Each and every dollar dedicated to creationist propaganda makes it easier to further the creationist agenda and to introduce creationism and other religious doctrines in schools.
If we give them no money - if nobody visits, they earn nothing. With time, we can make stupidity, if not painful, at least expensive.
Sorry. What I wanted to say is that they never, ever, should think it's acceptable to truncate data that does not fit.
It's kind of the "On Error Resume Next" mentality that some VB programmers use (or used, 10 years ago, when I still worked at consulting for companies whose in-house programs were too buggy).
Playing the Devil's advocate, any given pixel on a MacBook display can show 2 ^ 18 distinct colours, which is to say about 262 thousand or a little less than what a human eye can distinguish, but any given _contiguous_screen_region_of_sufficient_size_ can display a lot more dithered colours, so, the display has that advertised number of pixels _and_ is capable of displaying several million colours _at_the_same_time_.
Not that I find it pretty that Apple and the rest of the industry (I doubt my HP notebook has a much better display) keep doing it.
As an useful example on how communication and interpersonal skills (like properly phrasing a question so not to trigger the troll flag), Theo (of *BSD fame) may not really be the [insert opinion here] he appears to be. I think his lack of empathy and, therefore, his inability to properly communicate and to relate to other people is to account for his reputation.
If that does not ring a bell, perhaps I may suggest you to seek a good shrink in your area to help you understand and deal with this "non-problem".
BTW, if you read this, Theo, it is valid suggestion. I strongly advise you to take it.
The question mark indicated a question. The rest of the question implied a "I know better" attitude that is utterly unwelcome just about everywhere.
It would take you about 10 seconds of googling to find out most drivers on Linux are loadable drivers and are not compiled into the kernel, but belong to what we call the kernel because they are maintained together and released as a single set of files.
It's easy to imagine a Microsoft exec saying "Nice driver they made for your hardware. But it would be a shame if the device didn't work under Windows anymore, wouldn't it?"
Microsoft never did a big kernel switch. They had two different kernels - the "classic" one that was in 3.x with win32s and 9x up to Windows Me and the NT that started back in the 3.1 days and went up to the 2000, in two different product lines. They co-existed in different products for many years. What MS did is that they kept more or less the same look and feel on two different kernels and they simply EOL-ed the older one when they came out with Windows XP.
They never did a kernel transition within the same OS.
They only discontinued the 9x line when they thought XP was good enough for playing games.
Which, by the way, is pretty much what XP is good for;-)
US$ 40 million for a power grid made of superconducting cryogenic cables is incredibly cheap, considering it's the Minipax^H^H^H^H^H^H^H Homeland Security Agency that's funding part of it.
Anyway, if this demonstrates that superconducting cable is a more efficient way to build a power grid (don't forget to count the energy that's required to keep it superconducting), I am all in. Maybe a project this scale is what's needed for the technology to be refined and really take off.
Just imagine the bad sci-fi movie in that a terrorist uses an EMP attached to the superconducting grid to generate a time-warp that sends New York a million years into the past...
So yes, I think it would be hard to use DRM to block foreign news any more than they can be blocked without DRM. Feel free to enlighten me on how it could be done, though.
By mandating that all players sold in your market also obey decency flags that can only be set by the Ministry of Decency should be enough of an example. Adding mandatory self-updates to the equipment would forbid anyone to tamper with such controls. Making it a crime to tamper with them would also not help. IIRC, TiVos can be instructed to erase recordings after a certain amount of time. Truth is DRM by itself does not provide means to do it, but laws can be crafted in such a way as to use DRM as a tool to do it and, since the machines are not really under your control, they can do pretty much anything they want.
You know, democracy was developed in a time without any audio or video recording equipment... I'd be more concerned with self-disintegrating paper, introduced in the name of recycling, that would disintegrate in 20 years... Once somebody proposes that, then you know you should be scared (and I'm not saying this merely as a joke...).
