That's pretty damn cool. It reminds me of scene completion, which is another take on the same idea - combining images from Flickr to create new images according to a brief sketch.
You can request footage of yourself from private cameras using data protection laws.
Anyway, no need to worry for two reasons:
This is a story in the Daily Fail. So it's practically guaranteed to be wrong, made up, exaggurated etc.
Despite that the story makes it quite clear that the system doesn't have any cameras today. The dude is trying to sign up businesses to his plan. Obviously you can't just plug into random CCTV cameras, that'd be insane - the owners have to opt them in. Good luck with that!
Another thought that hit me, and it was my first thought, was "compromised XBox360s joining botnets."
You need to read up on the Xbox 360 security system. You can't compromise an Xbox 360 that's connected to Live, it's not possible. There are no botnets of Xbox 360 machines, and it's highly unlikely there ever will be.
Not if the attacker is using a botnet, unless TFA means the number of friend requests a user can receive.
Having a botnet doesn't do much good if you can't automatically create accounts. As you need an XBox to sign up for an Xbox Live account restricting the number of requests an account can send is trivial.
They aren't really "ads" in the traditional sense. X360 has a sort of featured games/videos section, and some videos you can opt to download for trailers. It's promotional but it's a stretch to call them an ad - as the op says, they're actually useful.
That's partly because most indie games aren't that good. Now XBL has a ratings system it'll be easier for the good stuff to float to the top. That said, the expensive part of writing a commercial XBL Arcade or retail game isn't the SDK/dev unit. The expensive part is paying for (and passing) the QA Microsoft put all games through. They test games like crazy which is why console games basically never crash or seriously glitch out.
His enthusiasm can be summed up as "OOXML works, ODF doesn't". I mean, ODF didn't even bother to properly specify spreadsheet formula functions, right? How can anybody take such a "standard" seriously? There was never a choice between OOXML and ODF because ODF as a usable specification did not exist.
Stallman and the FSF have a history of just making shit up when they decide they don't like something. Look at some of the stuff they've written about trusted computing for examples of this. Miguel is right - the guy plays fast and loose with the facts repeatedly. I see lots of people praising his consistency in this discussion. Well guess what - it's easy to have a consistent position over a long period of time when you flatly refuse to accept factual reality.
The point of.NET is not cross platform compatibility. The point is to make it easy for developers to write software - and it succeeds at that. The whole reason Miguel created Mono is that the free software community didn't and still hasn't created any productivity boosting tool equivalent in power.
Um, no. Microsoft don't provide any Windows built in game copy protection, yet nearly every game provides their own. If you think Microsoft refusing to implement DRM would have meant no DRM you're crazy.
I think the argument is that they aren't really needing to do extra work to support the Pre. The extra work they're doing is deliberately un-supporting Palm, which achieves nothing other than annoying the customers of their competition. The real question is why bother with emulating an iPod at all, surely there's a better way for Palm to do this. There must be some quite compelling reason to go with this protocol over some other solution.
I'm pretty sure there's nothing illegal about the Pre telling iTunes it's an iPod. After all, Internet Explorer claims to be Mozilla, and Google Chrome claims to be Chrome, Safari, AND Mozilla! Now if iTunes started issuing firmware hash challenges to iPods, Palm would be stuck - to answer the challenges they'd have to ship a copy of the iTunes firmware which other than being very large would also be illegal. But I guess Apple can't easily update every iPod to support that retroactively.
Really, I've really got to wonder what Apples long term strategy is here. The constant stream of stories like this have to be causing recruitment issues if nothing else. They're already being questioned by the US Govt over the Google Voice issue and now they're apparently issuing updates intended only to break interop? Despite many rumors that is something I don't recall Microsoft ever doing.
Again, unless we're talking an HPC application, I'm calling bullshit. I can't think of a language these days that won't produce sufficient performance for most applications. After all, the vast majority of performance for any application comes from the selection of appropriate algorithms/data structures. And for those few cases where performance needs to be carefully tuned, most languages provide an FFI that can be used to call out to C or some other lower-level language.
Frankly, you sound like somebody who has done a CS course but not written much real world, shipping software.
