The DMCA provides no protection for signing keys, because they cannot be used to circumvent any copy protection (and in fact TI doesn't even have copy protection on its OS, nor would it be productive to include it).
The EFF has some stuff to read to read on the issue; specifically, the EFF's letter to TI regarding the takedown notice explains in detail why the DMCA does not apply here.
In this case, the keys do not protect the OS. Instead, the keys are merely used to sign the OS so that the calculator will run it.
In other words, encryption keys may be covered under the DMCA, but that's not the issue here. TI issued takedowns for signing keys, which are not used for DRM purposes, nor are they used to circumvent anything.
The other interesting issue here is that TI does not suffer any financial loss by having the signing keys be public knowledge - it does not charge for the calculator OSes, and before you can use a third-party OS you already have to have purchased a TI calculator.
In other words, TI has zero financial incentive to try to suppress these signing keys.
You've just increased the number of inodes your OS takes by a significant amount.
What, you're going to release a distro cd made up entirely of FatELF binaries? That seems a bad idea in more ways than one. Your libraries will be no different; you're not going to create FatELF-based library files, are you?
Also remember, a significant portion of the files included in any distro are configuration files (or other data files); these are not duplicated. This is even more true of third-party software.
FatELF doesn't even seem useful for OSes (except maybe LiveCDs) because the package manager is going to select arch-specific packages anyway. Point being, your "built-in" OS executables will only have one file anyway, since FatELF is redundant for them, meaning FatELF is only useful for third-party software. But when your third-party software has two dozen data files or more, what's another inode or three for an extra binary and a script to choose which one you run?
(For example, Valve's Source servers do this. If you're concerned about inode counts on a game server, you need to upgrade your game server.)
I guess that leads to another (genuine) question: under what circumstances would the majority of people be concerned about inode counts?
(The reason behind that question is this - if the vast majority of people should never care about inode counts, then it's kind of pointless to push FatELF on everyone using inode counts as a reason.)
Also, you still aren't solving the bigger problem, which is libraries.
As far as I know, FatELF doesn't solve that either, so it's basically irrelevant in the context of this discussion. I could be wrong, though; how, exactly, would FatELF solve this better than separate compiled executables can solve it?
(FatELF isn't much different than separate compiled executables, except that the loader chooses which chunk of the fat binary to run for this arch. You still have the same linking "problems", which I'm not convinced are actually problems.)
"If you didn't know to check for "x86_64" in 1998, your otherwise-functional i386 version won't be run, and the script fails."
... but FatELF has the same problem. A FatELF binary doesn't compile a new version of itself whenever it encounters a new arch; it merely selects among the existing bundled binaries to see which one is appropriate.
Now, sure, some backwards-compatible architectures might have issues with the script, but the solution to that is simple - add a shell/library/kernel/whatever function to determine "are you $ARCH-compatible?" and expose that to scripts.
So that particular problem, at least, is solvable by something far simpler than FatELF.
You realize that the "censorship" is disabled by default, right? That you have to specifically enable the "Hate/Discrimination" filter before it will filter anything? It's not censorship if you voluntarily enable it.
Furthermore, you realize that the "hate/discrimination" tags are user-submitted, and can be out-voted by other users, right?
Did you even read the article you linked to? Way to pass on the FUD.
You know you can turn that off, right? It takes all of thirty seconds.
As for this claim:
They censor certain domains and redirect others.
... I've never seen that happen while using OpenDNS, so I don't know what you're talking about. You can deliberately enable content filtering, but that's opt-in; by default it lets everything through.
So... what domains does OpenDNS routinely censor or redirect without permission? Do you know of any, or are you just making things up as you go?
I'm not asking for dedicated, guaranteed bandwidth. I'm asking for truth in advertising.
Besides, you don't need dedicated, guaranteed bandwidth to run an internet business from home, and that's what this whole conversation is about. You just need internet - and for many internet-based businesses, dialup doesn't cut it, but cable does.
On that I totally agree with you. I generally use the highest quality setting that the streaming service offers; if they offered higher quality than they do now, I'd use it. (I'd also probably upgrade my internet connection if necessary.)
Now that I'm at home and can actually *see* my indicators, I see that it's KBps, not kbps. 300KBps is about 2.35 Mbps. That seems more in line with what other people are saying. (Yes, I watch in HD mode when possible.)
