Heh. I'm thinking about switching from Comcast to Uverse because it looks like I can get substantially similar service (minus some channels I don't use) for half as much. I've been trying to figure out what I'm missing that will make it actually not a savings.
Yeah, but everyone keeps saying the benefit of liquid petroleum fuel is ease of transport, fueling, and energy density. If we can get the total energy expenses down to parity with long distance electrical transmission, we can get much of the best of both worlds: reduced net carbon emission (because you're sucking it from the air, not the ground), fast fill ups, and not as much need to beef up the energy grid.
There has not been a working commercial plant, but there has been a working thorium reactor. Oak Ridge had one running for 15 thousand hours. But the folks running the AEC wanted plutonium, so they shut down thorium research in 1973. wikipedia has more detail.
The only part of the DMCA complaint that is under penalty of perjury is the person doing the filing claiming they are an authorized agent of the purported complainer. I, as a non-Sony employee, could not issue takedown requests in Sony's name (probably because folks might deliberately file bogus requests in order to make Sony look dumb. Er). The rest of it is all on "good faith belief" basis, which is nigh impossible to disprove. In particular "our computer system says it looks like a match, and the system hasn't been shown to be totally ridiculously inaccurate yet" qualifies, so far at least.
Now, the rebuttal to the complaint, that is if I recall correctly under penalty of perjury. But the takedown request is utterly safe to file.
This is essentially what IEX does (described in this article); putting in the delay does have an effect because it increases the minimum turnaround time which reduces the "fast" players' ability to see what the "slow" players are doing and react. Or, put another way, a 1ms round trip can do 100 times as many transactions as a 100ms round trip, but a 501ms round trip is not that much better than a 600ms round trip.
I assume you're addressing the "seeing the same file uploaded again" part, since you seem to agree with what it would take to avoid noticing that. The thing is, most file hosting sites I've heard of either (a) don't claim to be deduplicating stuff or... well, really, aside from Dropbox, I don't know of a file sharing site that claims to deduplicate. Most of them don't see a need to release implementation details like that. Dropbox, however, is known to compare incoming files to existing files, with extreme accuracy, at which point keeping a blacklist of items which should not be distributable is trivial and the *AA could easily assert (and sound sort of reasonable) that it's their responsibility to do so. They may have even said as much to Dropbox already. And since Dropbox wants to stay out of court and in business, and it is such an easy step to take to cut down on distribution of questionable materials (not storage, please note) there we go.
Can we also start focusing on "description sufficient for someone skilled in the art to duplicate the invention"? I think that would also take a bite out of vagueness. "I built that according to your instructions and it doesn't do that other stuff at all."
While I dislike many things about the DMCA, I believe you are in error. The DMCA text (yeah, it's a weird place to find it, but it's what I found first) specifies that you can't decrypt a work _protected by this title_ (which is to say, Title 17, which contains copyright law). Once it's out of copyright, it's no longer protected and the prohibition no longer applies.
It's more of a repackage of rsync, hooked into a daemon that watches the filesystem. I still wouldn't call it trivial, or I'd have my own version working between my fileserver and my laptop already.
While the text does not specifically address reuploading that I see, I think it would be very hard to make a case that seeing the same file uploaded again would not constitute being "aware of facts or circumstances from which infringing activity is apparent" [sec 202, squiggle 512 (c) "INFORMATION RESIDING ON SYSTEMS OR NETWORKS AT DIRECTION OF USERS" (1)(A)(ii)]. However, in the absence of a specific statement, I don't see that designing a system to avoid being able to prevent uploading would be illegal. It would probably get questioned and even accused of being contributory, and it would likely have to avoid deduplication or any other automatic comparison of files, which would increase costs, but I don't see that it would be illegal.
Negligible. (I calculated in another post that the odds of a hash collision for SHA-1 and a trillion files was about 1 in 2^79; I have since learned that they actually use SHA-256, so make that 1 in 2^175).
If you think it's worth worrying about then it's a legitimate concern for you; I wouldn't worry about it.
