Actually, IIRC, the DMCA has statutes/statements in it about how the law isn't retroactive (IOW: if someone started circumventing an access control method prior to the law taking effect, the DMCA couldn't be used against them). If your friend indeed started working on this prior to the enactment of the DMCA, I don't think the BNetD guys have anything to worry about.
It might be worth looking into, sure, it's a backwards way to avoid the DMCA affecting their work, but it sure beats folding like they have.
I think you missed the 'networking effect' he mentioned-- sure, those 7,000 people tell 100 of their friends/relatives/associates, then THOSE 100 people tell THEIR friends/relatives/associates (so now we've got 700,000 people each telling another 100 people about this injustice-- in case you lost count at step 3, that's 70,000,000), repeat this again, and you get my point. The only thing is, people who agree with the point *MUST* be persistant or it'll NEVER work. That's the only weakness I see in his idea, it's hard to convince people of something, let alone get them to act upon it... (which is part of the reason you'll likely never see an uprising against any modern government, for example; not only do the laws forbid it (militia's can't legally be formed anymore) but people just don't care (I'm speaking of the USA here, obviously laws elsewhere may differ)).
It COULD be done though, it'd just take a ton of coaching and prodding and reminding.
I've never played a Blizzard game, but I'm wondering of Battle.Net is at all like Westwood's (now EA's) crap setup for online play-- if you were to look at it, you'd swear Westwood Online (or as they call it, WOL) was built on top of an IRC server of some sort. The ONLY way to play internet based TCP/IP games is to use their login system and meet with a friend on WOL. You can't just enter an IP address/port and connect, it's gotta be through them. LAN-based play is limited to IPX/SPX gaming, NO LAN TCP/IP support.
Is this the way Battle.Net is? And why aren't more companies like id, allowing people to JUST PLAY THE WAY THEY WANT TO? (FYI: id has master servers that provide a list of public games you can connect to, but you can also enter an IP/port and connect to a private game (be it on a LAN, or someplace on the internet).)
I'm just curious if someone can clarify the differences.. and if they are the same, I wonder if someones written a replacement for WOL. =)
Using his idea, I'd presume that the algorithm would be implemented in some sort of hardware (a chip on the board inside the drive, for a PC), not part of the firmware. The studios are no doubt livid (no pun intended) over DeCSS, so I really wouldn't put it past them to think of every angle...
1) True, the genie is literally out of the bottle once a disc is stolen or misplaced.
2) False, they can still track down non-stolen or misplaced copies to the source to prosecute those distributing pirated copies (or, less illegal, people who do what the studios don't WANT you to do, rip the movie, but for fair-use purposes (a backup, part of it used in a school report, etc)).
Just because someone breaks into your house doesn't mean you put the baseball bat down once they've touched your prized vase-- the MPAA certainly has many more fish to fry than a lost/stolen copy that lead to pirating a movie on the internet.
Wasn't suggesting that it was-- but the article title suggested this was some writable DVD format that would work in existing players. Clearly this is not the case. And in fact the entire issue of burning a Blu-ray disc isn't discussed at all in either the BBC article or the press release-- yet the title of the/. story suggests this as competition to DVD-R/DVD+R/DVD-RAM, which for now, it most certainly is not (not even slightly compatible, beyond physical dimensions)...
But if the watermarking is irreversible (which, really, is the entire goal of watermarking content) then as soon as that footage hits the internet, they can (and likely will try) to track down whomever bought the disc in question (thus enters the unique ID, identifying beyond a shadow of a doubt who bought the disc, and whom to prosecute).
Yeah but this is definately a process they can automate-- if they can take the time to generate a Unique ID, they can certainly take the additional few minutes to encode the movie/video/audio with a watermark of some sort. Realize that they won't be re-encoding the content (MPEG2 compression is quite time consuming, if this were the case I'd agree with you), but watermarking content should be quite fast. All they need to do is change their production methodologies and ramp up, and they'll probably be able to churn out discs with 100% unique content at nearly the same price as current DVD movies...