How many desktop apps does the average person use that are written in Java? The answer is.... none. If you attempt to write a useful desktop app in Java you will soon understand why. Of course, the language is turing complete. It is also designed in such a way that memory usage/working set size is dramatically greater than the equivalent program written in C++, and it's close to impossible to fix this.
About a month ago, I wrote some code that processed a large quantity of data (it was locating tree-like structures within a large graph). The initial version was written in Python because the implementation I inherited was written in Python. However its runtime was measured in hours. I rewrote it in C++ and it now completes in around five minutes. The previous version was impossible to work on: a combination of terrible performance and weak/loose typing meant that if there was a typo near the end of the algorithm, I might only find out about it 6 hours later. This is clearly impractical.
Final example, read up on the Chandler project. One of the biggest problems they had was performance. Python is [a] single threaded [b] interpreted [c] a memory pig and [d] damn near impossible to optimize or compile. This is death by a thousand cuts.... it really doesn't matter how smart your algorithms are if every time you want to do some non-trivial thing your gui stops responding for a few seconds. Works great for simple scripts though.
I don't know if LTE will ever be 100% deployed. LTE chips are backwards compatible with 3G and 2G, right? And they can roam freely between them. LTE makes most sense to deploy in the home first, as a WiFi replacement, and then by upgrading macrocells in urban areas where smartphone traffic is stressing the local infrastructure. Whether LTE gets deployed to suburban or rural areas I can't say - the bandwidth/latency upgrades probably aren't important there for people outside.
LTE won't replace your DSL line - it'll use it. I mean, do you directly plug into your modems ethernet port? That configuration is rare, from what I've seen. Much more common is home wifi. LTE will replace your current wifi router box with a different wireless router box, and you won't know much difference except that if you open up your 4G equipped netbook and start a fast download, then walk outside down to the local park, the only thing that'll happen is your download gets slower. No connection interruptions, no new IP address, no changing in billing providers.... it'll all be consistent.
If you read the letter, it's clear that this non-"solution" is being proposed exactly because the rights holders are refusing to allow their HD content to be broadcast without some kind of DRM. I'm pretty surprised anyone agreed to some kind of nonsense like hiding Huffman tables, but as this is being used only as a stick to make the set top box/TV manufacturers implement stronger client-side DRM, who knows, perhaps it will actually discourage casual piracy. Or perhaps the BBC will have to change the no-encryption principles on which Freeview is based or be banished to a non-HD ghetto.
Well, firstly the UK has 65 million citizens. That's pretty big actually. About 1/5th the population of the United States.
Secondly the UK was the first country to deploy digital TV anywhere in the world. So TV [set top box] manufacturers are used to dealing with this market. A lot of the finnickey details of how it works were hammered out during the initial UK deployment experience.
Thirdly, no, the BBC are not the only supplier of TV programming in the British market. There are two major competitors platform wise - Sky and cable. Freeview is the digital terrestrial platform in the UK (ie, broadcast from poles on hills). It used to be an open consortium run by the BBC and ITV, itself a consortium of independent terrestrial broadcasters, however, ITV have spent the last 10 years showing the world how not to run a TV company so eventually it became infeasible to continue like that and the BBC took complete control to stop the platform imploding.
So what about Sky and cable? Firstly, Sky is owned by News Corp. News Corp also own DirecTV. Murdoch has spent massive amounts of money developing largely "unbreakable" DRM for his satellite TV companies. Whilst I believe it's possible to "break" the DRM by setting up a PC to emulate a set-top box and then, say, broadcasting the result over the internet, there are obviously bandwidth and single-point-of-failure problems with that which make it especially infeasible for HD material. There is I believe no way to decrypt the transmissions without a Sky controlled smart card.
I don't know if Sky set top boxes require HDCP or the like but it wouldn't surprise me, so, HD broadcasts especially of live events are not likely to suffer piracy thanks to Sky any time soon.
The sitation with cable I know less about, but presume it's similar. Freeview is sort of unique in that because it's intended as the non-corporate controlled platform, it doesn't encrypt the broadcasts. This worked OK for a long time because piracy of SD material was seen as a TV company problem. But now content rights holders are apparently refusing to license HD content to insecure platforms. They can do that because the majority are secure and Freeview is the odd one out!
