>>122
The difference is that you can convince a jury that apparently random data is very likely encrypted, but the argument that ``somewhere in the computer there might be some encrypted data but we haven't found any traces of it yet'' doesn't really work. Unless you're in a witch trial or something.
Not only is that obvious bullshit (Why would you have the software if you weren't trying to hide anything? Even if you haven't actually hidden anything, no jury is going to buy that), steganography is the definition of security through obscurity. If they have your algorithm (in the form of the software you used), they have the data you tried to hide.
I'm just saying that you can get in deep shit just by writing free software.
Which is an entirely irrelevant point.