In the aftermath of gas drilling micro-disasters (the nature of gas drilling results in localized environmental damage, but when it happens it is a disaster for those nearby), I'm guessing increasing regulation is going to increase the costs of gas drilling.
There's a moratorium on shale gas drilling (specifically on well stimulation by hydrofracturing, but no one is going to drill a well they can't frack) in New York State after the rampant water contamination incidents all over Pennsylvania. For example, the groundwater in Dimock, PA became undrinkable within a year or so of the commencement of drilling. People can actually light their tap water on fire now.
Gas is not a long-term option, and in fact, it looks like the way it is being drilled now is going to have severe long-term environmental consequences (it already has in many drilling areas). Nuclear is a long-term investment.
I'm referring to UMTS1700 support. Only a few carriers around the world support UMTS1700.
(UMTS1700 is the main reason why manufacturers now have to make phones with different designs for TMo and AT&T - I have yet to see a single phone that supports 3G on both T-Mobile and AT&T.)
This is, in my opinion, why the optical networking industry crashed a few years ago. There are massive amounts of dark fiber along nearly every backbone connection (as the material costs of the fiber are far less than the costs of laying the fiber), and there wasn't enough last-mile capacity to keep all these super-backbones full.
As the gap between the backbone and the last mile increased, it became harder and harder to sell these new "omg-crazy-bandwidth-over-a-single-fiber" solutions.
Rumor has it they're possibly switching to Qualcomm baseband processor chipsets for the iPhone 5 for a variety of reasons. Note that Qualcomm does produce GSM/UMTS baseband chips, in fact a large portion of the GSM smartphone market uses various Qualcomm MSM silicon.
Once you're using Qualcomm MSM silicon for your GSM/UMTS solution, it's not too difficult to swap in the CDMA2000 version of that chip and get Sprint/Verizon support. Look at the large number of HTC GSM phones that have CDMA variants. (As to licensing fees - no matter what Qualcomm gets paid, and if you use their MSM chipsets you probably don't have to pay them patent royalties.)
While the CPU in the iPhone is an "Apple spec" CPU, most likely the radio chipset is from a third party.
Most such third parties offer roughly equivalent GSM and CDMA chipsets (In fact, probably 50%+ of the GSM/UMTS smartphone market is powered by Qualcomm chipsets, despite Qualcomm being the creator and backer of cdmaOne/CDMA2000), making it not too difficult to make a CDMA version of a phone and a GSM version.
Also, you show a fundamental lack of understanding of RF and communications engineering. New band = new antenna. Same bands but different protocol = no new antenna, just a different baseband processor. Therefore, adding T-Mobile's additional band is the least likely route for Apple to go.
Changing the iPhone from GSM/UMTS to CDMA2000 (note: UMTS uses a CDMA modulation scheme, one of the reasons Qualcomm is involved in that market despite them backing a competing protocol/air interface suite - they're good at making CDMA chipsets whether it is the UMTS protocol/air interface suite or CDMA2000) most likely entails: 1) Switch out the baseband processor chip. Note that as I said before, nearly all UMTS chipset providers also provide CDMA2000 chipsets that are very similar and require little engineering 2) Remove the now unneeded SIM slot (what was that about a packed device?)
That's about it
Note: Right now, Apple gets their radio chipsets from Infineon, but there is a lot of evidence pointing to them moving to Qualcomm for the 5th gen iPhone, with the ability to produce a CDMA version being one of a number of reasons, Infineon getting purchased by Intel being a bigger reason - see http://touchreviews.net/iphone-5-qualcomm-radio-chip/ - I don't know about Infineon chipsets, but it is VERY easy to make both a CDMA2000 and GSM/UMTS version of a phone if you use Qualcomm for your radio chipsets, which is why a large portion of HTC's GSM product line have near identical CDMA equivalents.
I don't think it means anything for Android. In fact, it could be good for Android.
