$60. No touchscreen- not even a little LCD like the RM-V60, but *every* key is programmable for each of 8 components (unlike some universals like one RCA that have very limited key override storage). 3 macro keys.
It has real buttons and you can set the volume keys to control the receiver regardless of component selection.
I control my Tivo with it and while I have to remember that TV/Video is thumbs-down, I don't have to look at it to find the button.
Simple & cheap. If you think you need something more complicated, like another poster said you probably need to get out more. I use a complicated computer all day. When I want to watch some TiVo or DVD, I'd prefer simplicity.
How many servers are subjected to shocks that will make the heatsink FALL OFF?
None. Unless you're in an earthquake zone, and then I rather suspect you have more important things to worry about.
I live less than 10 miles from the San Andreas and I'm more worried about an Athlon burning my house down than an earthquake knocking it down.
The house sits on steel reinforced concrete piers that go more than 15 feet into the sandstone, the house is bolted to the foundation and the framing has shear panels attached. The gas lines to the water heater and furnace are flexible so when the inertia of 30 gallons of water resists shaking, it won't snap the line.
We're less careful about the contents of the house, so an earthquake will give us fish on the floor swimming in shards of vases and my CRT. That should be about it, though. I hope if the computer falls over, though, that it doesn't cause a fire!
What Tom's Hardware didn't show is what the PC looks like 10 minutes after the heat sink is removed. Does the thermal runaway eventually burn itself out?
I wouldn't be worried about my PC being dead after an earthquake, but I'd rather it not take anything with it.
You're probably thinking of ISDN BRI, the Basic Rate Interface consisting of 2 B (bearer) channels of 56kbps or 64kpbs and one 16kbps D channel (signalling).
ISDN PRI or Primary Rate Interface is 23 B channels and one 64kbps D channel. 24 * 64 = 1.5mbps. Hmmm... T1!
I know it's common to call BRI simply ISDN, particularly in the residential context of this discussion, but people are trying to claim they know more than the broadband ISPs about how to achieve security, so using the proper terms might help.
The only ways around the patent are to find a completely different way of decoding/encoding the data (very unlikely, if the patent attorney did his job)
It might be hard, but finding patent workarounds is done. Let me give three examples:
1) Creative Technology, Ltd. was sick of paying Yamaha for their FM synthesizer chips which by the mid 90's weren't amazing tech. anymore(low transistor count) but were expensive due to a nice passel of patents on stuff developed at CCRMA @ Stanford. The 1st/main patent was expiring, so the Creative Advanced Technology Center came up with a workaround that skirted the remaining patents, saving $millions (I think it was even pin compatible so they didn't have to rev their PCBs).
2) There might be patents on Dolby's analog surround systems, but plenty of 2nd run theaters show cheap prints in "Surround Sound" and some VHS tapes are even encoded that way (there's a distinctive logo, but no info on the company behind this- anyone know?). I'm betting that phase-shifted matrixing is not protected anymore, but Pro Logic which is Dolby's extraction method that gives better channel separation through "steering" is.
3) Sony tried to skirt Dolby's perceptual coding patents when they developed ATRAC. There were also patents by Fraunhofer & probably others. They failed (at least Dolby's lawyers convinced Sony's lawyers they'd failed) and there's a Dolby logo (and Dolby license fees paid) on Minidisc machines.
First thing is to stop calling the open source decoder AC-3. Call it digital surround and avoid the trademark dispute. Ogg Vorbis managed to create a perceptual codec and avoid the patents (how!!???). Maybe it can be done for a Dolby Digital decoder. Probably not, though, as the method by which the channels are interleaved (I believe busy channels can "borrow" bandwidth from quiet ones) is protected.
Didn't IBM protect Microchannel with patents? The courts have repeatedly protected companies who reverse-engineer and make compatible peripherals, but if the interface (Intel Slot 1 and Apple Desktop Bus are other examples) contains any "innovations" it can be protected until it's obsolete. Adding copyright-able code to the peripheral that must be run before it can be "activated" (Nintendo cartridges) is another way of keeping out the compatible glommers-on.
