There might be some (more or less) hidden option to avoid any kind of system restore checkpoint and uninstall possibility before going ahead. That should save a lot, but I wouldn't call it a wise gamble.
Generally available system requirements for previous SPs have at least assumed that you go the "keep uninstall files" route, while also keeping local copies of all the new SP files, in addition to the actually installed copy.
On the other hand, the choices can be worded in such a way that it's really harder to choose the totally right one, than writing a free form answer that is seemingly right (because you don't expect an answer given in natural language to be totally complete and comprehensive without any logical loopholes). That is, you can test a higher level of knowledge by simplifying part of it through the inclusion of options.
One thing I really hated in school was essay questions where the teacher still had this strict template of what a "good" answer should include. It's applicable to some degree, but that requires imagination on the part of the teacher on how a student might interpret the question and what (s)he might choose to focus on, without even actively trying to misinterpret the question. In those cases, it could really hurt to have general knowledge and not only knowing the main points given in the textbook by heart.
A simple mp3 won't bring a system down on its own. If you add a BluRay disc or HDDVD (with not so good GPU acceleration) it will bring it down. Stuttering sound on a machine that's just coping with playing HD material, and handling network activity, woulnd't be so good. Just network and simple media don't tend to cause stuttering, but I'm really interested in hearing about a OS that doesn't stutter no matter what disk thrashing/network overloading/generally mean stuff you throw at it while watching. (Vista doesn't handle it properly either if you do enough of it, but saying it's not a problem, or that it's a problem that can be fixed by more optimization of the sound code, is ignorant.)
Another aspect here is that more of the Vista audio stack is in user mode, and detached from the interrupt handling. That's generally considered A Good Thing (microkernel thinking, huh?), but you still need to make sure that those user mode threads are scheduled.
The current solution is screwed up, but trying to solve the problem (because it is a real problem) isn't. Quicktime tends to stutter playing podcast mp3s on my MacBook when I do computations at 100 % CPU (even when reniced). Vista (and XP) doesn't. I'm sure I could tweak the Mac, but that's my quite personal experience on non-tweaked systems. That it happens at all indicates that there is a real problem that one can choose to fix, or ignore.
Undervolting, underclocking both CPU and GPU, and ideally a SSD (or 4,200 or MAYBE 5,400 RPM drive) would go a long way. A metal case would probably be best as well. Yeah, you're conducting the heat more, but you're also increasing the surface area. Considering what these things together will do to the total power used (down far below 20 W, more like 10 if you dim the light enough), I think most modern machines would do.
You forgot the part where sand is already SO2. (But, true, not all silicon in a chip is oxidized, so the problem in the original post was the assumption that a chip is sand or quartz. But I would dobut that the die itself will burn well compared to the thermal paste or just about anything else.)
KLM on-demand in-flight entertainment was great in one direction, fast forward, pause, anything.
Not so nice on the way back, about 1/3 of the seats had some kind of filesystem cross linking, the introduction menu was a piece of Shrek 2, any other movie would play the wrong one and then break. They tried to fix it by rebooting (making us all see that it was indeed that penguin-kernel running it), but it didn't work. This was back in 2005, but as others have already pointed out, it takes far more than just avoiding Redmond-ware to get a good system for that environment.
The major lines of mammals were already in existence long before the dinosaur extinction (65 mya). None of them were too large, but if the fossil record teaches us anything, changes in size can come on quite fast.
It's not compression as we know it, Jim. It's more like scaling on totally overcool steroids. The basic idea seems rather simple. I would even imagine you could get a bit of enhanced picture quality by coding simplified vector info on seams, and then doing a normal JPEG of a downscaled picture. That would be a quite contrived way to get a kind of VBR-like behavior in normal JPEG. One issue with JPEG is, after all, that redundancy is detected and handled on the block level, while this algorithm works along arbitrary paths.
I'm really impressed. Again, maybe not too hard to implement at first, but probably damn hard to get working perfectly, and I might just be ignorant (and I'm entitled too, it's far from my field of work), but I've not seen anyone doing it before.
Yes and no. If we're going to have a backlit screen anyway (even with LEDs), we can only gain so much by reducing the CPU consumption. Amdahl's law and all that. I think the summary is quite right in pointing out UMPCs and similar devices instead.
A really low-power Dothan or single-core Yonah will sure draw a few multiples of this beast, but they will do so while giving much better performance.