I'm assuming the 'unique ID' is written to a normally non-writable part of the disc (for Blu-ray writers/burners), so yes, this is part of it I imagine (the player/reader would validate that this unique ID existed and was properly initialized). The other side of it that I can see is media watermarking using the 'unique ID' as a key to encrypt/mark with. Watermarking, if done properly (and I'm not claiming to be an expert), would be irreversable and potentially trackable (ie: Joe Nobody buys 'The Fifth Element' on a shiny new Blu-Ray disc from his local Fred Meyer/Hastings/Sam Goody. Joe Nobody rips the movie and encodes it into a semi-high quality DiVX (or some other format, VCD, etc). The watermark survives the recompression, but Joe Nobody doesn't know this and distributes it on Gnutella/FreeNet/eDonkey2000. MPAA Lawyer downloads enough of the movie off of one of these file sharing utilities and uses an application to 1) extract the watermark, 2) correlate the watermark with the store the disc was sold at and 3) identify the individual to whom the disc was sold. MPAA lawyer dispatches law enforcement to Joe Nobody's place of residence.).
It's disconcerting, to say the least. And this is definately an all-new format, nothing in the press release or BBC article seem to indicate that the discs will work in existing DVD drives, so this is the studios/MPAA's "second chance" to get copy protection implemented correctly.. I'm sure after the DeCSS beating they got, they're definately looking at every possibility they have...
In the press release, they make mention of provisions for a unique ID (aka: a serial number) on each disc to help curb/stop piracy.. this, to me, is the media's biggest problem. I imagine that unlike CSS (which the studio's botched) they'll do the smart thing and use the unique ID to somehow watermark the data and/or video content of these new discs. Some might see this as good (if the studio's actually do the logical thing and allow fair use copying again, unlike DVD), but I can see a situation where the studio's turn this around and use it to track down offenders for individual prosecution. (Something that I've never seen them do, but when you've got these kinds of smoking guns (the watermark being found in some DiVX ripped copy on the net), you gotta wonder if they can really contain themselves from blasting people into the afterlife with their "lawyer death ray"...)
Otherwise I love the technology, I've been hearing about blue-laser technology and optical discs since I was a kid (I'm in my mid-20's now), it's good to see it finally coming of age.
No kidding, this is the one thing that's really been a let-down about DVD-- no support for *true* HDTV. We'll all be re-buying our DVD libraries when the new TV standard takes hold in a few years...
Actually, while you're right for the most part in what you say, I've read (both in stories here on/. and elsewhere) that MPEG-4, at the bitrates used by MPEG-2, outperforms MPEG-2. MPEG-4 though was designed for streaming (internet based) media, not the kind of stuff you'd see on DVD/Blu-ray discs.
Uh, I know.. that's what I was talking about, but currently no hardware review sites USE IT AS A BENCHMARK. I was saying it'd be interesting/funny if sites DID start using it.. eg: "This new Athlon/Pentium 5 20 GHz cranked out a valid reg code in UNDER a minute!". I think ya missed the point I was trying to make. That this would be used alongside the other standard magazine/webzine benchmarks like SiSoft Sandra and Quake III Arena time-demos.
The one you linked too IS NOT the one the article is talking about (though if you browse that same thread you linked to, someone replied with the correct app (or so I believe)). The file date on the file you linked to is sometime in August of last year (the reply with the correct one is date in February of THIS year however).
The correct one (again, I'm assuming here) appears to be written with Visual C++ (not sure which ver, but it links against MFC42.DLL). I agree about optimizations, I which this were open source code so I could take a look at it-- the most obvious optimization is one I mentioned elsewhere; the code isn't SMP-friendly. It has two threads, but only ONE thread actually does the brute force work (so if you look in Task Manager, on a dual-CPU system, it only uses 50% of the total processor power available).. in order to properly utilitize all of the resources available you need to run one copy for every CPU in your system. (Ideally the app would spawn a thread for each CPU, and set the thread affinity to an individual processor (1 - max processors available). This is the change I'd implement if I had the code.)