It seems like AT&T is doing everything they can to milk the iPhone exclusivity and not piss off Apple enough to start selling iPhones through other carriers. Just look at AT&T's piss poor rollout of Android-based handsets - One of the linked articles links to another article about three new Android phones coming out for AT&T from Motorola. Note that every single one of these "new" phones lags at least 4-6 months behind Verizon's Android offerings, and in terms of hardware specs, behind even some of AT&T's Windows Mobile offerings. (HVGA or QVGA screen? Are you kidding me? My Tilt 2 came out last October and has a full WVGA screen. Also note that nearly everyone including Motorola is moving to 1 GHz+ Snapdragon-based phones except these three low-speed offerings for AT&T.)
I'm beginning to be tempted to return to Verizon. I ditched Verizon for AT&T 3-4 years ago because I was sick of Verizon phones coming out 6-9 months after Sprint with numerous features removed (such as the Treo 650's lack of Bluetooth DUN on VZW). Now, a few years later, VZW has done a complete about-face (at least with Android) which completely shocked me, and AT&T is now doing the same crap Verizon did a few years ago in the name of milking the iPhone for all it's worth.
Once AT&T loses iPhone exclusivity, they're going to have to shape up with their Android offerings, otherwise Verizon will have both the iPhone and superior Android offerings for those that don't want an iPhone. Meanwhile, it is getting to the point where all AT&T has is the iPhone.
Actually this is the case for a lot of vehicles, usually related to accessibility of some plugs. As interior room for passengers becomes more heavily marketed, engine compartment space gets more and more cramped.
For example, my 1995 Chrysler LeBaron's front plugs took minutes to change, the rear plugs was an hour+ job due to having to remove the air intake to get at them. (I think it was a much shorter job if you had a lift and could access the vehicle from below.)
My parents' 1996 Dodge Caravan was awful - the dash of the vehicle overhangs a significant amount of the rear portion of the engine compartment. It's basically impossible to get at the rear bank of the spark plugs without coming in from below. (Again, probably a short job for a mechanic with a lift.)
I think my 2009 Subaru's plugs are similarly difficult to access - that engine compartment is REALLY cramped.
It's a pain in the ass, messy, results in you having to (in most cases) take the waste oil to a garage for proper disposal anyway.
To make it NOT a pain in the ass requires a lift (I mean a proper lift and not a jack), and the initial investment in that tool is more than the cost of lifetime oil changes for the car.
Oh, and keep in mind, most oil changes include the price of the oil (up to a point - unusually large engines that require lots of oil get charged extra) - I've managed to get oil changes for less than what I'd pay for the oil used in most stores.
Yup. I buy well price controlled stuff. I used to be willing to pay a little extra for the ability to return something instore, but at this point, their return policy sucks compared to most of the reputable online vendors (Amazon, Newegg, B&H, Adorama) - yes, in case you couldn't tell from the list, I shop for photographic equipment a lot.
For HDMI cables, pay the extra for the cables with 24AWG wires. The ultracheap ones use 28AWG wires. Even the 24AWG ones are only a few bucks for a 6 foot cable.
Honestly, this is one of the few things that Comcast does better than Time Warner. TW will just cut you off. They only will give you the time of the complaint filed against your IP, not the time of whatever traffic was generated, making it difficult to pin down the machine in question. (They gave me a time window that was while on vacation, leading me to suspect my file server/MythTV box had been rootkitted. It turns out that it took someone a few days to complain about the bot-generated spam - the Windows partition I only use for gaming on my desktop turned out to be the culprit.)
Google claims that the codecs they back (In this case, VP8) are patent-free.
But apparently Microsoft claimed the same thing about VC-1, and within a year, numerous patent holders stepped forward with claims that VC-1 infringed on a patent they held, and now there is a VC-1 patent pool.
So really, "Unencumbered by patents" really means "We don't think it's encumbered but no guarantees."
Problem is that, according to the analysis by the x264 developer (see the first independent analysis link), WebP is missing quite a few features that JPEG has, does not add any of the features JPEG is missing but people really want (like a lossy format that contains alpha capability - although admittedly, lossy compression of the alpha channel itself could cause some REALLY weird artifacts.)
It also, at least in the current state of the encoder, does not appear to perform any better than JPEG.
I think you greatly underestimate the advantage an ASIC has over a general purpose CPU (even the latest Intel CPUs with AES-NI) when it comes to crypto.
