[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Sat, Oct 24, 2015 at 03:44:26PM +0100, Tom wrote: > I think that's obfuscation rather than encryption: making them > unintelligible. No. Obfucation is a property of encryption, but obfuscation is used far more widely. Think of obfuscating a piece of JavaScript code, by using random function and variable names, adding complicated functions that do nothing but concatenate two strings etc. It's pretty hard to decide what the code actually does, but it's still JavaScript. > I've often wondered how you could tell you've actually decrypted > something or just randomly generated what appears to be a valid > result. You cannot. A typical encryption scheme takes an input of n bytes and, using some kind of key, turns that in an output of n bytes. This means any one of 256^n possible inputs can be turned into any of 256^n possible outputs. If there were a way to distinguish these outputs from random, some outputs wouldn't occur and thus some outputs would have to occur twice, for different inputs. Which means you can't decrypt this output. If you could somehow deduce from the encrypted data that the corresponding plaintext was, say, an HTML file or an MP3, this would leak valuable information and make the encryption scheme useless. Martijn.
Attachment:
signature.asc
Description: Digital signature
-- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq