But, hotpatching is supported, to a higher degree. (It's also listed as one of the new features in Vista SP1, or really, the support was kind of there, but as it is really a feature developed for the server version the code in Vista RTM was too bad to be used.)
They support terminal services, from what I can see. The real issue is that.NET is not supported, with all the dependencies that might mean. No.NET means no ASP.NET, no fancy SQL Server features, no PowerShell.
Desktop Nehalem seems to be on track for the first half of 2009, with some server/workstation/enthusiast late 2008. That is, it's almost perfectly mirroring Penryn, but a year later (and about 2.5 years after Conroe). I don't know where you got 2011 from...
I don't see that many myself... that's a pretty high number.
Most of mine lately are the windows RCP exploits and the exploit for old symantec overflows on port 2967. That, or I used to get a lot of traffic from SSH brute force attacks and malicious HTTP stuff... Do you mean RPC? (Not trolling, I was really confused and that's the explanation that makes most sense to mean.)
Re:One potential future advantage of AMD's technol
on
Is AMD Dead Yet?
·
· Score: 1
The stated response time of 3 ns is, however, only equivalent to 333 MHz, or 10 cycles at 3.0 GHz. To realize the benefits of shorter trace lengths to compensate for the added latency, the cache volume cannot be increased. It's also harder to share cache between cores, as that's another source of "required" trace length. Z-RAM is very interesting, but it's not obvious whether it can improve over SRAM at all places in the cache hierarchy.
The Phenom's a bit of a disappointment, and will probably remain so until/unless people start writing much more parallelisable code (until then, Intel's bigger L2 cache more than makes up for Phenom's "true" quad-core design). But AMD are fighting back on the GPU side - the HD 3870 X2 has had some great reviews, and in many games it's faster than an 8800 Ultra for sixty quid less. I'm not actually so sure on the "Phenom will be good as soon as parallelism gets better" thinking. The point is that even when you have a shared L2/fast interconnect/whatever, you would have really liked to stay in the L1. Simultaneous writes will mean locks and cause trouble (or added latency, even for seemingly lockfree structures), even if the cache is shared. The point is that good parallelization should not require frequent transactions at high bandwidth between threads, no matter how close together you put them.
I wonder what the low-frequency observators will make of the fact that the emissions from a body orbiting Sol has remarkable signals in the 50-60 Hz range, and that the linear components of the two distinctive signals changes distinctively with a 24-hour period. (or: not all countries use 60 Hz, you insensitive clod! Try growing up with 50 Hz PAL CRTs and you would also be touchy about it!)
Contrary to what others here have been saying, Dragon Naturally Speaking and similar programs are NOT "Natural Language Processing". They are, instead, language translaton programs. The former refers to extracting actual meaning, of some kind, from the input. The latter means little more than determining what words you said closely enough to perform some pre-programmed actions. The two things are worlds apart. There has been some real progress in the latter... but computers really don't "understand" language any better than they did decades ago. Do you have any specific source for that? I would say that most anything that fits as an ACL paper is also NLP of some sort, or at least that's the current trend in modifying the definition. With the advances of statistical systems in many fields where "real" content extraction was once thought to be the holy grail, the actual difference is blurred. I think that we will define away anything close to strong AI as "something else" for quite a long time, even at the point when it is obvious that the only thing keeping a specific piece of software from being strong would be an arbitrary limitation in input and output.
Existing cells in situ won't be changed. Injecting these in existing tissue will, supposedly, add into the empty spots of the cellular matrix and so could add a kind of rejuvenating effect. Any attempt to activate telomerase in general in existing tissue is outright dangerous, considering that it only takes a single cell out of a billion to seed a cancer.
I don't think that's the case. Graphics cards work on the same PCI-X buses that acceleration cards probably use lately. They use DMA to communicate with main memory without involving the processor. The VRAM might be optimised for writing, but it should be very possible to do calculations on the card, and get the results back. That's the whole point of the generalised GPGPU techniques. Nitpicking: PCI Express is not PCI-X. PCI-X was a derivative of the parallel PCI bus and never found in mainstream machines.
As it's all about VM handling, I'm not sure, maybe some platform might give a panic, maybe some will use a different codepath, but it seems to be high-level enough that it is common code for all architectures. However, the only parts that are specific for x86 and x64 in the actual exploit are really the exploit code that's called within the kernel. As it is inline assembler and calling-convention specific, one would need a separate set for each platform. This is just like a user-space buffer overflow needs a specific exploit per platform, although there might be a common bug.
