Never start an argument with a professional comedian.
Especially if the only "facts" you have is the fairy tale you keep telling yourself when you go to sleep at night, and you have a well-documented history of lying to the press.
How does the quote go? A comedian tells the truth that nobody wants to hear?
I seem to recall something similar happened with MIPS in the past (ironically, Imagination owns MIPS): MIPS's new design was too important to SGI, and MIPS was in danger of going out of business... so SGI bought MIPS to keep them alive.
It almost sounds like Apple decided that Imagination couldn't continue on its own for long, tried to buy it, the offer was refused, so Apple decided to leave them to their fate and moved on...
I don't know... it's vaguely interesting to see any carrier dumping every brand except one these days.
I can't help but wonder if it's just cheaper/easier for Virgin Mobile to reduce their overhead by going with one manufacturer -- they don't have to worry about juggling OS updates, etc. between multiple manufacturers & models.
I'm sure it's more complicated than that, but there is a value in having a narrower range of hardware & software to support.
Homomorphic encryption is well suited for this situation - you can even perform operations on the data stored in the cloud, and the cloud provider isn't able to eavesdrop on anything, because decryption isn't required.
Even without homomorphic encryption, there are plenty of HSM (hardware security module) devices which are widely used to handle encryption in a way that the cloud provider doesn't have access to the encryption key, nor do they have the ability to decrypt data.
The TL;DR version is that the AFL-CIO started campaigning for a 40-hour week in 1886. There was a workplace explosion three days after the AFL-CIO's announcement, killing several, and resulting in a few trials & executions. That brought the 40-hour work week into international news, where it remained.
The Ford Motor company famously introduced the policy in 1914, but wasn't the first company to do so. A couple of years later, a strike by railroad workers crippled the nation's commerce, so the government mandated they get overtime pay in 1916. (This is not unlike strikes by longshoremen in recent years).
Overall, the labor unions deserve most of the credit, for doggedly pursuing the idea and seeing progress for nearly sixty years.
Let's not forget that Wal*Mart is the same group of geniuses that brought us the laughably insecure CurrentC/MCX - and after that folded, they doubled down, and deployed it anyway as "Wal*Mart Pay".
And seriously? Complaining that your vendor uses AWS for their own business?!?
What's next, saying they'll penalize companies that use Ford delivery trucks?
So you want to compare the number of cars -- which cost $30-90k a phone, and lasts upwards of 20 years to a smartphone which costs $500-800 and lasts for three years? That's a false equivalence if there ever was one.
And you compare sales number for Blu-ray players the US alone vs worldwide sales for phones? Seriously? The PlayStation alone sold more than 80 million units -- to say nothing of the 60 million (and counting) PS4's, or the 30+ million Xbox One's. That doesn't count players sold by Samsung, LG, Sony, Panasonic, Yamaha, Toshiba, Phillips, Vizio, and dozens of Chinese brands.
Nor are you counting the number of normal DVD players which can queue up Netflix, Hulu, YouTube, and Amazon Prime video.
Home entertainment hardware Isn't often replaced every year, unlike a phone. I've bought 5 phones for every Blu-ray player I've owned - because the Disc player has more longevity. The situation is even more lopsided with cars.
There's a huge difference between the number of devices sold and the number of devices in use. It's not terribly surprising, but phones are disposable - it's often cheaper to replace the thing than repair. I sure as hell don't buy a new car every two years, but a phone? Sure... no problem.
Unsurprisingly, when long lifetime devices are designed, they plan for the long term, instead of what is trendy or shiny at the moment. Nobody is going to care about VP9 six years from now, but H.264 will still be common. MP3 will be with us forever...
You add support for codecs which will be around long term.
You realize there's a big world outside of the web browser, right? Take a look at the home theatre isle at your local electronics store; visit a camera shop. When I see a codec move into that market, I take it seriously.
It's not at all surprising to see Google's codec is mandatory on Google's Android OS and Google's WebRTC. That's a no brainer.
Apple has patents in the HEVC pool, and if memory serves, the late Steve Jobs lead the (failed) charge to try to get VP8 into the MPEG patent pool, so there may be bad blood with respect to the various incantations of VP8/9.
Google is supporting its codec, and Apple appears to be supporting the codec they worked on. AV1 is interesting, but VP9 is less so: Netflix's internal testing showed x.265 outperforming VP9 by an average of 20%.
