Set-top boxes may be used for streaming video, but they are irrelevant to a discussion of web standards for streaming video, because the set-top boxes will be their own closed little world and no-one will expect them to display web pages.
No. Content providers want to minimize the number of copies of videos.
Bullshit. One does not need a web browser to stream video. Many set top boxes with streaming capability have their own non-HTML interfaces for streaming videos; even the iPhone has a separate YouTube application instead of an HTML interface with HTML video.
And these existing devices already stream H.264 and VC-1 without a web browser.
The amount of features isn't important. What's important is if they're useful. And it would seem that many features in H.264 (like 8x16 blocks) are barely used/useful in practice.
But many features in H.264 are used in H.264 implementations today, giving it an advantage.
In fact, they could have used Ogg Vorbis, whose bitstream was fixed as of 2000 and whose encoder was finalised in 2002 - AAC is only just getting to Vorbis levels of quality.
You mean people should stop supporting things like mp3 just because you personally think it's too old?
Not only because it's old; because it's old and there are newer codes that are better. This is why, for example, Apple uses AAC, even though they could have just used MP3 in a DRM container (back when they still applied DRM to downloads).
New video codes have brought clearly perceptible improvements, which is why we've seen MPEG-2, MPEG-4 ASP, and MPEG-4 Part 10 AVC/H.264 within the past 15 years (and H.265 coming in a few years too). On the audio side, MP3 (and Dolby Digital/AC3) still see much use because they are both 'good enough' for what they are used for, but that doesn't mean we should avoid all progress.
Theora is perhaps better than H.263 and MPEG-2 (from the mid 90s), but does not come close to H.264/MPEG-4 AVC or VC-1. (The frozen Theora bitstream format is lacking many features found in H.264 and VC-1.) Results might be similar to H.263+/MPEG-4 ASP.
Note that there are many sites which perform misleading or flawed comparisons of the two; for example, they might compare the result from YouTube's H.264 encoder with a lossy source (which optimizes for encoding speed) to a locally ran Theora encode with a lossless source.
Since OS X 10.6 and Windows 7 come with H.264 decoding, and Windows 7 supports H.264 hardware decoding with compatible hardware from any source, I recommend sticking with H.264. (OS X 10.6's H.264 hardware decoding support appears to be limited to videos played in QuickTime X from MPEG4 or QuickTime container files on systems with nVidia 9400M GPUs or newer, even though Macs with capable GPUs started appearing in 2007.)
Or the result of C++ being fundamentally difficult on the compiler?
Some valid C++ code indeed requires more work to compile (if it can be compiled at all), but Qt? It does not use features like exceptions or templates...
And GCC C++ is known to not be the fastest C++ compiler...
I'm kinda joking, but the compilation time for complex kde stuff is humongous. It's one of the few things I quite like about Gnome stuff - it's pure C and is real rapid to build.
Is that the fault of KDE/Qt or is it a result of a slow C++ compiler?
Right... so there's absolutely no chance of any possible benefits whatsoever?
Perhaps if you are encrypting a single stream at a very high rate. But a server will be more likely serving many streams, which can already be performed on different cores.
Impossible that there could be any context switching overhead savings reachable, impossible that correct scheduling could reduce stalls in tight running loops by doing heavy cacheline busting processing elsewhere? Even with the cost of some latency, there's -no- way to improve throughput by parallel processing it? What if there was (somehow) nothing else running on the system/at all/?
Unless your time is worthless, it's probably cheaper to buy more CPUs.
Depends what else the system is doing, surely. Having several clients connected to a server, you want to free up the processor as much as possible for servery duties, any savings you make go to those when running full pelt, or convert to energy savings when you're not.
What does that have to do with anything? AES-NI enables AES implementations to encrypt and decrypt much faster than any network and storage system I've ever used can provide data. Parallelizing AES-NI would save energy only in the case when running multiple cores at partial load is more efficient than running just one core at near-full load (which could be the case if all cores run at the same clock speed), and would be faster only if your I/O systems are already faster than a single core can encrypt/decrypt with AES-NI.
Well, the Intel AES instructions would benefit even more from parallelized AES CTR mode pre-computation than straight multiple cores, so that doesn't invalidate what I'm saying at all.:-)
SunOS is SVR5 with customizations. What is SVR5? It's a kernel, a libc, and a set of applications and other associated libraries.
Solaris was based on SVR4, but that was almost two decades ago. Like modern BSD-derived systems, quite a bit has been rewritten since.
dtrace would require changes all over the kernel
Linux has systemtap. But dtrace has been incorporated to other operating systems already (such as Darwin)
zones' functionality is provided by a combination of KVM and colinux, depending on your particular goal.
The functionality provided by Zones is more like that provided by the OpenVZ patch to Linux or FreeBSD jails. KVM and coLinux run entire Linux kernels as guests, incurring significant overhead (over Zones, OpenVZ, or jails).
If you compare these numbers, you see that the US sees about 9 deaths per billion kilometers, and the UK sees 6.3 deaths.
I would be interested in seeing the US statistic broken down further - is there a significant difference between areas of the country with different weather, for example?
I honestly don't know how the US expects to maintain an adequate presences in STEM if the basic mathematics curriculum is so deficient. The notion of an able student of 18 completing 12 years of formal education without being able to differentiate seems very odd to me.
The able students generally do indeed complete 12 years of formal education knowing how to differentiate.
The "Walled Gardens" of the 1990s (AOL, CompuServe, The Microsoft Network, etc) were just value-added content layers on top of services provided by the Internet and all included access to the World Wide Web.