Your company may have the right to search the file drawer in your desk (every co. I've worked for has had that provision in the employment contract), but do they have the right to search your purse (backpack) sitting in it? Not in any company I've worked for.
To take this even further, should they be able to search your pockets? Your rectum (better use Memory Sticks for those secrets; Compact Flash cards are gonna hurt!), your urine (whoops, too late)? The crud under your fingernails?
The "wrapper of privacy" may seem arbitrary or silly, but don't you feel more comfortable having it? Wouldn't it be nice if it extended to your computer?
This may lead to an arms race of freenet traffic trying to look more and more like "normal" traffic and the filters and bounty hunters getting smarter. However, despite those that suggest we must develop the "cypherpunk net" in secret to keep it from being shut down, I think open source may be a big advantage. The filter developers cannot work together and will end up re-inventing wheels ad naseum.
We not only need a better underlying technology in freenet we need better content. Consider RMS's argument against the Lesser GPL: if the good stuff is under GPL, more people will be convinced to go the Free route. If there is a lot of good stuff and it's easier to get due to a decentralized approach (no slashdot effect) and EZ2USE clients are written, the network will have enough nodes to survive attacks.
How? Well, to start if you're an independent band with mp3s @ mp3.com, consider hosting 'em on a freenet server also. If you have a server, offer the space.
I don't know where peekabooty is going, but consider this: a freenet proxy that acts as a local webserver so you can browse the net using a web browser.
Could (something like) freenet bring the promise of multicast without requiring infrastructure upgrades?
We will need user controllable rate limiters on the server/clients. However, if there are enough nodes on the network and we include ideas like the 110% [1] download, this network will be just a background hum. Some kinds of packet address spoofing and maybe using TCP packets ('cause UDP might be filtered) but not ack/nak-ing them (treat 'em as UDP) might be necessary to hide server's identities, but this has to be balanced with making the traffic indistinguishable from regular surfing (or playing Quake). It must be impossible to find the owner of the server not only through the client's interface, but also using any tool available to an ISP or bounty hunter.
I haven't set up a freenet server at home 'cause I haven't convinced myself it won't fill up with kiddie porn. Besides the moral issues, I'd rather not have the stormtroopers breaking down my door. (so... a related thread is how do you build a truly anonymous network and not have it turn into a haven for criminals you don't like (vs. the liberal copyright users (DMCA criminals) you do))
[1] There was some noise recently about a technique where a file is turned into a sort of "hologram" that results in a slight increase in size (say 125%). The packets are downloaded using UDP and any collection of 110% or more of the orig. size (so less than all the packets) can reconstruct the original.
DVDs are 704x486 (they use non-square pixels and provide for a bit more overscan). Check out the CCIR-601 spec or Watkinson's Digital Video book. 640x480 is not quite DVD, but it's better than laserdisc.
Of course, resolution is probably not the problem in downloaded movies- compression artifacts are.
My reason for buying the real thing (CD-audio, CD-ROM software, DVD movies) is that you have more assurance that you're really getting what you paid for. My time isn't free so finding and downloading an mp3 with a glitch is no fun. I can't get my time back and that glitch might be a full-scale tone burst (ouch!) because of the way the codec works.
There was a recent sf book that suggested that software companies invented viruses as a form of copy protection. Very few CD-ROMs have gone out the door infected.
seriously, digital cable is low quality [...because...] the decoders just don't have the power to decode it well enough on the fly.
Nope. The boxes are built as cheaply as possible, but let me assure you that the MPEG-2 decoders in all of Scientific-Atlanta's and Motorola's boxes are fully capable of CCIR-601 resolution video. That's 704 horizontal pixels (they can actually handle 720, but 704 is an integer multiple of 32) on 486 interlaced lines (the other lines to get you to 525 are vertical blanking) by 29.97 frames per second. Incidentally, 704/(4/3) = 528 which is why DSS and DVD are often spec'd as having "over 500 lines of resolution." LOR are a measure of horizontal resolution using a square image.