So Apple isn't exactly crazy in deciding VP9 isn't worth its time; HEVC is the better codec, and is the ISO/ITU/MPEG standard.
If AV1 meets its goals, then it can give HEVC a run for its money - and I'm all for that. But for 2017 and likely most of 2018, HEVC is the codec to beat.
I did realize after posting that VP6 & 7 were On2's proprietary codecs before they were bought by Google; sorry for that.
And yet, I cannot plug a flash drive with my music in Opus format into my 2017 model car's stereo and expect it to play.
Nor can I plug it into my brand new Blu-ray player and hear music.
A big wide world exists outside the web browser, and in many of them, MP3 is the only game in town.
I'll be thrilled if BlueTooth's next iteration mandates Opus support for its A2DP profile. I suspect I'll be disappointed, as few device makers implement anything other than its only mandatory codec, SBC. (Though there is sporadic support for AAC or AptX, so I have some hope)
In the meantime, let's acknowledge that Apple hasn't joined the Alliance for Open Media [aomedia.org]. When will Apple join?
Why single out Apple? It's not like Sony, Samsung, Panasonic, Canon, Nikon, LG, Philips, HP, or Lenovo are members either..
It's silly to criticize Apple for adding support for a codec which "all of the above" are also supporting.
Honestly, I'm all for open codecs, but after waiting years to see Vorbis, Opus, Theora, and VP6-VP8 adoption stand at an effective Level of Zero, in spite of being competitive (or better) while not having a patent minefield, I've grown wary of the ever-growing next shiny thing that goes nowhere.
AFAIK, AV1 is the first industry-wide co-developed open codec, unlike others (VC-1, VC-2, Theora, and VP4-9). It combines components of Thor, daala, and VP9. It has real promise, and I look forward to see it.
At the end of the day, I wonder if any of the new codecs will make much headway vs H.264. H.264 is everywhere, and is "good enough" - much like how DVD has doggedly hung on inspire of the superiority of Blu-ray & HD streaming. We've seen something similar with audio codecs; MP3 is still a top format, in spite of the huge advances in coding efficiency with Opus, and the various flavors of AAC.
There are some human interface issues that remain a challenge for these 'partially autonomous' driving modes. This article mentions some of the challenges with the handoff between autonomous steering and manual (not Tesla specific).
This.
I don't have a Tesla, but my car has its own less-capable partially autonomous features. It's kind of hard to describe, other than there are corner cases which aren't handled very smoothly.
For example: a car directly to the side of me starts to drift into my lane - I notice, and start to move to avoid him, but the car decides I'm drifting out of my lane and applies force to steer me back into the center of the lane -- and into the other car.
In that situation the car is applying its own steering force. I can easily override the car, but it's difficult to predict when the car will realize I'm overriding it and hand off full control.
It's like having a kid randomly pull on the steering wheel during an emergency correction - staying on a straight line isn't so easy.
HEVC is out now, and has broad industry support from embedded hardware manufacturers (set top boxes, roku-a-likes, cameras, camcorders, etc.), as well as software players like Microsoft and Apple.
I feel that VP9 is a dead-end in many ways: It's a "standard" that really has only one controlling interest: Google. VP9 has virtually zero mindshare outside the Googleplex; I'm not aware of any dedicated cameras, camcorders, set top boxes, etc. that support VP9.
AV1, on the other hand, looks very compelling... it actually has broad industry support, from big players like Microsoft, Cisco, Netflix, Google, all the way down to silicon makers like Broadcom, Xilinx, RealTek, ARM, AMD, and NVIDIA.
AV1 stands a good chance of killing HEVC entirely (and they deserve it, given the licensing minefield that has to be negotiated to use HEVC)
While AV1 sounds awesome on paper, it's important to know that AV1 isn't completed yet, the bitstream is scheduled to be "frozen" in Q3 or Q4 of 2017, and even that doesn't mean the standard will be at version 1.0. Hardware that supports AV1 isn't expected until Q4 2018 at the earliest.
It's disingenuous to complain that Apple isn't going to include AV1 when it isn't - and won't be - ready before High Sierra.
Half of what I hear about Vulkan is that it's an "industry-standard" dumpster fire; it came out considerably after Metal and DX12, and still reeks of half-baked thoughts rushed to market.
