Cheaper to use a dedicated MP3 chip
on
New MP3 Portables
·
· Score: 1
why not just make a portable audio player that allows you to flash any type of audio decoder into it
Apparently, it's cheaper to build an MP3 player from a 4,000-gate 6502 processor plus a dedicated MP2/MP3 decoder than it is to include a general-purpose DSP that's fast enough for real-time playback.
Overwriting will AND the data
on
Tiny Boxen
·
· Score: 3, Informative
Clearing to 1s then blanking 0s would be more writes than just basic overwriting.
Flash memory is divided into sectors. When you erase a sector of flash memory, the whole sector becomes all 1's. The 100K writes figure refers to 100K successful erases of a given sector. "Just basic overwriting" would AND the written data into the existing data. I'm guessing that flash file systems take advantage of this somehow.
And Huffman encoding is about repeating patterns, not frequency of 1s and 0s without order taken into account. You take a long but frequent pattern and replace it with a shorter one, but then have to remap the shorter one etc.
Huffman maps fixed-length sequences of bits (usually 4, 8, or 16 at a time) to variable-length sequences of bits. If you have lots more 1's than 0's, then you'll get a lot of 1111, 1110, 1101, 1011, and 0111 nibbles, which can be reduced to shorter words. However, JPEG, MP3, and MPEG already have compression (including Huffman coding) in the bitstream, so re-compressing the data isn't going to help.
200K writes is still too few
on
Tiny Boxen
·
· Score: 3, Informative
The special logic involves checking to see if the flash bit contains the value (1/0) you want before writing to it.
Not exactly. Flash memory is written to by first erasing the sector to all 1's and then clearing the bits you want cleared.
With a 50% hit-rate 100k writes becomes 200k.
200,000 writes is still too few for a directory track.
data of a particular type (eg html or mpeg) tend to bias towards 1s or 0s.
HTML might bias slightly, but MPEG doesn't. If it did, you would be able to compress MPEG files further with the Huffman coding used in gzip. (You can't.)
There is something called fair use involved here, which palladium seeks to impinge upon.
Fair enough. Fair use (17 USC 107) does not guarantee to the owner of a copy the right to make an exact reproduction. A digital restrictions management system implemented on top of Palladium will never stop you from exercising fair use through two little analog holes on the back of your PC: the headphone connector and the video output connector. Watermark defeat devices are legal under the DMCA, 17 USC 1201(a)(2) and (b)(1), because the devices are marketed for their substantial non-infringing use of improving the quality of generic video signals.
Flash's wear pattern is different from HD's
on
Tiny Boxen
·
· Score: 1
Hard disks wear out too, most of them faster than flash memory
When a hard disk dies, the whole thing dies. On the other hand, flash dies one sector at a time. A hard disk will last over 100K writes to the directory tracks. On flash, without some kind of smart sector wear management, the sectors holding the root directory and free space bitmap/FAT will wear out first.
The stylesheet keeps it from rendering
on
Tiny Boxen
·
· Score: 1
When I try the Google cache, I get the HTML all right, but Mozilla doesn't render the page until either the stylesheet has loaded or the connection times out, which could be a whole minute.
If Flash memory gets just a little cheaper, you could have a serious desktop computer with no moving parts at all.
That'd be hard. A sector of a flash chip will wear out and turn into a "bad sector" after about 100,000 writes. The flash controller will have to have some sort of logic to treat repeated writes specially. Apparently, most modern CompactFlash cartridges' integrated controllers can do this; can anybody explain how such logic works?
And even though the blurb mentions that the CPU doesn't need a fan, wouldn't the power supply still need a fan? I can't get to the server that is hosting the article, and when I try to use the Google cache, it takes several minutes for Mozilla to realize that the real server won't respond to requests for the page's stylesheet.
and all the rest become pointless diversions that CANNOT be used while any Palladium app is currently in memory.
This is incorrect. Palladium applications can certainly run on the same machine at the same time that non-Palladium applications run. They just can't see each other's memory spaces.
And IIS starts serving locked documents by default.