AT&T Broadband uses the "Headend In The Sky" (HITS) which broadcasts at 352x480i. They also soften the picture before it hits the real-time-encoders so the digital artifacts will be less apparent (instead it's just soft). I agree that the saturation seems low too. I don't know why this is. If you look at HITS website (http://www.hits.com), you'll see they're running up to 12 channels per transponder (actually, the "Programming Lineup" page is 404-ing). A QAM64 channel (they modulate the digital stuff into a 6MHz carrier so it looks just like another analog channel to the transmission equipment) carries 27MBits/sec so that's under 3MBits/sec per channel. DVD's peak isn't much higher than this, but the compression isn't done in real time and can be hand-tweaked.
In the case of AT&T, they're cheaping out on the network, not the boxes because TCI got the network by buying a lot of small operators and never upgraded them. In Santa Cruz, we have a 350MHz plant so they have to really squeeze in order to get some new channels in. Time-Warner has been upgrading their plants for the last few years and have 750MHz and higher bandwidth installed. They started with HITS, but put up their own bird (Athena) 'cause the quality sucked. I don't know if it actually turned out better.
So, where do they save $$ on the boxes? MPEG-2 decoders aren't actually that expensive and they only need a couple meg of SDRAM.
Some of you have noticed that the boxes are slow changing channels & drawing graphics. Try 8MB SDRAM, less than 100MHz CPUs with teeny L1 and no L2 cache, running code directly out of 4MB of flash (mucho wait states) and only rudimentary graphics acceleration.
They could actually stream MPEG-2 right out of DVD players (after un-CSS-ing 'em) into the cable network, but I guess the studios wouldn't like that.
Oh- another way the cable co.'s are cheaping out is that (for PPV) they compress Dolby Pro Logic soundtracks (2 channels) using AC-3 (the codec can run on any no. of channels, not only 5+1). That Dolby logo on your box doesn't mean you're getting "real" Dolby Digital (meaning discrete surround channels to many people).
Maybe someday real-time-encoding will catch up with the offline stuff.
1st, don't assume S/PDIF is error-corrected. There's actually a bit in the subcodes in a S/PDIF or AES3 (AES/EBUG) stream that signifies invalid samples. The sender detects an ECC error but doesn't do the interpolation, leaving that up to the receiver/DAC. I don't know whether it's actually used in practice- lots of things that S/PDIF allows are not. (see the dat-heads mailing list archives or John Watkinson's 'The Art of Digital Audio'
2nd, it's time for some firmware hacking! A CD-ROM that circumvents one particular scheme is not my goal: a CD-ROM drive that moves much of the control up to the host is. Even if the microcontroller in the drive has its code (mask) ROMmed, most uCs have one-time-programmable or external-code versions. *Extracting* the code in order to dissassemble, reverse-engineer and re-write it might be a hassle, but there's a big population of drives out there to choose from. So... how 'bout suggestions purely from the good-in-general, popular/plentiful and reasonably cheap standpoint? We'll examine drives on the candidate list & pick one if we ever get a list of protected CDs.
A CD-R drive would probably be the best so we can write CDs with strange tricks in them too.
3rd, to add to the speculation about how this could work, consider this: once in a while a sample will exactly match what the preceding (and possibly following depending on algorithm) samples would have caused the interpolator to produce. If you replace it with a sample of a radically different value and use cheesed ECC you will get exactly the original audio (so audiophiles won't be able to hear it). Until, that is, another sample in the vicinity is made unreadable by scratching. Then the missing sample which could have been used to help interpolate the newly-wrecked sample isn't available and the interpolation may introduce distortion.
If it's a digital signal and it's encrypted, it can be broken.
But it hasn't been (yet). The cable box manufacturers learned from the DSS hacks and used better security. Even though the Motorola and Scientific-Atlanta boxes (used by AT&T Broadband, Time-Warner Cable, Charter, Videotron, Cogeco and just about every other US & Canadian cableco) have smartcard slots, they don't use them.
They do use some obscurity, but they don't rely on it- there's some real thought here (not just XORs!).