Were I calling the shots, I'd give Vulkan more time to polish out its issues.
For those that feel differently, there's even moltenvk which appears to layer Vulkan on top of Metal.
And, of course, many of the major graphics engines just abstract DirectX, Vulkan, Metal, and OpenGL away from the game developer anyway, so it's hard to see how it'll be a real issue.
I learned to type with a typewriter, and back then most printers had a single fixed-width font, so it was still considered proper form to double-space after a period.
It wasn't until a couple years ago I learned that it wasn't considered proper form to double-space; decades of muscle memory are hard to unlearn.
Nicotine itself is not a useful insecticide, primarily because it's so poisonous to non-target species: Avians & Mammals.
As Nicotine is absorbed through the skin easily, it's also quite easy for somebody using nicotine as an insecticide to poison themselves, as well as anybody else in the area. 30-60 mg can kill a human adult, though the LD50 is generally 500-1000 mg.
The neonicotinoids (similar to Nicotine) are extremely useful insecticides - as deadly to insects, but relatively nontoxic to birds & mammals.
Programmers will get bored and jump on some other fad
I agree with the sentiment in general, but Python is well past the point of being a fad -- maybe it would have been considered a fad in 2000, but since then we've already seen a number of fads come and go since then (PHP, Ruby on Rails, Groovy & Grails, Node.js, golang, and Scala to name a few).
Even Perl seems to have finished its time in the sun, and is on its way into obscurity.
Slashdot Mirror
Never start an argument with a professional comedian.
Especially if the only "facts" you have is the fairy tale you keep telling yourself when you go to sleep at night, and you have a well-documented history of lying to the press.
How does the quote go? A comedian tells the truth that nobody wants to hear?
I seem to recall something similar happened with MIPS in the past (ironically, Imagination owns MIPS): MIPS's new design was too important to SGI, and MIPS was in danger of going out of business... so SGI bought MIPS to keep them alive.
It almost sounds like Apple decided that Imagination couldn't continue on its own for long, tried to buy it, the offer was refused, so Apple decided to leave them to their fate and moved on...
I don't know... it's vaguely interesting to see any carrier dumping every brand except one these days.
I can't help but wonder if it's just cheaper/easier for Virgin Mobile to reduce their overhead by going with one manufacturer -- they don't have to worry about juggling OS updates, etc. between multiple manufacturers & models.
I'm sure it's more complicated than that, but there is a value in having a narrower range of hardware & software to support.
Unless iPhones are being given away then it's a bit more than $1.
I'm not sure if you've looked lately, but you can't get a new, subsidized iPhone from any major carrier in the US.
This is of historical interest to me, though: I used to use Virgin Mobile a decade ago. I left mainly because Sprint's network is lousy in my area.
So keep good backups and if you get hacked send the attackers the middle finger instead of bitcoin and just let them do whatever
Agreed. Because you have zero assurance that they'll do anything except take your money and release the data anyway.
It's not unlike when kidnappers ransom somebody, take the money, and then kill the victim.
Homomorphic encryption is well suited for this situation - you can even perform operations on the data stored in the cloud, and the cloud provider isn't able to eavesdrop on anything, because decryption isn't required.
Even without homomorphic encryption, there are plenty of HSM (hardware security module) devices which are widely used to handle encryption in a way that the cloud provider doesn't have access to the encryption key, nor do they have the ability to decrypt data.
A good article on the subject is at PolitiFact
The TL;DR version is that the AFL-CIO started campaigning for a 40-hour week in 1886. There was a workplace explosion three days after the AFL-CIO's announcement, killing several, and resulting in a few trials & executions. That brought the 40-hour work week into international news, where it remained.
The Ford Motor company famously introduced the policy in 1914, but wasn't the first company to do so. A couple of years later, a strike by railroad workers crippled the nation's commerce, so the government mandated they get overtime pay in 1916. (This is not unlike strikes by longshoremen in recent years).
Overall, the labor unions deserve most of the credit, for doggedly pursuing the idea and seeing progress for nearly sixty years.
So you want to compare the number of cars -- which cost $30-90k a phone
Correction: $30-90k per car, not per phone.
Let's not forget that Wal*Mart is the same group of geniuses that brought us the laughably insecure CurrentC/MCX - and after that folded, they doubled down, and deployed it anyway as "Wal*Mart Pay".