I think you would be surprised to see what kinds of draconian licensing schemes are still out there for special-purpose (and highly professional) software. It's the same class of products that can get away with a crappy installer and even possibly quite specific requirements of OS, since they have their market and the product, when it works, actually adds a lot to productivity.
And nothing stops you from implementing that in COM (where the "interface" aspect is very much a part of the concept) or.NET. In fact, that's the way ADO data access worked in Windows, to continue your example. That doesn't magically solve the issue of incompatible DB syntax. I think there were some plans on implementing a version of Gecko that exposed the MSHTML interfaces, but I don't think they ever got close to actually finishing it.
I thought Mozilla and other projects kept the real public resources (like the bug db) under some limited access, where relevant (exploitable) bugs tend to be locked-down until the fix is released. I don't say it's simple to join an inner circle through some social engineering, but it will and should be harder than just signing up in a form.
You are aware that the same code will be interpreted radically differently with and without a proper declaration, right? ("Quirks mode" in both IE and Firefox.)
That doesn't stop the loading page from making a request to the W3C servers, does it? Worse, many of the existing cache implementations won't cache an error result.
Things like simple directory sharing rapidly hit connection limits quickly on the desktop versions, and modifying file permissions has a very crippled interface on the desktop versions (plus a lot of other gripes). XP Pro has no crippled ACL GUI.
This is in the area of Really Bad Analogies, but bear with me anyway: Windows 2003 was the followup to XP. Quite a few users who got their hands (ahem...) on Windows 2003 think that it is a very solid workstation OS, handling a few situations better than XP. If blatantly ignoring the lack of drivers, XP64 (which is derived from the 2003 codebase) is also solid. More polish went into handling high-load cases and simple bugfixing, things that were never justified to backport to XP. (SP2 carried over some things and added a few, so until Windows 2003 SP1 was released, there were two clear forks again with non-overlapping features.)
What does this mean for Vista SP1? Well, there should be very little reason to use Windows 2008 as a desktop OS. One could imagine that some geek/pro user workloads (network/disk I/O, anyone?) might be improved. On the other hand, these changes should already be in the SP release candidates, and the reviews of those haven't shown any big changes. A practical concern would be that the platforms should be similar from now on, like in the W2K days. I guess that will make at least some hardware vendor developers happy. Maybe this will also mean that additional hotfixes more acutely needed for server scenarios will trickle down to Vista.
I would not be surprised if an enemy-agent is caught in act of cutting and that agent speaks arabic, especially Persian. I see that you have real inside knowledge about the cultural and ethnic structure of the region.
It's also common in mathematics - Sir Isaac Newton stole copiously from Huygens, Descartes, Hooke, and anyone else stupid enough to let him. Or perhaps not stupid - the only person to resist Newton's claim of ownership did die rather soon after. Ehm, Descartes died in 1650. Newton was born in 1643. As at least Hooke was actively contesting the rights of discovery, I am not sure what you mean by the death reference, as he lived to the age of 67, while the conflict arose far earlier. Newton was not amicable to everyone, but you certainly had quite a bit of hyperbole in your post.
As x64 eliminated the more rarely used x86 rings (IIRC), and the kernel protection of Vista is most encompassing on x64, it's quite clear that this source was not very reliable. Vista and 2K8 are siblings, but noone (in their right mind) at MS denies that 2K8 is based on 2K3.
Slashdot Mirror
But, hotpatching is supported, to a higher degree. (It's also listed as one of the new features in Vista SP1, or really, the support was kind of there, but as it is really a feature developed for the server version the code in Vista RTM was too bad to be used.)
They support terminal services, from what I can see. The real issue is that .NET is not supported, with all the dependencies that might mean. No .NET means no ASP.NET, no fancy SQL Server features, no PowerShell.
Desktop Nehalem seems to be on track for the first half of 2009, with some server/workstation/enthusiast late 2008. That is, it's almost perfectly mirroring Penryn, but a year later (and about 2.5 years after Conroe). I don't know where you got 2011 from...
Most of mine lately are the windows RCP exploits and the exploit for old symantec overflows on port 2967. That, or I used to get a lot of traffic from SSH brute force attacks and malicious HTTP stuff... Do you mean RPC? (Not trolling, I was really confused and that's the explanation that makes most sense to mean.)