Obviously I do know democracy has evolved without audio or video recording, but it also evolved without any technical means for mass surveillance. Such means exist now. Even the best equipped totalitarian state of the first half of the past century could not summon the vast amount of information that is a subpoena away now and, even if they had that data, they would not be able to sort through it. This is very possible now and has been for the last couple decades. Our IRS-equivalent in Brazil is capable of massive data-mining to spot tax-evasion. In São Paulo, cameras can recognize license plates and issue fines accordingly (people are not allowed to drive in certain days of the week at certain times according to the numbers in their license plates). This capabilities hint on what a determined government could do if laws allowed it.
While it may be impossible to revise the books you have on your shelf, it would be technically trivial to revise e-books in a library or the news published on a web-site. "Your edition of George Owell's 1984 is being updated. Please do not disconnect".
It is terribly unwise to ignore such possibilities until it is too late. Because, of course, by then, it will be too late.
And yes, I too would be terribly scared of self-disintegrating paper, but I am already terribly scared of computers I can't control.
And I don't see the problem of information control. Quite the opposite, if you have the freedom to view any news broadcast from all over the world with a click (well, that's reality even now, I think), there's no control. If people want information, they'll get it easily (well, at least here in the free world). If they don't want it, no DRM is going to make them want it.
Do you really think it would be hard to block your access to foreign news broadcasts via DRM?
The mere existence of this broadcast flag threatens your ability to record the present and document the past. It drives a nail through some of the more basic requirements for a democracy, which is the right and need to be and stay well informed.
After all, we've always been at war with Eastasia.
OK. You got the point. I could say "a valid, unpadded, bzip2 stream" and this argument would disappear.
I could also encode my voice with the document as a direct sample sequence, but then one could tweak the least significant bits of every sample in order to get a hash collision, but the kind of effort required to do that without touching any other hash we could generate out of the file (and we could generate a dozen of them) and still have something that could pass for my voice borders the impossible so close it would risk being deported back to the realm of the possible.
But we are basically saying the same thing - it is possible, but it is impossibly-hard to do it.
If hashes are evenly distributed and we are considering a 20 million byte file, you would get a collision every 2 ^ (20 million - 128) times you tweak the file in a unique way. My brain refuses to dedicate itself to make sure of this number, but 2 ^ 20e6 is a lot of work.
And using a program (postcript is a programming language) is cheating. Given different inputs (i.e. date, who is looking) the same program can display anything. One should always know what is being shown. Although the document is the same, the output when it runs on a PS interpreter is vastly different. One should never think of a program as a static document. And, IIRC, this programmability extends to PDF.
As pegr pointed out, there cannot be an unbreakable hash and hashes may only buy you time.
However as someone else pointed out, if you divulge further information on your file, while it won't make it impossible to have a different file with the same hash, it will make far less likely that someone can bring out a file with the same hash that also has that same length, is a valid bzip2 stream and, after decompression, has the same internal structure (is a valid OOo document). There is a finite number of files with a given length and thus, a finite number of hash collisions at that same file length. If none of the collisions is also a valid bzip2 stream that encodes a valid OOo file, you have, effectively, a pretty much as-good-as-it-gets almost-unforgeable dataset.
Of course, the vestigial mathematician inside me says that, in order to create that unforgeable hash plus the set of rules the file must conform to in order to be considered authentic, you will have to disclose a volume of information equal or larger than the file itself.
There are several ways to decrease the risk of a hash collision. The most obvious one is to publish both the SHA and, say, an MD5 of the same data. This way, anyone who wants a collision, will have to make a collision that happens in both algorithm.
You can also run the hash in your hard disk backwards too, so, unless your HD holds the longest palindrome ever, the same collision will not work.
You could also sha1 both even and odd bytes separately. You could also hash them separately with multiple algorithms. Well.. You get the idea.
I doubt many sha1 collisions will also even be CRC32 collisions.
The breakthrough will not come from batteries but most probably from improved processors, memory-based storage and better display efficiencies.
If you don't need Windows games, you could go with an open-source OS, which means anything ARM or MIPS-based could be fair game. I can easily imagine a asynchronous multi-core ARM processor that runs Linux fast enough for me as long as it has enough memory to keep Firefox happy. Just doing away with PC compatibility would increase efficiencies (imagine not having a vestigial IBM 5150 inside your chipset).