Except for perhaps MSN, these services included access to only parts of the Internet. CompuServe added Internet email access earlier than the others in 1989, and AOL added Usenet in 1993. Prodigy added a web browser (no sockets support) in 1994.
I don't think these services started offering real Internet (with TCP sockets support) until after the release of Windows 95.
Much the same reason IPv6 wasn't added to Windows until Vista even though IPv6's specifications were stable enough by the release of XP SP2 in 2005.
IPv6 wasn't enabled by default until Vista, but was included with XP from the beginning. (The version included with the original XP release was included as an unsupported preview.) MS also released experimental IPv6 implementations for NT4 and 2000.
Slashdot Mirror
No. Content providers want to minimize the number of copies of videos.
Bullshit. One does not need a web browser to stream video. Many set top boxes with streaming capability have their own non-HTML interfaces for streaming videos; even the iPhone has a separate YouTube application instead of an HTML interface with HTML video.
And these existing devices already stream H.264 and VC-1 without a web browser.
And each core is about four times slower than my 2.53 GHz Core 2 Duo. No thanks!
$50 per post, $10 per reply. Standard rates.
But many features in H.264 are used in H.264 implementations today, giving it an advantage.
Early tests show Apple AAC to be comparable to Vorbis.
No. codec/encoder/decoder refers to a specific implementation or implementations. "code" here refers to the encoding format itself.
Not only because it's old; because it's old and there are newer codes that are better. This is why, for example, Apple uses AAC, even though they could have just used MP3 in a DRM container (back when they still applied DRM to downloads).
New video codes have brought clearly perceptible improvements, which is why we've seen MPEG-2, MPEG-4 ASP, and MPEG-4 Part 10 AVC/H.264 within the past 15 years (and H.265 coming in a few years too).
On the audio side, MP3 (and Dolby Digital/AC3) still see much use because they are both 'good enough' for what they are used for, but that doesn't mean we should avoid all progress.
Theora is perhaps better than H.263 and MPEG-2 (from the mid 90s), but does not come close to H.264/MPEG-4 AVC or VC-1. (The frozen Theora bitstream format is lacking many features found in H.264 and VC-1.) Results might be similar to H.263+/MPEG-4 ASP.
The Ogg container also has some documented flaws.
Note that there are many sites which perform misleading or flawed comparisons of the two; for example, they might compare the result from YouTube's H.264 encoder with a lossy source (which optimizes for encoding speed) to a locally ran Theora encode with a lossless source.
Since OS X 10.6 and Windows 7 come with H.264 decoding, and Windows 7 supports H.264 hardware decoding with compatible hardware from any source, I recommend sticking with H.264. (OS X 10.6's H.264 hardware decoding support appears to be limited to videos played in QuickTime X from MPEG4 or QuickTime container files on systems with nVidia 9400M GPUs or newer, even though Macs with capable GPUs started appearing in 2007.)
Some valid C++ code indeed requires more work to compile (if it can be compiled at all), but Qt? It does not use features like exceptions or templates...
And GCC C++ is known to not be the fastest C++ compiler...
Is that the fault of KDE/Qt or is it a result of a slow C++ compiler?
Perhaps if you are encrypting a single stream at a very high rate. But a server will be more likely serving many streams, which can already be performed on different cores.
Unless your time is worthless, it's probably cheaper to buy more CPUs.
Sucks for you.
What does that have to do with anything? AES-NI enables AES implementations to encrypt and decrypt much faster than any network and storage system I've ever used can provide data. Parallelizing AES-NI would save energy only in the case when running multiple cores at partial load is more efficient than running just one core at near-full load (which could be the case if all cores run at the same clock speed), and would be faster only if your I/O systems are already faster than a single core can encrypt/decrypt with AES-NI.
Are your storage and network devices that fast?
If you want faster AES, just upgrade your CPU.
It's not that difficult to see the throttle position sensor values, via OBD-II with something like a ScanGauge.
My car's ECU reports the TPS values, and I don't think it even has electronic throttle control.
Slashdot's not using Arial, it's using the browser's default sans-serif font. For me, this is Helvetica.
Solaris was based on SVR4, but that was almost two decades ago. Like modern BSD-derived systems, quite a bit has been rewritten since.
Linux has systemtap. But dtrace has been incorporated to other operating systems already (such as Darwin)
The functionality provided by Zones is more like that provided by the OpenVZ patch to Linux or FreeBSD jails. KVM and coLinux run entire Linux kernels as guests, incurring significant overhead (over Zones, OpenVZ, or jails).
I would be interested in seeing the US statistic broken down further - is there a significant difference between areas of the country with different weather, for example?
Realtek hardware generally does not have the advanced hardware features found in the fancier Intel e1000(e) and Broadcom tg3.
The able students generally do indeed complete 12 years of formal education knowing how to differentiate.
Just add more error correction coding.
Except for perhaps MSN, these services included access to only parts of the Internet. CompuServe added Internet email access earlier than the others in 1989, and AOL added Usenet in 1993. Prodigy added a web browser (no sockets support) in 1994.
I don't think these services started offering real Internet (with TCP sockets support) until after the release of Windows 95.
IPv6 wasn't enabled by default until Vista, but was included with XP from the beginning. (The version included with the original XP release was included as an unsupported preview.) MS also released experimental IPv6 implementations for NT4 and 2000.
Correction: the iPod Touch runs the same software with the same restrictions