Think about it- if there weren't any PC DVD decryptors would we have DeCSS and other "unauthorized" decryptors today? *Maybe* if the key were in the clear (but it would be in ROM rather than in a DLL/EXE), but what if the key were embedded in a chip and that takes an encoded stream in and gives a clear one out. There is technology to scrape off the top of a chip and read its design or contents using an electron microscope, but there's also technology which uses a bunch of layers of metallization piled on top of the sensitive stuff (probably developed for Clipper;-).
Here's something on Mot's box (formerly GI):
"The fundamental elements of GI's approach include: (i) a secure, non-reusable, single die VLSI
custom decoder chip; (ii) a cryptographically secure mating verification scheme between the buried
secure processor and the renewable element (if and when renewable elements are installed); (iii)
battery backed-up volatile memory for secure storage in both the fixed and renewable security
elements; (iv) working key (control word) which changes several times per second; (v) use of proven
and strong cryptographic algorithms (e.g., DES and DES variants); and (vi) renewable security"
And here's something on S-A's:
"Scientific-Atlanta's PowerKEY System is the broadband industry's first CA system to support both
public key and secret key cryptography. PowerKEY's use of public key (RSA) cryptography allows it to
address the issues discussed above in a unique way that traditional secret key-only CA systems
cannot match."
"The PowerKEY CA system employs a multi-level key hierarchy. Control words are fast-changing keys
used to encrypt the services (video, audio, data). Mid-level keys called multi-session keys are used
to protect the control words so that they can not be discovered in transmission, except by
authorized units. The multi-session keys are sent to individual decoders using messages (EMMs) that
are encrypted with the RSA public key algorithm. These EMMs are also digitally signed by an
Entitlement Authority. "
(the original URLs are broken now, but look around for DigiCipher and PowerKEY if yer innerested)
The work on the connection between cable boxes and digital TV sets/decoders and on retail (Circuit City) cable boxes is mostly going on under the banner of OpenCable, run by Cable Labs. You'll buy a TV or cable box or D-VCR with a PCMCIA-like slot into which you'll plug a Point-of-Deployment (POD) module rented from the cable provider.
The generic box doesn't know DigiCipher or PowerKEY, but they don't want the POD to output a clear stream so it's re-encrypted using a generic system- 5c. The digital connection between a cable box and HDTV decoder might also be 5c-encrypted MPEG over firewire, but it also might be decompressed DVI with some other nasty "generic" (less proprietary than DigiCihper- more like CSS) encryption applied.
True, but the more money I receive from my writing, the less time I need to spend on a "real" job earning it, and therefore the more time I can spend writing.
No, the less money you spend, the less time you'll need to spend at a "real" job. Move to a garret in a cheap city and eat nothing but cold soup and you'll have plenty of time to write.
The Maori's claims don't seem to me to be out of whack to me. Legally, they may lose but the legal system is stacked in favor of industrial citizens in the first place.
From a *moral* point of view, how is protecting a trademark or copyright on a work of fiction or photograph created by an industrialist so different from what the Maori are attempting to do? Maybe you don't believe in copyrights, but even copyleft is based on copyright law and the notion that a creator of a work can have some say in its use.
Numerous examples of other cultures that have been pilfered from and allowed it have been posed, but let me pose one closer to the truth (as I see it): I create a game that's a work of fiction "drawing on several cultures." (even if we accept Lego's assertion, there are problems). I have characters named Charlie Brown and Asterix who drive deux cheveauxs, ride pedersen bikes drink coca-cola and wear dickies and apollo moon helmets.
If they instead rode a horse named Altivo, wore lederhosen and drank mead, I wouldn't have a legal problem. Why? I guess because the people that made that stuff up were too busy not dying of disease by 30 to bother with copyrights and trademarks.
Deriving work from that of your ancestors is still stealing, but your current culture is derived from the ancient one and icons and tales of yore are going to resonate with you. It's still a good idea to be respectful and accurate so you don't look like a boor or a moron.
The problem is that western culture has been through the mish-most of self-reexamination so much that creative works can become inbred and stale (cf Britney Spears). Some cross-pollination becomes necessary. The desire to appropriate and derive from other cultures is certainly understandable, but it's not yours (or Lego's) to take.