Memory-based storage devices also are much more power-efficient than their spinning disk counterparts and, if their latency and bandwidth are good enough, unused main memory could even be powered down. Right now, my computer reports about 50% of memory used for caches, so, about half of my DIMMs would not need power.
If non-volatile memory becomes fast and durable enough to be used as main memory, then we could simplify the design even further. Less stuff means less power.
Sony is already selling LED-backlit LCD laptops. LEDs require a lot less power than fluorescent backlights, so, again, there is room for improvement. Another major improvement could come from the OLPC project - their LCD requires even less power than a traditional one because they eliminate filters (which throw away 66% of the backlight) and split light in RGB components instead. The OLPC hardware also can take over many functions traditionally CPU-bound, saving power and pointing the way for future power-efficient laptops - I suspect the OLPC could already run for days on my current batteries.
The OLPC folks are also responsible for a lot of power-saving enhancements inside the Linux kernel itself and those could improve battery life of my _current_ computer without any extra hardware.
So, if done this way, most probably a laptop with the same form factor as mine could run for 3 to 10 times longer on the same batteries.
When I graduated as an engineer, I took an oath. I swore to protect people and not let the devices I design and build be used for evil. I have so far kept to my word.
I can't imagine how journalists, whose code of conduct obliges them to inform the people and to seek out the truth no matter how dark it is, feel when they are coerced into hiding their findings or not investigating some subjects.
Theirs is a job of the utmost importance to the defense of democracy and freedom and we rely on them to do their jobs.
Slashdot Mirror
Very good research?!
Where can it be found?
Each and every dollar dedicated to creationist propaganda makes it easier to further the creationist agenda and to introduce creationism and other religious doctrines in schools.
If we give them no money - if nobody visits, they earn nothing. With time, we can make stupidity, if not painful, at least expensive.
Sorry. What I wanted to say is that they never, ever, should think it's acceptable to truncate data that does not fit.
It's kind of the "On Error Resume Next" mentality that some VB programmers use (or used, 10 years ago, when I still worked at consulting for companies whose in-house programs were too buggy).
Sorry. I was talking about the truncation of data, not the other way around.
The first time I saw a VARCHAR2 I wanted to puke.
This should never, ever be even considered to ship enabled.
It is so horrid it should be forbidden and deemed illegal in any sane jurisdiction.
Somehow, I don't think the sets "people who think twice" and "people who buy Zune" share any meaningful intersection.
Playing the Devil's advocate, any given pixel on a MacBook display can show 2 ^ 18 distinct colours, which is to say about 262 thousand or a little less than what a human eye can distinguish, but any given _contiguous_screen_region_of_sufficient_size_ can display a lot more dithered colours, so, the display has that advertised number of pixels _and_ is capable of displaying several million colours _at_the_same_time_.
Not that I find it pretty that Apple and the rest of the industry (I doubt my HP notebook has a much better display) keep doing it.
As an useful example on how communication and interpersonal skills (like properly phrasing a question so not to trigger the troll flag), Theo (of *BSD fame) may not really be the [insert opinion here] he appears to be. I think his lack of empathy and, therefore, his inability to properly communicate and to relate to other people is to account for his reputation.
If that does not ring a bell, perhaps I may suggest you to seek a good shrink in your area to help you understand and deal with this "non-problem".
BTW, if you read this, Theo, it is valid suggestion. I strongly advise you to take it.
The question mark indicated a question. The rest of the question implied a "I know better" attitude that is utterly unwelcome just about everywhere.
It would take you about 10 seconds of googling to find out most drivers on Linux are loadable drivers and are not compiled into the kernel, but belong to what we call the kernel because they are maintained together and released as a single set of files.
It's easy to imagine a Microsoft exec saying "Nice driver they made for your hardware. But it would be a shame if the device didn't work under Windows anymore, wouldn't it?"
Sorry, but you have your facts wrong.