How about you RTFA: "The HDCP cipher is designed to be efficient when implemented in hardware, but it is terribly inefficient in software, primarily because it makes extensive use of bit operations. Our implementation uses bit-slicing to achieve high speeds by exploiting bit-level parallelism. We have created a few high-level routines to make it as easy as possible to implement HDCP, as shown in the following example. "
Same with Pentax (except for some of the entry-level Pentaxes, the upper level ones like the K10D/K20D/K-7/K-5 can use the onboard flash as a P-TTL master).
Whether or not the onboard flash contributes or not depends on a setting in the camera. Even when it's set to not contribute, it's more like "contribute at absolute minimum power". It's most obvious if there is a reflective surface in the scene that allows the onboard flash to be seen.
I'm actually shocked that the Canons which do have onboard flashes can't act as a master for wireless E-TTL. (From doing a little looking, at the very least the 7D supports this.)
Technically they are all infrared - the sensor on the slave flashes has peak sensitivity in the IR, and in most cases there's an IR-pass/visible-block filter over the sensor. However, there is enough IR content within a xenon strobe's output that in most situations (exception - stuff like the SU-800 and ST-E2), the camera sources the IR using pulses from the onboard flash.
I do hope the police officer involved does have disciplinary action. He pulled his gun prior to identification, and his vehicle did not appear to have any sort of police lights installed.
However, the person in question would have had an arrest record anyway for his reckless driving infraction. He's still going to get nailed hard by the courts, but at least now he's getting nailed for the right reasons.
WRONG. FTDI has built-in drivers for Linux but not most other OSes. It definately requires drivers in Windows.
The newer USB CDC ACM class, however, has built-in drivers on any modern OS, with the "worst case" situation being Windows where you need an INF (but not actual drivers) to tell Windows that that VID/PID is a CDC ACM device.
"So how is it that a new custom chipset with a new vendor id requires no drivers? Is there a standard for USB->serial that if you follow you don't need a driver? (and why do most cables not support it if there is?)"
As others have said - the CDC ACM class. I believe that class was standardized relatively recently in an attempt to address the fact that there are 23432743290 different USB-serial chipsets all of which required special drivers.
Not sure if XP has CDC-ACM support built in (might have been added in a service pack?), but in general, CDC-ACM is the "new kid on the block" and the legacy (PL230x and FTDI) approaches are still well entrenched.
The nice thing about the little popup flashes in most SLRs these days is that they can trigger a remote flash wirelessly.
(A modern take on the classic "optical slave" where the camera can actually control the remote flash's light output. Nikon calls this CLS, Canon calles it wireless E-TTL I think, Pentax calls it wireless P-TTL.)
Most SLRs can be put in a mode where the onboard flash is reduced so much in output that it doesn't contribute significantly to exposure at all. (It will still, however, contribute some.) It can also be allowed to contribute to exposure, causing it to act as a small fill flash.
One thing you didn't explicitly mention but have hinted at in past posts: Sometimes, instead of bringing a large reflective object (umbrella, softbox, etc), a photographer will use their environment. This is why highend flashes swivel/tilt - for example, to point the flash head at the ceiling. In this situation, the ceiling becomes a large diffuse light source. (Note: White ceiling required unless you're going for a funky effect. Black ceilings such as found at many concert venues are a pain in the ass. )
The computing power exists and is proven to work. The computer is known as "human brain".
Someone who understands photographic light (read Strobist's Lighting 101 as a good way to start understanding it) and doing stuff like what you described above becomes second nature.
Note: Regardless of computing method, non-white ceilings/walls greatly decrease the solution space of the problem.:(
This is only a marginal improvement (and may in many situations prove to be a non-improvement making those who understand photographic lighting even sicker).
The unfortunate fact is, we'll never see bounce flash coming from a cameraphone. There simply is not any way to get the energy/power requirements for bounce flash crammed into a cell phone. The patent above is, if anything, going in the exact opposite direction from bounce flash.
Slashdot Mirror
In the aftermath of gas drilling micro-disasters (the nature of gas drilling results in localized environmental damage, but when it happens it is a disaster for those nearby), I'm guessing increasing regulation is going to increase the costs of gas drilling.
