Being an HTPC, you would be able to configure this? I'm confused as to why an undesired conversion would take place, given that HTPCs tend to be custom built by the user.
You'd think this, but it isn't true. You can choose YCbCr from the video driver on some solutions, but testing shows what it is actually doing is converting from YCbCr to RGB for the framebuffer and then back to YCbCr for output! (See this thread for some additional information.) This, of course, is a lossy process since 8-bit quantization is used for all these values. There is no way to tell a HTPC to keep its dirty hands off your signal.
Granted, it is used on most current BD releases. However, "most" is not "all." To give one example, my "T2 Judgment Day: Skynet Edition" Blu-Ray disc is encoded in VC-1. This can easily be verified by ripping the original disc with AnyDVD HD or a similar program and then examining the resulting stream. So if I wanted to watch this on the Raspberry Pi, I would have to transcode it to H.264, losing quality along the way.
This article also confirms that the Raspberry Pi will likely be worthless as a media streaming device - at least for those of us who care about video quality.
It is now official that only H.264 (and MPEG-4) will be supported for hardware decoding. That means that playback of ripped Blu-rays that use VC-1 and MPEG-2 (not an inconsiderable number) is impossible, since the weak ARM CPU will almost certainly not be able to decode them in real-time at 1080p.
I'm sure the Pi will work fine for people who play back transcoded crap downloaded from TPB, but for anyone who actually cares about video quality, the lack of these essential codecs is a death sentence. We can only hope that at some point in the future a different (even if more expensive) model will ship with them enabled.
It's really a shame that there is currently no open-source-friendly SoC platform that supports HD video decoding and HDMI 1.3 audio bitstreaming in all its forms. HTPCs mess with the signal in all kinds of ways (YUV->RGB conversion is forced, even if you select YUV, it converts to RGB then converts back) so SoCs are really all that can provide decent quality. And the firmware on commercial media streamers is almost all crap.
I have several observations to make here. First, there are a lot of people who admire European trains, but have no idea how those are paid for. Sure, it'd be nice to have a US train paid for by European taxpayers, like how European trains are funded, but it's a wee bit unrealistic. So then the US would be stuck paying for US trains with hapless US taxpayers. That changes the US-oriented cost/benefit for such projects.
Then maybe US citizens need to start adopting a more reasonable attitude towards taxes. It's worth pointing out that the "no taxes EVAR!" crap is fairly recent, dating back only to the 1970s. Before that, taxes for infrastructure spending were fairly noncontroversial.
Throwing less money down the military rathole would also help US citizens get a better deal on their tax expenditures.
Except that will never happen. The important difference is that automakers don't want autodrive cars. It would mean dramatically fewer cars sold because individuals wouldn't own cars anymore. It's stupid to pay tens of thousands of dollars for something that sits idle 90% of the time. But cars that can drive themselves never need to be idle, they can constantly be picking people up and dropping them off. They can be busy 90% of the time. Which means that there only needs to be 10% cars in the world.
But most people don't want that. They want their own personal car, not a car shared with random strangers. You lose a lot of utility by sharing a vehicle; you can no longer keep your personal property inside, you can no longer customize the vehicle's interior to your liking, it's no longer available as you need it 24/7, and so forth.
You'd need a convincing cop car, and that isn't something you can just go out and buy.
Sure you can. A search on eBay for 'crown victoria police interceptor' currently lists 42 vehicles (plus thousands of accessories). Many of them still have the "push bar" on the front, and the original B&W paint job. All the malicious user would have to do would be to slap a "POLICE" sticker on the side, and a siren on the roof.
The issue is licensing the patents from MPEG-LA. The cost of licensing a codec is too high to license them all for every Pi sold: the cost of licensing AAC alone is 4% of the total price of the board. So, either there will be a hardware version that comes with all the codecs (and costs a lot more), or there will a software codec pack that you can pay to download.
The problem is that the Raspberry Pi's relatively weak ARM-based CPU is almost certainly not powerful enough to decode high-bitrate 1080p VC-1 or MPEG-2 without dropping frames. Hardware decoding is the only viable option. If that's fused off, then the Pi's usefulness as a media streamer drops dramatically.
I'm not sure why they bothered paying for an AAC license, since that almost certainly can be decoded in real-time by software, and an open-source implementation (ffmpeg) is readily available.
If the base-model Pi doesn't support VC-1 and MPEG-2, hopefully they'll consider releasing a more expensive version that supports all the codecs the SoC contains implementation for. Even if it shot the price up to the $75-$100 range it would still be cheaper than any decent dedicated streamer (or HTPC).