And now, they have stuff that provides a sensible approach to concurrency
Erm, if I read that article correctly, it means Apple finally discovered that cutting edge new technology, the "thread pool". Oh, and they hacked closures into Objective C. How is that not simply a ThreadPoolExector?
O3D does indeed use either D3D or GL depending on platform. O3D is slightly higher level yes, but not dramatically so (you still have to code up shaders etc).
I think we'll end up with both. NaCL integration is the key regardless.
Since canvas is already known territory (comparatively), and JavaScript is being optimized like crazy by all browser developers, I'd bet that you should expect to see WebGL picked up much faster than O3D. Developers that are already comfortable using canvas for some 2D representations will have only a small step to take to reach WebGL.
Well, there are a few major things in O3Ds favor (btw: full disclosure, I am a Googler but have no particular interest in 3D tech beyond enjoying its products).
Firstly, O3D is a plugin that can be installed in Internet Explorer today. The IE team hasn't joined in the sudden frenzy of webapp work that the other browser vendors have been doing. That means things not based on plugins will have limited impact if this doesn't change.
Secondly, O3D embeds V8. Consistent JavaScript performance is going to be an issue for any web-based 3D work. Although it's true that every browser has been getting faster, there are still big differences in what they can do. This may be resolved over time as all the different engines approach the theoretical limits of how fast you can make JavaScript, but until then, having a single known target engine with predictable performance is a win.
Thirdly, canvas vs SVG isn't really a good comparison. 2D graphics are rarely very complex in the same way 3D graphics are. If 2D graphics had the same data throughput, hardware acceleration and tool integration issues that 3D has then an SVG type model would probably look a lot more attractive than a canvas type model. As it happens, for most 2D work a simpler API works well. I remain sceptical that the same is true for 3D.
That said, I am by no means an expert. There is a fun 3D platformer game (Infinite Journey) based on the O3D plugin that I tried - if somebody puts together a webgl demo as compelling as that one is, I'll definitely have to rethink my position.
WiFi has a limited future anyway so who cares? The future is becoming increasingly clear... over the next 10-20 years most existing air protocols are likely to be phased out in favor of GSM LTE. LTE (and the "Advanced LTE" which is likely to become the actual deployed 4G technology) offer speeds in the hundreds of megabits/sec range and latency in the ~millisecond range. In fact LTE is very close to the theoretical limits of what is physically possible to do, speed wise. LTE is also being designed with support for femtocells in mind right from the start, in fact, there seems to be growing consensus that 4G mobile networks will primarily be deployed through LTE gateways in the home first with traditional cell-tower style macrocells coming much later.
LTE offers some compelling advantages over the mixed 3G/WiFi tech we use today. Firstly, authentication and billing are solved problems. WiFi is made significantly less useful by the way every public hotspot has its own random billing infrastructure, often with pages that don't work well on mobile devices. Because GSM/UMTS sim cards are secure devices, the same convenience that 3G offers today will be possible everywhere, with operators either paying for the ADSL backhaul on their own, merging with cable/DSL companies to become vertically integrated radio/landline companies, or simply paying people who run LTE femtocells for the cost of the backhaul.
Secondly, LTE is a natively IPv6 based protocol. That means that if you use an LTE/4G enabled NetBook in combination with a home femtocell, there won't be any crap related to WiFi NAT routers as long as you're connecting to an IPv6 site. The devices will probably be controlled and leased by the operators and so won't suffer the same featureitis that has made home internet so flaky and requires so many bizarre workarounds like UPnP today.
Thirdly, hand-off actually works in mobile protocols. 4G/LTE devices will be able to transparently hand-off from your personal home femtocell to a macrocell when you walk outside, to a 3G or even GPRS/2G cell if you roam out of range.... all without you even noticing. Try that with a WiFi based system!
Finally, the LTE protocols include support for true single channel multi-cast. For this reason it can not only replace 2G/3G and WiFi, but also digital terrestrial TV broadcasts, as well as digital and FM radio with no loss in spectrum efficiency due to needless retransmissions.
LTE + IPv6 is the most efficient and user-friendly way to use limited spectrum, period. 20 years from now other air protocols will seem like an anachronism.
