Doesn't work with BitLocker and a TPM chip. The key is kept in protected memory on the chip and only authenticated code can use it.

I don't think that's true. The passphrase (perhaps hashed?) pay only be in the TPM chip, but the actual cryto key used to decrypt disk sectors is in main memory, because the main CPU is used to do the decryption. There's nowhere near enough bandwidth to and from the TPM chip to let it do the actual disk encryption/decryption. There's not even enough bandwidth to ask the TPM for the key each time you want to do a disk transfer, and erase it from memory after the disk transfer is completed.

This means that software that extracts the encryption key from memory probably can't turn it back into the passphrase that the user enters, but if you have a copy of the disk and the key, you don't actually need that passphrase.

The TPM is not a high-performance device and doesn't do anything but give out the keys on (authenticated) request. What the software does with those keys is up to the software. If someone has privileged or physical access to the machine while the keys are in use, all bets are off.

Note that the 558 is not retriggerable. This led to bugs in joystick (or game paddle) reading in many Apple II programs. If you trigger the 558 to read input 0, then want to read input 1, do NOT just trigger it again. Make sure that input 1 has timed out first, before triggering it to be read. Otherwise you'll read an incorrect value. Or trigger once, then read all inputs you're interested in at the same time.

I, for one, welcome our old protozoan overlords.

In the 1970s, magnetic bubble memory was expected to be the next big thing in nonvolatile data storage, and there were commercial products from Hitachi, Intel, Rockwell, and TI. Commercial bubble memory devices were fabricated using garnet films, though there was research into the use of other materials.

Due to high cost, bubble memory was successful only in limited niches, so by the mid-1980s it was discontinued. Intel stopped development at the 4 Mbit level; I don't think the other vendors even pushed it that far. Late 1980s research results suggested the possibility of 64 Mbit devices. I suspect that the technology probably wouldn't have scaled much further anyhow.

More recently, IBM has been working on "racetrack memory", which works similarly to magnetic bubble memory.

Is ther e a way to send a PM on Slashdot? I've got the skills you need, and am available. You can email me at

The LEM wasn't capable of doing this in earth gravity. It was *extremely* specialized for use on the moon.

Why should only the criminal side of the malware equation get the benefits of open-source?

Some of my cells were travelling self-propelled at 5 km/h earlier today. In fact, all of the were.

Which is "better than an hour", 59 minutes or 61 minutes, and why is it better?

This is why we can't have nice things!

You have to do more than "make it available". Since it is being commercially distributed, and isn't accompanied with the complete corresponding machine-readable source code, they need to satisfy section 3b of the GPLv2:

Can anyone who has a Xoom confirm whether it came with such a written offer?

As others have pointed out, this only applies to any GPL'd components of the software, which includes the Linux kernel but little else.

It's now obvious that it's really Goldman Sachs that is the Bank of Evil.

That would be way beyond "epic". I'm not sure that a suitable adjective for that level of fail has even been invented.

I've got 800bip nine-track tapes from a PDP-10 system from 1969. They're still readable on an HP 88780A SCSI tape drive. There are text files from 1968. The text is all in ASCII, though it's the 1963 standard, which has up-arrow and left-arrow in place of caret and underscore of the 1967 and later standards.

The Story of Sputnik