There's a moratorium on shale gas drilling (specifically on well stimulation by hydrofracturing, but no one is going to drill a well they can't frack) in New York State after the rampant water contamination incidents all over Pennsylvania. For example, the groundwater in Dimock, PA became undrinkable within a year or so of the commencement of drilling. People can actually light their tap water on fire now.
Gas is not a long-term option, and in fact, it looks like the way it is being drilled now is going to have severe long-term environmental consequences (it already has in many drilling areas). Nuclear is a long-term investment.
I'm referring to UMTS1700 support. Only a few carriers around the world support UMTS1700.
(UMTS1700 is the main reason why manufacturers now have to make phones with different designs for TMo and AT&T - I have yet to see a single phone that supports 3G on both T-Mobile and AT&T.)
This is, in my opinion, why the optical networking industry crashed a few years ago. There are massive amounts of dark fiber along nearly every backbone connection (as the material costs of the fiber are far less than the costs of laying the fiber), and there wasn't enough last-mile capacity to keep all these super-backbones full.
As the gap between the backbone and the last mile increased, it became harder and harder to sell these new "omg-crazy-bandwidth-over-a-single-fiber" solutions.
Rumor has it they're possibly switching to Qualcomm baseband processor chipsets for the iPhone 5 for a variety of reasons. Note that Qualcomm does produce GSM/UMTS baseband chips, in fact a large portion of the GSM smartphone market uses various Qualcomm MSM silicon.
Once you're using Qualcomm MSM silicon for your GSM/UMTS solution, it's not too difficult to swap in the CDMA2000 version of that chip and get Sprint/Verizon support. Look at the large number of HTC GSM phones that have CDMA variants. (As to licensing fees - no matter what Qualcomm gets paid, and if you use their MSM chipsets you probably don't have to pay them patent royalties.)
While the CPU in the iPhone is an "Apple spec" CPU, most likely the radio chipset is from a third party.
Most such third parties offer roughly equivalent GSM and CDMA chipsets (In fact, probably 50%+ of the GSM/UMTS smartphone market is powered by Qualcomm chipsets, despite Qualcomm being the creator and backer of cdmaOne/CDMA2000), making it not too difficult to make a CDMA version of a phone and a GSM version.
Also, you show a fundamental lack of understanding of RF and communications engineering. New band = new antenna. Same bands but different protocol = no new antenna, just a different baseband processor. Therefore, adding T-Mobile's additional band is the least likely route for Apple to go.
Changing the iPhone from GSM/UMTS to CDMA2000 (note: UMTS uses a CDMA modulation scheme, one of the reasons Qualcomm is involved in that market despite them backing a competing protocol/air interface suite - they're good at making CDMA chipsets whether it is the UMTS protocol/air interface suite or CDMA2000) most likely entails:
1) Switch out the baseband processor chip. Note that as I said before, nearly all UMTS chipset providers also provide CDMA2000 chipsets that are very similar and require little engineering
2) Remove the now unneeded SIM slot (what was that about a packed device?)
That's about it
Note: Right now, Apple gets their radio chipsets from Infineon, but there is a lot of evidence pointing to them moving to Qualcomm for the 5th gen iPhone, with the ability to produce a CDMA version being one of a number of reasons, Infineon getting purchased by Intel being a bigger reason - see http://touchreviews.net/iphone-5-qualcomm-radio-chip/ - I don't know about Infineon chipsets, but it is VERY easy to make both a CDMA2000 and GSM/UMTS version of a phone if you use Qualcomm for your radio chipsets, which is why a large portion of HTC's GSM product line have near identical CDMA equivalents.
I don't think it means anything for Android. In fact, it could be good for Android.
It seems like AT&T is doing everything they can to milk the iPhone exclusivity and not piss off Apple enough to start selling iPhones through other carriers. Just look at AT&T's piss poor rollout of Android-based handsets - One of the linked articles links to another article about three new Android phones coming out for AT&T from Motorola. Note that every single one of these "new" phones lags at least 4-6 months behind Verizon's Android offerings, and in terms of hardware specs, behind even some of AT&T's Windows Mobile offerings. (HVGA or QVGA screen? Are you kidding me? My Tilt 2 came out last October and has a full WVGA screen. Also note that nearly everyone including Motorola is moving to 1 GHz+ Snapdragon-based phones except these three low-speed offerings for AT&T.)