PCI is still at 33 MHz in most consumer systems. If the NIC and sound chip are connected through PCI and not PCIe or some other chipset interconncet, that's a possible bottleneck. Your number of 7 % would kind of match the drop seen. I still don't think it makes sense overall, but your debunking has serious problems as well.
It sounds like what is actually happening is that you're disabling and reenabling Aero on the fly. This basically means that all window manangement has to be shut down and reinitialized. Don't ask me WHY Media Center needs that to go fullscreen, but it does. (So does some games when run fullscreen.) Playing just about anything else fullscreen on one monitor and working on the other presents no problems to me.
So, I agree that this is broken, but it's not to due to bad threading but due to the fact that the seemingly simple operation you try to perform is actually one of the most complex ones the UI can do at all.
No, there is still a difference. Something created by man can be incredibly hard to really break, like most public-private key systems when handled properly. DRM is fundamentally broken. Something that is possible is to keep the actual key in hardware and even never let it leave a single chip. All communication to that chip can be done with encryption secured in another manner. If you make key-sniffing off the bus impossible, and make sure to keep the leakage and heat characteristics of the chip non-informative, you have a pretty secure solution.
But, still, it's not as simple as that they don't have the same resources that NSA do to throw at the problem. The real issue is that the problem they try to solve is fundamentally harder.
To be fair, you can get a fair deal of fertilizer out of urban sewage processing as well, but there is one issue (although I guess it's quite different in different regions): a lot of farmers (or regulatory bodies) don't like the trace amounts of cadmium and so on you'll find in the otherwise biological waste. I would imagine that the ammonia from this process isn't magically concentrated and pure before it reaches the lake. Extraction and separation to get it pure enough to sell could possibly even turn out to be more expensive than even turning it back into nitrogen and hydrogen...
It's actually more like the default. The description is just that, a description. You need a proper description, but it's still the actual claims that define your invention. The description can sometimes be used to interpret those claims.
One issue is that several (Java-based) applications tend to roll their own version of the JRE for "convenience". On the other hand, those apps hopefully do not use Java sandboxing to load arbitrary code safely from the net.
On the other hand, a broken partition table due to a random hardware error (or any other bug causing a write there) would probably not be detected until the next reboot anyway.
On the other hand, from glancing over his calculations, the point seems to be that the welfare created is dependent on the total number of copies ('copies' + 'originals') of the works in existence, and some arguments that new works are less valuable as there already exist so many of them. Again, this was only a very quick browsing through the article, but what I would argue that the total welfare is also dependent on the total number of "valuable" works in existence. His argument seems to be that the greater distribution (in itself, not as a source of new creation) cancels out the decline in new production quite easily.
Of course it's good to use proper models so one doesn't reach a conclusion which contradicts the stated targets, but his statement of the target doesn't seem to value cultural diversity. The main issue is that the model for creating 'originals' still doesn't reflect his own observation, that digital technology makes any distribution after the first, "real" creation, relatively trivial.
I would expect it to be much more noticeable if you have Visual Studio, and all of.NET 1.0,.NET 1.1,.NET 2.0 installed. (Visual Studio adding several large assemblies, and separate ngenning for each framework version.).NET 3.0 will also add a lot of assemblies.
If you run x64 Windows, then you'll probably run into even more duplicate work.
So, I would expect most W2K machines won't have VS2005 and certainly not.NET 3.0. This will make the NGen execution much shorter.
Now, if someone could just describe the finger-arm reflexes needed to make a first comment post and implement that in some kind of program or robot thingy...
That's why, even in urban areas, DSL is winning big over dedicated copper. No new cables over the last mile wins over any other solution, for existing buildings.
Slashdot Mirror
Generally available system requirements for previous SPs have at least assumed that you go the "keep uninstall files" route, while also keeping local copies of all the new SP files, in addition to the actually installed copy.
One thing I really hated in school was essay questions where the teacher still had this strict template of what a "good" answer should include. It's applicable to some degree, but that requires imagination on the part of the teacher on how a student might interpret the question and what (s)he might choose to focus on, without even actively trying to misinterpret the question. In those cases, it could really hurt to have general knowledge and not only knowing the main points given in the textbook by heart.
Another aspect here is that more of the Vista audio stack is in user mode, and detached from the interrupt handling. That's generally considered A Good Thing (microkernel thinking, huh?), but you still need to make sure that those user mode threads are scheduled.