NoScript? I'll laugh if it turns out this problem is caused by NoScript or ad blockers. First rule of supporting a complex website - tell users to switch these tools off, clear their cache, cookies and try again (also, privacy proxies/porn filters)
No competent computer user likes to be told to do this routine sort of thing, but the unhappy fact is that there are a lot of people out there that are somewhere between total n00b and web expert, who use tools that screw around with website contents in flight and then can't figure out that it breaks things. I've had to clean up NoScript created messes before. The number of support complaints it created was amazing.
Slashdot Mirror
That's pretty damn cool. It reminds me of scene completion, which is another take on the same idea - combining images from Flickr to create new images according to a brief sketch.
You can request footage of yourself from private cameras using data protection laws.
Anyway, no need to worry for two reasons:
You need to read up on the Xbox 360 security system. You can't compromise an Xbox 360 that's connected to Live, it's not possible. There are no botnets of Xbox 360 machines, and it's highly unlikely there ever will be.
Having a botnet doesn't do much good if you can't automatically create accounts. As you need an XBox to sign up for an Xbox Live account restricting the number of requests an account can send is trivial.
They aren't really "ads" in the traditional sense. X360 has a sort of featured games/videos section, and some videos you can opt to download for trailers. It's promotional but it's a stretch to call them an ad - as the op says, they're actually useful.
That's partly because most indie games aren't that good. Now XBL has a ratings system it'll be easier for the good stuff to float to the top. That said, the expensive part of writing a commercial XBL Arcade or retail game isn't the SDK/dev unit. The expensive part is paying for (and passing) the QA Microsoft put all games through. They test games like crazy which is why console games basically never crash or seriously glitch out.
His enthusiasm can be summed up as "OOXML works, ODF doesn't". I mean, ODF didn't even bother to properly specify spreadsheet formula functions, right? How can anybody take such a "standard" seriously? There was never a choice between OOXML and ODF because ODF as a usable specification did not exist.
Stallman and the FSF have a history of just making shit up when they decide they don't like something. Look at some of the stuff they've written about trusted computing for examples of this. Miguel is right - the guy plays fast and loose with the facts repeatedly. I see lots of people praising his consistency in this discussion. Well guess what - it's easy to have a consistent position over a long period of time when you flatly refuse to accept factual reality.
The point of .NET is not cross platform compatibility. The point is to make it easy for developers to write software - and it succeeds at that. The whole reason Miguel created Mono is that the free software community didn't and still hasn't created any productivity boosting tool equivalent in power.
Um, no. Microsoft don't provide any Windows built in game copy protection, yet nearly every game provides their own. If you think Microsoft refusing to implement DRM would have meant no DRM you're crazy.
I think the argument is that they aren't really needing to do extra work to support the Pre. The extra work they're doing is deliberately un-supporting Palm, which achieves nothing other than annoying the customers of their competition. The real question is why bother with emulating an iPod at all, surely there's a better way for Palm to do this. There must be some quite compelling reason to go with this protocol over some other solution.
I'm pretty sure there's nothing illegal about the Pre telling iTunes it's an iPod. After all, Internet Explorer claims to be Mozilla, and Google Chrome claims to be Chrome, Safari, AND Mozilla! Now if iTunes started issuing firmware hash challenges to iPods, Palm would be stuck - to answer the challenges they'd have to ship a copy of the iTunes firmware which other than being very large would also be illegal. But I guess Apple can't easily update every iPod to support that retroactively.
Really, I've really got to wonder what Apples long term strategy is here. The constant stream of stories like this have to be causing recruitment issues if nothing else. They're already being questioned by the US Govt over the Google Voice issue and now they're apparently issuing updates intended only to break interop? Despite many rumors that is something I don't recall Microsoft ever doing.
ChromeFrame isn't activated unless the website asks for it. So you were just testing the reliability of IE6, not Chrome.
A picture is worth a thousand words .....
Frankly, you sound like somebody who has done a CS course but not written much real world, shipping software.
How many desktop apps does the average person use that are written in Java? The answer is .... none. If you attempt to write a useful desktop app in Java you will soon understand why. Of course, the language is turing complete. It is also designed in such a way that memory usage/working set size is dramatically greater than the equivalent program written in C++, and it's close to impossible to fix this.