Oh well.. maybe I'll get bored and disassemble it.
This key generator isn't multi-threaded-- if you want to get the most out of this app (ie: generate a ton of keys) run ONE instance for each CPU in your system (so if you're a lucky bastard running a quad boxen, run four copies). Otherwise it'll just peg one of your CPU's and you'll miss out on (conceivably) twice the possible keys (or the same number of keys in half the time). You might also want to run Task Manager (if you're running NT/2000/XP) and change each processes affinity to a single CPU (keeping the code (and cache, presumably) limited to one CPU per instance).
I can see PC review webzines using this thing to benchmark how fast the latest processor is...
...unless I don't understand the story, this is a product key generator, not a circumvention of WPA. IE: This thing generates keys for when you have to install the OS; not the response string for when you have to actually ACTIVATE the thing (over the phone, for example).
Something that generated valid install keys AND produced working results for the over-the-phone activation would be much more newsworthy. (If I'm wrong on this though, let me know because maybe I don't understand the way in which MS generates responses to the WPA info/hash it's sent.)
I looked at it, and if you open the file in a text editor, you can see (towards the end of the file, well, more than halfway in atleast) that there's a few hardcoded keys in unicode format. It's definately not generating them on-the-fly. The other poster in this thread however appears to have the real file (the one talked about in this article and The Register's story). Plus the file date is sometime this month, vs. the other files date of sometime last August (2001).
My biggest gripe with some of my former employers was the lack of involvement in the design phase (eg: setting realistic goals, and not imaginary or impossible goals). By the same token, setting reasonable time-frames for completion of various tasks is another issue I've butted heads with management on-- a prime example is when I explicitly stated the project at hand would take 4 months to complete (longer with QA work). I was overruled and told that the entire project, with QA, could be completed in 3 months. Needless to say the project went beyond that limit and much complaining was heard from the management types (instead of realizing they were wrong, they took us aside and told us we weren't doing good enough-- somehow they thought this would speed things up).
Development takes time, and most geeks aren't like Scotty in Star Trek-- we don't multiply our estimates by 2 to make ourselves look like miracle workers when we get it done in half the time.
Right... and this invalidates the idea that $3000 for something that takes $300 to produce is overkill how exactly? Atleast the other poster gave some reasonable ideas (R&D and so forth) but you'd kind of think those would be tallied into the final cost of production. (And would pay for themselves if the units sold in quantity at even 1.5x the production (manufacture) cost.)
One would assume that the $300 cited for manufacture would INCLUDE such things as R&D and Administrative expenses. Otherwise, you're not really telling people how much they cost. (And this is really besides the point-- the things you list, while accurate for a startup, don't really apply to a company like Intel with a lot of cash to burn on innovation). I'm still missing the justification of those rather large prices.
Yah, but how MANY does it have? The IA-64 architecture, as I recall, calls for 256 such registers. I sincerely doubt your calculator has that many (let alone the redundant sets that no doubt exist internally to speed up execution). It is dumb to say it's the first processor with 64-bit registers, though..
Going from $300 to $3000 per processor at retail seems a bit extreme if you ask me. And don't mod this as a flame, I actually LIKE Intel's work, but it's a joke how much they charge for their processors compared to how much it costs to make them.
About the LZW patent and Unisys, doesn't their patent expire in 2005? (I went and found the patent (using the number provided from the Unisys page you linked to in the parent) and noted that the patent was granted in 1985.) Of course this doesn't affect MPEG4 I imagine (has a patent even been awarded yet for that?) but I'm just curious when Unisys's monopoly will end.;)
How will this affect things such as DivX which use MPEG4 in their CODEC(s)? Wouldn't such a fee system preclude them from giving away the encoder/decoder (or atleast the encoder)?
