeked or B) Simply serve as a host for some special access libraries. This is a very parallel machine with a lot of quirks (graphics RAM divided into 16 32MB chunks for example) (read the article about the difficulty of programming the Playstation 2 earlier on/.) and to get full performance out of it, its going to take specialized software. Not only are most off the shelf apps just not designed for 16 way operation, but none are designed with the peculiarities of the PS2 hardware in mind. More likely than not, Sony will probably get some companies to port special versions of their software and more likely than not will use a custom OS for the machine. (Though the choice of OS really doesn't matter.) As for performance, specially designed software will scream on this machine. 3D is a very easy task to split up between multiple processors and thus, the 16 way architecture will result in at least 10X the performance of a PS2. To put that into perspective, a PS2 is a good deal more powerful than a GeForce2 GTS, probably on par (or exceeding) Intergraph's Wildcat 4210 (the fastest PC based OpenGL accelerator.) Now at 10x that performance, you've got a machine easily capable of trashing most SGIs. Its not a Reality Engine, but if Sony can pull this off in a sane price it should be quite a machine. Secondly, has anybody notice the amount of embedded RAM on this thing? 32MB per chip! Not only is that one hell of a jump from 4MB on the PS2, but I'm wondering how they got a manufacturing process to handle all those transistors without having 1% yields.
Its called SMP. Maybe you've heard of it? 3D rendering is hideosly parallizable. You've got a few hundred million vertecies laid out in memory, and you can process them all in any order. Its a cinch to divy that up between the processors. Then, after geometry, you can do a lot of the lighting in parallel. Or if you're doing raytracing, you could assign a ray to each processor. As for rendering, you can simply do the SLI type thing 3DFx does. I mean putting many rendering and geometry engines together is nothing new. Just take a look at some of SGI's RealityEngine machines.
Mainly because the things designed to do floating point. Its kind of like how those 64 CPU Hitachi vector units end up in the top few flots along side 512CPU SGI units.
No, I think the only good integration between GNOME and KDE will come when both can share the same config files, are binary compatible, when I can use GIMP and KDevelop without loading two sets of very huge libraries, and when KDE objects can be embedded into a GNOME container, which is actually another object contained in KDE container and being shared via CORBA from a GNOME server running in my closet. Lastly, it has to do all this without increasing the bloat of either environment from "obscene" to "Windows 2000."
But what kind of desktops? The 4% figure quoted are those who'd use UNIX anyway. People on workstations looking to switch from NT, or people switching from another brand of UNIX because of their cost. However, in the desktop market I'm talking about (mainstream business desktops running NT or 98 and consumer desktops) the figure is next to nil. As for Linux vs. NT, they're no chance unless Linux is totally redone.
A) NT is as stable as Linux on a consumer box. Given that games crash more often than the OS, gamers won't notice increased stability. Plus, consumer boxes are usually powered down every night, so NTs "slow death" not as noticible. B) NT (W2K that is) still has a vastly superior user experience. Easy to access help and a pleasing format (not README text file strewn all over the system.) It is easy to configure (Plug and Play works better in Win2K than it does even in 98, and generally detects devices and installs drivers correctly the first time. A sound card install in Win2K requires but inserting a driver disk. The same thing in Linux requires knowing what an IRQ is. The interface is consistant, and games don't need hacking to get working.
C) Win2K is faster and has about the same memory requirements. First, Win2K is faster because DirectX allows an application to take over the system. Also, OpenGL on NT is very mature and well developed. Plus theres more vid card support in NT. The memory requirements for both are about the same (atrociously high for both) because a productive Linux machine often requires installation of multiple redundant libraries (Netscape for example requires compat-libs in Mandrake) and the installation of both KDE and GNOME. (Since each environent has some killer apps.)
D) The developer experience on Windows is a lot better. First, Win2K has Visual Studio, a tool that puts documentation a click away, and has source browsing, auto-completion, and an array of really good debugging tools. While Linux may have makefiles and gcc, Visual Studio can use makefiles too. DirectX is a godsend, because it offers game developers DOS-like access to hardware (game developers really don't want ANY OS on the system) while still enabling them to access hardware acceleration transparently. Plus, Direct3D has more core features, is just as fast as OpenGL, and lets you take advantage of features that only a few cards support (like vertex skinning on the Radeon and GeForce2) without having to write code to utilize the propriotary extensions for each card. (IE. In D3D you write two code cases, accelerated vertex skinning is available, or not. In OpenGL you write many cases, acceleration not available, the NVIDIA extension, the ATI extension, etc.)
The problem with Linux is that it doesn't really offer anything compelling over Win2K (aside from being free) to compete with it in the consumer arena. As for Intel supporting Linux, it is not due to any technical merit of Linux. Montery is nowhere near being ready, WinNT 64 is a joke, so the only thing Intel can fall back on is to support Linux as it will probably be the only one that will run on Itanium at its debut. Intel has supported BeOS in the past (to showcase its processors) but that doesn't mean they will support it for a long time. Soon as Montery comes out, Linux support will wane unless Intel has some business reason to keep Linux around.
In Windows you can click help, and get a nicely formatted HTML help file, with links and searchability. Some even have pictures to help you find a menu or what not. MS has innovated even more (yea, they DO do that) by having a centralized help file that manufactures can add their help files to. Now, one stop help whatever you're using. Sure it takes away the freedom to app writes to include their own style of help, but their freedom doesn't matter, now does it.
Lost half of the post, here it is in its entirety.
Not true. All technology is born in labs, not in kitchens. If you conquer the labs (and Linux is moving pretty fast in this direction), all new features (including new desktop ideas) will be yours eventually. All concepts eventually die, and even OLE will. And why won't the Linux desktop be next? >>>>>> Huh? As far as I can see, all new technology these days seems to be coming from consumer technology. The whole reason the PC market exists is because of the PC's utility as a business machine. Increasingly, the PC market is also being driven by consumers. The technology itself is being driven by games, mainly. As for labs, I don't know of any major labratories which use Linux. However, I do know that there are more features to be found in consumer and business space than anywhere else. Looking for the most efficient interface? Dominate the business market. These people spend hours each day in front of a computer, and they're the ones who need the most efficient interface. Want the easiest interface? Dominate the newbie product market. Want the fastest 3D, dominate the games market, where the need to wring massive performance out of small budgets drives the market faster than even SGI's machines ever did. Want the best security, dominate the business market where people need transparent access to documents without other people getting access too. Almost none of the cool tech that has come out has come because of the needs of the labratory. These days, consumers drive the entire computing inudstry.
As for Linux users attitude - that's not Linux fault. Choose better friends for yourself:) Same goes for your 10-years-old canned Linux myths ("no docs", "hard to install", "too many choices", etc., etc.) Believe me, every word you say here was said and proven false years ago. Please don't start this again. >>>>>>> No docs: Microsoft has beautifully done HTML help files that are easily searchable, include pictures to explain things, and cover each feature of an application. (The DirectX docs in particular are probably one of the finset examples of detailed API documentation I've seen) Linux has: README's. Hard to Install: This problem is related to the to many choices problem. Unless you want Mandrake installing 1.4 gigs of stuff on your harddrive, you've got to custom install packages. Then, you've got to wade through multiple redundant packages. "What the hell is the difference between gcc and egcs?" "why the hell to I need the C shell, I never USE the C shell." "Why can't I uninstall groff without man breaking?" The problem with the packages is that there is too much cruft that Linux apps depend on. Then you've got to partition your drive. Last time I installed, Win98, you didn't need to know what a partition was. Then, unless you want a ton of useless services on your machine, you've got to read up on each and disable the ones you don't need. "Samba? What they hell is this thing running Samba automatically for?" Finally, you've got to run sndconfig, which as often as not asks you for IRQs and DMAs. If you've gotten through that, you've got to go through hell everytime you want to install an app. "What do you mean this thing uses glibc2.1.2 what the hell's a glibc?" Want to upgrade your desktop? Quit out of X, download a dozen RPMS, and rpm -Ui --force --nodeps them. Why --nodeps? Because 50% of time KDE manages to depend on a package supplied within the package you're trying to install. Sure a lot of these theoretically don't happen, but A) They're never all in the same distro. B) Even if they're in the distro, it results you losing functionality. You can run Windows at 90% with no tweeking, Linux maybe 60%. Doesn't it simpy make MORE sense to let people install what they need rather than installing everything and making people wade through the mess getting rid of cruft? C) It doesn't work 50% of the time. Sure KDE is supposed to install right of the bat, but ask anybody who uses the NVIDIA drivers and can't get Qt-GL to install, and they'll tell you it isn't all its cracked up to be. To many choices: Lets see, two major incompatible versions of KDE, GNOME, three versions (incompatible) of libc, two versions of libstdc++, motif, gawk, mawk, pgcc, gcc, C-shell, Zshell, bash, etc, etc. Even worse, all the apps require different versions of each, so I have them all loaded at the same freaking time. I don't know about you, but I'm feeling pretty overwhelmed. (And pissed at all the resource sucking redundency.) Plus, years ago, Linux didn't even have KDE or the super simple (relativly) installers. Why was it proven then?
Not true. All technology is born in labs, not in kitchens. If you conquer the labs (and Linux is moving pretty fast in this direction), all new features (including new desktop ideas) will be yours eventually. All concepts eventually die, and even OLE will. And why won't the Linux desktop be next? >>>>>> Huh? As far as I can see, all new technology these days seems to be coming from consumer technology. The whole reason the PC market exists is because of the PC's utility as a business machine. Increasingly, the PC market is also being driven by consumers. The technology itself is being driven by games, mainly. As for labs, I don't know of any major labratories which use Linux. However, I do know that there are more features to be found in consumer and business space than anywhere else. Looking for the most efficient interface? Dominate the business market. These people spend hours each day in front of a computer, and they're the ones who need the most efficient interface. Want the easiest interface? Dominate the newbie product market. Want the fastest 3D, dominate the games market, where the need to wring massive performance out of small budgets drives the market faster than even SGI's machines ever did. Want the best security, dominate the business market where people need transparent access to documents without other people getting access too. Almost none of the cool tech that has come out has come because of the needs of the labratory. These days, consumers drive the entire computing inudstry.
As for Linux users attitude - that's not Linux fault. Choose better friends for yourself:) Same goes for your 10-years-old canned Linux myths ("no docs", "hard to install", "too many choices", etc., etc.) Believe me, every word you say here was said and proven false years ago. Please don't start this again. >>>>>>> No docs: Microsoft has beautifully done HTML help files that are easily searchable, include pictures to explain things, and cover each feature of an application. (The DirectX docs in particular are probably one of the finset examples of detailed API documentation I've seen) Linux has: README's. Hard to Install: This problem is related to the to many choices problem. Unless you want Mandrake installing 1.4 gigs of stuff on your harddrive, you've got to custom install packages. Then, you've got to wade through multiple redundant packages. "What the hell is the difference between gcc and egcs?" "why the hell to I need the C shell, I never USE the C shell." "Why can't I uninstall groff without man breaking?" The problem with the packages is that there is too much cruft that Linux apps depend on. Then you've got to partition your drive. Last time I installed, Win98, you didn't need to know what a partition was. Then, unless you want a ton of useless services on your machine, you've got to read up on each and disable the ones you don't need. "Samba? What they hell is this thing running Samba automatically for?" Finally, you've got to run sndconfig, which as often as not asks you for IRQs and DMAs. If you've gotten through that, you've got to go through hell everytime you want to install an app. "What do you mean this thing uses glibc2.1.2 what the hell's a glibc?" Want to upgrade your desktop? Quit out of X, download a dozen RPMS, and rpm -Ui --force --nodeps them. Why --nodeps? Because 50% of time KDE manages to depend on a package supplied within the package you're trying to install. Sure a lot of these theoretically don't happen, but A) They're never all in the same distro. B) Even if they're in the distro, it results you losing functionality. You can run Windows at 90% with no tweeking, Linux maybe 60%. Doesn't it simpy make MORE sense to let people install what they need rather than installing everything and making people wade through the mess getting rid of cruft?
However, KDE is meant for a broad acceptance. The whole reason why KDE (and GNOME) have done so well is the fact that they resemble Windows. Everyone uses Windows, and the best interface is the Win98 interface simply because 300-something million people already know it.
Slashdot Mirror
The problem is that most systems ship with DMA off. It causes problems in some configurations, and is usually a liability for the vendor.
You're kidding right? Not only is it commercial "payware" but it is "software that'll cost you an arm and a leg and you have to buy a new computer to use." Hell, this thing costs more than a lot of CARS! If you can find a GNU animation tool that comes anywhere within the same solar system as this thing, then you've found a really dedicated OSS guy because he just gave away a few million dollars. Geez, Linux users and the whole "perspective" problem.
Well, it just makes their life harder. Maybe if companies started requiring a certain distribution, we'd put an end to all the incompatibility between distributions. I mean, I don't seriously think they're going to allow you to use any "Bob's custom distro" here. This is a company who certifies the hardware you're running the thing on, and for something as quirky as the various Linux distributions, I'm sure they'll narrow it down to one or two that they'll support.
I think he meant that Softimage contains too much GL code. Although it still doesn't make any sense.
A) You're six month figure is ridiculous. Maybe in time, but nowhere near six months. /. story and eat it. GeForce support is nice, but without support from the big guys (Evans and Sutherland, 3Dlabs, etc) Linux GL will have a hard time.
B) What does a render farm have anything to do with it? And you MIGHT choose BSD, since it supposedly does better under that kind of load. Either way, the 3D modeler part won't actually ever run on the render farm, just the backend renderer.
C) If Intergraph supports the WildCat on Linux within 6 months, I'll print out an entire
D) I'll argue the better platform part. For me, NT tends to be extremely stable, and I for one prefer all the great tools available on NT over their less than spectacular Linux counterparts.
Its a port of a program from a major company to Linux. Why WOULDN'T LinuxDot report on this? (Not that there's anything wrong with being Linux oriented) Of course, if this results in a port of SoftImage, it will be a very funny thing. SoftImage, Blender, and Moonlight renderer are all programs from UI hell. After taking a look at the Maya demo, you'll find that SoftImage's UI is less functional. (Not to mention butt ugly)
Now if MS said that, /. would be up in arms! The truth is, that comment (write your own interface) is just an excuse for a crappy product. That's fine if the guys are OSS guys who do this for free, but if I pay for a product, I want the interface done correctly dammit! I mean you can replace the GUI in Windows even though its not OSS (its called modularity) but hell, its MS's product, I paid for it, and I want THEM to do it.
What exactly does the "Welcome Crackers" setting do? I mean what does it turn of? ... Oh, wait. Never mind, I didn't say that. My box isn't on that setting. Really, its not... You can't get an IP from a /. posting, can you?
Oh come on. You know if they reviewed Debian /. would be up in arms because they forgot to upgrade some "critical package that just recieved bug fixes this morning, and it REALLY works better and would have changed the outcome entierly!"
Now you can go back. Win2K doesn't require reboots anymore. Seriously, though, BeOS (and other microkernels like the HURD) is the king of non-rebooting. You can replace drivers on the fly. Hell, I reinstalled the GUI (Tracker and Deskbar) while I was still browsing /.! Just pops up a message telling you that you've been upgraded.
People don't have time to check with the maker of every program in the distro. If there's a problem in Solaris, I know I can just easily to go Sun and ask about it. Or I know that if a security hole is found, it will be on Sun's website soon as its discovered. People expect the same of a Linux vendor. (Hell, even MS takes responsibility once they want to admit a security flaw.) They don't expect to root through all the different vendors, and they really shouldn't be expected too. After all, the vendor IS selling a seperate OS as far as most people are concerned. It certainly is a seperate product, and the vendor is responsible for the product in its entirity. Some vendors do that job better than other.
To put it into perspective, what would you say if your computer's RAM went south and Dell told you to go ask Micron because they're the ones who made the RAM? I can hear to obsceneties now.
A) You're point is utterly irrelevant. I'm saying that KDE and GNOME should use fully compatible object libraries so I don't have to load both at the same time. That makes the system LESS bloated, not more so. I don't care if its in X, or there is a seperate library entierly, they should be together. Also, objects are very usefull for desktop applications. The problem with the current X model is that it allows too much diversity. For those people who don't only run one application, all this diversity leads to a lot of bloat and redundency.
B) Windows seems to have gotten the handle of objects pretty well. My point is that I want to be able to use GIMP object from within KDE apps. Its a bad idea to build up two libraries of objects, because it not only is duplicated effort, but almost always a person will need an object not availabe on their native platform, and thus have to run both GNOME AND KDE.
C)The "old fashioned way" is what made Linux so bloated to begin with. Here is a clean kernel with top notch components, competent libraries, and some great desktop environemnts, but simply because of the traditional way of doing things and all the numerous libraries they result in, you've got something that approaches the size of Win2K. And its truly a waste. And linking the old fashioned way leads to bugs as well. Take DirectX. It has been nearly rewritten and has 8 versions, each of which introduce new features, and still it manages to stay compatible with all the different applications that use it. And not only that, it does so without resorting to kludges like compatibility libraries and wrappers. Until I see something like that on Linux, the object approach seems to have an edge.
As for BeOS, it doesn't HAVE an object model. I'm not advocating it at all. If anything, I'm advocating COM. It might be messy to code for, but its fast, fairly easy to use once you understand it, and really redefines ways of making libraries.
Its stupid to use BeOS in its current state as a webserver. It can be done (especially since BeTips is running the experimental BONE networking) but why? BeTips simply runs the site on BeOS, because
A) That's his personal machine and he uses BeOS on it, and
B) He wants to show support for the platform.
The GScube is not designed for web serving. Its got small caches and unspectacular integer performance. Additionally, you're essentially wasting the 512MB or graphics RAM and the 16 graphics processors. Its like buying an SGI Infinate Reality to do webserving. Sure its doable (maybe it won't be, the GScube might use a custom OS) but you'd be wasting your money on the additional graphics hardware, and you'd be getting less performance than you would have from a much cheaper machine that was designed to handle webserving.
That's always bugged me. Why in hell are all daemons running automatically? If I want to set up a Sendmail server, I've got a bigger job ahead than installing the RPM and setting it to run. However, those who aren't running sendmail, have to go and disable it. I'm guessing less than 50% of Linux machines run Sendmail, so why inconvenience the majority, while making the job insignificantly easier for the minority?
I'd just like to nitpick on their definition of who's an early adopter. Recently, Suse and Mandrake have been much more agressive about introducing new features than has RedHat. Currently, Suse 6.4 comes with both ResierFS and XFree 4.0. Mandrake 7.1 does as well. However, RedHat has neither.
First things first. They probably won't run Linux on this thing. Even if they do, it will be either /.) and to get full performance out of it, its going to take specialized software. Not only are most off the shelf apps just not designed for 16 way operation, but none are designed with the peculiarities of the PS2 hardware in mind. More likely than not, Sony will probably get some companies to port special versions of their software and more likely than not will use a custom OS for the machine. (Though the choice of OS really doesn't matter.) As for performance, specially designed software will scream on this machine. 3D is a very easy task to split up between multiple processors and thus, the 16 way architecture will result in at least 10X the performance of a PS2. To put that into perspective, a PS2 is a good deal more powerful than a GeForce2 GTS, probably on par (or exceeding) Intergraph's Wildcat 4210 (the fastest PC based OpenGL accelerator.) Now at 10x that performance, you've got a machine easily capable of trashing most SGIs. Its not a Reality Engine, but if Sony can pull this off in a sane price it should be quite a machine. Secondly, has anybody notice the amount of embedded RAM on this thing? 32MB per chip! Not only is that one hell of a jump from 4MB on the PS2, but I'm wondering how they got a manufacturing process to handle all those transistors without having 1% yields.
A) Heavily tweeked or
B) Simply serve as a host for some special access libraries. This is a very parallel machine with a lot of quirks (graphics RAM divided into 16 32MB chunks for example) (read the article about the difficulty of programming the Playstation 2 earlier on
Its called SMP. Maybe you've heard of it? 3D rendering is hideosly parallizable. You've got a few hundred million vertecies laid out in memory, and you can process them all in any order. Its a cinch to divy that up between the processors. Then, after geometry, you can do a lot of the lighting in parallel. Or if you're doing raytracing, you could assign a ray to each processor. As for rendering, you can simply do the SLI type thing 3DFx does. I mean putting many rendering and geometry engines together is nothing new. Just take a look at some of SGI's RealityEngine machines.
Mainly because the things designed to do floating point. Its kind of like how those 64 CPU Hitachi vector units end up in the top few flots along side 512CPU SGI units.
No, it's something designed to model and render CG movies with. Who in their right mind would use it for a webserver?
No, I think the only good integration between GNOME and KDE will come when both can share the same config files, are binary compatible, when I can use GIMP and KDevelop without loading two sets of very huge libraries, and when KDE objects can be embedded into a GNOME container, which is actually another object contained in KDE container and being shared via CORBA from a GNOME server running in my closet. Lastly, it has to do all this without increasing the bloat of either environment from "obscene" to "Windows 2000."
But what kind of desktops? The 4% figure quoted are those who'd use UNIX anyway. People on workstations looking to switch from NT, or people switching from another brand of UNIX because of their cost. However, in the desktop market I'm talking about (mainstream business desktops running NT or 98 and consumer desktops) the figure is next to nil. As for Linux vs. NT, they're no chance unless Linux is totally redone.
A) NT is as stable as Linux on a consumer box. Given that games crash more often than the OS, gamers won't notice increased stability. Plus, consumer boxes are usually powered down every night, so NTs "slow death" not as noticible.
B) NT (W2K that is) still has a vastly superior user experience. Easy to access help and a pleasing format (not README text file strewn all over the system.) It is easy to configure (Plug and Play works better in Win2K than it does even in 98, and generally detects devices and installs drivers correctly the first time. A sound card install in Win2K requires but inserting a driver disk. The same thing in Linux requires knowing what an IRQ is. The interface is consistant, and games don't need hacking to get working.
C) Win2K is faster and has about the same memory requirements. First, Win2K is faster because DirectX allows an application to take over the system. Also, OpenGL on NT is very mature and well developed. Plus theres more vid card support in NT. The memory requirements for both are about the same (atrociously high for both) because a productive Linux machine often requires installation of multiple redundant libraries (Netscape for example requires compat-libs in Mandrake) and the installation of both KDE and GNOME. (Since each environent has some killer apps.)
D) The developer experience on Windows is a lot better. First, Win2K has Visual Studio, a tool that puts documentation a click away, and has source browsing, auto-completion, and an array of really good debugging tools. While Linux may have makefiles and gcc, Visual Studio can use makefiles too. DirectX is a godsend, because it offers game developers DOS-like access to hardware (game developers really don't want ANY OS on the system) while still enabling them to access hardware acceleration transparently. Plus, Direct3D has more core features, is just as fast as OpenGL, and lets you take advantage of features that only a few cards support (like vertex skinning on the Radeon and GeForce2) without having to write code to utilize the propriotary extensions for each card. (IE. In D3D you write two code cases, accelerated vertex skinning is available, or not. In OpenGL you write many cases, acceleration not available, the NVIDIA extension, the ATI extension, etc.)