Slashdot Mirror
http://www.sel.sony.com/SEL/consumer/ss5/home/acce ssories/universalremote/rm-vl900.shtml
$60. No touchscreen- not even a little LCD like the RM-V60, but *every* key is programmable for each of 8 components (unlike some universals like one RCA that have very limited key override storage). 3 macro keys.
It has real buttons and you can set the volume keys to control the receiver regardless of component selection.
I control my Tivo with it and while I have to remember that TV/Video is thumbs-down, I don't have to look at it to find the button.
Simple & cheap. If you think you need something more complicated, like another poster said you probably need to get out more. I use a complicated computer all day. When I want to watch some TiVo or DVD, I'd prefer simplicity.
How many servers are subjected to shocks that will make the heatsink FALL OFF?
None. Unless you're in an earthquake zone, and then I rather suspect you have more important things to worry about.
I live less than 10 miles from the San Andreas and I'm more worried about an Athlon burning my house down than an earthquake knocking it down.
The house sits on steel reinforced concrete piers that go more than 15 feet into the sandstone, the house is bolted to the foundation and the framing has shear panels attached. The gas lines to the water heater and furnace are flexible so when the inertia of 30 gallons of water resists shaking, it won't snap the line.
We're less careful about the contents of the house, so an earthquake will give us fish on the floor swimming in shards of vases and my CRT. That should be about it, though. I hope if the computer falls over, though, that it doesn't cause a fire!
What Tom's Hardware didn't show is what the PC looks like 10 minutes after the heat sink is removed. Does the thermal runaway eventually burn itself out?
I wouldn't be worried about my PC being dead after an earthquake, but I'd rather it not take anything with it.
Hey, cable (and DSL) is way faster than ISDN.
R I.html
ISDN isn't a technology, but a tariff structure.
You're probably thinking of ISDN BRI, the Basic Rate Interface consisting of 2 B (bearer) channels of 56kbps or 64kpbs and one 16kbps D channel (signalling).
ISDN PRI or Primary Rate Interface is 23 B channels and one 64kbps D channel. 24 * 64 = 1.5mbps. Hmmm... T1!
I know it's common to call BRI simply ISDN, particularly in the residential context of this discussion, but people are trying to claim they know more than the broadband ISPs about how to achieve security, so using the proper terms might help.
http://www.bell-labs.com/technology/access/ISDN-B
The only ways around the patent are to find a completely different way of decoding/encoding the data (very unlikely, if the patent attorney did his job)
It might be hard, but finding patent workarounds is done. Let me give three examples:
1) Creative Technology, Ltd. was sick of paying Yamaha for their FM synthesizer chips which by the mid 90's weren't amazing tech. anymore(low transistor count) but were expensive due to a nice passel of patents on stuff developed at CCRMA @ Stanford. The 1st/main patent was expiring, so the Creative Advanced Technology Center came up with a workaround that skirted the remaining patents, saving $millions (I think it was even pin compatible so they didn't have to rev their PCBs).
2) There might be patents on Dolby's analog surround systems, but plenty of 2nd run theaters show cheap prints in "Surround Sound" and some VHS tapes are even encoded that way (there's a distinctive logo, but no info on the company behind this- anyone know?). I'm betting that phase-shifted matrixing is not protected anymore, but Pro Logic which is Dolby's extraction method that gives better channel separation through "steering" is.
3) Sony tried to skirt Dolby's perceptual coding patents when they developed ATRAC. There were also patents by Fraunhofer & probably others. They failed (at least Dolby's lawyers convinced Sony's lawyers they'd failed) and there's a Dolby logo (and Dolby license fees paid) on Minidisc machines.
First thing is to stop calling the open source decoder AC-3. Call it digital surround and avoid the trademark dispute. Ogg Vorbis managed to create a perceptual codec and avoid the patents (how!!???). Maybe it can be done for a Dolby Digital decoder. Probably not, though, as the method by which the channels are interleaved (I believe busy channels can "borrow" bandwidth from quiet ones) is protected.
Didn't IBM protect Microchannel with patents? The courts have repeatedly protected companies who reverse-engineer and make compatible peripherals, but if the interface (Intel Slot 1 and Apple Desktop Bus are other examples) contains any "innovations" it can be protected until it's obsolete. Adding copyright-able code to the peripheral that must be run before it can be "activated" (Nintendo cartridges) is another way of keeping out the compatible glommers-on.
