And, if you are in a busy region with a nearby warehouse, I almost always get the items within 2 days even without Prime. I do have Prime, because occasionally I get something that comes from across the country and I appreciate the consistency of knowing when it should arrive.
Oddly enough, sometimes I order things I expect in two days and make plans around it; only to have Amazon deliver the next day (once I had an item the same day when I ordered very early in the morning). I prefer consistent times over anything else.
I've tested all these attacks myself, and with a good directional antenna with a high transmit power the attacker can be pretty damn far away from you.
Even if you lower your router's power output (a very good first step to mitigate this attack), his directional antenna will allow him to pick up fainter signals.
Disable 2.4GHz if you can, and just use 5Ghz as there are far fewer high powered directional antenna available. The 5GHz signal also doesn't propagate as far.
If you find the location he's coming from, you can shield that with foil.
The solution for wireless could be a TCP congestion control change, such as Westwood+ which accounts for bandwidth by delay rather than dropped packets.
But even better is a simple proxy setup. The proxy handles the request at the AP for the client, and retransmits can occur over the much faster wireless link.
It's mostly a cost issue, since only recent APs are powerful enough to run a local caching proxy.
And pretty much everything you said is true in some sense. Given not so superb equipment for mixing, recording, and playback, simply having the slight room for aliasing filters and frequency information can improve the final product that gets output at 16/44.
But as the article says, if you do it right the first time, there's really nothing to be had going for more than 16 bits, and 44kHz, it should encapsulate the entire range of human hearing in any normal situation.
I'm really glad the article was posted. It cleared up some of my misconceptions.
And now I know the final product at 16/44 is just fine if done right.
This is true, but modern branch predictors are pretty good. Sandy Bridge reportedly is correct more than 80% of the time. So it really depends if having fast on chip DRAM (but a smaller amount) is more valuable than having L1, L2, L3, and RAM caches at differing speeds with their own predictors.
I've heard anecdotal evidence (so take with a grain of salt) that doing stuff on ReFS is much faster.
Keep in mind this initial release is for servers only, and NOT for boot volumes, so it'll be a while (half a decade or more) before it trickles down into most desktops/laptops.
If Microsoft implements it right, it should be faster than NTFS.
Since it's copy on write, you can batch together random writes into a single linear write (while still maintaining consistency). They also mention having 3 allocators depending on the size of data to be written (because a one size fits all allocator is worse than 3 tuned to data size).
And considering it gets rid of some rarely used NTFS features, it also stands to be faster because it doesn't have to support as much.
Are you sure that's saving you power? Sometimes you have to initialize the device in order to activate the power saving features. It happens in Windows as well with wireless cards. The default state is "max power."
I agree generally, like AMD's bulldozer hitting 8GHz on a single core before failing to the limits of physics (even with extreme cooling). I'm assuming nobody will never be able to get more than 1 or 2 cores active (out of 8) while getting to 8GHz on that architecture.
But these days, the chips run in multiple clock domains. I believe the Intel chips are separated by a base clock, L3 Clock, Core clocks, RAM clocks, and bus clocks. The architectures are moving ever toward asynchronous operation in order to pack billion upon billion of transistors on a package without having to synchronize them all the time.
Microsoft is doing the right thing here. They are dropping the antiquated Win32 API and its bloat, in favor of a new universal WinRT API that targets both ARM and X86. Furthermore, it consolidates everything (Silverlight, Win32, WPF,.Net) into a cohesive API that you should be easily able to port to. If you cannot easily port to it, you're probably designing some custom business app that has and never will upgrade. Sorry but the new Windows isn't for you.
People have been asking Microsoft to the drop the bloat for some time. The security has been a nightmare because they've had to maintain the old unsecure model for the sake of compatibility. This clean break allows them to fix the permission model so that each App asks and gets only the permissions it needs during the install (like Android).
I like this change. Apps will now have to focus on doing one thing really well (and being able to connect and share with other apps) instead of being one-stop bloatware packages. Have you seen new contract API that facilitates this? It seems to follow the UNIX philosophy of doing a single task and being to pipe that output to anywhere.