About a month ago, I wrote some code that processed a large quantity of data (it was locating tree-like structures within a large graph). The initial version was written in Python because the implementation I inherited was written in Python. However its runtime was measured in hours. I rewrote it in C++ and it now completes in around five minutes. The previous version was impossible to work on: a combination of terrible performance and weak/loose typing meant that if there was a typo near the end of the algorithm, I might only find out about it 6 hours later. This is clearly impractical.
Final example, read up on the Chandler project. One of the biggest problems they had was performance. Python is [a] single threaded [b] interpreted [c] a memory pig and [d] damn near impossible to optimize or compile. This is death by a thousand cuts .... it really doesn't matter how smart your algorithms are if every time you want to do some non-trivial thing your gui stops responding for a few seconds. Works great for simple scripts though.
I don't know if LTE will ever be 100% deployed. LTE chips are backwards compatible with 3G and 2G, right? And they can roam freely between them. LTE makes most sense to deploy in the home first, as a WiFi replacement, and then by upgrading macrocells in urban areas where smartphone traffic is stressing the local infrastructure. Whether LTE gets deployed to suburban or rural areas I can't say - the bandwidth/latency upgrades probably aren't important there for people outside.
LTE won't replace your DSL line - it'll use it. I mean, do you directly plug into your modems ethernet port? That configuration is rare, from what I've seen. Much more common is home wifi. LTE will replace your current wifi router box with a different wireless router box, and you won't know much difference except that if you open up your 4G equipped netbook and start a fast download, then walk outside down to the local park, the only thing that'll happen is your download gets slower. No connection interruptions, no new IP address, no changing in billing providers .... it'll all be consistent.
If you read the letter, it's clear that this non-"solution" is being proposed exactly because the rights holders are refusing to allow their HD content to be broadcast without some kind of DRM. I'm pretty surprised anyone agreed to some kind of nonsense like hiding Huffman tables, but as this is being used only as a stick to make the set top box/TV manufacturers implement stronger client-side DRM, who knows, perhaps it will actually discourage casual piracy. Or perhaps the BBC will have to change the no-encryption principles on which Freeview is based or be banished to a non-HD ghetto.
Well, firstly the UK has 65 million citizens. That's pretty big actually. About 1/5th the population of the United States.
Secondly the UK was the first country to deploy digital TV anywhere in the world. So TV [set top box] manufacturers are used to dealing with this market. A lot of the finnickey details of how it works were hammered out during the initial UK deployment experience.
Thirdly, no, the BBC are not the only supplier of TV programming in the British market. There are two major competitors platform wise - Sky and cable. Freeview is the digital terrestrial platform in the UK (ie, broadcast from poles on hills). It used to be an open consortium run by the BBC and ITV, itself a consortium of independent terrestrial broadcasters, however, ITV have spent the last 10 years showing the world how not to run a TV company so eventually it became infeasible to continue like that and the BBC took complete control to stop the platform imploding.
So what about Sky and cable? Firstly, Sky is owned by News Corp. News Corp also own DirecTV. Murdoch has spent massive amounts of money developing largely "unbreakable" DRM for his satellite TV companies. Whilst I believe it's possible to "break" the DRM by setting up a PC to emulate a set-top box and then, say, broadcasting the result over the internet, there are obviously bandwidth and single-point-of-failure problems with that which make it especially infeasible for HD material. There is I believe no way to decrypt the transmissions without a Sky controlled smart card.
I don't know if Sky set top boxes require HDCP or the like but it wouldn't surprise me, so, HD broadcasts especially of live events are not likely to suffer piracy thanks to Sky any time soon.
The sitation with cable I know less about, but presume it's similar. Freeview is sort of unique in that because it's intended as the non-corporate controlled platform, it doesn't encrypt the broadcasts. This worked OK for a long time because piracy of SD material was seen as a TV company problem. But now content rights holders are apparently refusing to license HD content to insecure platforms. They can do that because the majority are secure and Freeview is the odd one out!
Erm, if I read that article correctly, it means Apple finally discovered that cutting edge new technology, the "thread pool". Oh, and they hacked closures into Objective C. How is that not simply a ThreadPoolExector?
Well, Google/HTC sold devices with root for a while, the "developer phones" .... so it can happen.