But even so, you'd have to watch a heck of a lot of streaming video to hit the 250GB limit:
(250 GB) / (2.35 Mbps) = 242.73 hours
242.73 hours / 31 days = 7.83 hours/day
That's a lot of streaming video. The majority of customers aren't going to have issues with that.
That said, I feel I should reply to this:
You have absolutely no complaints about the video quality and cannot understand why anybody would?
I'm not sure where you got that idea; I was merely pointing out that under current conditions, Netflix and Hulu aren't going to run the majority of users over the download limit. I said nothing of quality or even desirability of existing video.
I would certainly not be opposed to higher quality video (along with higher download limits).
Wow. You didn't even read the last paragraph of my post. I specifically addressed "business"-class Comcast connections. (Hint: they're the same, just with relaxed limits; particularly, you're still hosed if the network is congested.)
I don't see why you'd do that; the "throttling" (read: lowered priority) only occurs when there's network congestion, so restricting your connection to 70% means you'll never get 100%, whereas leaving it alone means you'll get more than 70% indefinitely, as long as the network is not congested - and if the network is congested, then realistically you're not going to get 70% anyway.
They don't actively throttle you for 15 minutes after you go 100% for 15 minutes; they only lower your connection's priority until you go below 50% for 15 minutes. If there's no network congestion, you're not even going to notice.
Trouble is, you can't usually tell legit traffic from illegal traffic, unless you're willing to assume "BitTorrent" == "illegal" - and even then, you're screwing eleven million World of Warcraft customers next time there's a patch.
Having read Comcast's ToS carefully (because I'm a subscriber), you can host a game server provided that the following things are true:
- you are not hosting the game on a standalone server meant exclusively to act as a server - it is for personal use only - your usage does not interfere with other customers' experience
Hosting a StarCraft game meets those requirements; hooking a pair of 1U rack servers up to your router and charging rent for Counter-Strike slots does not.
(IANAL. This post is my own interpretation of Comcast's ToS. If you're actually concerned, you should ask a lawyer to review their ToS with you.)
That's better than what Comcast pulled on my dad a few years ago. He signed up for some digital cable package where he basically got everything-plus-the-kitchen-sink-channel for like $80/month for two years. After the two years, they reconfigured his account so that he'd get the same channels, but not at the rates he'd get if he got them as a package. In other words, instead of giving him the "movie package" containing HBO, Showtime, etc., they put each movie channel on as an a-la-carte addition, meaning instead of $10 for the package, he paid $5 for each channel. Do that for five or six different packages, and he got his 25th bill - for $200.
I think he paid $80 for that month, then downgraded to the lowest basic cable package he could get.
It's hardly impossible to enforce. All they need is a federal webpage where you can submit both the false DMCA takedown and the evidence that you have rights to the work in question; in most of these false claims cases, it's fairly obvious what's going on, and doesn't require much effort to determine.
For example, Texas Instruments is trying to use DMCA notices to get people to take down some OS signing keys for their calculators. These are not covered by any portion of the DMCA, because a) you cannot copyright a pair of numbers, and b) knowing the numbers does not allow you to circumvent anything. The EFF is representing the guys who received the notices.
But do you think TI will get penalized for the false DMCA claims? Of course not. Therein lies the problem. These things don't result in penalties even when they're already in the legal system, so what incentive do media companies have to actually check the validity of their takedown notices?
Oh, I agree; I'm just saying that if they throttle my game download, then they're making me wait a lot longer to start playing the game. That's not so much a big deal when it's a new purchase; you can wait until tomorrow to play your game.
But when you sit down with the intention to play $MMO only to find that there's a new 500MB patch you have to download first... throttling your patch download pretty much kills your gaming session.
So I'm not referring to download caps when I talk about buying games, and I'm not talking about quantity so much as speed; if I buy a game, I want to play it sooner, not later, and "OMG PIRATES" is a stupid reason to throttle my legitimate download.
I only have the 768 kbit/s service, so streaming video really does use more than 70% of my bandwidth.
I glance at my network utilization widget on my keyboard's LCD screen once in a while while I'm watching Hulu and Netflix (which I watch nearly constantly when I'm at home). I rarely see it go above 300kbps, except maybe while it's doing the initial buffering for 10 seconds or so - and that's including whatever bandwidth my online games are using on the other monitor.
Does anyone have any concrete measurements of how much bandwidth Hulu or Netflix use?