Your company may have the right to search the file drawer in your desk (every co. I've worked for has had that provision in the employment contract), but do they have the right to search your purse (backpack) sitting in it? Not in any company I've worked for.
To take this even further, should they be able to search your pockets? Your rectum (better use Memory Sticks for those secrets; Compact Flash cards are gonna hurt!), your urine (whoops, too late)? The crud under your fingernails?
The "wrapper of privacy" may seem arbitrary or silly, but don't you feel more comfortable having it? Wouldn't it be nice if it extended to your computer?
...widespread adoption of systems like Freenet...
Hear hear!
This may lead to an arms race of freenet traffic trying to look more and more like "normal" traffic and the filters and bounty hunters getting smarter. However, despite those that suggest we must develop the "cypherpunk net" in secret to keep it from being shut down, I think open source may be a big advantage. The filter developers cannot work together and will end up re-inventing wheels ad naseum.
We not only need a better underlying technology in freenet we need better content. Consider RMS's argument against the Lesser GPL: if the good stuff is under GPL, more people will be convinced to go the Free route. If there is a lot of good stuff and it's easier to get due to a decentralized approach (no slashdot effect) and EZ2USE clients are written, the network will have enough nodes to survive attacks.
How? Well, to start if you're an independent band with mp3s @ mp3.com, consider hosting 'em on a freenet server also. If you have a server, offer the space.
I don't know where peekabooty is going, but consider this: a freenet proxy that acts as a local webserver so you can browse the net using a web browser.
Could (something like) freenet bring the promise of multicast without requiring infrastructure upgrades?
We will need user controllable rate limiters on the server/clients. However, if there are enough nodes on the network and we include ideas like the 110% [1] download, this network will be just a background hum. Some kinds of packet address spoofing and maybe using TCP packets ('cause UDP might be filtered) but not ack/nak-ing them (treat 'em as UDP) might be necessary to hide server's identities, but this has to be balanced with making the traffic indistinguishable from regular surfing (or playing Quake). It must be impossible to find the owner of the server not only through the client's interface, but also using any tool available to an ISP or bounty hunter.
I haven't set up a freenet server at home 'cause I haven't convinced myself it won't fill up with kiddie porn. Besides the moral issues, I'd rather not have the stormtroopers breaking down my door. (so... a related thread is how do you build a truly anonymous network and not have it turn into a haven for criminals you don't like (vs. the liberal copyright users (DMCA criminals) you do))
[1] There was some noise recently about a technique where a file is turned into a sort of "hologram" that results in a slight increase in size (say 125%). The packets are downloaded using UDP and any collection of 110% or more of the orig. size (so less than all the packets) can reconstruct the original.
DVDs are 704x486 (they use non-square pixels and provide for a bit more overscan). Check out the CCIR-601 spec or Watkinson's Digital Video book. 640x480 is not quite DVD, but it's better than laserdisc.
Of course, resolution is probably not the problem in downloaded movies- compression artifacts are.
My reason for buying the real thing (CD-audio, CD-ROM software, DVD movies) is that you have more assurance that you're really getting what you paid for. My time isn't free so finding and downloading an mp3 with a glitch is no fun. I can't get my time back and that glitch might be a full-scale tone burst (ouch!) because of the way the codec works.
There was a recent sf book that suggested that software companies invented viruses as a form of copy protection. Very few CD-ROMs have gone out the door infected.
seriously, digital cable is low quality [...because...] the decoders just don't have the power to decode it well enough on the fly.
Nope. The boxes are built as cheaply as possible, but let me assure you that the MPEG-2 decoders in all of Scientific-Atlanta's and Motorola's boxes are fully capable of CCIR-601 resolution video. That's 704 horizontal pixels (they can actually handle 720, but 704 is an integer multiple of 32) on 486 interlaced lines (the other lines to get you to 525 are vertical blanking) by 29.97 frames per second. Incidentally, 704/(4/3) = 528 which is why DSS and DVD are often spec'd as having "over 500 lines of resolution." LOR are a measure of horizontal resolution using a square image.
