On 30 May 1996 07:56:26 pdt, EALLENSMITH@ocelot.Rutgers.EDU wrote:
From: IN%"mclow@owl.csusm.edu" "Marshall Clow" 29-MAY-1996 19:18:31.60
* Why encrypt before compression? If the encryption is any good, then the encrypted data won't compress much at all. However, compression before encryption has its own problems.
What problems does compression before encryption have? It at least seems to work for PGP.
The most significant would be the presence of a compression header (what has the data telling how to decompress a datastream). If you leave this in (for instance, assuming you were to use PKZIP (STUPID!) you would have PK as the first two bytes EVERY TIME!) you give your codebreaker a HUGE advantage. You can bypass this by removing ANY information which isn't message specific and standardizing on one compression method. Alternately, you could use your session key to encrypt a length word and then pad the start (and end) with a string of random characters. This Message Was Sent With An UNREGISTERED Version Of PMMail. Please Encourage Its Author To Register Their Copy Of PMMail. For More Information About PMMail And SouthSide Software's Other Products, Contact http://www.southsoft.com. /* From Chris Adams <adamsc@io-online.com> on a Warped PC running a proudly unregistered (for now) PMMAIL 1.5! The Enigman Group - We do Web Pages! */ This Message Was Sent With An UNREGISTERED Version Of PMMail. Please Encourage Its Author To Register Their Copy Of PMMail. For More Information About PMMail And SouthSide Software's Other Products, Contact http://www.southsoft.com.