Slashdot Mirror
WARNING: IANAL
Actually, IIRC, the DMCA has statutes/statements in it about how the law isn't retroactive (IOW: if someone started circumventing an access control method prior to the law taking effect, the DMCA couldn't be used against them). If your friend indeed started working on this prior to the enactment of the DMCA, I don't think the BNetD guys have anything to worry about.
It might be worth looking into, sure, it's a backwards way to avoid the DMCA affecting their work, but it sure beats folding like they have.
I think you missed the 'networking effect' he mentioned-- sure, those 7,000 people tell 100 of their friends/relatives/associates, then THOSE 100 people tell THEIR friends/relatives/associates (so now we've got 700,000 people each telling another 100 people about this injustice-- in case you lost count at step 3, that's 70,000,000), repeat this again, and you get my point. The only thing is, people who agree with the point *MUST* be persistant or it'll NEVER work. That's the only weakness I see in his idea, it's hard to convince people of something, let alone get them to act upon it... (which is part of the reason you'll likely never see an uprising against any modern government, for example; not only do the laws forbid it (militia's can't legally be formed anymore) but people just don't care (I'm speaking of the USA here, obviously laws elsewhere may differ)).
It COULD be done though, it'd just take a ton of coaching and prodding and reminding.
I've never played a Blizzard game, but I'm wondering of Battle.Net is at all like Westwood's (now EA's) crap setup for online play-- if you were to look at it, you'd swear Westwood Online (or as they call it, WOL) was built on top of an IRC server of some sort. The ONLY way to play internet based TCP/IP games is to use their login system and meet with a friend on WOL. You can't just enter an IP address/port and connect, it's gotta be through them. LAN-based play is limited to IPX/SPX gaming, NO LAN TCP/IP support.
Is this the way Battle.Net is? And why aren't more companies like id, allowing people to JUST PLAY THE WAY THEY WANT TO? (FYI: id has master servers that provide a list of public games you can connect to, but you can also enter an IP/port and connect to a private game (be it on a LAN, or someplace on the internet).)
I'm just curious if someone can clarify the differences.. and if they are the same, I wonder if someones written a replacement for WOL. =)
Using his idea, I'd presume that the algorithm would be implemented in some sort of hardware (a chip on the board inside the drive, for a PC), not part of the firmware. The studios are no doubt livid (no pun intended) over DeCSS, so I really wouldn't put it past them to think of every angle...
You're half right--
1) True, the genie is literally out of the bottle once a disc is stolen or misplaced.
2) False, they can still track down non-stolen or misplaced copies to the source to prosecute those distributing pirated copies (or, less illegal, people who do what the studios don't WANT you to do, rip the movie, but for fair-use purposes (a backup, part of it used in a school report, etc)).
Just because someone breaks into your house doesn't mean you put the baseball bat down once they've touched your prized vase-- the MPAA certainly has many more fish to fry than a lost/stolen copy that lead to pirating a movie on the internet.
Wasn't suggesting that it was-- but the article title suggested this was some writable DVD format that would work in existing players. Clearly this is not the case. And in fact the entire issue of burning a Blu-ray disc isn't discussed at all in either the BBC article or the press release-- yet the title of the /. story suggests this as competition to DVD-R/DVD+R/DVD-RAM, which for now, it most certainly is not (not even slightly compatible, beyond physical dimensions)...
But if the watermarking is irreversible (which, really, is the entire goal of watermarking content) then as soon as that footage hits the internet, they can (and likely will try) to track down whomever bought the disc in question (thus enters the unique ID, identifying beyond a shadow of a doubt who bought the disc, and whom to prosecute).
Yeah but this is definately a process they can automate-- if they can take the time to generate a Unique ID, they can certainly take the additional few minutes to encode the movie/video/audio with a watermark of some sort. Realize that they won't be re-encoding the content (MPEG2 compression is quite time consuming, if this were the case I'd agree with you), but watermarking content should be quite fast. All they need to do is change their production methodologies and ramp up, and they'll probably be able to churn out discs with 100% unique content at nearly the same price as current DVD movies...