Which AOLers behind a Gecko browser (such as Compuserve users, Mac AOL users, and AOL 8.0 users who have turned on Gecko) cannot see. AOL has a lot of influence on the Web; if AOLers can't see your pages, you lose 30% of your market right there.
"this document must be locked to open it on a computer using Palladium"... Think about what happens when all programs must be Palladium-safe to run.
Then the custom applications developed for use by the U.S. federal government will stop working, and Congress will repeal any CBDTPA-like legislation that it may have passed. I don't see anything on Microsoft's page stating that Windows will prohibit applications that don't use Palladium.dll from running on a Palladium-capable computer in the foreseeable future.
I think our only hope is the USA legislating itself into irrelevance and other nations picking up the slack and giving them the finger.
Why has the population of Greece declined over the last few weeks?
And what's to say non-TCPA equipment, other than expensive studio gear, will be made illegal?
And what's to say studio equipment that lets you sign your own recordings will necessarily be prohibitively expensive for home-studio use? The laws of the fifty U.S. states already provide ways to deal with those who fraudulently sign recordings they don't own.
You don't have to be a Kreskin to see that *LPFM is dying. In the United States, applications for new low-power FM radio stations must be filed electronically within a five-day filing window. The FCC hasn't released any new filing window dates in 17 months. For all practical purposes, *LPFM is dead.
After which the lawyer will steal the idea and start his own radio station
Bar associations license attorneys in the U.S. and can take away the license of a lawyer who does something unprofessional such as stealing a client's idea for a business and then either competing with the client or shutting out the client with a government-granted monopoly.
How will it prevent virtual machine from interacting with virtual hardware that doesn't happen to enforce any restrictions?
Windows will detect "VMware Brand PC without Palladium support" and refuse to load the Palladium driver, and Palladium apps will refuse to open locked documents. However, AOL's Winamp media player won't need a Palladium driver to play.ogg,.mp3, or any other unlocked format that independent recording artists publish (PROVIDED that they have licensed the underlying musical composition).
Law 3) Require DRM hardware to ONLY run DRM-compliant software (not too hard to imagine).
Neither TCPA nor Palladium does this.
That's the end of legal free (and Free) software in the USA.
The federal government uses Free software. The news media use Free software (largely in BSD and Linux based web servers). Heck, two-thirds of the Big Nine media publishers (MPAA studios and RIAA labels), such as AOL Time Warner, Sony, BMG, Fox, Paramount, and Universal, run free web server software such as Apache or AOLserver on their web sites. (Disney and EMI run IIS, and MGM runs Netscape Enterprise Server.)
AOL Time Warner's "America Online" service wants the biggest market it can get, including older computers that were manufactured before the Itanium and Opteron processors will have become popular. I don't see AOL requiring Palladium in the foreseeable future, except perhaps for some premium Warner content.
companies that want to excercise their "IP" could simple make it a "you want what we have, then you have to turn it on". Negating the difference between "opt-in" and "opt-out".
Perhaps that could be true for movies given the current cost of feature film production, but for music in the popular genres, the RIAA labels have a lot of competition from independent bands who publish some of their singles via unlocked documents such as.ogg and.mp3 files. (Being independent has drawbacks: you don't get Clear Channel airplay, and licensing musical works to perform becomes expensive.) So what if opting out of Palladium means opting out of the RIAA?
What I don't get is what AMD says about Opting-in. Will we really be able to just "turn it off" in CMOS or something?
Yes. If you turn off Palladium in your BIOS's setup program, then there will be only one effect: you won't be able to open locked documents. Publishers make the choice whether or not to lock a document. For instance, an independent recording artist could still publish unlocked.ogg files via HTTP or via Gnutella (PROVIDED that he has licensed the underlying musical works).
Hollywood and the music industry are lobbying hard to make DRM mandatory in all new devices
Once the TCPA system becomes more widespread, Hollywood will have less room to bitch because there will be a Secure Memory Space(tm)(patent) in the most popular consumer operating system, and Hollywood studios will be able to provide Video On Demand services within that space.
and existing laws here and in the US make it a crime to switch it off.
