I'd like to see a CUDA implementation. With calculation speedups from 10x to hundreds of times over host-based processors for some algorithms on just one GeForce 8800GTS/GTX card (with 128 stream processors working simultaneously.. 8x the cores, even if lower clocked) this could get interesting very quickly.
Do you think that they subscribe in order to order to take part in a world located in the US? Think again. And this time engage a brain cell.
Yes. Yes, they do actually. I'm an SL player/"resident", and I realized when I signed up where Linden Labs is located. Second Life may well be a virtual world, but the company and servers that allow the virtual world to exist are still subject to the laws where they are physically located. Were they not headquartered in one of the more open minded areas of the US, who knows what might have happened to them already. It's an interesting ideal to make a virtual world completely separated from RL, but even ignoring moral issues there's a very real economic link between the two. That alone ensures that it will never happen.
I think the problem is that companies have a tendency to view people strictly as consumers, which I suppose is what they have to do if they want to survive. These "consumers" have portions of their lives that don't involve spending money though, and that's just as true in second life. Unless the company's product enhances the social aspect of SL somehow, or provides something actually useful in the gameplay, I don't see people wanting to view what is effectively a dull marketing campaign unless they're already interested in the product. How many people go out looking for pages of banner ads?
As an active Second Life 'resident', I'd actually be happy to see marketers give up on it. Perhaps they could get rid of the ad farms too, although I wouldn't hold my breath on either. SL in a lot of ways has strong parallels to the web as I remember it in the Mosaic/early Netscape years... The client's not really finished, network bandwidth hasn't quite caught up with its demands yet, effective advertising mechanisms aren't really well defined, and the emphasis on user-supplied content initially confuses many who are just looking for something to play with. I have a feeling that we're going to be using SL or something like it in 10 years or so though, in the same way that we're using web browsers now. Marketers are no doubt thinking along these lines.
On the other hand, the Linux client doesn't have voice support yet. It won't have it until the voice sw vendor gives them an object file or shared library build for the platform, which will happen who knows when. For those of us using the Linux viewer, there's still a good excuse.
FWIW, half the time I play SL with sound turned off - never mind voice. There's no real need for listening to the sound when you're just building objects.
Linux can certainly be used as an RTOS, in fact I've configured it for several embedded systems, but it's also relatively easy to use as a general purpose OS. If you stick to ARINC 653 (you're right on that of course, it's been a while since I was involved in spec'ing the CPCI boards) a full TCP/IP stack is basically out, amongst other niceties. You can do a cut down one, in our case with a special network card, but not everything works within the 653 spec according to the Green Hills staff with whom I've spoken. POSIX and ARINC 653 are really intended for different classes of problems.
Xen gives you a complete virtual machine, similar to but perhaps not quite as heavy as VMware, so you get a guest kernel inside the primary kernel. Memory and most other resources are partitioned deterministically, and something like a forkbomb - assuming you didn't reduce the process' maximum number of processes, modified in/etc/security/limits.conf in many distros - would only affect the virtual machine. CPU resources are not quite as deterministic, since they rely on both the host and target kernel's scheduling algorithms, but if you need something like that you're probably going to use a real-time patched kernel assuming you go with Linux as the target.
Integrity is an RTOS platform, not a general purpose OS. I've worked with their ARINC 673 product a bit, much standard UNIX functionality would break the guarantees made by an ARINC-compliant OS so it's just not present. Xen is a close enough approximation if you just want to partition the system off without using ARINC 673, but in order to get the same sort of certifications as Integrity (or VxWorks' ARINC 673 product for that matter) all the code involved with Linux - kernel, userspace etc. - would need a line-by-line code review, probable changes, and sign off.
I would say Bill has had an overall negative impact on the world Your biases are not a useful metric of ability. There is a computer in nearly every home in the US, and Windows has a lot to do with that. This alone contributes to the number of people who have been introduced to Computer Science, and arguably does more than any one person could have done. The Linux kernel has allowed for a complete GNU system to be released before the Hurd kernel was completed, giving a useful UNIX development environment to people who previously may never have had access to a full UNIX system. These people have produced software which has had a profound impact on Computer Science, by introducing many more people into the field. Despite my dislike for Microsoft products in general, I can see the profound good that their software has done to the field. What have you produced, beyond cookies? I make a mean batch of brownies myself, let's keep at least in the same general field.