If your content gets taken down that way, you just send YouTube (or whatever provider it is) a counter-claim, and they will just as meekly put it back up.
... except that they don't.
As I've mentioned elsewhere, I don't think the DMCA is in and of itself a bad thing (for the most part; I too disagree with the anti-circumvention tools prohibition), but rather the obvious lack of ability or desire to actually enforce the consequences of a false DMCA claim. I don't think I've heard of anyone actually getting slapped with any sort of punishment at all for a false claim, but if memory serves a false claim is supposed to be punished by a fine (at the very least).
More to the point, I'd much rather have no DMCA at all than have the DMCA without enforcement of punishment for false claims.
we both live in the greater seattle area. 2 weeks ago no basic cable channels would come in at all, they just said you need a box on them.
As it happens, I live in South King County... but if it happened just two weeks ago, then I may not have noticed. I unhooked that tv a month ago. I apologize for making assumptions:)
I've been seeing a lot of Comcast outages recently, too. Usually they're brief, no more than a minute or two, but once it died for a good fifteen minutes. Doesn't make for a good online gaming experience when your connection dies at random:/
I called Verizon and asked that they record my desire to sign up for FiOS (I'm in Kent near the golf course on Meeker Street). My address isn't even in their computer system, so they can't even make a note of my request.
only thing i dind't like is it took the guy 4 hours to install, when my apartment already had the wiring done.
I've seen fiber installations before. It's fun watching fiber get spliced at 300x magnification:)
Was it your apartment that was already wired with fiber, or your building? If your apartment wasn't wired already, then your building having been wired before is irrelevant; they had to run fiber from the junction box to your apartment directly.
Dunno why it would take four hours though. Seems kinda weird. Maybe the installers had nowhere to go next?
Slashdot Mirror
No, you're still misunderstanding.
The DMCA provides no protection for signing keys, because they cannot be used to circumvent any copy protection (and in fact TI doesn't even have copy protection on its OS, nor would it be productive to include it).
The EFF has some stuff to read to read on the issue; specifically, the EFF's letter to TI regarding the takedown notice explains in detail why the DMCA does not apply here.
In this case, the keys do not protect the OS. Instead, the keys are merely used to sign the OS so that the calculator will run it.
In other words, encryption keys may be covered under the DMCA, but that's not the issue here. TI issued takedowns for signing keys, which are not used for DRM purposes, nor are they used to circumvent anything.
The other interesting issue here is that TI does not suffer any financial loss by having the signing keys be public knowledge - it does not charge for the calculator OSes, and before you can use a third-party OS you already have to have purchased a TI calculator.
In other words, TI has zero financial incentive to try to suppress these signing keys.
You've just increased the number of inodes your OS takes by a significant amount.
What, you're going to release a distro cd made up entirely of FatELF binaries? That seems a bad idea in more ways than one. Your libraries will be no different; you're not going to create FatELF-based library files, are you?
Also remember, a significant portion of the files included in any distro are configuration files (or other data files); these are not duplicated. This is even more true of third-party software.
FatELF doesn't even seem useful for OSes (except maybe LiveCDs) because the package manager is going to select arch-specific packages anyway. Point being, your "built-in" OS executables will only have one file anyway, since FatELF is redundant for them, meaning FatELF is only useful for third-party software. But when your third-party software has two dozen data files or more, what's another inode or three for an extra binary and a script to choose which one you run?
(For example, Valve's Source servers do this. If you're concerned about inode counts on a game server, you need to upgrade your game server.)
I guess that leads to another (genuine) question: under what circumstances would the majority of people be concerned about inode counts?
(The reason behind that question is this - if the vast majority of people should never care about inode counts, then it's kind of pointless to push FatELF on everyone using inode counts as a reason.)
Also, you still aren't solving the bigger problem, which is libraries.
As far as I know, FatELF doesn't solve that either, so it's basically irrelevant in the context of this discussion. I could be wrong, though; how, exactly, would FatELF solve this better than separate compiled executables can solve it?
(FatELF isn't much different than separate compiled executables, except that the loader chooses which chunk of the fat binary to run for this arch. You still have the same linking "problems", which I'm not convinced are actually problems.)
"If you didn't know to check for "x86_64" in 1998, your otherwise-functional i386 version won't be run, and the script fails."
... but FatELF has the same problem. A FatELF binary doesn't compile a new version of itself whenever it encounters a new arch; it merely selects among the existing bundled binaries to see which one is appropriate.
