That topic has *lots* (too much, probably) of info on this. There are several different designs. Personally I would think that the wood-based ones developed there are better (and they are the most common). I am designing a CCFL-lighted(one tube per arrow which light up permanently, when you step, or when you're SUPPOSED to step depending on mode), microcontrolled (PIC18F4550 based), pressure-sensor-based (no moving parts, at least not substantially moving), adjustable (you can set up how much pressure trips the state machine and registers a step), triple-system (USB,XBOX,PS2) wood DDR pad (check out the last couple of pages on that thread).
I'd be happy to be the poster. When *my* 256kbit DSL was "upgraded" (along with all other customers') to 512k, instead, mine dropped to ~80kbit. The service agreement only guarantees 10% of the advertised speed (BS!), so they refused to fix it. It took an email forwarded to the top directives of the ISP to get a proper engineer on the phone, who fixed the misconfiguration within two minutes. No one's going to refund me for my almost-dialup speeds for a month, or for the endless non-toll-free support calls (which they made sure you waited 15 minutes each).
Of course, it was a network-level cap (or misconfiguration causing a cap) somewhere on their servers, not the sync with the DSLAM (that was fine), so I had no definite proof (and they kept arguing that it was just congestion). Right, except the little bar graph on their own website's speed test was about two pixels wide for my DSL.
he means _graphing_ the curve. Which, if SVG doesn't natively support as a curve type, definitely involves converting to bezier. Except that I doubt many icons give a damn what curve type you are using. All they have to do is make sure it looks good.
by loading it into Audacity. Set Spectrogram size to 4096 on Preferences, maximum freq to ~2500Hz. Load up Deep Note from the USPTO website, stretch the audio tracks, and enjoy. It actually seems the description is pretty accurate - at first frequencies drift randomly, and then they converge to a set of fixed ones.
The software is very annoying to install, it even requires a specific component for that test. I'm on a windows machine now anyway, otherwise I would not have taken the test (my usual PC runs Gentoo amd64). It's not even flash (which I can run inside a 32bit chroot under linux), it's shockwave.
that's called steganography. You can embed data in the LSBs of sound samples, or add "noise" to an image and embed info on it. There's plenty of software to do that already.
The DVD drive will (if locked) refuse to decode and authenticate DVDs of another region. But nothing prevents libdvdcss from getting the raw encrypted sectors out and cracking the encryption itself. It currently first tries the normal method with some calculated player keys, then a brute force on the disc key, then a crypto attack on the title key. This last attack works on unauthenticated drives, and is also OK for raw encrypted VOBs that have been ripped as-is to HDD.
That applies to xbox. But what about PS2? Basic PS2 mods contain no cracked BIOS, only minor patches that are applied to the onboard BIOS. More advanced mods contain a heck of a lot more code that gets injected, but no cracked BIOS (usually, the code itself gets run and patches the BIOS in RAM, but it contains no BIOS code itself).
At least on the PS3 (and before on the PS2) the point is not having several static threads doing work, but putting all the units in the system to work at once. So while the SPEs (similar to VUs on the PS2) churn out some 3d vertex translation list and do physics, the cpu can do game logic or some other stuff. The best optimization is definitely having everything running 100% at once, although I suspect what will usually happen (which makes life easier and is acceptable too) is that units operate in sync but simultaneously (e.g. the main cpu starts everything else each frame, and the all work at once.) There will be some wasted time as some unit will finish sooner, but being kept in sync simplifies programming. It is not like threads in a PC which operate asynchronously (unless you explicitly sync)
and the ps3 will play ps2 and ps1 games, and it also has more than enough power to emulate everything ps1 and below (hell they might even make the ps2 support full emulation, although I suspect hardware aid)
this is just another case of the satellite TV smartcard case. Find a vulnerability, and use it. Or use power glitches and the like. It just takes one person with access to microchip dissection equipment to read the whole thing out for us.
Because file transfer (or rather, the initiation thereof) is just information exchange between clients, and AFAIK the server does not need to know about that. I'd bet you can get the whiteboard or chess plugins in tkabber to work too.