And seriously? Complaining that your vendor uses AWS for their own business?!?
What's next, saying they'll penalize companies that use Ford delivery trucks?
So you want to compare the number of cars -- which cost $30-90k a phone, and lasts upwards of 20 years to a smartphone which costs $500-800 and lasts for three years? That's a false equivalence if there ever was one.
And you compare sales number for Blu-ray players the US alone vs worldwide sales for phones? Seriously? The PlayStation alone sold more than 80 million units -- to say nothing of the 60 million (and counting) PS4's, or the 30+ million Xbox One's. That doesn't count players sold by Samsung, LG, Sony, Panasonic, Yamaha, Toshiba, Phillips, Vizio, and dozens of Chinese brands.
Nor are you counting the number of normal DVD players which can queue up Netflix, Hulu, YouTube, and Amazon Prime video.
Home entertainment hardware Isn't often replaced every year, unlike a phone. I've bought 5 phones for every Blu-ray player I've owned - because the Disc player has more longevity. The situation is even more lopsided with cars.
There's a huge difference between the number of devices sold and the number of devices in use. It's not terribly surprising, but phones are disposable - it's often cheaper to replace the thing than repair. I sure as hell don't buy a new car every two years, but a phone? Sure... no problem.
Unsurprisingly, when long lifetime devices are designed, they plan for the long term, instead of what is trendy or shiny at the moment. Nobody is going to care about VP9 six years from now, but H.264 will still be common. MP3 will be with us forever...
You add support for codecs which will be around long term.
Nearly everything copies from a few of Shakespeare's plays - and even they were derivitive.
The trend hasn't slowed down for 400 years, so I doubt things will change.
Selection bias, much? You're clearly cherry-picking your data.
You can point to the videos and photos shared and iCloud and say that Apple dominates the market; that doesn't make it so.
What about virtually every recent Blu-ray and DVD player, which can stream Hulu and Amazon video?
Or automotive media players for the Kids in the back?
why do you keep insisting that the embedded market doesn't matter? It dwarfs the PC/phone/computer market.
You realize there's a big world outside of the web browser, right? Take a look at the home theatre isle at your local electronics store; visit a camera shop. When I see a codec move into that market, I take it seriously.
It's not at all surprising to see Google's codec is mandatory on Google's Android OS and Google's WebRTC. That's a no brainer.
Apple has patents in the HEVC pool, and if memory serves, the late Steve Jobs lead the (failed) charge to try to get VP8 into the MPEG patent pool, so there may be bad blood with respect to the various incantations of VP8/9.
Google is supporting its codec, and Apple appears to be supporting the codec they worked on. AV1 is interesting, but VP9 is less so: Netflix's internal testing showed x.265 outperforming VP9 by an average of 20%.
So Apple isn't exactly crazy in deciding VP9 isn't worth its time; HEVC is the better codec, and is the ISO/ITU/MPEG standard.
If AV1 meets its goals, then it can give HEVC a run for its money - and I'm all for that. But for 2017 and likely most of 2018, HEVC is the codec to beat.
I did realize after posting that VP6 & 7 were On2's proprietary codecs before they were bought by Google; sorry for that.
And yet, I cannot plug a flash drive with my music in Opus format into my 2017 model car's stereo and expect it to play.
Nor can I plug it into my brand new Blu-ray player and hear music.
A big wide world exists outside the web browser, and in many of them, MP3 is the only game in town.
I'll be thrilled if BlueTooth's next iteration mandates Opus support for its A2DP profile. I suspect I'll be disappointed, as few device makers implement anything other than its only mandatory codec, SBC. (Though there is sporadic support for AAC or AptX, so I have some hope)
In the meantime, let's acknowledge that Apple hasn't joined the Alliance for Open Media [aomedia.org]. When will Apple join?
Why single out Apple? It's not like Sony, Samsung, Panasonic, Canon, Nikon, LG, Philips, HP, or Lenovo are members either..
It's silly to criticize Apple for adding support for a codec which "all of the above" are also supporting.
Honestly, I'm all for open codecs, but after waiting years to see Vorbis, Opus, Theora, and VP6-VP8 adoption stand at an effective Level of Zero, in spite of being competitive (or better) while not having a patent minefield, I've grown wary of the ever-growing next shiny thing that goes nowhere.