The stated response time of 3 ns is, however, only equivalent to 333 MHz, or 10 cycles at 3.0 GHz. To realize the benefits of shorter trace lengths to compensate for the added latency, the cache volume cannot be increased. It's also harder to share cache between cores, as that's another source of "required" trace length. Z-RAM is very interesting, but it's not obvious whether it can improve over SRAM at all places in the cache hierarchy.
I'm not actually so sure on the "Phenom will be good as soon as parallelism gets better" thinking. The point is that even when you have a shared L2/fast interconnect/whatever, you would have really liked to stay in the L1. Simultaneous writes will mean locks and cause trouble (or added latency, even for seemingly lockfree structures), even if the cache is shared. The point is that good parallelization should not require frequent transactions at high bandwidth between threads, no matter how close together you put them.
They did a showcase on their website, the key itself is not a display. You only move a transparent keycap.
I wonder what the low-frequency observators will make of the fact that the emissions from a body orbiting Sol has remarkable signals in the 50-60 Hz range, and that the linear components of the two distinctive signals changes distinctively with a 24-hour period. (or: not all countries use 60 Hz, you insensitive clod! Try growing up with 50 Hz PAL CRTs and you would also be touchy about it!)
Existing cells in situ won't be changed. Injecting these in existing tissue will, supposedly, add into the empty spots of the cellular matrix and so could add a kind of rejuvenating effect. Any attempt to activate telomerase in general in existing tissue is outright dangerous, considering that it only takes a single cell out of a billion to seed a cancer.
You say this despite the fact that the official RTM build seems to have a build date of Jan 18?
Nitpicking: PCI Express is not PCI-X. PCI-X was a derivative of the parallel PCI bus and never found in mainstream machines.
You are right, vmsplice should reduce the number of separate copy operations and buffers.
As it's all about VM handling, I'm not sure, maybe some platform might give a panic, maybe some will use a different codepath, but it seems to be high-level enough that it is common code for all architectures. However, the only parts that are specific for x86 and x64 in the actual exploit are really the exploit code that's called within the kernel. As it is inline assembler and calling-convention specific, one would need a separate set for each platform. This is just like a user-space buffer overflow needs a specific exploit per platform, although there might be a common bug.
I think you would be surprised to see what kinds of draconian licensing schemes are still out there for special-purpose (and highly professional) software. It's the same class of products that can get away with a crappy installer and even possibly quite specific requirements of OS, since they have their market and the product, when it works, actually adds a lot to productivity.
And nothing stops you from implementing that in COM (where the "interface" aspect is very much a part of the concept) or .NET. In fact, that's the way ADO data access worked in Windows, to continue your example. That doesn't magically solve the issue of incompatible DB syntax. I think there were some plans on implementing a version of Gecko that exposed the MSHTML interfaces, but I don't think they ever got close to actually finishing it.
I thought Mozilla and other projects kept the real public resources (like the bug db) under some limited access, where relevant (exploitable) bugs tend to be locked-down until the fix is released. I don't say it's simple to join an inner circle through some social engineering, but it will and should be harder than just signing up in a form.
You are aware that the same code will be interpreted radically differently with and without a proper declaration, right? ("Quirks mode" in both IE and Firefox.)
It is not a NEW kernel. It is the one that went into Windows Server 2008 final, rather than a pre-beta 3 build (a.k.a. Vista RTM).
What does this mean for Vista SP1? Well, there should be very little reason to use Windows 2008 as a desktop OS. One could imagine that some geek/pro user workloads (network/disk I/O, anyone?) might be improved. On the other hand, these changes should already be in the SP release candidates, and the reviews of those haven't shown any big changes. A practical concern would be that the platforms should be similar from now on, like in the W2K days. I guess that will make at least some hardware vendor developers happy. Maybe this will also mean that additional hotfixes more acutely needed for server scenarios will trickle down to Vista.
I see that you have real inside knowledge about the cultural and ethnic structure of the region.
As x64 eliminated the more rarely used x86 rings (IIRC), and the kernel protection of Vista is most encompassing on x64, it's quite clear that this source was not very reliable. Vista and 2K8 are siblings, but noone (in their right mind) at MS denies that 2K8 is based on 2K3.