It's not quite the same thing, but the lottery is still more likely. Assume a trillion files (1e12). Assume Sha-1, for 160 bit hashes. Then the probability of a collision is less than or equal to 1e24/2 * 1/2^160. 1e3 is _rougly_ 2^10 so call it 2^80/2 * 1/2^160, with a final result of about 1/2^79. Your odds of winning the powerball on one ticket are more on the order of 1 in a couple of billion (2 * 1e9, call it 2^31) so you're still vastly more likely to win the lottery than find a collision in a trillion files.
How does the existence of invalid DMCA requests cause it to be in Dropbox's interest to expend more manpower handling DMCA requests than necessary? You think even if they did it one request at a time by hand it would prevent fake DMCA requests from getting processed? Not happening. Their safe harbor is based on them processing the request on the assumption that it's legit. You, the file holder, can then challenge that and it's no longer their problem. But the initial request is gonna get honored.
Netflix is also not exceptionally good at keeping their anime disc collections complete. I added Kiddy Grade to my queue and about half the discs went into "save for later" land. *sigh*
even some stuff that used to be english-only (Soul Eater comes to mind) now has subtitles available. (I asked the Funimation rep at a convention about language options on netflix at one time, and he said "we really are working on it"; looks like they got it together.)
Oh, it's been a net loss for quite a while. Video has a pretty good back catalog, but wanna place bets on how much of the stuff you liked reading as a kid is still in print? And it won't be public domain until your grandkids are grown, if then.
The end users (collectively) pay for more than just the last mile; they pay for the whole thing. Yes, if everyone stopped downloading from netflix at primetime and downloaded the equivalent amount of data from 500 different sources, it's still the same amount of data and it's still going to overload ATT's unmaintained pipes. But ATT wouldn't have a single target to try to gouge, so they'd pretty much have to recover improvement costs from their usual sources - their customers.
I don't really think they should charge the user more. I think they should man up and make the network improvements they've been collecting money for for the past decade or more. But I think if they do need to charge more, they should just do it and not try a pr-spin end run that winds up not saving their customers anything.
Slashdot Mirror
Heh. I'm thinking about switching from Comcast to Uverse because it looks like I can get substantially similar service (minus some channels I don't use) for half as much. I've been trying to figure out what I'm missing that will make it actually not a savings.
This. Increase costs, and the point of maximum profit may increase even if the number of subscribers at that point decreases.
or if the shirt had a telescope on it.
Yeah, but everyone keeps saying the benefit of liquid petroleum fuel is ease of transport, fueling, and energy density. If we can get the total energy expenses down to parity with long distance electrical transmission, we can get much of the best of both worlds: reduced net carbon emission (because you're sucking it from the air, not the ground), fast fill ups, and not as much need to beef up the energy grid.
There has not been a working commercial plant, but there has been a working thorium reactor. Oak Ridge had one running for 15 thousand hours. But the folks running the AEC wanted plutonium, so they shut down thorium research in 1973. wikipedia has more detail.
The only part of the DMCA complaint that is under penalty of perjury is the person doing the filing claiming they are an authorized agent of the purported complainer. I, as a non-Sony employee, could not issue takedown requests in Sony's name (probably because folks might deliberately file bogus requests in order to make Sony look dumb. Er). The rest of it is all on "good faith belief" basis, which is nigh impossible to disprove. In particular "our computer system says it looks like a match, and the system hasn't been shown to be totally ridiculously inaccurate yet" qualifies, so far at least.
Now, the rebuttal to the complaint, that is if I recall correctly under penalty of perjury. But the takedown request is utterly safe to file.
And it also compresses the data really really well :)
This is essentially what IEX does (described in this article); putting in the delay does have an effect because it increases the minimum turnaround time which reduces the "fast" players' ability to see what the "slow" players are doing and react. Or, put another way, a 1ms round trip can do 100 times as many transactions as a 100ms round trip, but a 501ms round trip is not that much better than a 600ms round trip.
Because they also needed functionality that TCP didn't suit, like ICMP and UDP, and didn't want to duplicate all of the stuff in IP for each of them.
