Mikero, I think you kind of missed the point.

In ADDITION to the meaning you quoted, 'to decimate' can also mean 'to cause great destruction' (with 'great' potentially much larger than 10%). Check a dictionary: http://www.answers.com/decimation

I find it surprising that they push their horrible memory stick standard even onto their professional gear. I wish it would just die.

@Joshua Topolsky

No underscores in new usernames? Did I miss something? jgpuff just looks stupid now.

Full credit to Ross for being reasonable about it though. There's basically no-one else on the internet decent enough to admit to their mistakes.

I pity whoever is paying you $50k pa. to inaccurately proofread articles on someone else's gadget blog

You wouldn't be slow if they had a turbo button which could boost them to 66MHz

Sorry, but I predict failure.

I had various build-your-own type construction kits as a child (including a camera, albeit a film one) and thorougly enjoyed them. I think they're a great way to get kids into engineering which, naturally, is the best discipline in the world.

The thing that bothered me about some of the kits of my childhood (and also this kit) is that you often didn't really know what was going on: all the clever stuff happens in a black plastic square and you never get to build that bit yourself. It sometimes feels a bit false to just screw a PCB into a plastic shell.

It's the processing gubbins that are more likely to be upgradable anyway. Any major changes to the sensor (like the ability to shoot very high resolution high-rate video) is likely to require a processing upgrade too. Why not just replace the whole body?

Well spotted, Fro.

I suppose there are two ways around this:

1. Don't decrypt the volume on the drive: copy it to a computer first. Of course, when you copy the re-encrypted volume back, it may not necessarily overwrite the original volume, leaving an attacker with two versions of the same volume. The may weaken the encryption, but only very slightly (as far as my understanding goes).

2. Make the volume the same size as the USB drive, then the re-encryption process will definitely overwrite any decrypted information that was previously on the drive, wear levelling or not.

I really know nothing about this, but as far as I know the wear levelling mechanisms are often built into the flash chips themselves...if this is true, it would take a pretty dedicated and well-funded attacker to bypass the mechanisms and access the memory cells directly.


Please, correct me if any of the above is wrong.