That is kinda pointless. The point of RAID5 is to survive a crash of ONE store. You've got two on a single drive. Drive crashes, lose all your data. To use RAID5 properly you need 3 physical drives.
The color subcarrier has to be between about 2Mhz and 4Mhz to prevent interference with the luminance signal and the audio carrier (4.5Mhz above video). 3.58 is chosen as a good number giving little interference and a very small response in old B/W sets, while still having enough bandwidth for color and not getting into the audio subcarrier space.
The old line frequency is 15,750Hz. Then you get the audio carrier, 4.5Mhz, and divide by 286 which gives 15,734.2657Hz. 286 is the even number that results in the closest approxmation to the old frequency. The frame and field numbers are calculated as you said. Then, the color frequency is the 455th harmonic of half the line frequency, or 15,734.2657/2*455 = 3579545Hz. 455 is the odd number that gets it closest to 3.58Mhz. All this helps minimize interference.
Some of my numbers are slightly different from the parent's, but those are just rounding errors.
Actually it seems to me like the engineers figured out "aha, we'll just remove the key" and not realize that (due to the way the keyboard is wired up and the way the software scans it) it is possible to make it think you pressed other keys. I figure they wanted to save themselves the hassle of changing the controller chip design, or they were just lazy or too stupid.
1 2
| |
A-B-3
| |
C-D-4
| |
Take a keyscanning algorith that works scanning left-to-right columns and up-to-down rows, that decodes the first key detected as pressed and ignores the rest. Take a keyboard matrix as shown above, with no isolation diodes. Press keys B,C,D. Watch how the connections 3-2,2-4,4-1 also create a 3-1 connection. Now the calculator just thought you pressed A. Depending on the details of it, similar stuff can happen. For example, if the thing worked by switching inputs and outputs e.g. sending current to all columns and watching for the active row, then sending current to all rows and watching to the active column, two keys (B and C) would be enough to activate all the rows and columns in the previous matrix. The calculator checks the first it finds and voila, it happily performs the funcion assigned to "A".
that is basically what I do. I also printed out hexdumps of the encrypted file and keys, just in case my drive is wiped for whatever reason. I figure I can remember the passphrase.
Actually due to a flaw in the scripts I use to handle it, I lost the passwords once. Took me 2 seconds to grep the whole HDD for one of them I remembered, and found a stray copy of the plaintext file somewhere (close to the beginning even. Probably stored as inline data in reiserfs nodes, instead of bulk secors) from some previous decoding. I guess I'll have to zero the file out next time. Now my scripts always make a backup of the encrypted file that never gets erased (only overwritten), just in case.
Using your bank account password in WP is like giving it to them (in fact, it is. Don't count on the fact that the passwords are hashed, as this proves).
Do you trust Wikipedia to keep your bank account secure? I do not. Same way as I do not trust any website any of my important passwords. That means all of them except one or two which I sometimes repeat only for online forums/websites/etc which wouldn't be a fatal loss if they were revealed.
So you went wikipedia the password you use on X. Tough luck. You should have directly *assumed* that they were going to steal you account everywhere else and see if it pays to use a different password. THAT is the only way of keeping things secure. And if you used a different password and they revealed it? So what? It's their site, they can do whatever they want with it.
I am NOT saying that wikipedia revealed anyone's password directly, I am just presenting worst-case scenarios. Not that revealing passwords is ethical or even legal (I am not sure about the laws governing passwords stored on website databases), but it is always a possibility.
Slashdot Mirror
http://www.ddrfreak.com/phpBB2/viewtopic.php?t=592 54
That topic has *lots* (too much, probably) of info on this. There are several different designs. Personally I would think that the wood-based ones developed there are better (and they are the most common). I am designing a CCFL-lighted(one tube per arrow which light up permanently, when you step, or when you're SUPPOSED to step depending on mode), microcontrolled (PIC18F4550 based), pressure-sensor-based (no moving parts, at least not substantially moving), adjustable (you can set up how much pressure trips the state machine and registers a step), triple-system (USB,XBOX,PS2) wood DDR pad (check out the last couple of pages on that thread).