Computer science (despite the misnomer) is the science of computation. Do you know what science is? On what do you base your assertion that they do not understand the methodology? Have you polled them yourself? Opinion is not a useful metric, and the fact is you really have no idea aside from your own opinions and bias.
the code he has written, and the design decisions he has made shows quite clearly that he is not someone to look to for a CS luminary Further opinion. Regardless, we're not talking about whether these people are luminaries, although they have had a known profound impact. We're talking about your assertion that they are mediocre, and that's frankly indefensible.
It's certainly not undue credit; whether you consider them the top programmers of our time or not, they are well beyond the label 'mediocre'. Each has contributed to the general good, and your biases are not a useful metric of ability. Knuth may have done more for pure computer science than the rest of the people you listed combined, yet that does not make them mediocre.
Name anything either Bill or Linus has EVER done that could in any way be construed as demonstrating that they have solid CS skills.
CS skills? You keep using those words. I don't think they mean what you think they mean.
Bill Gates is an excellent marketer. Linus Torvalds is an excellent manager. Both also have solid computer science skills. Although surely they were in the right place at the right time, to call either mediocre is either stroking your own ego or putting your head in the sand.
Whether Linus would have failed Andy's class depends on whether he tailored his code to the assignment; it's just part of going to college, and eventually applies to the real world as well. Based on Linus' management of the Linux kernel, I'd say he has a reasonable understanding of what to expect from people. Thus, I'd say he would have done quite well in Andy's class. Neither of us can say for certain however.
With x86 processors you typically have had to emulate the Ring 0 (supervisor mode) instructions, since these supervisor-level instructions could cause the virtualized system to break out of its sandbox and inadvertantly (or advertantly?) crash your machine. In a way, it's like gaining root or admin privileges in the OS. These newer virtualization instructions provide a secondary "Ring 0" that can be used to trap these instructions in hardware, rather than looking for the opcodes yourself. In theory these instructions should allow faster and possibly more stable emulation. In practice VMware runs a bit slower with these in 32 bit mode at the moment, although that will no doubt change in the future. It requires them in 64 bit mode however. Xen uses them to good effect, and is able to run unmodified operating systems only because of these instructions.
Seems fine here... Scrolling is smooth and the thumbwheel doesn't slow it down. What is your video card and which driver are you running? A crappy driver will slow down everything that scrolls text mixed with images, not just Firefox. For reference, I'm running a 32-bit Linux distro on an AMD 4200+ dual core, with 2gb RAM, a PCIE GeForce 6600 and NVidia's 9755 drivers. FF 2 runs fine (although it's eating 260mb RAM) with 26 tabs open. Come to think of it, I guess that's 10MB per tab...
Ever buy an OS from Dell with the PC? You're generally better off buying it separately. The restore discs I've seen detect whether the PC you're booting it on is a Dell (i.e. not VMware). The installer bails out if it isn't, and installs extra junk you probably don't want if it is. I'd go the Ubuntu route even if I did want to run Vista, and just buy the OS later.
There was a 6 button Genesis controller, which frankly was quite nice. I still have one around, the D-Pad just sort of floats there; it's very responsive. I particularly like the way Sega rounded the Genesis controllers to fit better into one's hands. The sharper angles on the NES and SNES controllers were rather uncomfortable to hold in comparison. The Wii's remote seems to have fallen back in that direction a bit, but it's not so thin. I haven't noticed the same cramping that came with the NES and SNES controllers, so that (along with not needing to hold both hands close together in most games) may be the reason.
It's been more expensive in San Francisco FWIW, $4.20 a gallon for the least expensive stuff at one gas station. We're already past $3.92/gallon, not that I think SUVs are ideal commuting vehicles. People who drive them for no good reason (there are good reasons, mind you, but most people don't need them) deserve the higher prices. Hopefully we'll see a shift in the types of vehicles people are driving as prices continue to rise.