O3D does indeed use either D3D or GL depending on platform. O3D is slightly higher level yes, but not dramatically so (you still have to code up shaders etc). I think we'll end up with both. NaCL integration is the key regardless.
AMPS was only decommissioned last year, right? That dates from the 80s, so 20 years seems reasonable.
Well, there are a few major things in O3Ds favor (btw: full disclosure, I am a Googler but have no particular interest in 3D tech beyond enjoying its products).
Firstly, O3D is a plugin that can be installed in Internet Explorer today. The IE team hasn't joined in the sudden frenzy of webapp work that the other browser vendors have been doing. That means things not based on plugins will have limited impact if this doesn't change.
Secondly, O3D embeds V8. Consistent JavaScript performance is going to be an issue for any web-based 3D work. Although it's true that every browser has been getting faster, there are still big differences in what they can do. This may be resolved over time as all the different engines approach the theoretical limits of how fast you can make JavaScript, but until then, having a single known target engine with predictable performance is a win.
Thirdly, canvas vs SVG isn't really a good comparison. 2D graphics are rarely very complex in the same way 3D graphics are. If 2D graphics had the same data throughput, hardware acceleration and tool integration issues that 3D has then an SVG type model would probably look a lot more attractive than a canvas type model. As it happens, for most 2D work a simpler API works well. I remain sceptical that the same is true for 3D.
That said, I am by no means an expert. There is a fun 3D platformer game (Infinite Journey) based on the O3D plugin that I tried - if somebody puts together a webgl demo as compelling as that one is, I'll definitely have to rethink my position.
WiFi has a limited future anyway so who cares? The future is becoming increasingly clear ... over the next 10-20 years most existing air protocols are likely to be phased out in favor of GSM LTE. LTE (and the "Advanced LTE" which is likely to become the actual deployed 4G technology) offer speeds in the hundreds of megabits/sec range and latency in the ~millisecond range. In fact LTE is very close to the theoretical limits of what is physically possible to do, speed wise. LTE is also being designed with support for femtocells in mind right from the start, in fact, there seems to be growing consensus that 4G mobile networks will primarily be deployed through LTE gateways in the home first with traditional cell-tower style macrocells coming much later.
LTE offers some compelling advantages over the mixed 3G/WiFi tech we use today. Firstly, authentication and billing are solved problems. WiFi is made significantly less useful by the way every public hotspot has its own random billing infrastructure, often with pages that don't work well on mobile devices. Because GSM/UMTS sim cards are secure devices, the same convenience that 3G offers today will be possible everywhere, with operators either paying for the ADSL backhaul on their own, merging with cable/DSL companies to become vertically integrated radio/landline companies, or simply paying people who run LTE femtocells for the cost of the backhaul.
Secondly, LTE is a natively IPv6 based protocol. That means that if you use an LTE/4G enabled NetBook in combination with a home femtocell, there won't be any crap related to WiFi NAT routers as long as you're connecting to an IPv6 site. The devices will probably be controlled and leased by the operators and so won't suffer the same featureitis that has made home internet so flaky and requires so many bizarre workarounds like UPnP today.
Thirdly, hand-off actually works in mobile protocols. 4G/LTE devices will be able to transparently hand-off from your personal home femtocell to a macrocell when you walk outside, to a 3G or even GPRS/2G cell if you roam out of range .... all without you even noticing. Try that with a WiFi based system!
Finally, the LTE protocols include support for true single channel multi-cast. For this reason it can not only replace 2G/3G and WiFi, but also digital terrestrial TV broadcasts, as well as digital and FM radio with no loss in spectrum efficiency due to needless retransmissions.
LTE + IPv6 is the most efficient and user-friendly way to use limited spectrum, period. 20 years from now other air protocols will seem like an anachronism.
NoScript? I'll laugh if it turns out this problem is caused by NoScript or ad blockers. First rule of supporting a complex website - tell users to switch these tools off, clear their cache, cookies and try again (also, privacy proxies/porn filters)
No competent computer user likes to be told to do this routine sort of thing, but the unhappy fact is that there are a lot of people out there that are somewhere between total n00b and web expert, who use tools that screw around with website contents in flight and then can't figure out that it breaks things. I've had to clean up NoScript created messes before. The number of support complaints it created was amazing.