The current solution is screwed up, but trying to solve the problem (because it is a real problem) isn't. Quicktime tends to stutter playing podcast mp3s on my MacBook when I do computations at 100 % CPU (even when reniced). Vista (and XP) doesn't. I'm sure I could tweak the Mac, but that's my quite personal experience on non-tweaked systems. That it happens at all indicates that there is a real problem that one can choose to fix, or ignore.
Undervolting, underclocking both CPU and GPU, and ideally a SSD (or 4,200 or MAYBE 5,400 RPM drive) would go a long way. A metal case would probably be best as well. Yeah, you're conducting the heat more, but you're also increasing the surface area. Considering what these things together will do to the total power used (down far below 20 W, more like 10 if you dim the light enough), I think most modern machines would do.
You forgot the part where sand is already SO2. (But, true, not all silicon in a chip is oxidized, so the problem in the original post was the assumption that a chip is sand or quartz. But I would dobut that the die itself will burn well compared to the thermal paste or just about anything else.)
Your PIII cannot saturate Gigabit Ethernet, I can almost promise you that. (On what bus would you connect the NIC?)
Not so nice on the way back, about 1/3 of the seats had some kind of filesystem cross linking, the introduction menu was a piece of Shrek 2, any other movie would play the wrong one and then break. They tried to fix it by rebooting (making us all see that it was indeed that penguin-kernel running it), but it didn't work. This was back in 2005, but as others have already pointed out, it takes far more than just avoiding Redmond-ware to get a good system for that environment.
The major lines of mammals were already in existence long before the dinosaur extinction (65 mya). None of them were too large, but if the fossil record teaches us anything, changes in size can come on quite fast.
I'm really impressed. Again, maybe not too hard to implement at first, but probably damn hard to get working perfectly, and I might just be ignorant (and I'm entitled too, it's far from my field of work), but I've not seen anyone doing it before.
A really low-power Dothan or single-core Yonah will sure draw a few multiples of this beast, but they will do so while giving much better performance.
Yeah, I suggest an immediate ban on yoghurt.
PCI is still at 33 MHz in most consumer systems. If the NIC and sound chip are connected through PCI and not PCIe or some other chipset interconncet, that's a possible bottleneck. Your number of 7 % would kind of match the drop seen. I still don't think it makes sense overall, but your debunking has serious problems as well.
100 Mbit for what medium? Cat5? So, tell me again, how many repeaters are you going to place per mile?
Do you remember NT4 to 2000? Vista has a new graphics driver model, so did 2000.
So, I agree that this is broken, but it's not to due to bad threading but due to the fact that the seemingly simple operation you try to perform is actually one of the most complex ones the UI can do at all.
But, still, it's not as simple as that they don't have the same resources that NSA do to throw at the problem. The real issue is that the problem they try to solve is fundamentally harder.
To be fair, you can get a fair deal of fertilizer out of urban sewage processing as well, but there is one issue (although I guess it's quite different in different regions): a lot of farmers (or regulatory bodies) don't like the trace amounts of cadmium and so on you'll find in the otherwise biological waste. I would imagine that the ammonia from this process isn't magically concentrated and pure before it reaches the lake. Extraction and separation to get it pure enough to sell could possibly even turn out to be more expensive than even turning it back into nitrogen and hydrogen...
It's actually more like the default. The description is just that, a description. You need a proper description, but it's still the actual claims that define your invention. The description can sometimes be used to interpret those claims.
One issue is that several (Java-based) applications tend to roll their own version of the JRE for "convenience". On the other hand, those apps hopefully do not use Java sandboxing to load arbitrary code safely from the net.
On the other hand, a broken partition table due to a random hardware error (or any other bug causing a write there) would probably not be detected until the next reboot anyway.
Of course it's good to use proper models so one doesn't reach a conclusion which contradicts the stated targets, but his statement of the target doesn't seem to value cultural diversity. The main issue is that the model for creating 'originals' still doesn't reflect his own observation, that digital technology makes any distribution after the first, "real" creation, relatively trivial.
Just wait for NGen to complete. The issues with the patch failing to install might warrant a patch, though.
If you run x64 Windows, then you'll probably run into even more duplicate work.
So, I would expect most W2K machines won't have VS2005 and certainly not .NET 3.0. This will make the NGen execution much shorter.
Now, if someone could just describe the finger-arm reflexes needed to make a first comment post and implement that in some kind of program or robot thingy...
That's why, even in urban areas, DSL is winning big over dedicated copper. No new cables over the last mile wins over any other solution, for existing buildings.