That's the case only if you included the redacted items in the PDF to begin with. If you create a new PDF from your source document instead of modifying the PDF, you should be fine.
What are the odds that Dell might have their own repositories for their customers? That would give them the ability to add Dell-specific software, which they'd no doubt be very happy to do.
That's a good idea. I'd like to see it taken further, that is, a project to create new software technology and obtain patents on it. Any patented technologies would be explicitly licensed for open source usage. Commercial entities could license the developed technologies on the "inventor's" cost terms, thus funding the creation of newer technologies for open source projects and giving something back to the "inventor" as well. That would probably interest more people, create a stream of revenue, build a patent portfolio of safe technologies for non-fictional people, and involve those who are not necessarily coders.
I wouldn't be surprised if Dell's new position on Linux had a lot to do with this particular threat. Linux will now be offered to people who before didn't even know they had a choice, by one of the largest PC vendors on the planet. That's a powerful statement of support from Dell. It's good to keep this fact in perspective to be sure, it can't be argued that Microsoft will fall simply because of this. Do not however make the mistake of denying that this is happening at all. It will have an impact when people can buy an inexpensive Linux PC from Dell for less than the cost of a Windows system, and other PC manufacturers will start to feel more pressure to provide similar offerings.
As for XP, well, it's both obvious why Dell wants to sell it yet sad that MS doesn't want them doing it. Vista runs poorly if at all on existing hardware. You need significantly more RAM and a recent video card to run it at the same speed with the Aeroglass widget decorations from our experience. Many laptops need not apply. IT departments in larger organizations also do not modify their supported configurations so quickly, no matter what Microsoft wants. You need your third party software such as VPN clients working on the new platforms. Pushing customers to the platform before they are ready is not a good long term business move. I agree that it's not a death knell, but it is a shot in the arm for any of their competitors who are willing to make transitions more gradual.
In my experience, WPA works fine for me under Linux. I'm not using Ubuntu, but rather Mandriva 2007. The wpa_supplicant takes care of it nicely. The interfaces aren't where they need to be yet to be sure, but that comes with time. That's one of the nice things about Linux; as technologies become widespread, they become supported. Certainly there is always a need for new people to push software forward, hardware will continue to change and the OS must be updated to allow for this. Still, Linux doesn't need to conquer the desktop market any more than Apple does. There are plenty of kernel developers now, and there will likely continue to be in the future based on the current size of its community. The fact that it exists as a completely Free UNIX reimplementation, with the GPL license providing for those who prefer it, has garnered it a large enough userbase to sustain itself.
Back to the topic, that's not to say there's no place for Solaris. It's a fine OS, great for servers but a bit weak on the desktop based on my experience with my Ultra 40/Solaris 10. On a typical PC however, I'll take Linux any day. It does what I need it to, and tends to support the most cutting edge free *NIX tools first. It also tends to absorb useful features from other operating systems over time. Have a look at OpenVZ for example; Solaris 10-style containers already exist for Linux. Every time someone lists a set of features Linux needs, they end up being implemented either by someone who wants it, or by a company that decides they need it. The world's clearly richer for it. There would probably never even be an open source Solaris were this not the case.
Out of curiosity, how would they know who bought Doom 3 for Linux? I bought it, but it was a Windows box and they've never heard from me about it. If they did hear from me, it'd likely be a complaint that they need to vary the gameplay a bit. Great as it is, there's only so much Super Turbo Turkey Puncher can do to get past the fact that I've played the same basic game over and over again for the past twelve years or so.
IIRC they were disappointed in the sales of Quake III Arena, but we have Doom 3 and Quake 4 on Linux. It doesn't appear to have stopped them from porting their new titles. As long as they keep porting, I'll keep buying.
They very well might, although Japan at least already has its own RIAA-like organization. I have my share of Japanese music, mostly on CDs that I pick up on my occasional travels there mixed with a bit that I find here, and it's been interesting to see that various forms of copy protection have been on Japanese CDs for quite some time. Certainly longer than here in the US, and without most of the furor. Not that it actually stops me from ripping them to MP3s mind you, but it's there.