Reading the paper, it seems the proposed protocol for key exchange forces a wait time of 17ms, and then hashes the packet to ensure it doesn't get modified (forcing the use of slots and keeping the air open during attack).
The only problem I see is that you could easily use this mechanism to effectively DoS the network by making it wait for the CTS packets constantly while the protocol rejects the bad check-summed packets.
But I guess that's a minor flaw since it's already trivial to DoS wireless networks in general.
Here's to hoping this actually gets widely implemented.
You say it creates a recovery partition? But when I run it from an existing Snow Leopard install, my existing Snow leopard partition size remains unchanged. I don't have an extra partition after installing Lion.
As an example, I've loaded a custom build of Tomato firmware onto my WRT54GL router. This router is considered out dated, slow, and lacking in RAM. And yet, I've got a new firmware for it that gives me a 2.6.22 kernel (originally 2.4) and IPv6 support.
Now my ISP (U-Verse) doesn't have any mention of IPv6 support, but the provided gateway does have much more powerful specs. In theory it should be a much more capable box, but their crapware firmware doesn't unlock any of its potential.
So basically, we would need the FCC to open up the airway for more frequencies, thus reducing collisions and allowing switching when one channel is in use.
AFAIK, almost every wireless device in my house is around 2.4GHz.
I'm surprised Half-Life wasn't on that list. It was and still is regarded in the gaming industry as one of the best games ever made during that era. I think it, and Half-Life 2 have some of the highest combined scores by reviewers.
It also spawned a whole scene of modding that produced Counter-Strike, among other games, which today is still of the most played games in the world.
Steam is slightly better than the absolute worst of the DRM systems yes, but it still can take all your games away if you so much as look at it funny. Even SecuROM or StarForce can't do that. If you travel with a laptop, I suggest you buy your games somewhere else. Steam has a tendency to lock accounts that log in from multiple IP addresses. 'Cause, you know, that's a crime.
I call bullshit on that. I regularly use Steam from school, my apartment, and my parents house and on many different machines. My brother also logs into my account to access my game libarary and vice versa.
Really the only ban is for cheating on VAC secured servers and games. Even if you get banned for cheating, it's still possible to play on non VAC secured games and servers.
For a while now Linux has had built in ACPI power governors, and they seem to work across all Pentium and AMD processors. The ondemand governor is default and automatically bursts to 100% Frequency and tries to quickly scale back to the minimum frequency to save power.
Of course you can also control and select the governors (ondemand, conservative, user set frequency, and performance)
I can even patch the kernel to undervolt my cpu per frequency to really squeeze out extra battery life.
AFAIK, ACPI power states are well supported on modern processors (everything ACPI related works on my pentium-m).
But of course all of that functionality is built into Vista and Windows 7. The ASUS utility just provides a different interface for it, and perhaps choses better defaults for their own hardware.
Anandtech has done tests and found using Windows 7's "battery mode" does improve battery life by mostly capping the CPU frequency.
Where Linux falls short is controlling the power of other systems, like wireless, usb, and graphics cards. Most of the power saving features are just not implemented for Linux, yet.
OS X also does a very good job getting extended battery life compared to Windows 7 on the same hardware. Of course apple gets to control all the hardware and make sure the power systems work. Microsoft does really well for how many platforms it support. Linux is getting there.
All the multi-touch bits were just added to the Xorg stack. And Intel is working on a multi-touch library for common gestures.
The software is just barely getting there, so it will be a while.
NoScript allows per site javascript blocking. And flash blocking. And XSS protection.
In combination with adblock+ my web surfing is much safer and faster.
http://noscript.net/
Personally I usually set it to allow javascript from the site itself (top-level), but block external javascript. That makes most pages work. Also disable the annoying pop-up telling you it blocked something.
It's actually porous, like a fine mesh or fine screen door. It works ok, but eventually it wears down from gardening and other maintenance requiring your to cut a hole in it. Some weeds can still get through it (we have tough spiked weeds in the desert).