I know that the Raspberry Pi is specifically advertised as supporting hardware decoding of H.264 up to 1080p30 at up to 40 Mbps. What I want to know is if it also supports VC-1 and MPEG-2 decoding at the same resolutions and data rates. I know that the underlying SoC has this capability, but will it be blocked or omitted from the SDK for licensing/patent reasons? Any of these three codecs can be found on Blu-Rays, and transcoding the rips to H.264 would reduce the quality.
Also, what about bitstreaming the HD audio codecs (Dolby TrueHD and DTS-HD MA) over HDMI 1.3? I know Raspberry Pi didn't want to pay for audio decoding licenses, but simply sending the raw bitstream to a receiver over the HDMI link shouldn't present any licensing issues (and is the best quality method to use anyway).
For the Raspberry Pi to be a good media streamer, it needs to be able to do these things.
at what point did the police get the power to randomly stop someone to frisk them
The 1968 Supreme Court case of Terry v. Ohio allows the police to stop and frisk suspects upon "reasonable suspicion." Of course, this is a vaguely defined term, and in practice means whatever the police want it to mean.
Since both CoW and ZFS's copy work a lot like RAID0 (as far as I can tell), I'd expect them to be similar in this respect, however ZFS also does checksum tests and NTFS doesn't BUT I don't know if ReFS will or not.
As mentioned previously, one of our design goals was to detect and correct corruption. This not only ensures data integrity, but also improves system availability and online operation. Thus, all ReFS metadata is check-summed at the level of a B+ tree page, and the checksum is stored independently from the page itself. This allows us to detect all forms of disk corruption, including lost and misdirected writes and bit rot (degradation of data on the media). In addition, we have added an option where the contents of a file are check-summed as well.
This is definitely good news; for too long ZFS has been the only file system to take data integrity seriously, and it's nice to see more options start to become available.
Blue-collar workers get paid more in Alaska than in most other states (in large part because so few people want to live there!) Also, the oil reserves are the property of the state, so not only are there no state taxes, but residents actually get a rebate check from the Permanent Fund Dividend (it was $1,174 per person last year).
The reason we hand out drivers' licenses at the drop of a hat is that in most parts of the US, driving is a necessary part of being an independent adult. That is due to the physical layout of our country and is not going to change in the forseeable future. What will happen eventually is that automatic driving systems like the ones currently being tested by Google and other companies will become ready for public consumption, and by a decade or so after their initial release, all new cars will have them. Then we can change the laws to make it much harder to get a license allowing manual driving, while everyone else (including children and drunks) can use the automated driving system to get from Point A to Point B.
These days, most webmasters have stopped caring if their sites look good in IE6. It is IE8 that is currently the lowest common denominator of the Web. Microsoft's failure to port their modern browsers to Windows XP means that we are basically stuck not being able to use CSS3 and other advanced HTML features until after 2014.
Other then microsoft's bogus tests where they specifically rig it to be 99.9%, every test I have seen has shown IE9 to meet 40-60% of html 5, while chrome and FF 80-100%.
I'm going to have to call [citation needed] on this one. All the reviews I've seen from non-MS sources (e.g. Tom's Hardware browser sweepstakes) indicate that IE9 has good compatibility with modern HTML5 features.
Atoms are the weakest x86 chips currently on the market, by a substantial margin. The 32nm refresh won't change this, since the core isn't changing and clock speeds aren't being increased much if at all. It will, at least, fix the lack of 1080p video support. Still, it's hard to justify an Atom processor for even the most casual user when AMD Bobcat E-350 motherboards are available for $100 or so from good brands. And for a few tens of dollars more than that, you can get a decent if plain micro-ATX H61 motherboard and Sandy Bridge Celeron that beats the Bobcat by a wide margin. You pay a bit more money for a massive jump in performance, and Atoms are so weak that even casual users will see performance problems when trying to browse sites with heavy JavaScript, edit photos, play simple Flash games, or perform other basic tasks.
At $80-$100, Atom motherboards just don't make any sense. Intel needs to be selling the things at Raspberry Pi price levels – $25 or $35. With their manufacturing tech, they could still make a profit at that level, considering how tiny the die size is.
Because if they overcharge you, or keep billing you after it was supposed to be cancelled, you can easily do a chargeback with credit cards. You can't do that with your bank account.
So, are Google staff going through and flagging these manually, or do they have some kind of algorithm to detect whether a photo has a raised middle finger in it? I wouldn't be surprised if it was the latter, considering this is Google we're talking about. It opens up some interesting new Image Search possibilities...