I'd be happy to be the poster. When *my* 256kbit DSL was "upgraded" (along with all other customers') to 512k, instead, mine dropped to ~80kbit. The service agreement only guarantees 10% of the advertised speed (BS!), so they refused to fix it. It took an email forwarded to the top directives of the ISP to get a proper engineer on the phone, who fixed the misconfiguration within two minutes. No one's going to refund me for my almost-dialup speeds for a month, or for the endless non-toll-free support calls (which they made sure you waited 15 minutes each).
Of course, it was a network-level cap (or misconfiguration causing a cap) somewhere on their servers, not the sync with the DSLAM (that was fine), so I had no definite proof (and they kept arguing that it was just congestion). Right, except the little bar graph on their own website's speed test was about two pixels wide for my DSL.
he means _graphing_ the curve. Which, if SVG doesn't natively support as a curve type, definitely involves converting to bezier. Except that I doubt many icons give a damn what curve type you are using. All they have to do is make sure it looks good.
by loading it into Audacity. Set Spectrogram size to 4096 on Preferences, maximum freq to ~2500Hz. Load up Deep Note from the USPTO website, stretch the audio tracks, and enjoy. It actually seems the description is pretty accurate - at first frequencies drift randomly, and then they converge to a set of fixed ones.
third and last, 25ms. I'm pretty sure down here it starts to get close to the technical limit. I'll get some recordings of the waveform.
Down to 29ms, and I'm sure I can go lower.
I got 31. Pretty sure I can do better the second time.
The software is very annoying to install, it even requires a specific component for that test. I'm on a windows machine now anyway, otherwise I would not have taken the test (my usual PC runs Gentoo amd64). It's not even flash (which I can run inside a 32bit chroot under linux), it's shockwave.
Anyway, my score was 31ms. Age 15 here.
that's called steganography. You can embed data in the LSBs of sound samples, or add "noise" to an image and embed info on it. There's plenty of software to do that already.
You don't have to remove the platters, just the case. That won't change the alignment.
I once disasembled a 100MB drive I had lying around. NO amount of dust would make it headcrash. I was so disappointed.
The DVD drive will (if locked) refuse to decode and authenticate DVDs of another region. But nothing prevents libdvdcss from getting the raw encrypted sectors out and cracking the encryption itself. It currently first tries the normal method with some calculated player keys, then a brute force on the disc key, then a crypto attack on the title key. This last attack works on unauthenticated drives, and is also OK for raw encrypted VOBs that have been ripped as-is to HDD.
That applies to xbox. But what about PS2? Basic PS2 mods contain no cracked BIOS, only minor patches that are applied to the onboard BIOS. More advanced mods contain a heck of a lot more code that gets injected, but no cracked BIOS (usually, the code itself gets run and patches the BIOS in RAM, but it contains no BIOS code itself).
So are PS2 mods (flashed properly) legal?
At least on the PS3 (and before on the PS2) the point is not having several static threads doing work, but putting all the units in the system to work at once. So while the SPEs (similar to VUs on the PS2) churn out some 3d vertex translation list and do physics, the cpu can do game logic or some other stuff. The best optimization is definitely having everything running 100% at once, although I suspect what will usually happen (which makes life easier and is acceptable too) is that units operate in sync but simultaneously (e.g. the main cpu starts everything else each frame, and the all work at once.) There will be some wasted time as some unit will finish sooner, but being kept in sync simplifies programming. It is not like threads in a PC which operate asynchronously (unless you explicitly sync)
It isn't.
http://i.tuxgames.com/lgpcomp/
and the ps3 will play ps2 and ps1 games, and it also has more than enough power to emulate everything ps1 and below (hell they might even make the ps2 support full emulation, although I suspect hardware aid)
this is just another case of the satellite TV smartcard case. Find a vulnerability, and use it. Or use power glitches and the like. It just takes one person with access to microchip dissection equipment to read the whole thing out for us.
so, hack the player firmware, find out where the key is stored, read (if symmetric) or replace with your own (if public/private type). Done.
uh-huh. Now how the hell did you post a link without the [domain.com] thing afterwards?