I would now go with a natural alternative considering you end up having to replace it anyways. A thick weed resistant lawn helps more than anything.
Aero offloads the GUI onto your graphics card if it is capable of DirectX 9. It provides a faster, tear free interface, and if you notice DWM.exe (Desktop Window Manager) uses only 0-1% of CPU during use.
If you disable Aero and fall back to GDI, DWM.exe will disappear, and explorer.exe instead takes the load, usually using 1-5% of my CPU (at least on this machine).
In general, you should get better performance if you have a decent video card. If you are using the desktop anyways, why not utilize the GPU?
A couple of considerations:
1. Vista uses more GPU ram with each window. If you have a shared memory GPU, it's conceivable that it would be too slow when you start opening many windows. Or if your GPU just doesn't have a lot of RAM.
2. Maybe your GPU isn't as power efficient as using the CPU for rendering the windows. Battery life could be affected.
3. Windows 7 with driver model 1.1 uses a constant amount of GPU ram for any amount of windows (steaming in textures instead of keeping them loaded). It also re-enables GDI 2D HW acceleration which was disabled in Vista, but available in XP. Windows 7 also accelerates Cleartype text.
From the pictures, this device appears to be a made up of a ARM board with lots of usb ports. Internally there are two usb ports, and if I'm right that's a wireless usb dongle, and a SD card in a usb dongle for storage.
The demo device also has a Dell keyboard, the same as the one on my 4 year old Inspiron 700m. What this guy did is put together a compelling package using various parts as a demo, AFAIK. I wonder if laptop manufactures share a lot of parts? (besides the intel centrino chipset)
it's getting even worse by crossing the 1.0TB barrier. Not only are they using the base10 to 2 conversion to cheat you out of space, but they are advertising this drive as 1.5TB when it's actiually 1500TB. That's 36GB missing. Still though his drive was worth it for it's low GB per dollar ratio.
Slashdot Mirror
And, if you are in a busy region with a nearby warehouse, I almost always get the items within 2 days even without Prime. I do have Prime, because occasionally I get something that comes from across the country and I appreciate the consistency of knowing when it should arrive.
Oddly enough, sometimes I order things I expect in two days and make plans around it; only to have Amazon deliver the next day (once I had an item the same day when I ordered very early in the morning). I prefer consistent times over anything else.
I've tested all these attacks myself, and with a good directional antenna with a high transmit power the attacker can be pretty damn far away from you.
Even if you lower your router's power output (a very good first step to mitigate this attack), his directional antenna will allow him to pick up fainter signals.
Disable 2.4GHz if you can, and just use 5Ghz as there are far fewer high powered directional antenna available. The 5GHz signal also doesn't propagate as far.
If you find the location he's coming from, you can shield that with foil.
The solution for wireless could be a TCP congestion control change, such as Westwood+ which accounts for bandwidth by delay rather than dropped packets.
But even better is a simple proxy setup. The proxy handles the request at the AP for the client, and retransmits can occur over the much faster wireless link.
It's mostly a cost issue, since only recent APs are powerful enough to run a local caching proxy.
And pretty much everything you said is true in some sense. Given not so superb equipment for mixing, recording, and playback, simply having the slight room for aliasing filters and frequency information can improve the final product that gets output at 16/44.
But as the article says, if you do it right the first time, there's really nothing to be had going for more than 16 bits, and 44kHz, it should encapsulate the entire range of human hearing in any normal situation.
I'm really glad the article was posted. It cleared up some of my misconceptions.
And now I know the final product at 16/44 is just fine if done right.
This is true, but modern branch predictors are pretty good. Sandy Bridge reportedly is correct more than 80% of the time. So it really depends if having fast on chip DRAM (but a smaller amount) is more valuable than having L1, L2, L3, and RAM caches at differing speeds with their own predictors.
I'm guessing it depends on the application.
I've heard anecdotal evidence (so take with a grain of salt) that doing stuff on ReFS is much faster.
Keep in mind this initial release is for servers only, and NOT for boot volumes, so it'll be a while (half a decade or more) before it trickles down into most desktops/laptops.
