Of course they contain info on the file. A hash for each piece. On the other hand, they contain NO information on who has the file, only a reference back to the tracker, which provides that information on demand over HTTP. Any torrent on TPB only contains the TPB web server address, and hashes based on the original file(s).
It also contains the address back to the tracker, which keeps a current list of nodes carrying the real file. TPB is both an indexer (carrying torrent files) and a tracker (keeping track of the peers that are currently available). BitTorrent uses SHA-1, with one hash value for each "piece", the size of which might quite a lot, but it's often in the range 256k-4M. The peers only exchange complete pieces, checking that the result they receive mathces the SHA-1 specified in the torrent. A normal MD5 (or SHA-1) for the complete file would mean that you couldn't do any checking until the complete file had been received, which wouldn't fit too well with the intent.
In addition, of course, it's possible to generate MD5 collisions these days. Although it's still not possible to generate data with a specific hash, you can generate two sets of data with the same hash easily.
The point is that the automated summary is just that, automated. And there HAS been legal precedence around commercial plagiarism involving summaries -- it's not immediately ok just because you've not copied everything.
If the manual summary was detailed enough that you could read the original book/see the original movie and notice if any one frame/sentence was missing or had been replaced, it would obviously be rather detailed...
Hashing would not be illegal. The key is redistribution (and, for that matter, intent). Your hint is actually contradicting your point of view, as the very fact that a summary is a work on its own makes it more acceptable, while the hash is just that - a one-way transformation of the original work. There is no new originality, just a massive lossy compression. The amount of data is still large enough to make a collision by accident next to impossible.
You can just as well say that it's hard to draw the line between mp3 and.torrent. If you downsample far enough and lower the bitrate enough, what's left is a (bad) hashing method.
What is your reasoning behind the.torrent not being copyrighted? You're not allowed to summarize the works of others in any manner in text (although much is allowed), so why should you be allowed to do it in binary? True, this "summary" is basically only useful to make sure you have the correct file, but it is still very directly based on a copyrighted work, with no added value of any kind, it's not like a review of the original work, for example.
"The projectile is rugged enough that it can land on anything without damage."
Without damage to WHAT? But, ok, there are lots of sparsely populated places, in addition to open sea, but it's not like you can just ignore the damage at return, even if it's not your projectile that is damaged.
I think a significant aspect here is the size needed to get a reasonably large volume/area with somewhat consistent perceived gravity (preferably close to 1 g, as well). We simply don't build things that large. When we do, I don't see any big technical problems, but of course one can start to consider proper docking protocols and so on at that point. (Problematic if anyone gets the crazy idea to try to dock the rotating parts of two art-grav vessels with a mismatch in angular velocity.)
If it's anything like the tax in most European countries, it goes hand in hand with the regulation to allow personal copying (including sharing with friends) of songs (note: not software) you have a valid copy of yourself. A result of this is that this copying should at least approximately correspond to the actual sales, while it's quite obvious that downloads can vary far more, and it would therefore be far harder to create a distribution plan that would be "fair".
I got over 800 kbps for a long time while flying Stockholm <=> Chicago a month ago. I think Skype connected just fine, but I was in no need to make any calls. The latency is certainly a problem, but I'm not sure if the normal in-flight phone is any better in this regard, as I think that is also done by satellite uplink on longhaul flights, which should be the most significant part of the delay anyway.
What "surviving lines" are you talking about? Do you subscribe to the theory of (some) RNA viruses actually being highly modified descendants to that primordial life, or something else?
Anyway, one form of life that lacks DNA, and is red, is of course just that - mammalian red blood cells. No nucleus and no mitochondria. And they make a pretty red. I guess an Indian Dracula was hit by a lightning strike (or a flying alien garlic) in mid-air.
What's the point of doing a file type assocation or registering a codec, if only the app/codec itself is able to read it? Some of your examples are typical reasons for why the stuff you install should have global access, but global access is inherently dangerous...