I'm beginning to be tempted to return to Verizon. I ditched Verizon for AT&T 3-4 years ago because I was sick of Verizon phones coming out 6-9 months after Sprint with numerous features removed (such as the Treo 650's lack of Bluetooth DUN on VZW). Now, a few years later, VZW has done a complete about-face (at least with Android) which completely shocked me, and AT&T is now doing the same crap Verizon did a few years ago in the name of milking the iPhone for all it's worth.
Once AT&T loses iPhone exclusivity, they're going to have to shape up with their Android offerings, otherwise Verizon will have both the iPhone and superior Android offerings for those that don't want an iPhone. Meanwhile, it is getting to the point where all AT&T has is the iPhone.
Actually this is the case for a lot of vehicles, usually related to accessibility of some plugs. As interior room for passengers becomes more heavily marketed, engine compartment space gets more and more cramped.
For example, my 1995 Chrysler LeBaron's front plugs took minutes to change, the rear plugs was an hour+ job due to having to remove the air intake to get at them. (I think it was a much shorter job if you had a lift and could access the vehicle from below.)
My parents' 1996 Dodge Caravan was awful - the dash of the vehicle overhangs a significant amount of the rear portion of the engine compartment. It's basically impossible to get at the rear bank of the spark plugs without coming in from below. (Again, probably a short job for a mechanic with a lift.)
I think my 2009 Subaru's plugs are similarly difficult to access - that engine compartment is REALLY cramped.
I wouldn't call it dead fucking easy.
It's a pain in the ass, messy, results in you having to (in most cases) take the waste oil to a garage for proper disposal anyway.
To make it NOT a pain in the ass requires a lift (I mean a proper lift and not a jack), and the initial investment in that tool is more than the cost of lifetime oil changes for the car.
Oh, and keep in mind, most oil changes include the price of the oil (up to a point - unusually large engines that require lots of oil get charged extra) - I've managed to get oil changes for less than what I'd pay for the oil used in most stores.
Now PS3 firmware updates - those are dead easy.
Yup. I buy well price controlled stuff. I used to be willing to pay a little extra for the ability to return something instore, but at this point, their return policy sucks compared to most of the reputable online vendors (Amazon, Newegg, B&H, Adorama) - yes, in case you couldn't tell from the list, I shop for photographic equipment a lot.
monoprice.com is where I go for my cables.
For HDMI cables, pay the extra for the cables with 24AWG wires. The ultracheap ones use 28AWG wires. Even the 24AWG ones are only a few bucks for a 6 foot cable.
Honestly, this is one of the few things that Comcast does better than Time Warner. TW will just cut you off. They only will give you the time of the complaint filed against your IP, not the time of whatever traffic was generated, making it difficult to pin down the machine in question. (They gave me a time window that was while on vacation, leading me to suspect my file server/MythTV box had been rootkitted. It turns out that it took someone a few days to complain about the bot-generated spam - the Windows partition I only use for gaming on my desktop turned out to be the culprit.)
Google claims that the codecs they back (In this case, VP8) are patent-free.
But apparently Microsoft claimed the same thing about VC-1, and within a year, numerous patent holders stepped forward with claims that VC-1 infringed on a patent they held, and now there is a VC-1 patent pool.
So really, "Unencumbered by patents" really means "We don't think it's encumbered but no guarantees."
To play devil's advocate - If they are using a high-quality (low compression ratio) JPEG as a starting point, that isn't too bad.
E.g. start with a quality 95 JPEG, then:
Generate a quality 75 (small) JPEG
Generate a webP image of the same size
Compare
Problem is that, according to the analysis by the x264 developer (see the first independent analysis link), WebP is missing quite a few features that JPEG has, does not add any of the features JPEG is missing but people really want (like a lossy format that contains alpha capability - although admittedly, lossy compression of the alpha channel itself could cause some REALLY weird artifacts.)
It also, at least in the current state of the encoder, does not appear to perform any better than JPEG.
Yeah, especially since no one on /. RTFAs... Since linked articles do cover this question.
Can you say:
Virtual Boy?
Bet the Virtual Boy 2 will be as much of a failure as the original. It was difficult to use and gave people headaches.