Maybe the problem is that 5000 employees are just too many to supervise effectively in a monolithic fashion. Isn't there a management practice that treats one larger company as an array of smaller companies for organizational purposes?
(Incidentally, 6 people is the size of my IT department, not the whole library! We have about 100 employees in all, give or take a few.)
As a wise man once said, with great power comes great responsibility.
If we want the power to say "No" to users who are doing unsecure things, we have the corresponding responsibility to provide an easy-to-use substitute in a reasonable time frame.
Once everyone else starts seeing IT as "the department of no," or as unapproachable "high priests" (as a previous article said), the clock is ticking. Other employees now perceive IT as the enemy and will try to work around us by whatever means they can. And if these enemies include upper management, the outsourcing of the IT department won't be far behind.
I work as a Database/Web Administrator in a small (6-person) IT department in a public library system. Until about 6 months ago, I was doing general IT support, and still do from time to time; we're not hung up on formal job descriptions too much with a department this small. Do we sometimes advise people not to do things for security reasons? Yes. We've had to prohibit a handful of specific bad practices (generic logins) because of PCI compliance. But this is not the primary focus of our work. The primary focus of our work is helping other people to do their work more effectively. And this means providing solutions, not withholding them. It means if someone wants to do something insecure, we try to find out WHY they want to do it, and come up with a way to make things as convenient for them as possible. I have personally written multiple scripts to make peoples' jobs easier. (Example: on one occasion, I noticed that staff were manually running circulation totals from self-check units each morning. So I offered to automate this process, which saves them 5-10 minutes a day.) Because everyone knows us, and knows we will do what we can to help them, we have the credibility to draw the line where it matters. Many IT departments have forfeited this credibility, or never had it in the first place. IT should be an important part of the business, a strategic partner with a voice at the table - not a bunch of antisocial BOFHs in the back room.
Unfortunately, the damage goes beyond that. You can't effectively use GCC (MinGW) to build most Windows applications, not only because these applications are full of Visual C++isms, but also because many newer APIs (notably Direct2D and DirectWrite) are not currently supported under MinGW at all.
Slashdot Mirror
RGB to YUV is lossless in both directions.
Only if you're working in infinite-precision floating point. In the real world, this is a lossy conversion.
Being an HTPC, you would be able to configure this? I'm confused as to why an undesired conversion would take place, given that HTPCs tend to be custom built by the user.
You'd think this, but it isn't true. You can choose YCbCr from the video driver on some solutions, but testing shows what it is actually doing is converting from YCbCr to RGB for the framebuffer and then back to YCbCr for output! (See this thread for some additional information.) This, of course, is a lossy process since 8-bit quantization is used for all these values. There is no way to tell a HTPC to keep its dirty hands off your signal.
You DO realise that Blu-ray's are encoded in h.264 High-profile, right?
You DO realise that H.264 is only one of the three supported codecs under the Blu-Ray standard, right?
Granted, it is used on most current BD releases. However, "most" is not "all." To give one example, my "T2 Judgment Day: Skynet Edition" Blu-Ray disc is encoded in VC-1. This can easily be verified by ripping the original disc with AnyDVD HD or a similar program and then examining the resulting stream. So if I wanted to watch this on the Raspberry Pi, I would have to transcode it to H.264, losing quality along the way.
This article also confirms that the Raspberry Pi will likely be worthless as a media streaming device - at least for those of us who care about video quality.
It is now official that only H.264 (and MPEG-4) will be supported for hardware decoding. That means that playback of ripped Blu-rays that use VC-1 and MPEG-2 (not an inconsiderable number) is impossible, since the weak ARM CPU will almost certainly not be able to decode them in real-time at 1080p.
I'm sure the Pi will work fine for people who play back transcoded crap downloaded from TPB, but for anyone who actually cares about video quality, the lack of these essential codecs is a death sentence. We can only hope that at some point in the future a different (even if more expensive) model will ship with them enabled.
It's really a shame that there is currently no open-source-friendly SoC platform that supports HD video decoding and HDMI 1.3 audio bitstreaming in all its forms. HTPCs mess with the signal in all kinds of ways (YUV->RGB conversion is forced, even if you select YUV, it converts to RGB then converts back) so SoCs are really all that can provide decent quality. And the firmware on commercial media streamers is almost all crap.
That's the primary reason that Wisconsin and Florida backed out of high speed rail projects.
The primary reason Wisconsin and Florida backed out of high-speed rail projects is that they elected right-wing crackpots in 2010.
I have several observations to make here. First, there are a lot of people who admire European trains, but have no idea how those are paid for. Sure, it'd be nice to have a US train paid for by European taxpayers, like how European trains are funded, but it's a wee bit unrealistic. So then the US would be stuck paying for US trains with hapless US taxpayers. That changes the US-oriented cost/benefit for such projects.