The problem with Linux is that it doesn't really offer anything compelling over Win2K (aside from being free) to compete with it in the consumer arena. As for Intel supporting Linux, it is not due to any technical merit of Linux. Montery is nowhere near being ready, WinNT 64 is a joke, so the only thing Intel can fall back on is to support Linux as it will probably be the only one that will run on Itanium at its debut. Intel has supported BeOS in the past (to showcase its processors) but that doesn't mean they will support it for a long time. Soon as Montery comes out, Linux support will wane unless Intel has some business reason to keep Linux around.
In Windows you can click help, and get a nicely formatted HTML help file, with links and searchability. Some even have pictures to help you find a menu or what not. MS has innovated even more (yea, they DO do that) by having a centralized help file that manufactures can add their help files to. Now, one stop help whatever you're using. Sure it takes away the freedom to app writes to include their own style of help, but their freedom doesn't matter, now does it.
Lost half of the post, here it is in its entirety.
:)
Not true. All technology is born in labs, not in kitchens. If you conquer the labs (and Linux is moving pretty fast in this direction), all new features (including new desktop ideas) will be yours eventually. All concepts eventually die, and even OLE will. And why won't the Linux desktop be next?
>>>>>>
Huh? As far as I can see, all new technology these days seems to be coming from consumer technology. The whole reason the PC market exists is because of the PC's utility as a business machine. Increasingly, the PC market is also being driven by consumers. The technology itself is being driven by games, mainly. As for labs, I don't know of any major labratories which use Linux. However, I do know that there are more features to be found in consumer and business space than anywhere else. Looking for the most efficient interface? Dominate the business market. These people spend hours each day in front of a computer, and they're the ones who need the most efficient interface. Want the easiest interface? Dominate the newbie product market. Want the fastest 3D, dominate the games market, where the need to wring massive performance out of small budgets drives the market faster than even SGI's machines ever did. Want the best security, dominate the business market where people need transparent access to documents without other people getting access too. Almost none of the cool tech that has come out has come because of the needs of the labratory. These days, consumers drive the entire computing inudstry.
As for Linux users attitude - that's not Linux fault. Choose better friends for yourself
Same goes for your 10-years-old canned Linux myths ("no docs", "hard to install", "too many choices", etc., etc.) Believe me, every word you say here was said and proven false years ago. Please don't start this again.
>>>>>>>
No docs: Microsoft has beautifully done HTML help files that are easily searchable, include pictures to explain things, and cover each feature of an application. (The DirectX docs in particular are probably one of the finset examples of detailed API documentation I've seen) Linux has: README's.
Hard to Install: This problem is related to the to many choices problem. Unless you want Mandrake installing 1.4 gigs of stuff on your harddrive, you've got to custom install packages. Then, you've got to wade through multiple redundant packages. "What the hell is the difference between gcc and egcs?" "why the hell to I need the C shell, I never USE the C shell." "Why can't I uninstall groff without man breaking?" The problem with the packages is that there is too much cruft that Linux apps depend on. Then you've got to partition your drive. Last time I installed, Win98, you didn't need to know what a partition was. Then, unless you want a ton of useless services on your machine, you've got to read up on each and disable the ones you don't need. "Samba? What they hell is this thing running Samba automatically for?" Finally, you've got to run sndconfig, which as often as not asks you for IRQs and DMAs. If you've gotten through that, you've got to go through hell everytime you want to install an app. "What do you mean this thing uses glibc2.1.2 what the hell's a glibc?" Want to upgrade your desktop? Quit out of X, download a dozen RPMS, and rpm -Ui --force --nodeps them. Why --nodeps? Because 50% of time KDE manages to depend on a package supplied within the package you're trying to install. Sure a lot of these theoretically don't happen, but
A) They're never all in the same distro.
B) Even if they're in the distro, it results you losing functionality. You can run Windows at 90% with no tweeking, Linux maybe 60%. Doesn't it simpy make MORE sense to let people install what they need rather than installing everything and making people wade through the mess getting rid of cruft?