I think you greatly underestimate the advantage an ASIC has over a general purpose CPU (even the latest Intel CPUs with AES-NI) when it comes to crypto.
How about you RTFA:
"The HDCP cipher is designed to be efficient when implemented in hardware, but it is terribly inefficient in software, primarily because it makes extensive use of bit operations. Our implementation uses bit-slicing to achieve high speeds by exploiting bit-level parallelism. We have created a few high-level routines to make it as easy as possible to implement HDCP, as shown in the following example. "
Same with Pentax (except for some of the entry-level Pentaxes, the upper level ones like the K10D/K20D/K-7/K-5 can use the onboard flash as a P-TTL master).
Whether or not the onboard flash contributes or not depends on a setting in the camera. Even when it's set to not contribute, it's more like "contribute at absolute minimum power". It's most obvious if there is a reflective surface in the scene that allows the onboard flash to be seen.
I'm actually shocked that the Canons which do have onboard flashes can't act as a master for wireless E-TTL. (From doing a little looking, at the very least the 7D supports this.)
Technically they are all infrared - the sensor on the slave flashes has peak sensitivity in the IR, and in most cases there's an IR-pass/visible-block filter over the sensor. However, there is enough IR content within a xenon strobe's output that in most situations (exception - stuff like the SU-800 and ST-E2), the camera sources the IR using pulses from the onboard flash.
I do hope the police officer involved does have disciplinary action. He pulled his gun prior to identification, and his vehicle did not appear to have any sort of police lights installed.
However, the person in question would have had an arrest record anyway for his reckless driving infraction. He's still going to get nailed hard by the courts, but at least now he's getting nailed for the right reasons.
WRONG. FTDI has built-in drivers for Linux but not most other OSes. It definately requires drivers in Windows.
The newer USB CDC ACM class, however, has built-in drivers on any modern OS, with the "worst case" situation being Windows where you need an INF (but not actual drivers) to tell Windows that that VID/PID is a CDC ACM device.
"So how is it that a new custom chipset with a new vendor id requires no drivers? Is there a standard for USB->serial that if you follow you don't need a driver? (and why do most cables not support it if there is?)"
As others have said - the CDC ACM class. I believe that class was standardized relatively recently in an attempt to address the fact that there are 23432743290 different USB-serial chipsets all of which required special drivers.
Not sure if XP has CDC-ACM support built in (might have been added in a service pack?), but in general, CDC-ACM is the "new kid on the block" and the legacy (PL230x and FTDI) approaches are still well entrenched.
The nice thing about the little popup flashes in most SLRs these days is that they can trigger a remote flash wirelessly.
(A modern take on the classic "optical slave" where the camera can actually control the remote flash's light output. Nikon calls this CLS, Canon calles it wireless E-TTL I think, Pentax calls it wireless P-TTL.)
Most SLRs can be put in a mode where the onboard flash is reduced so much in output that it doesn't contribute significantly to exposure at all. (It will still, however, contribute some.) It can also be allowed to contribute to exposure, causing it to act as a small fill flash.
One thing you didn't explicitly mention but have hinted at in past posts: Sometimes, instead of bringing a large reflective object (umbrella, softbox, etc), a photographer will use their environment. This is why highend flashes swivel/tilt - for example, to point the flash head at the ceiling. In this situation, the ceiling becomes a large diffuse light source. (Note: White ceiling required unless you're going for a funky effect. Black ceilings such as found at many concert venues are a pain in the ass. )
My guess is that Apple's algorithm is more likely to target fixate on the foreground subject, creating the exact opposite effect.
The computing power exists and is proven to work. The computer is known as "human brain".
Someone who understands photographic light (read Strobist's Lighting 101 as a good way to start understanding it) and doing stuff like what you described above becomes second nature.
Note: Regardless of computing method, non-white ceilings/walls greatly decrease the solution space of the problem. :(
This is only a marginal improvement (and may in many situations prove to be a non-improvement making those who understand photographic lighting even sicker).
The unfortunate fact is, we'll never see bounce flash coming from a cameraphone. There simply is not any way to get the energy/power requirements for bounce flash crammed into a cell phone. The patent above is, if anything, going in the exact opposite direction from bounce flash.