Because file transfer (or rather, the initiation thereof) is just information exchange between clients, and AFAIK the server does not need to know about that. I'd bet you can get the whiteboard or chess plugins in tkabber to work too.
That is kinda pointless. The point of RAID5 is to survive a crash of ONE store. You've got two on a single drive. Drive crashes, lose all your data. To use RAID5 properly you need 3 physical drives.
Except 3.579545Mhz, the color subcarrier.
The color subcarrier has to be between about 2Mhz and 4Mhz to prevent interference with the luminance signal and the audio carrier (4.5Mhz above video). 3.58 is chosen as a good number giving little interference and a very small response in old B/W sets, while still having enough bandwidth for color and not getting into the audio subcarrier space.
The old line frequency is 15,750Hz. Then you get the audio carrier, 4.5Mhz, and divide by 286 which gives 15,734.2657Hz. 286 is the even number that results in the closest approxmation to the old frequency. The frame and field numbers are calculated as you said. Then, the color frequency is the 455th harmonic of half the line frequency, or 15,734.2657/2*455 = 3579545Hz. 455 is the odd number that gets it closest to 3.58Mhz. All this helps minimize interference.
Some of my numbers are slightly different from the parent's, but those are just rounding errors.
Actually it seems to me like the engineers figured out "aha, we'll just remove the key" and not realize that (due to the way the keyboard is wired up and the way the software scans it) it is possible to make it think you pressed other keys. I figure they wanted to save themselves the hassle of changing the controller chip design, or they were just lazy or too stupid.
1 2
| |
A-B-3
| |
C-D-4
| |
Take a keyscanning algorith that works scanning left-to-right columns and up-to-down rows, that decodes the first key detected as pressed and ignores the rest. Take a keyboard matrix as shown above, with no isolation diodes. Press keys B,C,D. Watch how the connections 3-2,2-4,4-1 also create a 3-1 connection. Now the calculator just thought you pressed A. Depending on the details of it, similar stuff can happen. For example, if the thing worked by switching inputs and outputs e.g. sending current to all columns and watching for the active row, then sending current to all rows and watching to the active column, two keys (B and C) would be enough to activate all the rows and columns in the previous matrix. The calculator checks the first it finds and voila, it happily performs the funcion assigned to "A".
that is basically what I do. I also printed out hexdumps of the encrypted file and keys, just in case my drive is wiped for whatever reason. I figure I can remember the passphrase.
Actually due to a flaw in the scripts I use to handle it, I lost the passwords once. Took me 2 seconds to grep the whole HDD for one of them I remembered, and found a stray copy of the plaintext file somewhere (close to the beginning even. Probably stored as inline data in reiserfs nodes, instead of bulk secors) from some previous decoding. I guess I'll have to zero the file out next time. Now my scripts always make a backup of the encrypted file that never gets erased (only overwritten), just in case.
Using your bank account password in WP is like giving it to them (in fact, it is. Don't count on the fact that the passwords are hashed, as this proves).
Do you trust Wikipedia to keep your bank account secure? I do not. Same way as I do not trust any website any of my important passwords. That means all of them except one or two which I sometimes repeat only for online forums/websites/etc which wouldn't be a fatal loss if they were revealed.
So you went wikipedia the password you use on X. Tough luck. You should have directly *assumed* that they were going to steal you account everywhere else and see if it pays to use a different password. THAT is the only way of keeping things secure. And if you used a different password and they revealed it? So what? It's their site, they can do whatever they want with it.
I am NOT saying that wikipedia revealed anyone's password directly, I am just presenting worst-case scenarios. Not that revealing passwords is ethical or even legal (I am not sure about the laws governing passwords stored on website databases), but it is always a possibility.