I assume you're addressing the "seeing the same file uploaded again" part, since you seem to agree with what it would take to avoid noticing that. The thing is, most file hosting sites I've heard of either (a) don't claim to be deduplicating stuff or... well, really, aside from Dropbox, I don't know of a file sharing site that claims to deduplicate. Most of them don't see a need to release implementation details like that. Dropbox, however, is known to compare incoming files to existing files, with extreme accuracy, at which point keeping a blacklist of items which should not be distributable is trivial and the *AA could easily assert (and sound sort of reasonable) that it's their responsibility to do so. They may have even said as much to Dropbox already. And since Dropbox wants to stay out of court and in business, and it is such an easy step to take to cut down on distribution of questionable materials (not storage, please note) there we go.
Can we also start focusing on "description sufficient for someone skilled in the art to duplicate the invention"? I think that would also take a bite out of vagueness. "I built that according to your instructions and it doesn't do that other stuff at all."
While I dislike many things about the DMCA, I believe you are in error. The DMCA text (yeah, it's a weird place to find it, but it's what I found first) specifies that you can't decrypt a work _protected by this title_ (which is to say, Title 17, which contains copyright law). Once it's out of copyright, it's no longer protected and the prohibition no longer applies.
It's more of a repackage of rsync, hooked into a daemon that watches the filesystem. I still wouldn't call it trivial, or I'd have my own version working between my fileserver and my laptop already.
While the text does not specifically address reuploading that I see, I think it would be very hard to make a case that seeing the same file uploaded again would not constitute being "aware of facts or circumstances from which infringing activity is apparent" [sec 202, squiggle 512 (c) "INFORMATION RESIDING ON SYSTEMS OR NETWORKS AT DIRECTION OF USERS" (1)(A)(ii)]. However, in the absence of a specific statement, I don't see that designing a system to avoid being able to prevent uploading would be illegal. It would probably get questioned and even accused of being contributory, and it would likely have to avoid deduplication or any other automatic comparison of files, which would increase costs, but I don't see that it would be illegal.
It's not quite the same thing, but the lottery is still more likely. Assume a trillion files (1e12). Assume Sha-1, for 160 bit hashes. Then the probability of a collision is less than or equal to 1e24/2 * 1/2^160. 1e3 is _rougly_ 2^10 so call it 2^80/2 * 1/2^160, with a final result of about 1/2^79. Your odds of winning the powerball on one ticket are more on the order of 1 in a couple of billion (2 * 1e9, call it 2^31) so you're still vastly more likely to win the lottery than find a collision in a trillion files.
How does the existence of invalid DMCA requests cause it to be in Dropbox's interest to expend more manpower handling DMCA requests than necessary? You think even if they did it one request at a time by hand it would prevent fake DMCA requests from getting processed? Not happening. Their safe harbor is based on them processing the request on the assumption that it's legit. You, the file holder, can then challenge that and it's no longer their problem. But the initial request is gonna get honored.
yes, but you believe in accepting risk. Many unfortunately believe that the duty of government is to eliminate risk (to them and theirs, at least).
You can't get arrested for being dangerous.
but apparently you can be punished for it, without charges or trial.
I forget which series I was looking at, but I remember it started with season 2. WTF?
Netflix is also not exceptionally good at keeping their anime disc collections complete. I added Kiddy Grade to my queue and about half the discs went into "save for later" land. *sigh*
even some stuff that used to be english-only (Soul Eater comes to mind) now has subtitles available. (I asked the Funimation rep at a convention about language options on netflix at one time, and he said "we really are working on it"; looks like they got it together.)
what would you consider a real media center package? I'm looking for good options to build my system with :)
Oh, it's been a net loss for quite a while. Video has a pretty good back catalog, but wanna place bets on how much of the stuff you liked reading as a kid is still in print? And it won't be public domain until your grandkids are grown, if then.
The end users (collectively) pay for more than just the last mile; they pay for the whole thing. Yes, if everyone stopped downloading from netflix at primetime and downloaded the equivalent amount of data from 500 different sources, it's still the same amount of data and it's still going to overload ATT's unmaintained pipes. But ATT wouldn't have a single target to try to gouge, so they'd pretty much have to recover improvement costs from their usual sources - their customers.
I don't really think they should charge the user more. I think they should man up and make the network improvements they've been collecting money for for the past decade or more. But I think if they do need to charge more, they should just do it and not try a pr-spin end run that winds up not saving their customers anything.