Now, sure, some backwards-compatible architectures might have issues with the script, but the solution to that is simple - add a shell/library/kernel/whatever function to determine "are you $ARCH-compatible?" and expose that to scripts.
So that particular problem, at least, is solvable by something far simpler than FatELF.
I found Flameeyes' post regarding FatELF to be enlightening.
I'll have to try it. Thanks.
You realize that the "censorship" is disabled by default, right? That you have to specifically enable the "Hate/Discrimination" filter before it will filter anything? It's not censorship if you voluntarily enable it.
Furthermore, you realize that the "hate/discrimination" tags are user-submitted, and can be out-voted by other users, right?
Did you even read the article you linked to? Way to pass on the FUD.
They hijack NXDOMAIN results to provide ads.
You know you can turn that off, right? It takes all of thirty seconds.
As for this claim:
They censor certain domains and redirect others.
... I've never seen that happen while using OpenDNS, so I don't know what you're talking about. You can deliberately enable content filtering, but that's opt-in; by default it lets everything through.
So... what domains does OpenDNS routinely censor or redirect without permission? Do you know of any, or are you just making things up as you go?
I'm not asking for dedicated, guaranteed bandwidth. I'm asking for truth in advertising.
Besides, you don't need dedicated, guaranteed bandwidth to run an internet business from home, and that's what this whole conversation is about. You just need internet - and for many internet-based businesses, dialup doesn't cut it, but cable does.
On that I totally agree with you. I generally use the highest quality setting that the streaming service offers; if they offered higher quality than they do now, I'd use it. (I'd also probably upgrade my internet connection if necessary.)
Now that I'm at home and can actually *see* my indicators, I see that it's KBps, not kbps. 300KBps is about 2.35 Mbps. That seems more in line with what other people are saying. (Yes, I watch in HD mode when possible.)
But even so, you'd have to watch a heck of a lot of streaming video to hit the 250GB limit:
(250 GB) / (2.35 Mbps) = 242.73 hours
242.73 hours / 31 days = 7.83 hours/day
That's a lot of streaming video. The majority of customers aren't going to have issues with that.
That said, I feel I should reply to this:
You have absolutely no complaints about the video quality and cannot understand why anybody would?
I'm not sure where you got that idea; I was merely pointing out that under current conditions, Netflix and Hulu aren't going to run the majority of users over the download limit. I said nothing of quality or even desirability of existing video.
I would certainly not be opposed to higher quality video (along with higher download limits).
Wow. You didn't even read the last paragraph of my post. I specifically addressed "business"-class Comcast connections. (Hint: they're the same, just with relaxed limits; particularly, you're still hosed if the network is congested.)
I don't see why you'd do that; the "throttling" (read: lowered priority) only occurs when there's network congestion, so restricting your connection to 70% means you'll never get 100%, whereas leaving it alone means you'll get more than 70% indefinitely, as long as the network is not congested - and if the network is congested, then realistically you're not going to get 70% anyway.
They don't actively throttle you for 15 minutes after you go 100% for 15 minutes; they only lower your connection's priority until you go below 50% for 15 minutes. If there's no network congestion, you're not even going to notice.
No streaming video service I'm aware of will give you video faster than the 70% limit...
My connection is a 12Mbps connection. 70% of that is 8Mbps. Neither Hulu nor Netflix go near that.
Trouble is, you can't usually tell legit traffic from illegal traffic, unless you're willing to assume "BitTorrent" == "illegal" - and even then, you're screwing eleven million World of Warcraft customers next time there's a patch.
Even though downloading content is (currently) legal in Canada (we pay a tax on storage media).
Writable media in the U.S. is taxed for that reason, too, but that doesn't make it legal to download RIAA-owned music over BitTorrent.
In other words, I don't think you should test that one in court.
Having read Comcast's ToS carefully (because I'm a subscriber), you can host a game server provided that the following things are true:
- you are not hosting the game on a standalone server meant exclusively to act as a server
- it is for personal use only
- your usage does not interfere with other customers' experience
Hosting a StarCraft game meets those requirements; hooking a pair of 1U rack servers up to your router and charging rent for Counter-Strike slots does not.
(IANAL. This post is my own interpretation of Comcast's ToS. If you're actually concerned, you should ask a lawyer to review their ToS with you.)
I dunno about two shows at once, but I regularly watch Netflix/Hulu on one monitor and play a game on the other...