If Microsoft implements it right, it should be faster than NTFS.
Since it's copy on write, you can batch together random writes into a single linear write (while still maintaining consistency). They also mention having 3 allocators depending on the size of data to be written (because a one size fits all allocator is worse than 3 tuned to data size).
And considering it gets rid of some rarely used NTFS features, it also stands to be faster because it doesn't have to support as much.
Are you sure that's saving you power? Sometimes you have to initialize the device in order to activate the power saving features. It happens in Windows as well with wireless cards. The default state is "max power."
I agree generally, like AMD's bulldozer hitting 8GHz on a single core before failing to the limits of physics (even with extreme cooling). I'm assuming nobody will never be able to get more than 1 or 2 cores active (out of 8) while getting to 8GHz on that architecture.
But these days, the chips run in multiple clock domains. I believe the Intel chips are separated by a base clock, L3 Clock, Core clocks, RAM clocks, and bus clocks. The architectures are moving ever toward asynchronous operation in order to pack billion upon billion of transistors on a package without having to synchronize them all the time.
Microsoft is doing the right thing here. They are dropping the antiquated Win32 API and its bloat, in favor of a new universal WinRT API that targets both ARM and X86. Furthermore, it consolidates everything (Silverlight, Win32, WPF, .Net) into a cohesive API that you should be easily able to port to. If you cannot easily port to it, you're probably designing some custom business app that has and never will upgrade. Sorry but the new Windows isn't for you.
People have been asking Microsoft to the drop the bloat for some time. The security has been a nightmare because they've had to maintain the old unsecure model for the sake of compatibility. This clean break allows them to fix the permission model so that each App asks and gets only the permissions it needs during the install (like Android).
I like this change. Apps will now have to focus on doing one thing really well (and being able to connect and share with other apps) instead of being one-stop bloatware packages. Have you seen new contract API that facilitates this? It seems to follow the UNIX philosophy of doing a single task and being to pipe that output to anywhere.
Reading the paper, it seems the proposed protocol for key exchange forces a wait time of 17ms, and then hashes the packet to ensure it doesn't get modified (forcing the use of slots and keeping the air open during attack).
The only problem I see is that you could easily use this mechanism to effectively DoS the network by making it wait for the CTS packets constantly while the protocol rejects the bad check-summed packets.
But I guess that's a minor flaw since it's already trivial to DoS wireless networks in general.
Here's to hoping this actually gets widely implemented.
Is there an advantage to a reformat?
You say it creates a recovery partition? But when I run it from an existing Snow Leopard install, my existing Snow leopard partition size remains unchanged. I don't have an extra partition after installing Lion.
As an example, I've loaded a custom build of Tomato firmware onto my WRT54GL router. This router is considered out dated, slow, and lacking in RAM. And yet, I've got a new firmware for it that gives me a 2.6.22 kernel (originally 2.4) and IPv6 support.
Now my ISP (U-Verse) doesn't have any mention of IPv6 support, but the provided gateway does have much more powerful specs. In theory it should be a much more capable box, but their crapware firmware doesn't unlock any of its potential.
Thanks for that informative post!
So basically, we would need the FCC to open up the airway for more frequencies, thus reducing collisions and allowing switching when one channel is in use.
AFAIK, almost every wireless device in my house is around 2.4GHz.
I'm surprised Half-Life wasn't on that list. It was and still is regarded in the gaming industry as one of the best games ever made during that era. I think it, and Half-Life 2 have some of the highest combined scores by reviewers.
It also spawned a whole scene of modding that produced Counter-Strike, among other games, which today is still of the most played games in the world.
The Olympics happened?
http://googleblog.blogspot.com/2010/03/searching-for-gold-during-games.html
Steam is slightly better than the absolute worst of the DRM systems yes, but it still can take all your games away if you so much as look at it funny. Even SecuROM or StarForce can't do that. If you travel with a laptop, I suggest you buy your games somewhere else. Steam has a tendency to lock accounts that log in from multiple IP addresses. 'Cause, you know, that's a crime.