Slashdot Mirror
http://www.thewebsiteisdown.com
I'd like to see a CUDA implementation. With calculation speedups from 10x to hundreds of times over host-based processors for some algorithms on just one GeForce 8800GTS/GTX card (with 128 stream processors working simultaneously.. 8x the cores, even if lower clocked) this could get interesting very quickly.
Maybe, but that's just to throw off the authorities.
Yes. Yes, they do actually. I'm an SL player/"resident", and I realized when I signed up where Linden Labs is located. Second Life may well be a virtual world, but the company and servers that allow the virtual world to exist are still subject to the laws where they are physically located. Were they not headquartered in one of the more open minded areas of the US, who knows what might have happened to them already. It's an interesting ideal to make a virtual world completely separated from RL, but even ignoring moral issues there's a very real economic link between the two. That alone ensures that it will never happen.
I think the problem is that companies have a tendency to view people strictly as consumers, which I suppose is what they have to do if they want to survive. These "consumers" have portions of their lives that don't involve spending money though, and that's just as true in second life. Unless the company's product enhances the social aspect of SL somehow, or provides something actually useful in the gameplay, I don't see people wanting to view what is effectively a dull marketing campaign unless they're already interested in the product. How many people go out looking for pages of banner ads?
As an active Second Life 'resident', I'd actually be happy to see marketers give up on it. Perhaps they could get rid of the ad farms too, although I wouldn't hold my breath on either. SL in a lot of ways has strong parallels to the web as I remember it in the Mosaic/early Netscape years... The client's not really finished, network bandwidth hasn't quite caught up with its demands yet, effective advertising mechanisms aren't really well defined, and the emphasis on user-supplied content initially confuses many who are just looking for something to play with. I have a feeling that we're going to be using SL or something like it in 10 years or so though, in the same way that we're using web browsers now. Marketers are no doubt thinking along these lines.
On the other hand, the Linux client doesn't have voice support yet. It won't have it until the voice sw vendor gives them an object file or shared library build for the platform, which will happen who knows when. For those of us using the Linux viewer, there's still a good excuse.
FWIW, half the time I play SL with sound turned off - never mind voice. There's no real need for listening to the sound when you're just building objects.
Linux can certainly be used as an RTOS, in fact I've configured it for several embedded systems, but it's also relatively easy to use as a general purpose OS. If you stick to ARINC 653 (you're right on that of course, it's been a while since I was involved in spec'ing the CPCI boards) a full TCP/IP stack is basically out, amongst other niceties. You can do a cut down one, in our case with a special network card, but not everything works within the 653 spec according to the Green Hills staff with whom I've spoken. POSIX and ARINC 653 are really intended for different classes of problems.
/etc/security/limits.conf in many distros - would only affect the virtual machine. CPU resources are not quite as deterministic, since they rely on both the host and target kernel's scheduling algorithms, but if you need something like that you're probably going to use a real-time patched kernel assuming you go with Linux as the target.
Xen gives you a complete virtual machine, similar to but perhaps not quite as heavy as VMware, so you get a guest kernel inside the primary kernel. Memory and most other resources are partitioned deterministically, and something like a forkbomb - assuming you didn't reduce the process' maximum number of processes, modified in
Integrity is an RTOS platform, not a general purpose OS. I've worked with their ARINC 673 product a bit, much standard UNIX functionality would break the guarantees made by an ARINC-compliant OS so it's just not present. Xen is a close enough approximation if you just want to partition the system off without using ARINC 673, but in order to get the same sort of certifications as Integrity (or VxWorks' ARINC 673 product for that matter) all the code involved with Linux - kernel, userspace etc. - would need a line-by-line code review, probable changes, and sign off.
this is pointless
Yes, yes it is. You're presenting opinion and stating it as fact.
I would say Bill has had an overall negative impact on the world
Your biases are not a useful metric of ability. There is a computer in nearly every home in the US, and Windows has a lot to do with that. This alone contributes to the number of people who have been introduced to Computer Science, and arguably does more than any one person could have done. The Linux kernel has allowed for a complete GNU system to be released before the Hurd kernel was completed, giving a useful UNIX development environment to people who previously may never have had access to a full UNIX system. These people have produced software which has had a profound impact on Computer Science, by introducing many more people into the field. Despite my dislike for Microsoft products in general, I can see the profound good that their software has done to the field. What have you produced, beyond cookies? I make a mean batch of brownies myself, let's keep at least in the same general field.