AFAIK, AV1 is the first industry-wide co-developed open codec, unlike others (VC-1, VC-2, Theora, and VP4-9). It combines components of Thor, daala, and VP9. It has real promise, and I look forward to see it.
At the end of the day, I wonder if any of the new codecs will make much headway vs H.264. H.264 is everywhere, and is "good enough" - much like how DVD has doggedly hung on inspire of the superiority of Blu-ray & HD streaming. We've seen something similar with audio codecs; MP3 is still a top format, in spite of the huge advances in coding efficiency with Opus, and the various flavors of AAC.
There are some human interface issues that remain a challenge for these 'partially autonomous' driving modes. This article mentions some of the challenges with the handoff between autonomous steering and manual (not Tesla specific).
This.
I don't have a Tesla, but my car has its own less-capable partially autonomous features. It's kind of hard to describe, other than there are corner cases which aren't handled very smoothly.
For example: a car directly to the side of me starts to drift into my lane - I notice, and start to move to avoid him, but the car decides I'm drifting out of my lane and applies force to steer me back into the center of the lane -- and into the other car.
In that situation the car is applying its own steering force. I can easily override the car, but it's difficult to predict when the car will realize I'm overriding it and hand off full control.
It's like having a kid randomly pull on the steering wheel during an emergency correction - staying on a straight line isn't so easy.
HEVC is out now, and has broad industry support from embedded hardware manufacturers (set top boxes, roku-a-likes, cameras, camcorders, etc.), as well as software players like Microsoft and Apple.
I feel that VP9 is a dead-end in many ways: It's a "standard" that really has only one controlling interest: Google. VP9 has virtually zero mindshare outside the Googleplex; I'm not aware of any dedicated cameras, camcorders, set top boxes, etc. that support VP9.
AV1, on the other hand, looks very compelling... it actually has broad industry support, from big players like Microsoft, Cisco, Netflix, Google, all the way down to silicon makers like Broadcom, Xilinx, RealTek, ARM, AMD, and NVIDIA.
AV1 stands a good chance of killing HEVC entirely (and they deserve it, given the licensing minefield that has to be negotiated to use HEVC)
While AV1 sounds awesome on paper, it's important to know that AV1 isn't completed yet, the bitstream is scheduled to be "frozen" in Q3 or Q4 of 2017, and even that doesn't mean the standard will be at version 1.0. Hardware that supports AV1 isn't expected until Q4 2018 at the earliest.
It's disingenuous to complain that Apple isn't going to include AV1 when it isn't - and won't be - ready before High Sierra.
Half of what I hear about Vulkan is that it's an "industry-standard" dumpster fire; it came out considerably after Metal and DX12, and still reeks of half-baked thoughts rushed to market.
Were I calling the shots, I'd give Vulkan more time to polish out its issues.
For those that feel differently, there's even moltenvk which appears to layer Vulkan on top of Metal.
And, of course, many of the major graphics engines just abstract DirectX, Vulkan, Metal, and OpenGL away from the game developer anyway, so it's hard to see how it'll be a real issue.
I had the same problem with LaTeX and LyX. Took ages to get used to...
It is two when using a typewriter. :-)
I learned to type with a typewriter, and back then most printers had a single fixed-width font, so it was still considered proper form to double-space after a period.
It wasn't until a couple years ago I learned that it wasn't considered proper form to double-space; decades of muscle memory are hard to unlearn.
No man can serve two masters. You're either committed to your marriage, or your mouse.
If you try to do both, expect to be burned.
Nicotine itself is not a useful insecticide, primarily because it's so poisonous to non-target species: Avians & Mammals.
As Nicotine is absorbed through the skin easily, it's also quite easy for somebody using nicotine as an insecticide to poison themselves, as well as anybody else in the area. 30-60 mg can kill a human adult, though the LD50 is generally 500-1000 mg.
The neonicotinoids (similar to Nicotine) are extremely useful insecticides - as deadly to insects, but relatively nontoxic to birds & mammals.
Programmers will get bored and jump on some other fad
I agree with the sentiment in general, but Python is well past the point of being a fad -- maybe it would have been considered a fad in 2000, but since then we've already seen a number of fads come and go since then (PHP, Ruby on Rails, Groovy & Grails, Node.js, golang, and Scala to name a few).
Even Perl seems to have finished its time in the sun, and is on its way into obscurity.