I call bullshit on that. I regularly use Steam from school, my apartment, and my parents house and on many different machines. My brother also logs into my account to access my game libarary and vice versa.
Really the only ban is for cheating on VAC secured servers and games. Even if you get banned for cheating, it's still possible to play on non VAC secured games and servers.
For a while now Linux has had built in ACPI power governors, and they seem to work across all Pentium and AMD processors. The ondemand governor is default and automatically bursts to 100% Frequency and tries to quickly scale back to the minimum frequency to save power.
Of course you can also control and select the governors (ondemand, conservative, user set frequency, and performance)
I can even patch the kernel to undervolt my cpu per frequency to really squeeze out extra battery life.
AFAIK, ACPI power states are well supported on modern processors (everything ACPI related works on my pentium-m).
But of course all of that functionality is built into Vista and Windows 7. The ASUS utility just provides a different interface for it, and perhaps choses better defaults for their own hardware.
Anandtech has done tests and found using Windows 7's "battery mode" does improve battery life by mostly capping the CPU frequency.
Where Linux falls short is controlling the power of other systems, like wireless, usb, and graphics cards. Most of the power saving features are just not implemented for Linux, yet.
OS X also does a very good job getting extended battery life compared to Windows 7 on the same hardware. Of course apple gets to control all the hardware and make sure the power systems work. Microsoft does really well for how many platforms it support. Linux is getting there.
All the multi-touch bits were just added to the Xorg stack. And Intel is working on a multi-touch library for common gestures. The software is just barely getting there, so it will be a while.
NoScript allows per site javascript blocking. And flash blocking. And XSS protection. In combination with adblock+ my web surfing is much safer and faster. http://noscript.net/ Personally I usually set it to allow javascript from the site itself (top-level), but block external javascript. That makes most pages work. Also disable the annoying pop-up telling you it blocked something.
It applies against 2.6.30 (Just look at the patch file headers). Or you could grab zen-sources which includes a choice between the latest CFS or BFS.
It's actually porous, like a fine mesh or fine screen door. It works ok, but eventually it wears down from gardening and other maintenance requiring your to cut a hole in it. Some weeds can still get through it (we have tough spiked weeds in the desert). I would now go with a natural alternative considering you end up having to replace it anyways. A thick weed resistant lawn helps more than anything.
Aero offloads the GUI onto your graphics card if it is capable of DirectX 9. It provides a faster, tear free interface, and if you notice DWM.exe (Desktop Window Manager) uses only 0-1% of CPU during use.
If you disable Aero and fall back to GDI, DWM.exe will disappear, and explorer.exe instead takes the load, usually using 1-5% of my CPU (at least on this machine).
In general, you should get better performance if you have a decent video card. If you are using the desktop anyways, why not utilize the GPU?
A couple of considerations:
1. Vista uses more GPU ram with each window. If you have a shared memory GPU, it's conceivable that it would be too slow when you start opening many windows. Or if your GPU just doesn't have a lot of RAM.
2. Maybe your GPU isn't as power efficient as using the CPU for rendering the windows. Battery life could be affected.
3. Windows 7 with driver model 1.1 uses a constant amount of GPU ram for any amount of windows (steaming in textures instead of keeping them loaded). It also re-enables GDI 2D HW acceleration which was disabled in Vista, but available in XP. Windows 7 also accelerates Cleartype text.
From the pictures, this device appears to be a made up of a ARM board with lots of usb ports. Internally there are two usb ports, and if I'm right that's a wireless usb dongle, and a SD card in a usb dongle for storage. The demo device also has a Dell keyboard, the same as the one on my 4 year old Inspiron 700m. What this guy did is put together a compelling package using various parts as a demo, AFAIK. I wonder if laptop manufactures share a lot of parts? (besides the intel centrino chipset)
I meant 1500GB.
it's getting even worse by crossing the 1.0TB barrier. Not only are they using the base10 to 2 conversion to cheat you out of space, but they are advertising this drive as 1.5TB when it's actiually 1500TB. That's 36GB missing. Still though his drive was worth it for it's low GB per dollar ratio.