Computer science (despite the misnomer) is the science of computation.
Do you know what science is? On what do you base your assertion that they do not understand the methodology? Have you polled them yourself? Opinion is not a useful metric, and the fact is you really have no idea aside from your own opinions and bias.
the code he has written, and the design decisions he has made shows quite clearly that he is not someone to look to for a CS luminary
Further opinion. Regardless, we're not talking about whether these people are luminaries, although they have had a known profound impact. We're talking about your assertion that they are mediocre, and that's frankly indefensible.
It's certainly not undue credit; whether you consider them the top programmers of our time or not, they are well beyond the label 'mediocre'. Each has contributed to the general good, and your biases are not a useful metric of ability. Knuth may have done more for pure computer science than the rest of the people you listed combined, yet that does not make them mediocre.
Name anything either Bill or Linus has EVER done that could in any way be construed as demonstrating that they have solid CS skills.
CS skills? You keep using those words. I don't think they mean what you think they mean.
Bill Gates is an excellent marketer. Linus Torvalds is an excellent manager. Both also have solid computer science skills. Although surely they were in the right place at the right time, to call either mediocre is either stroking your own ego or putting your head in the sand.
Whether Linus would have failed Andy's class depends on whether he tailored his code to the assignment; it's just part of going to college, and eventually applies to the real world as well. Based on Linus' management of the Linux kernel, I'd say he has a reasonable understanding of what to expect from people. Thus, I'd say he would have done quite well in Andy's class. Neither of us can say for certain however.
With x86 processors you typically have had to emulate the Ring 0 (supervisor mode) instructions, since these supervisor-level instructions could cause the virtualized system to break out of its sandbox and inadvertantly (or advertantly?) crash your machine. In a way, it's like gaining root or admin privileges in the OS. These newer virtualization instructions provide a secondary "Ring 0" that can be used to trap these instructions in hardware, rather than looking for the opcodes yourself. In theory these instructions should allow faster and possibly more stable emulation. In practice VMware runs a bit slower with these in 32 bit mode at the moment, although that will no doubt change in the future. It requires them in 64 bit mode however. Xen uses them to good effect, and is able to run unmodified operating systems only because of these instructions.
Seems fine here... Scrolling is smooth and the thumbwheel doesn't slow it down. What is your video card and which driver are you running? A crappy driver will slow down everything that scrolls text mixed with images, not just Firefox. For reference, I'm running a 32-bit Linux distro on an AMD 4200+ dual core, with 2gb RAM, a PCIE GeForce 6600 and NVidia's 9755 drivers. FF 2 runs fine (although it's eating 260mb RAM) with 26 tabs open. Come to think of it, I guess that's 10MB per tab...
Ever buy an OS from Dell with the PC? You're generally better off buying it separately. The restore discs I've seen detect whether the PC you're booting it on is a Dell (i.e. not VMware). The installer bails out if it isn't, and installs extra junk you probably don't want if it is. I'd go the Ubuntu route even if I did want to run Vista, and just buy the OS later.
There was a 6 button Genesis controller, which frankly was quite nice. I still have one around, the D-Pad just sort of floats there; it's very responsive. I particularly like the way Sega rounded the Genesis controllers to fit better into one's hands. The sharper angles on the NES and SNES controllers were rather uncomfortable to hold in comparison. The Wii's remote seems to have fallen back in that direction a bit, but it's not so thin. I haven't noticed the same cramping that came with the NES and SNES controllers, so that (along with not needing to hold both hands close together in most games) may be the reason.
It's been more expensive in San Francisco FWIW, $4.20 a gallon for the least expensive stuff at one gas station. We're already past $3.92/gallon, not that I think SUVs are ideal commuting vehicles. People who drive them for no good reason (there are good reasons, mind you, but most people don't need them) deserve the higher prices. Hopefully we'll see a shift in the types of vehicles people are driving as prices continue to rise.
That's the case only if you included the redacted items in the PDF to begin with. If you create a new PDF from your source document instead of modifying the PDF, you should be fine.