Then maybe US citizens need to start adopting a more reasonable attitude towards taxes. It's worth pointing out that the "no taxes EVAR!" crap is fairly recent, dating back only to the 1970s. Before that, taxes for infrastructure spending were fairly noncontroversial.
Throwing less money down the military rathole would also help US citizens get a better deal on their tax expenditures.
Except that will never happen. The important difference is that automakers don't want autodrive cars. It would mean dramatically fewer cars sold because individuals wouldn't own cars anymore. It's stupid to pay tens of thousands of dollars for something that sits idle 90% of the time. But cars that can drive themselves never need to be idle, they can constantly be picking people up and dropping them off. They can be busy 90% of the time. Which means that there only needs to be 10% cars in the world.
But most people don't want that. They want their own personal car, not a car shared with random strangers. You lose a lot of utility by sharing a vehicle; you can no longer keep your personal property inside, you can no longer customize the vehicle's interior to your liking, it's no longer available as you need it 24/7, and so forth.
You'd need a convincing cop car, and that isn't something you can just go out and buy.
Sure you can. A search on eBay for 'crown victoria police interceptor' currently lists 42 vehicles (plus thousands of accessories). Many of them still have the "push bar" on the front, and the original B&W paint job. All the malicious user would have to do would be to slap a "POLICE" sticker on the side, and a siren on the roof.
The issue is licensing the patents from MPEG-LA. The cost of licensing a codec is too high to license them all for every Pi sold: the cost of licensing AAC alone is 4% of the total price of the board. So, either there will be a hardware version that comes with all the codecs (and costs a lot more), or there will a software codec pack that you can pay to download.
The problem is that the Raspberry Pi's relatively weak ARM-based CPU is almost certainly not powerful enough to decode high-bitrate 1080p VC-1 or MPEG-2 without dropping frames. Hardware decoding is the only viable option. If that's fused off, then the Pi's usefulness as a media streamer drops dramatically.
I'm not sure why they bothered paying for an AAC license, since that almost certainly can be decoded in real-time by software, and an open-source implementation (ffmpeg) is readily available.
If the base-model Pi doesn't support VC-1 and MPEG-2, hopefully they'll consider releasing a more expensive version that supports all the codecs the SoC contains implementation for. Even if it shot the price up to the $75-$100 range it would still be cheaper than any decent dedicated streamer (or HTPC).
I know that the Raspberry Pi is specifically advertised as supporting hardware decoding of H.264 up to 1080p30 at up to 40 Mbps. What I want to know is if it also supports VC-1 and MPEG-2 decoding at the same resolutions and data rates. I know that the underlying SoC has this capability, but will it be blocked or omitted from the SDK for licensing/patent reasons? Any of these three codecs can be found on Blu-Rays, and transcoding the rips to H.264 would reduce the quality.
Also, what about bitstreaming the HD audio codecs (Dolby TrueHD and DTS-HD MA) over HDMI 1.3? I know Raspberry Pi didn't want to pay for audio decoding licenses, but simply sending the raw bitstream to a receiver over the HDMI link shouldn't present any licensing issues (and is the best quality method to use anyway).
For the Raspberry Pi to be a good media streamer, it needs to be able to do these things.
"I am just a businessman, giving the people what they want,"
"All I do is satisfy a public demand."
Both are quotes from Al Capone
Was he wrong? After all, we repealed Prohibition in 1933 precisely because the public demand in question created people like Al Capone.
Changing file names to gibberish would also annoy and inconvenience legitimate users. Why should people have to rename after each download?
at what point did the police get the power to randomly stop someone to frisk them
The 1968 Supreme Court case of Terry v. Ohio allows the police to stop and frisk suspects upon "reasonable suspicion." Of course, this is a vaguely defined term, and in practice means whatever the police want it to mean.
Since both CoW and ZFS's copy work a lot like RAID0 (as far as I can tell), I'd expect them to be similar in this respect, however ZFS also does checksum tests and NTFS doesn't BUT I don't know if ReFS will or not.
According to the original Building Windows 8 post, checksums are indeed supported:
This is definitely good news; for too long ZFS has been the only file system to take data integrity seriously, and it's nice to see more options start to become available.
Blue-collar workers get paid more in Alaska than in most other states (in large part because so few people want to live there!) Also, the oil reserves are the property of the state, so not only are there no state taxes, but residents actually get a rebate check from the Permanent Fund Dividend (it was $1,174 per person last year).