Not exactly. The Palladium and TCPA systems simply provide a way to lock down data such that only specific applications running on one machine can use it. In order for Palladium or TCPA to actually restrict anything, the content provider must make the choice to lock down the data (conforming Compact Discs are not considered locked down). This doesn't give the RIAA labels an absolute oligopoly, as it's still possible for artists to Not Lock Down(tm) their.ogg files.
The public TCPA information stresses that only TCPA apps will use the TCPA memory space. Microsoft's Palladium materials make the same claim. And you'll apparently be able to turn off the systems in the BIOS setup, which will have only one effect: apps that use those systems will throw up an alert box to the effect "The locked document 'Love Me Now.wma' could not be opened because Palladium was not found." They do NOT force all documents to be locked documents.
Palladium only provides a secure space inside your machine in which software can run. If you turn off Palladium, or you don't run any apps that import palladium.dll, then Palladium has no effect on your use of the Internet. WinMX and Gnucleus will continue to let you trade copies of live recordings, especially of bands who have authorized it.
black holes evaporate after approximately 10^60 to 10^100 years, due to hawking radiation.
The article claims that the black holes will eat the universe after about 10^10 years, which isn't near long enough for M.C. Hawking to have eaten the black holes.
The point of an Ask Slashdot is to ask the general readership about their experiences in an area. In this case, it's quite likely that somebody will step forward and help out: "When I was in your situation, I talked to an attorney, and she told me to do this and this and that and that, but on this line of the contract, I was screwed".
When you delete a file, NOS/NDS, flags the file as "deleted", and timestamps it. As you reuse disk space, the space used by the oldest deleted file (regadless of owner) is removed first. Once the system uses any part of a deleted file, you cannot recover it.
Windows 95's Recycle Bin does exactly that. Mac OS version 7.0 had something a bit more primitive: whenever a file is copied onto a disk, if it won't fit, the shell "empties the trash" (purges all deleted files on all mounted volumes) first.
why not just make a portable audio player that allows you to flash any type of audio decoder into it
Apparently, it's cheaper to build an MP3 player from a 4,000-gate 6502 processor plus a dedicated MP2/MP3 decoder than it is to include a general-purpose DSP that's fast enough for real-time playback.
Clearing to 1s then blanking 0s would be more writes than just basic overwriting.
Flash memory is divided into sectors. When you erase a sector of flash memory, the whole sector becomes all 1's. The 100K writes figure refers to 100K successful erases of a given sector. "Just basic overwriting" would AND the written data into the existing data. I'm guessing that flash file systems take advantage of this somehow.
And Huffman encoding is about repeating patterns, not frequency of 1s and 0s without order taken into account. You take a long but frequent pattern and replace it with a shorter one, but then have to remap the shorter one etc.
Huffman maps fixed-length sequences of bits (usually 4, 8, or 16 at a time) to variable-length sequences of bits. If you have lots more 1's than 0's, then you'll get a lot of 1111, 1110, 1101, 1011, and 0111 nibbles, which can be reduced to shorter words. However, JPEG, MP3, and MPEG already have compression (including Huffman coding) in the bitstream, so re-compressing the data isn't going to help.
The special logic involves checking to see if the flash bit contains the value (1/0) you want before writing to it.
Not exactly. Flash memory is written to by first erasing the sector to all 1's and then clearing the bits you want cleared.
With a 50% hit-rate 100k writes becomes 200k.
200,000 writes is still too few for a directory track.
data of a particular type (eg html or mpeg) tend to bias towards 1s or 0s.
HTML might bias slightly, but MPEG doesn't. If it did, you would be able to compress MPEG files further with the Huffman coding used in gzip. (You can't.)
There is something called fair use involved here, which palladium seeks to impinge upon.
Fair enough. Fair use (17 USC 107) does not guarantee to the owner of a copy the right to make an exact reproduction. A digital restrictions management system implemented on top of Palladium will never stop you from exercising fair use through two little analog holes on the back of your PC: the headphone connector and the video output connector. Watermark defeat devices are legal under the DMCA, 17 USC 1201(a)(2) and (b)(1), because the devices are marketed for their substantial non-infringing use of improving the quality of generic video signals.