What are the odds that Dell might have their own repositories for their customers? That would give them the ability to add Dell-specific software, which they'd no doubt be very happy to do.
That's a good idea. I'd like to see it taken further, that is, a project to create new software technology and obtain patents on it. Any patented technologies would be explicitly licensed for open source usage. Commercial entities could license the developed technologies on the "inventor's" cost terms, thus funding the creation of newer technologies for open source projects and giving something back to the "inventor" as well. That would probably interest more people, create a stream of revenue, build a patent portfolio of safe technologies for non-fictional people, and involve those who are not necessarily coders.
I wouldn't be surprised if Dell's new position on Linux had a lot to do with this particular threat. Linux will now be offered to people who before didn't even know they had a choice, by one of the largest PC vendors on the planet. That's a powerful statement of support from Dell. It's good to keep this fact in perspective to be sure, it can't be argued that Microsoft will fall simply because of this. Do not however make the mistake of denying that this is happening at all. It will have an impact when people can buy an inexpensive Linux PC from Dell for less than the cost of a Windows system, and other PC manufacturers will start to feel more pressure to provide similar offerings.
As for XP, well, it's both obvious why Dell wants to sell it yet sad that MS doesn't want them doing it. Vista runs poorly if at all on existing hardware. You need significantly more RAM and a recent video card to run it at the same speed with the Aeroglass widget decorations from our experience. Many laptops need not apply. IT departments in larger organizations also do not modify their supported configurations so quickly, no matter what Microsoft wants. You need your third party software such as VPN clients working on the new platforms. Pushing customers to the platform before they are ready is not a good long term business move. I agree that it's not a death knell, but it is a shot in the arm for any of their competitors who are willing to make transitions more gradual.
In my experience, WPA works fine for me under Linux. I'm not using Ubuntu, but rather Mandriva 2007. The wpa_supplicant takes care of it nicely. The interfaces aren't where they need to be yet to be sure, but that comes with time. That's one of the nice things about Linux; as technologies become widespread, they become supported. Certainly there is always a need for new people to push software forward, hardware will continue to change and the OS must be updated to allow for this. Still, Linux doesn't need to conquer the desktop market any more than Apple does. There are plenty of kernel developers now, and there will likely continue to be in the future based on the current size of its community. The fact that it exists as a completely Free UNIX reimplementation, with the GPL license providing for those who prefer it, has garnered it a large enough userbase to sustain itself.
Back to the topic, that's not to say there's no place for Solaris. It's a fine OS, great for servers but a bit weak on the desktop based on my experience with my Ultra 40/Solaris 10. On a typical PC however, I'll take Linux any day. It does what I need it to, and tends to support the most cutting edge free *NIX tools first. It also tends to absorb useful features from other operating systems over time. Have a look at OpenVZ for example; Solaris 10-style containers already exist for Linux. Every time someone lists a set of features Linux needs, they end up being implemented either by someone who wants it, or by a company that decides they need it. The world's clearly richer for it. There would probably never even be an open source Solaris were this not the case.
Try "The Godfather: Blackhand Edition" before giving up. It's quite an experience, and I expect to see more like this with Scarface on the way.
Out of curiosity, how would they know who bought Doom 3 for Linux? I bought it, but it was a Windows box and they've never heard from me about it. If they did hear from me, it'd likely be a complaint that they need to vary the gameplay a bit. Great as it is, there's only so much Super Turbo Turkey Puncher can do to get past the fact that I've played the same basic game over and over again for the past twelve years or so.
IIRC they were disappointed in the sales of Quake III Arena, but we have Doom 3 and Quake 4 on Linux. It doesn't appear to have stopped them from porting their new titles. As long as they keep porting, I'll keep buying.
They very well might, although Japan at least already has its own RIAA-like organization. I have my share of Japanese music, mostly on CDs that I pick up on my occasional travels there mixed with a bit that I find here, and it's been interesting to see that various forms of copy protection have been on Japanese CDs for quite some time. Certainly longer than here in the US, and without most of the furor. Not that it actually stops me from ripping them to MP3s mind you, but it's there.