I'm assuming the 'unique ID' is written to a normally non-writable part of the disc (for Blu-ray writers/burners), so yes, this is part of it I imagine (the player/reader would validate that this unique ID existed and was properly initialized). The other side of it that I can see is media watermarking using the 'unique ID' as a key to encrypt/mark with. Watermarking, if done properly (and I'm not claiming to be an expert), would be irreversable and potentially trackable (ie: Joe Nobody buys 'The Fifth Element' on a shiny new Blu-Ray disc from his local Fred Meyer/Hastings/Sam Goody. Joe Nobody rips the movie and encodes it into a semi-high quality DiVX (or some other format, VCD, etc). The watermark survives the recompression, but Joe Nobody doesn't know this and distributes it on Gnutella/FreeNet/eDonkey2000. MPAA Lawyer downloads enough of the movie off of one of these file sharing utilities and uses an application to 1) extract the watermark, 2) correlate the watermark with the store the disc was sold at and 3) identify the individual to whom the disc was sold. MPAA lawyer dispatches law enforcement to Joe Nobody's place of residence.).
It's disconcerting, to say the least. And this is definately an all-new format, nothing in the press release or BBC article seem to indicate that the discs will work in existing DVD drives, so this is the studios/MPAA's "second chance" to get copy protection implemented correctly.. I'm sure after the DeCSS beating they got, they're definately looking at every possibility they have...
In the press release, they make mention of provisions for a unique ID (aka: a serial number) on each disc to help curb/stop piracy.. this, to me, is the media's biggest problem. I imagine that unlike CSS (which the studio's botched) they'll do the smart thing and use the unique ID to somehow watermark the data and/or video content of these new discs. Some might see this as good (if the studio's actually do the logical thing and allow fair use copying again, unlike DVD), but I can see a situation where the studio's turn this around and use it to track down offenders for individual prosecution. (Something that I've never seen them do, but when you've got these kinds of smoking guns (the watermark being found in some DiVX ripped copy on the net), you gotta wonder if they can really contain themselves from blasting people into the afterlife with their "lawyer death ray"...)
Otherwise I love the technology, I've been hearing about blue-laser technology and optical discs since I was a kid (I'm in my mid-20's now), it's good to see it finally coming of age.
No kidding, this is the one thing that's really been a let-down about DVD-- no support for *true* HDTV. We'll all be re-buying our DVD libraries when the new TV standard takes hold in a few years...
Actually, while you're right for the most part in what you say, I've read (both in stories here on /. and elsewhere) that MPEG-4, at the bitrates used by MPEG-2, outperforms MPEG-2. MPEG-4 though was designed for streaming (internet based) media, not the kind of stuff you'd see on DVD/Blu-ray discs.
I'd burn it before it had a chance to breed.
Uh, I know.. that's what I was talking about, but currently no hardware review sites USE IT AS A BENCHMARK. I was saying it'd be interesting/funny if sites DID start using it.. eg: "This new Athlon/Pentium 5 20 GHz cranked out a valid reg code in UNDER a minute!". I think ya missed the point I was trying to make. That this would be used alongside the other standard magazine/webzine benchmarks like SiSoft Sandra and Quake III Arena time-demos.
The one you linked too IS NOT the one the article is talking about (though if you browse that same thread you linked to, someone replied with the correct app (or so I believe)). The file date on the file you linked to is sometime in August of last year (the reply with the correct one is date in February of THIS year however).
The correct one (again, I'm assuming here) appears to be written with Visual C++ (not sure which ver, but it links against MFC42.DLL). I agree about optimizations, I which this were open source code so I could take a look at it-- the most obvious optimization is one I mentioned elsewhere; the code isn't SMP-friendly. It has two threads, but only ONE thread actually does the brute force work (so if you look in Task Manager, on a dual-CPU system, it only uses 50% of the total processor power available).. in order to properly utilitize all of the resources available you need to run one copy for every CPU in your system. (Ideally the app would spawn a thread for each CPU, and set the thread affinity to an individual processor (1 - max processors available). This is the change I'd implement if I had the code.)