Well, I've never come to terms with the one in Office, at all. The one in OO.org is absolutely usable (as long as you stick to writing in the coded form). It works fine for the occasional formula, but I would of course switch to some TeX-related environment if I really wanted a nice layout. It's just to scratch some things down, where I want some more than plain text, but that's it.
Well, two chip packages will always be more expensive than one. Any PIII could easily be put into a dual-processor system, but the fraction of SMP users was always very low. A normal P4 can't do SMP on its own, Intel chose to make that a differentiator for Xeon. The same thing basically applies to SLI for video -- the affordable way to get 4 cores will not be some "4x4" scheme, but rather just the K8L, in a single package. Likewise, the mainstream users will generally just get a more powerful single video card of a later generation, or even integrated graphics of comparable power.
Even in disks, where the benefits are quite obvious, and the overall limits of storage density, rotation speed and number of platters in one drive are limiting, RAID is not commonplace for normal users. You get far less than 100 % more for over 100 % the cost.
(ok, I'm writing this on a dual-mon, dual-core, quad-HD system, after finally upgrading from a similar setup, but with two distinct PIIIs. That only proves that I'm not a normal user. The point is still that higher integration, not more discrete parts, has been the overall trend. There is no reason to believe that should not continue.)
To be fair, I think that Office 2003 (at least) is "smarter" regarding not redoing changs. Some OO.org releases, on the other hand... I actually noticed this when starting to take lecture notes in OO.org, to use the far superior formula editing there, so I went back and forth trying a couple of samples like this one.
The point is that it's not the app that tries to perform a disallowed action that gives you the "proceed" window. If you compile your own code, or run a shell script, you will still be stopped unless you confirm it. The user is still an admin of sorts, but it's admin with an additional permission filter. (Kind of like being admin, but then normally running with an additional group membership carrying a lot of "Deny" permissions.)
Considering that this figure is an optimized version of XP "hibernation", it should matter far more what the workset was when turning the machine off, and how quick the HD is.
Have a look at how DirectX was really implemented in 9x vs NT vs 2000/XP. (NT4 being the stepchild with a very crippled DirectX support ever released.) It may have been redistributed as an addon, but it talked directly to the kernel drivers, included its own thunking in Win32K.sys (on 2000/XP) and so on. If we compare it with for instance the.NET Framework the difference is very important --.NET doesn't allow you to do anything with the underlying OS that wasn't already exposed in Win32, while that is most of the point of DirectX.
DirectX 10 is a big change, including the fact that DirectX 10 hardware is much of an all-or-nothing scenario, you either provide the full set of features, or you are stuck in a DirectX 9 mode. Considering that, and that there are NO DirectX 10 compliant GPUs right now, I would say that the lack of a backport is not the most significant reason for a slow adoption.
You would naturally need some kind of workspace to store the results. I think that a normal shader is not supposed to write data back to from where it was read, so there can be some quite significant "read, multiply, write to new location" going on.
Slashdot Mirror
Of course they contain info on the file. A hash for each piece. On the other hand, they contain NO information on who has the file, only a reference back to the tracker, which provides that information on demand over HTTP. Any torrent on TPB only contains the TPB web server address, and hashes based on the original file(s).
In addition, of course, it's possible to generate MD5 collisions these days. Although it's still not possible to generate data with a specific hash, you can generate two sets of data with the same hash easily.
If the manual summary was detailed enough that you could read the original book/see the original movie and notice if any one frame/sentence was missing or had been replaced, it would obviously be rather detailed...
You can just as well say that it's hard to draw the line between mp3 and .torrent. If you downsample far enough and lower the bitrate enough, what's left is a (bad) hashing method.
What is your reasoning behind the .torrent not being copyrighted? You're not allowed to summarize the works of others in any manner in text (although much is allowed), so why should you be allowed to do it in binary? True, this "summary" is basically only useful to make sure you have the correct file, but it is still very directly based on a copyrighted work, with no added value of any kind, it's not like a review of the original work, for example.