The reason we hand out drivers' licenses at the drop of a hat is that in most parts of the US, driving is a necessary part of being an independent adult. That is due to the physical layout of our country and is not going to change in the forseeable future. What will happen eventually is that automatic driving systems like the ones currently being tested by Google and other companies will become ready for public consumption, and by a decade or so after their initial release, all new cars will have them. Then we can change the laws to make it much harder to get a license allowing manual driving, while everyone else (including children and drunks) can use the automated driving system to get from Point A to Point B.
Then the vendors of that software need to be kicked in the ass hard.
These days, most webmasters have stopped caring if their sites look good in IE6. It is IE8 that is currently the lowest common denominator of the Web. Microsoft's failure to port their modern browsers to Windows XP means that we are basically stuck not being able to use CSS3 and other advanced HTML features until after 2014.
Other then microsoft's bogus tests where they specifically rig it to be 99.9%, every test I have seen has shown IE9 to meet 40-60% of html 5, while chrome and FF 80-100%.
I'm going to have to call [citation needed] on this one. All the reviews I've seen from non-MS sources (e.g. Tom's Hardware browser sweepstakes) indicate that IE9 has good compatibility with modern HTML5 features.
Atoms are the weakest x86 chips currently on the market, by a substantial margin. The 32nm refresh won't change this, since the core isn't changing and clock speeds aren't being increased much if at all. It will, at least, fix the lack of 1080p video support. Still, it's hard to justify an Atom processor for even the most casual user when AMD Bobcat E-350 motherboards are available for $100 or so from good brands. And for a few tens of dollars more than that, you can get a decent if plain micro-ATX H61 motherboard and Sandy Bridge Celeron that beats the Bobcat by a wide margin. You pay a bit more money for a massive jump in performance, and Atoms are so weak that even casual users will see performance problems when trying to browse sites with heavy JavaScript, edit photos, play simple Flash games, or perform other basic tasks.
At $80-$100, Atom motherboards just don't make any sense. Intel needs to be selling the things at Raspberry Pi price levels – $25 or $35. With their manufacturing tech, they could still make a profit at that level, considering how tiny the die size is.
Why are you even using credit cards to pay bills?
Because if they overcharge you, or keep billing you after it was supposed to be cancelled, you can easily do a chargeback with credit cards. You can't do that with your bank account.
So, are Google staff going through and flagging these manually, or do they have some kind of algorithm to detect whether a photo has a raised middle finger in it? I wouldn't be surprised if it was the latter, considering this is Google we're talking about. It opens up some interesting new Image Search possibilities...
Maybe the problem is that 5000 employees are just too many to supervise effectively in a monolithic fashion. Isn't there a management practice that treats one larger company as an array of smaller companies for organizational purposes?
(Incidentally, 6 people is the size of my IT department, not the whole library! We have about 100 employees in all, give or take a few.)
As a wise man once said, with great power comes great responsibility.
If we want the power to say "No" to users who are doing unsecure things, we have the corresponding responsibility to provide an easy-to-use substitute in a reasonable time frame.
Once everyone else starts seeing IT as "the department of no," or as unapproachable "high priests" (as a previous article said), the clock is ticking. Other employees now perceive IT as the enemy and will try to work around us by whatever means they can. And if these enemies include upper management, the outsourcing of the IT department won't be far behind.
I work as a Database/Web Administrator in a small (6-person) IT department in a public library system. Until about 6 months ago, I was doing general IT support, and still do from time to time; we're not hung up on formal job descriptions too much with a department this small. Do we sometimes advise people not to do things for security reasons? Yes. We've had to prohibit a handful of specific bad practices (generic logins) because of PCI compliance. But this is not the primary focus of our work. The primary focus of our work is helping other people to do their work more effectively. And this means providing solutions, not withholding them. It means if someone wants to do something insecure, we try to find out WHY they want to do it, and come up with a way to make things as convenient for them as possible. I have personally written multiple scripts to make peoples' jobs easier. (Example: on one occasion, I noticed that staff were manually running circulation totals from self-check units each morning. So I offered to automate this process, which saves them 5-10 minutes a day.) Because everyone knows us, and knows we will do what we can to help them, we have the credibility to draw the line where it matters. Many IT departments have forfeited this credibility, or never had it in the first place. IT should be an important part of the business, a strategic partner with a voice at the table - not a bunch of antisocial BOFHs in the back room.
Unfortunately, the damage goes beyond that. You can't effectively use GCC (MinGW) to build most Windows applications, not only because these applications are full of Visual C++isms, but also because many newer APIs (notably Direct2D and DirectWrite) are not currently supported under MinGW at all.