;-)
Microsoft never did a big kernel switch. They had two different kernels - the "classic" one that was in 3.x with win32s and 9x up to Windows Me and the NT that started back in the 3.1 days and went up to the 2000, in two different product lines. They co-existed in different products for many years. What MS did is that they kept more or less the same look and feel on two different kernels and they simply EOL-ed the older one when they came out with Windows XP.
They never did a kernel transition within the same OS.
They only discontinued the 9x line when they thought XP was good enough for playing games.
Which, by the way, is pretty much what XP is good for
It must be US$ 40 million a mile.
US$ 40 million for a power grid made of superconducting cryogenic cables is incredibly cheap, considering it's the Minipax^H^H^H^H^H^H^H Homeland Security Agency that's funding part of it.
Anyway, if this demonstrates that superconducting cable is a more efficient way to build a power grid (don't forget to count the energy that's required to keep it superconducting), I am all in. Maybe a project this scale is what's needed for the technology to be refined and really take off.
Just imagine the bad sci-fi movie in that a terrorist uses an EMP attached to the superconducting grid to generate a time-warp that sends New York a million years into the past...
By mandating that all players sold in your market also obey decency flags that can only be set by the Ministry of Decency should be enough of an example. Adding mandatory self-updates to the equipment would forbid anyone to tamper with such controls. Making it a crime to tamper with them would also not help. IIRC, TiVos can be instructed to erase recordings after a certain amount of time. Truth is DRM by itself does not provide means to do it, but laws can be crafted in such a way as to use DRM as a tool to do it and, since the machines are not really under your control, they can do pretty much anything they want.
You know, democracy was developed in a time without any audio or video recording equipment... I'd be more concerned with self-disintegrating paper, introduced in the name of recycling, that would disintegrate in 20 years... Once somebody proposes that, then you know you should be scared (and I'm not saying this merely as a joke...).Obviously I do know democracy has evolved without audio or video recording, but it also evolved without any technical means for mass surveillance. Such means exist now. Even the best equipped totalitarian state of the first half of the past century could not summon the vast amount of information that is a subpoena away now and, even if they had that data, they would not be able to sort through it. This is very possible now and has been for the last couple decades. Our IRS-equivalent in Brazil is capable of massive data-mining to spot tax-evasion. In São Paulo, cameras can recognize license plates and issue fines accordingly (people are not allowed to drive in certain days of the week at certain times according to the numbers in their license plates). This capabilities hint on what a determined government could do if laws allowed it.
While it may be impossible to revise the books you have on your shelf, it would be technically trivial to revise e-books in a library or the news published on a web-site. "Your edition of George Owell's 1984 is being updated. Please do not disconnect".
It is terribly unwise to ignore such possibilities until it is too late. Because, of course, by then, it will be too late.
And yes, I too would be terribly scared of self-disintegrating paper, but I am already terribly scared of computers I can't control.
Do you really think it would be hard to block your access to foreign news broadcasts via DRM?
The mere existence of this broadcast flag threatens your ability to record the present and document the past. It drives a nail through some of the more basic requirements for a democracy, which is the right and need to be and stay well informed.
After all, we've always been at war with Eastasia.
While I agree lookalike characters should not be allowed, what you can or cannot type is entirely dependent on your operating environment.
the same planet we heard about a couple weeks back, that is 50% more massive than Earth and with 1.5 times our gravity?
OK. You got the point. I could say "a valid, unpadded, bzip2 stream" and this argument would disappear.
I could also encode my voice with the document as a direct sample sequence, but then one could tweak the least significant bits of every sample in order to get a hash collision, but the kind of effort required to do that without touching any other hash we could generate out of the file (and we could generate a dozen of them) and still have something that could pass for my voice borders the impossible so close it would risk being deported back to the realm of the possible.
But we are basically saying the same thing - it is possible, but it is impossibly-hard to do it.
If hashes are evenly distributed and we are considering a 20 million byte file, you would get a collision every 2 ^ (20 million - 128) times you tweak the file in a unique way. My brain refuses to dedicate itself to make sure of this number, but 2 ^ 20e6 is a lot of work.