Without damage to WHAT? But, ok, there are lots of sparsely populated places, in addition to open sea, but it's not like you can just ignore the damage at return, even if it's not your projectile that is damaged.
I think a significant aspect here is the size needed to get a reasonably large volume/area with somewhat consistent perceived gravity (preferably close to 1 g, as well). We simply don't build things that large. When we do, I don't see any big technical problems, but of course one can start to consider proper docking protocols and so on at that point. (Problematic if anyone gets the crazy idea to try to dock the rotating parts of two art-grav vessels with a mismatch in angular velocity.)
I somehow imagine that the pylon construction could be far more expensive than the power plant part.
If it's anything like the tax in most European countries, it goes hand in hand with the regulation to allow personal copying (including sharing with friends) of songs (note: not software) you have a valid copy of yourself. A result of this is that this copying should at least approximately correspond to the actual sales, while it's quite obvious that downloads can vary far more, and it would therefore be far harder to create a distribution plan that would be "fair".
I got over 800 kbps for a long time while flying Stockholm <=> Chicago a month ago. I think Skype connected just fine, but I was in no need to make any calls. The latency is certainly a problem, but I'm not sure if the normal in-flight phone is any better in this regard, as I think that is also done by satellite uplink on longhaul flights, which should be the most significant part of the delay anyway.
Anyway, one form of life that lacks DNA, and is red, is of course just that - mammalian red blood cells. No nucleus and no mitochondria. And they make a pretty red. I guess an Indian Dracula was hit by a lightning strike (or a flying alien garlic) in mid-air.
What's the point of doing a file type assocation or registering a codec, if only the app/codec itself is able to read it? Some of your examples are typical reasons for why the stuff you install should have global access, but global access is inherently dangerous...
Thanks for trying to help, anyway.
Even in disks, where the benefits are quite obvious, and the overall limits of storage density, rotation speed and number of platters in one drive are limiting, RAID is not commonplace for normal users. You get far less than 100 % more for over 100 % the cost.
(ok, I'm writing this on a dual-mon, dual-core, quad-HD system, after finally upgrading from a similar setup, but with two distinct PIIIs. That only proves that I'm not a normal user. The point is still that higher integration, not more discrete parts, has been the overall trend. There is no reason to believe that should not continue.)
To be fair, I think that Office 2003 (at least) is "smarter" regarding not redoing changs. Some OO.org releases, on the other hand... I actually noticed this when starting to take lecture notes in OO.org, to use the far superior formula editing there, so I went back and forth trying a couple of samples like this one.
The point is that it's not the app that tries to perform a disallowed action that gives you the "proceed" window. If you compile your own code, or run a shell script, you will still be stopped unless you confirm it. The user is still an admin of sorts, but it's admin with an additional permission filter. (Kind of like being admin, but then normally running with an additional group membership carrying a lot of "Deny" permissions.)
Considering that this figure is an optimized version of XP "hibernation", it should matter far more what the workset was when turning the machine off, and how quick the HD is.
DirectX 10 is a big change, including the fact that DirectX 10 hardware is much of an all-or-nothing scenario, you either provide the full set of features, or you are stuck in a DirectX 9 mode. Considering that, and that there are NO DirectX 10 compliant GPUs right now, I would say that the lack of a backport is not the most significant reason for a slow adoption.
You are rubber, I am glue.
It can stop your paranoia caused by all those glitches in the matrix you would otherwise assume as the only reasonable explanation.
And those missing things are just about my only complaints... On good displays, I find serifs highly superior.
Actually, last thing I heard most BSD forks are actually undead by now, ready to become our new free-as-in-not-RMS overlords.
Yeah, immediate the transfer of the "immediate party" button from the dorm room to the cockpit context.
You would naturally need some kind of workspace to store the results. I think that a normal shader is not supposed to write data back to from where it was read, so there can be some quite significant "read, multiply, write to new location" going on.
Intel?