AT&T Broadband uses the "Headend In The Sky" (HITS) which broadcasts at 352x480i. They also soften the picture before it hits the real-time-encoders so the digital artifacts will be less apparent (instead it's just soft). I agree that the saturation seems low too. I don't know why this is. If you look at HITS website (http://www.hits.com), you'll see they're running up to 12 channels per transponder (actually, the "Programming Lineup" page is 404-ing). A QAM64 channel (they modulate the digital stuff into a 6MHz carrier so it looks just like another analog channel to the transmission equipment) carries 27MBits/sec so that's under 3MBits/sec per channel. DVD's peak isn't much higher than this, but the compression isn't done in real time and can be hand-tweaked.
In the case of AT&T, they're cheaping out on the network, not the boxes because TCI got the network by buying a lot of small operators and never upgraded them. In Santa Cruz, we have a 350MHz plant so they have to really squeeze in order to get some new channels in. Time-Warner has been upgrading their plants for the last few years and have 750MHz and higher bandwidth installed. They started with HITS, but put up their own bird (Athena) 'cause the quality sucked. I don't know if it actually turned out better.
So, where do they save $$ on the boxes? MPEG-2 decoders aren't actually that expensive and they only need a couple meg of SDRAM.
Some of you have noticed that the boxes are slow changing channels & drawing graphics. Try 8MB SDRAM, less than 100MHz CPUs with teeny L1 and no L2 cache, running code directly out of 4MB of flash (mucho wait states) and only rudimentary graphics acceleration.
They could actually stream MPEG-2 right out of DVD players (after un-CSS-ing 'em) into the cable network, but I guess the studios wouldn't like that.
Oh- another way the cable co.'s are cheaping out is that (for PPV) they compress Dolby Pro Logic soundtracks (2 channels) using AC-3 (the codec can run on any no. of channels, not only 5+1). That Dolby logo on your box doesn't mean you're getting "real" Dolby Digital (meaning discrete surround channels to many people).
Maybe someday real-time-encoding will catch up with the offline stuff.
1st, don't assume S/PDIF is error-corrected. There's actually a bit in the subcodes in a S/PDIF or AES3 (AES/EBUG) stream that signifies invalid samples. The sender detects an ECC error but doesn't do the interpolation, leaving that up to the receiver/DAC. I don't know whether it's actually used in practice- lots of things that S/PDIF allows are not. (see the dat-heads mailing list archives or John Watkinson's 'The Art of Digital Audio'
2nd, it's time for some firmware hacking! A CD-ROM that circumvents one particular scheme is not my goal: a CD-ROM drive that moves much of the control up to the host is. Even if the microcontroller in the drive has its code (mask) ROMmed, most uCs have one-time-programmable or external-code versions. *Extracting* the code in order to dissassemble, reverse-engineer and re-write it might be a hassle, but there's a big population of drives out there to choose from. So... how 'bout suggestions purely from the good-in-general, popular/plentiful and reasonably cheap standpoint? We'll examine drives on the candidate list & pick one if we ever get a list of protected CDs.
A CD-R drive would probably be the best so we can write CDs with strange tricks in them too.
3rd, to add to the speculation about how this could work, consider this: once in a while a sample will exactly match what the preceding (and possibly following depending on algorithm) samples would have caused the interpolator to produce. If you replace it with a sample of a radically different value and use cheesed ECC you will get exactly the original audio (so audiophiles won't be able to hear it). Until, that is, another sample in the vicinity is made unreadable by scratching. Then the missing sample which could have been used to help interpolate the newly-wrecked sample isn't available and the interpolation may introduce distortion.
If it's a digital signal and it's encrypted, it can be broken.
;-).
But it hasn't been (yet). The cable box manufacturers learned from the DSS hacks and used better security. Even though the Motorola and Scientific-Atlanta boxes (used by AT&T Broadband, Time-Warner Cable, Charter, Videotron, Cogeco and just about every other US & Canadian cableco) have smartcard slots, they don't use them.