Hard disks wear out too, most of them faster than flash memory
When a hard disk dies, the whole thing dies. On the other hand, flash dies one sector at a time. A hard disk will last over 100K writes to the directory tracks. On flash, without some kind of smart sector wear management, the sectors holding the root directory and free space bitmap/FAT will wear out first.
When I try the Google cache, I get the HTML all right, but Mozilla doesn't render the page until either the stylesheet has loaded or the connection times out, which could be a whole minute.
If Flash memory gets just a little cheaper, you could have a serious desktop computer with no moving parts at all.
That'd be hard. A sector of a flash chip will wear out and turn into a "bad sector" after about 100,000 writes. The flash controller will have to have some sort of logic to treat repeated writes specially. Apparently, most modern CompactFlash cartridges' integrated controllers can do this; can anybody explain how such logic works?
And even though the blurb mentions that the CPU doesn't need a fan, wouldn't the power supply still need a fan? I can't get to the server that is hosting the article, and when I try to use the Google cache, it takes several minutes for Mozilla to realize that the real server won't respond to requests for the page's stylesheet.
and all the rest become pointless diversions that CANNOT be used while any Palladium app is currently in memory.
This is incorrect. Palladium applications can certainly run on the same machine at the same time that non-Palladium applications run. They just can't see each other's memory spaces.
And IIS starts serving locked documents by default.
Which AOLers behind a Gecko browser (such as Compuserve users, Mac AOL users, and AOL 8.0 users who have turned on Gecko) cannot see. AOL has a lot of influence on the Web; if AOLers can't see your pages, you lose 30% of your market right there.
"this document must be locked to open it on a computer using Palladium" ... Think about what happens when all programs must be Palladium-safe to run.
Then the custom applications developed for use by the U.S. federal government will stop working, and Congress will repeal any CBDTPA-like legislation that it may have passed. I don't see anything on Microsoft's page stating that Windows will prohibit applications that don't use Palladium.dll from running on a Palladium-capable computer in the foreseeable future.
I think our only hope is the USA legislating itself into irrelevance and other nations picking up the slack and giving them the finger.
Why has the population of Greece declined over the last few weeks?
Firewire is a proprietary trademark from Apple Computer. There's some sort of IEE number
Sony calls its implementations i.Link, but most companies selling 1394 devices license the "FireWire" trademark from Apple.
And what's to say non-TCPA equipment, other than expensive studio gear, will be made illegal?
And what's to say studio equipment that lets you sign your own recordings will necessarily be prohibitively expensive for home-studio use? The laws of the fifty U.S. states already provide ways to deal with those who fraudulently sign recordings they don't own.
You don't have to be a Kreskin to see that *LPFM is dying. In the United States, applications for new low-power FM radio stations must be filed electronically within a five-day filing window. The FCC hasn't released any new filing window dates in 17 months. For all practical purposes, *LPFM is dead.
After which the lawyer will steal the idea and start his own radio station
Bar associations license attorneys in the U.S. and can take away the license of a lawyer who does something unprofessional such as stealing a client's idea for a business and then either competing with the client or shutting out the client with a government-granted monopoly.
How will it prevent virtual machine from interacting with virtual hardware that doesn't happen to enforce any restrictions?
Windows will detect "VMware Brand PC without Palladium support" and refuse to load the Palladium driver, and Palladium apps will refuse to open locked documents. However, AOL's Winamp media player won't need a Palladium driver to play .ogg, .mp3, or any other unlocked format that independent recording artists publish (PROVIDED that they have licensed the underlying musical composition).
Law 3) Require DRM hardware to ONLY run DRM-compliant software (not too hard to imagine).
Neither TCPA nor Palladium does this.
That's the end of legal free (and Free) software in the USA.