If you want to know more, you should look at the DOCSIS specs. But I really doubt that you want to do that :-) Unless you have trouble sleeping.
Wouldn't that be a little like trying to cure a cocaine addiction by doing more cocaine?
That's better than what Comcast pulled on my dad a few years ago. He signed up for some digital cable package where he basically got everything-plus-the-kitchen-sink-channel for like $80/month for two years. After the two years, they reconfigured his account so that he'd get the same channels, but not at the rates he'd get if he got them as a package. In other words, instead of giving him the "movie package" containing HBO, Showtime, etc., they put each movie channel on as an a-la-carte addition, meaning instead of $10 for the package, he paid $5 for each channel. Do that for five or six different packages, and he got his 25th bill - for $200.
I think he paid $80 for that month, then downgraded to the lowest basic cable package he could get.
One would assume Comcast is smart enough to see that sort of thing happening and stop it.
It's hardly impossible to enforce. All they need is a federal webpage where you can submit both the false DMCA takedown and the evidence that you have rights to the work in question; in most of these false claims cases, it's fairly obvious what's going on, and doesn't require much effort to determine.
For example, Texas Instruments is trying to use DMCA notices to get people to take down some OS signing keys for their calculators. These are not covered by any portion of the DMCA, because a) you cannot copyright a pair of numbers, and b) knowing the numbers does not allow you to circumvent anything. The EFF is representing the guys who received the notices.
But do you think TI will get penalized for the false DMCA claims? Of course not. Therein lies the problem. These things don't result in penalties even when they're already in the legal system, so what incentive do media companies have to actually check the validity of their takedown notices?
Oh, I agree; I'm just saying that if they throttle my game download, then they're making me wait a lot longer to start playing the game. That's not so much a big deal when it's a new purchase; you can wait until tomorrow to play your game.
But when you sit down with the intention to play $MMO only to find that there's a new 500MB patch you have to download first... throttling your patch download pretty much kills your gaming session.
So I'm not referring to download caps when I talk about buying games, and I'm not talking about quantity so much as speed; if I buy a game, I want to play it sooner, not later, and "OMG PIRATES" is a stupid reason to throttle my legitimate download.
I only have the 768 kbit/s service, so streaming video really does use more than 70% of my bandwidth.
I glance at my network utilization widget on my keyboard's LCD screen once in a while while I'm watching Hulu and Netflix (which I watch nearly constantly when I'm at home). I rarely see it go above 300kbps, except maybe while it's doing the initial buffering for 10 seconds or so - and that's including whatever bandwidth my online games are using on the other monitor.
Does anyone have any concrete measurements of how much bandwidth Hulu or Netflix use?
If your content gets taken down that way, you just send YouTube (or whatever provider it is) a counter-claim, and they will just as meekly put it back up.
... except that they don't.
As I've mentioned elsewhere, I don't think the DMCA is in and of itself a bad thing (for the most part; I too disagree with the anti-circumvention tools prohibition), but rather the obvious lack of ability or desire to actually enforce the consequences of a false DMCA claim. I don't think I've heard of anyone actually getting slapped with any sort of punishment at all for a false claim, but if memory serves a false claim is supposed to be punished by a fine (at the very least).
More to the point, I'd much rather have no DMCA at all than have the DMCA without enforcement of punishment for false claims.
we both live in the greater seattle area. 2 weeks ago no basic cable channels would come in at all, they just said you need a box on them.
As it happens, I live in South King County... but if it happened just two weeks ago, then I may not have noticed. I unhooked that tv a month ago. I apologize for making assumptions :)
I've been seeing a lot of Comcast outages recently, too. Usually they're brief, no more than a minute or two, but once it died for a good fifteen minutes. Doesn't make for a good online gaming experience when your connection dies at random :/
I called Verizon and asked that they record my desire to sign up for FiOS (I'm in Kent near the golf course on Meeker Street). My address isn't even in their computer system, so they can't even make a note of my request.
only thing i dind't like is it took the guy 4 hours to install, when my apartment already had the wiring done.
I've seen fiber installations before. It's fun watching fiber get spliced at 300x magnification :)
Was it your apartment that was already wired with fiber, or your building? If your apartment wasn't wired already, then your building having been wired before is irrelevant; they had to run fiber from the junction box to your apartment directly.
Dunno why it would take four hours though. Seems kinda weird. Maybe the installers had nowhere to go next?