They do use some obscurity, but they don't rely on it- there's some real thought here (not just XORs!).
Think about it- if there weren't any PC DVD decryptors would we have DeCSS and other "unauthorized" decryptors today? *Maybe* if the key were in the clear (but it would be in ROM rather than in a DLL/EXE), but what if the key were embedded in a chip and that takes an encoded stream in and gives a clear one out. There is technology to scrape off the top of a chip and read its design or contents using an electron microscope, but there's also technology which uses a bunch of layers of metallization piled on top of the sensitive stuff (probably developed for Clipper
Here's something on Mot's box (formerly GI):
"The fundamental elements of GI's approach include: (i) a secure, non-reusable, single die VLSI custom decoder chip; (ii) a cryptographically secure mating verification scheme between the buried secure processor and the renewable element (if and when renewable elements are installed); (iii) battery backed-up volatile memory for secure storage in both the fixed and renewable security elements; (iv) working key (control word) which changes several times per second; (v) use of proven and strong cryptographic algorithms (e.g., DES and DES variants); and (vi) renewable security"
And here's something on S-A's:
"Scientific-Atlanta's PowerKEY System is the broadband industry's first CA system to support both public key and secret key cryptography. PowerKEY's use of public key (RSA) cryptography allows it to address the issues discussed above in a unique way that traditional secret key-only CA systems cannot match."
"The PowerKEY CA system employs a multi-level key hierarchy. Control words are fast-changing keys used to encrypt the services (video, audio, data). Mid-level keys called multi-session keys are used to protect the control words so that they can not be discovered in transmission, except by authorized units. The multi-session keys are sent to individual decoders using messages (EMMs) that are encrypted with the RSA public key algorithm. These EMMs are also digitally signed by an Entitlement Authority. "
(the original URLs are broken now, but look around for DigiCipher and PowerKEY if yer innerested)
The work on the connection between cable boxes and digital TV sets/decoders and on retail (Circuit City) cable boxes is mostly going on under the banner of OpenCable, run by Cable Labs. You'll buy a TV or cable box or D-VCR with a PCMCIA-like slot into which you'll plug a Point-of-Deployment (POD) module rented from the cable provider.
The generic box doesn't know DigiCipher or PowerKEY, but they don't want the POD to output a clear stream so it's re-encrypted using a generic system- 5c. The digital connection between a cable box and HDTV decoder might also be 5c-encrypted MPEG over firewire, but it also might be decompressed DVI with some other nasty "generic" (less proprietary than DigiCihper- more like CSS) encryption applied.
True, but the more money I receive from my writing, the less time I need to spend on a "real" job earning it, and therefore the more time I can spend writing.
No, the less money you spend, the less time you'll need to spend at a "real" job. Move to a garret in a cheap city and eat nothing but cold soup and you'll have plenty of time to write.
From a *moral* point of view, how is protecting a trademark or copyright on a work of fiction or photograph created by an industrialist so different from what the Maori are attempting to do? Maybe you don't believe in copyrights, but even copyleft is based on copyright law and the notion that a creator of a work can have some say in its use.
Numerous examples of other cultures that have been pilfered from and allowed it have been posed, but let me pose one closer to the truth (as I see it): I create a game that's a work of fiction "drawing on several cultures." (even if we accept Lego's assertion, there are problems). I have characters named Charlie Brown and Asterix who drive deux cheveauxs, ride pedersen bikes drink coca-cola and wear dickies and apollo moon helmets.
If they instead rode a horse named Altivo, wore lederhosen and drank mead, I wouldn't have a legal problem. Why? I guess because the people that made that stuff up were too busy not dying of disease by 30 to bother with copyrights and trademarks.
Deriving work from that of your ancestors is still stealing, but your current culture is derived from the ancient one and icons and tales of yore are going to resonate with you. It's still a good idea to be respectful and accurate so you don't look like a boor or a moron.
The problem is that western culture has been through the mish-most of self-reexamination so much that creative works can become inbred and stale (cf Britney Spears). Some cross-pollination becomes necessary. The desire to appropriate and derive from other cultures is certainly understandable, but it's not yours (or Lego's) to take.