Oh well.. maybe I'll get bored and disassemble it.
This key generator isn't multi-threaded-- if you want to get the most out of this app (ie: generate a ton of keys) run ONE instance for each CPU in your system (so if you're a lucky bastard running a quad boxen, run four copies). Otherwise it'll just peg one of your CPU's and you'll miss out on (conceivably) twice the possible keys (or the same number of keys in half the time). You might also want to run Task Manager (if you're running NT/2000/XP) and change each processes affinity to a single CPU (keeping the code (and cache, presumably) limited to one CPU per instance).
I can see PC review webzines using this thing to benchmark how fast the latest processor is...
...unless I don't understand the story, this is a product key generator, not a circumvention of WPA. IE: This thing generates keys for when you have to install the OS; not the response string for when you have to actually ACTIVATE the thing (over the phone, for example).
Something that generated valid install keys AND produced working results for the over-the-phone activation would be much more newsworthy. (If I'm wrong on this though, let me know because maybe I don't understand the way in which MS generates responses to the WPA info/hash it's sent.)
I looked at it, and if you open the file in a text editor, you can see (towards the end of the file, well, more than halfway in atleast) that there's a few hardcoded keys in unicode format. It's definately not generating them on-the-fly. The other poster in this thread however appears to have the real file (the one talked about in this article and The Register's story). Plus the file date is sometime this month, vs. the other files date of sometime last August (2001).
My biggest gripe with some of my former employers was the lack of involvement in the design phase (eg: setting realistic goals, and not imaginary or impossible goals). By the same token, setting reasonable time-frames for completion of various tasks is another issue I've butted heads with management on-- a prime example is when I explicitly stated the project at hand would take 4 months to complete (longer with QA work). I was overruled and told that the entire project, with QA, could be completed in 3 months. Needless to say the project went beyond that limit and much complaining was heard from the management types (instead of realizing they were wrong, they took us aside and told us we weren't doing good enough-- somehow they thought this would speed things up).
Development takes time, and most geeks aren't like Scotty in Star Trek-- we don't multiply our estimates by 2 to make ourselves look like miracle workers when we get it done in half the time.
Right... and this invalidates the idea that $3000 for something that takes $300 to produce is overkill how exactly? Atleast the other poster gave some reasonable ideas (R&D and so forth) but you'd kind of think those would be tallied into the final cost of production. (And would pay for themselves if the units sold in quantity at even 1.5x the production (manufacture) cost.)
One would assume that the $300 cited for manufacture would INCLUDE such things as R&D and Administrative expenses. Otherwise, you're not really telling people how much they cost. (And this is really besides the point-- the things you list, while accurate for a startup, don't really apply to a company like Intel with a lot of cash to burn on innovation). I'm still missing the justification of those rather large prices.
Yah, but how MANY does it have? The IA-64 architecture, as I recall, calls for 256 such registers. I sincerely doubt your calculator has that many (let alone the redundant sets that no doubt exist internally to speed up execution). It is dumb to say it's the first processor with 64-bit registers, though..
And this justifies the mark-up HOW exactly?
Going from $300 to $3000 per processor at retail seems a bit extreme if you ask me. And don't mod this as a flame, I actually LIKE Intel's work, but it's a joke how much they charge for their processors compared to how much it costs to make them.
About the LZW patent and Unisys, doesn't their patent expire in 2005? (I went and found the patent (using the number provided from the Unisys page you linked to in the parent) and noted that the patent was granted in 1985.) Of course this doesn't affect MPEG4 I imagine (has a patent even been awarded yet for that?) but I'm just curious when Unisys's monopoly will end. ;)
How will this affect things such as DivX which use MPEG4 in their CODEC(s)? Wouldn't such a fee system preclude them from giving away the encoder/decoder (or atleast the encoder)?