And using a program (postcript is a programming language) is cheating. Given different inputs (i.e. date, who is looking) the same program can display anything. One should always know what is being shown. Although the document is the same, the output when it runs on a PS interpreter is vastly different. One should never think of a program as a static document. And, IIRC, this programmability extends to PDF.
As pegr pointed out, there cannot be an unbreakable hash and hashes may only buy you time.
However as someone else pointed out, if you divulge further information on your file, while it won't make it impossible to have a different file with the same hash, it will make far less likely that someone can bring out a file with the same hash that also has that same length, is a valid bzip2 stream and, after decompression, has the same internal structure (is a valid OOo document). There is a finite number of files with a given length and thus, a finite number of hash collisions at that same file length. If none of the collisions is also a valid bzip2 stream that encodes a valid OOo file, you have, effectively, a pretty much as-good-as-it-gets almost-unforgeable dataset.
Of course, the vestigial mathematician inside me says that, in order to create that unforgeable hash plus the set of rules the file must conform to in order to be considered authentic, you will have to disclose a volume of information equal or larger than the file itself.
Make that 20 years and your chances of reading a valid Excel 2007 file in Excel 2027 are every bit as small as a hash collision.
"What is a 'file', granpa?"
There are several ways to decrease the risk of a hash collision. The most obvious one is to publish both the SHA and, say, an MD5 of the same data. This way, anyone who wants a collision, will have to make a collision that happens in both algorithm.
You can also run the hash in your hard disk backwards too, so, unless your HD holds the longest palindrome ever, the same collision will not work.
You could also sha1 both even and odd bytes separately. You could also hash them separately with multiple algorithms. Well.. You get the idea.
I doubt many sha1 collisions will also even be CRC32 collisions.
They already do. They call it Geode http://www.amd.com/us-en/ConnectivitySolutions/Pro ductInformation/0,,50_2330_9863,00.html.
The breakthrough will not come from batteries but most probably from improved processors, memory-based storage and better display efficiencies.
If you don't need Windows games, you could go with an open-source OS, which means anything ARM or MIPS-based could be fair game. I can easily imagine a asynchronous multi-core ARM processor that runs Linux fast enough for me as long as it has enough memory to keep Firefox happy. Just doing away with PC compatibility would increase efficiencies (imagine not having a vestigial IBM 5150 inside your chipset).
Memory-based storage devices also are much more power-efficient than their spinning disk counterparts and, if their latency and bandwidth are good enough, unused main memory could even be powered down. Right now, my computer reports about 50% of memory used for caches, so, about half of my DIMMs would not need power.
If non-volatile memory becomes fast and durable enough to be used as main memory, then we could simplify the design even further. Less stuff means less power.
Sony is already selling LED-backlit LCD laptops. LEDs require a lot less power than fluorescent backlights, so, again, there is room for improvement. Another major improvement could come from the OLPC project - their LCD requires even less power than a traditional one because they eliminate filters (which throw away 66% of the backlight) and split light in RGB components instead. The OLPC hardware also can take over many functions traditionally CPU-bound, saving power and pointing the way for future power-efficient laptops - I suspect the OLPC could already run for days on my current batteries.
The OLPC folks are also responsible for a lot of power-saving enhancements inside the Linux kernel itself and those could improve battery life of my _current_ computer without any extra hardware.
So, if done this way, most probably a laptop with the same form factor as mine could run for 3 to 10 times longer on the same batteries.
Dammit. You destroyed my little troll with reality...
When I graduated as an engineer, I took an oath. I swore to protect people and not let the devices I design and build be used for evil. I have so far kept to my word.
I can't imagine how journalists, whose code of conduct obliges them to inform the people and to seek out the truth no matter how dark it is, feel when they are coerced into hiding their findings or not investigating some subjects.
Theirs is a job of the utmost importance to the defense of democracy and freedom and we rely on them to do their jobs.
Chances are he would be arrested at the airport and sent to Guantanamo or some clandestine prison because of his connections with terrorism.
Of course, the connections won't be disclosed as it would damage the ongoing investigation on his crimes.