C) It doesn't work 50% of the time. Sure KDE is supposed to install right of the bat, but ask anybody who uses the NVIDIA drivers and can't get Qt-GL to install, and they'll tell you it isn't all its cracked up to be.
To many choices: Lets see, two major incompatible versions of KDE, GNOME, three versions (incompatible) of libc, two versions of libstdc++, motif, gawk, mawk, pgcc, gcc, C-shell, Zshell, bash, etc, etc. Even worse, all the apps require different versions of each, so I have them all loaded at the same freaking time. I don't know about you, but I'm feeling pretty overwhelmed. (And pissed at all the resource sucking redundency.) Plus, years ago, Linux didn't even have KDE or the super simple (relativly) installers. Why was it proven then?
Not true. All technology is born in labs, not in kitchens. If you conquer the labs (and Linux is moving pretty fast in this direction), all new features (including new desktop ideas) will be yours eventually. All concepts eventually die, and even OLE will. And why won't the Linux desktop be next?
:)
>>>>>>
Huh? As far as I can see, all new technology these days seems to be coming from consumer technology. The whole reason the PC market exists is because of the PC's utility as a business machine. Increasingly, the PC market is also being driven by consumers. The technology itself is being driven by games, mainly. As for labs, I don't know of any major labratories which use Linux. However, I do know that there are more features to be found in consumer and business space than anywhere else. Looking for the most efficient interface? Dominate the business market. These people spend hours each day in front of a computer, and they're the ones who need the most efficient interface. Want the easiest interface? Dominate the newbie product market. Want the fastest 3D, dominate the games market, where the need to wring massive performance out of small budgets drives the market faster than even SGI's machines ever did. Want the best security, dominate the business market where people need transparent access to documents without other people getting access too. Almost none of the cool tech that has come out has come because of the needs of the labratory. These days, consumers drive the entire computing inudstry.
As for Linux users attitude - that's not Linux fault. Choose better friends for yourself
Same goes for your 10-years-old canned Linux myths ("no docs", "hard to install", "too many choices", etc., etc.) Believe me, every word you say here was said and proven false years ago. Please don't start this again.
>>>>>>>
No docs: Microsoft has beautifully done HTML help files that are easily searchable, include pictures to explain things, and cover each feature of an application. (The DirectX docs in particular are probably one of the finset examples of detailed API documentation I've seen) Linux has: README's.
Hard to Install: This problem is related to the to many choices problem. Unless you want Mandrake installing 1.4 gigs of stuff on your harddrive, you've got to custom install packages. Then, you've got to wade through multiple redundant packages. "What the hell is the difference between gcc and egcs?" "why the hell to I need the C shell, I never USE the C shell." "Why can't I uninstall groff without man breaking?" The problem with the packages is that there is too much cruft that Linux apps depend on. Then you've got to partition your drive. Last time I installed, Win98, you didn't need to know what a partition was. Then, unless you want a ton of useless services on your machine, you've got to read up on each and disable the ones you don't need. "Samba? What they hell is this thing running Samba automatically for?" Finally, you've got to run sndconfig, which as often as not asks you for IRQs and DMAs. If you've gotten through that, you've got to go through hell everytime you want to install an app. "What do you mean this thing uses glibc2.1.2 what the hell's a glibc?" Want to upgrade your desktop? Quit out of X, download a dozen RPMS, and rpm -Ui --force --nodeps them. Why --nodeps? Because 50% of time KDE manages to depend on a package supplied within the package you're trying to install. Sure a lot of these theoretically don't happen, but
A) They're never all in the same distro.
B) Even if they're in the distro, it results you losing functionality. You can run Windows at 90% with no tweeking, Linux maybe 60%. Doesn't it simpy make MORE sense to let people install what they need rather than installing everything and making people wade through the mess getting rid of cruft?
However, KDE is meant for a broad acceptance. The whole reason why KDE (and GNOME) have done so well is the fact that they resemble Windows. Everyone uses Windows, and the best interface is the Win98 interface simply because 300-something million people already know it.