The federal government uses Free software. The news media use Free software (largely in BSD and Linux based web servers). Heck, two-thirds of the Big Nine media publishers (MPAA studios and RIAA labels), such as AOL Time Warner, Sony, BMG, Fox, Paramount, and Universal, run free web server software such as Apache or AOLserver on their web sites. (Disney and EMI run IIS, and MGM runs Netscape Enterprise Server.)
once theyre in place make aol/msn/etc require it
AOL Time Warner's "America Online" service wants the biggest market it can get, including older computers that were manufactured before the Itanium and Opteron processors will have become popular. I don't see AOL requiring Palladium in the foreseeable future, except perhaps for some premium Warner content.
companies that want to excercise their "IP" could simple make it a "you want what we have, then you have to turn it on". Negating the difference between "opt-in" and "opt-out".
Perhaps that could be true for movies given the current cost of feature film production, but for music in the popular genres, the RIAA labels have a lot of competition from independent bands who publish some of their singles via unlocked documents such as .ogg and .mp3 files. (Being independent has drawbacks: you don't get Clear Channel airplay, and licensing musical works to perform becomes expensive.) So what if opting out of Palladium means opting out of the RIAA?
What I don't get is what AMD says about Opting-in. Will we really be able to just "turn it off" in CMOS or something?
Yes. If you turn off Palladium in your BIOS's setup program, then there will be only one effect: you won't be able to open locked documents. Publishers make the choice whether or not to lock a document. For instance, an independent recording artist could still publish unlocked .ogg files via HTTP or via Gnutella (PROVIDED that he has licensed the underlying musical works).
Hollywood and the music industry are lobbying hard to make DRM mandatory in all new devices
Once the TCPA system becomes more widespread, Hollywood will have less room to bitch because there will be a Secure Memory Space(tm)(patent) in the most popular consumer operating system, and Hollywood studios will be able to provide Video On Demand services within that space.
and existing laws here and in the US make it a crime to switch it off.
Not exactly. The Palladium and TCPA systems simply provide a way to lock down data such that only specific applications running on one machine can use it. In order for Palladium or TCPA to actually restrict anything, the content provider must make the choice to lock down the data (conforming Compact Discs are not considered locked down). This doesn't give the RIAA labels an absolute oligopoly, as it's still possible for artists to Not Lock Down(tm) their .ogg files.
The public TCPA information stresses that only TCPA apps will use the TCPA memory space. Microsoft's Palladium materials make the same claim. And you'll apparently be able to turn off the systems in the BIOS setup, which will have only one effect: apps that use those systems will throw up an alert box to the effect "The locked document 'Love Me Now.wma' could not be opened because Palladium was not found." They do NOT force all documents to be locked documents.
and the looming, inevitable rise of pallidium
Palladium only provides a secure space inside your machine in which software can run. If you turn off Palladium, or you don't run any apps that import palladium.dll, then Palladium has no effect on your use of the Internet. WinMX and Gnucleus will continue to let you trade copies of live recordings, especially of bands who have authorized it.
black holes evaporate after approximately 10^60 to 10^100 years, due to hawking radiation.
The article claims that the black holes will eat the universe after about 10^10 years, which isn't near long enough for M.C. Hawking to have eaten the black holes.
You can't patent something in the US if it has been patented elsewhere
Yes you can. According to 17 USC 102, he who files a foreign patent has twelve months to file a U.S. patent.
This is as bad as the MP3 whiners. Want free music? Make some, and give it away.
Not if some songwriter comes out of the woodwork and claims that "Your song is substantially similar to my song; therefore, I'll see you in court."
The point of an Ask Slashdot is to ask the general readership about their experiences in an area. In this case, it's quite likely that somebody will step forward and help out: "When I was in your situation, I talked to an attorney, and she told me to do this and this and that and that, but on this line of the contract, I was screwed".
When you delete a file, NOS/NDS, flags the file as "deleted", and timestamps it. As you reuse disk space, the space used by the oldest deleted file (regadless of owner) is removed first. Once the system uses any part of a deleted file, you cannot recover it.
Windows 95's Recycle Bin does exactly that. Mac OS version 7.0 had something a bit more primitive: whenever a file is copied onto a disk, if it won't fit, the shell "empties the trash" (purges all deleted files on all mounted volumes) first.