Difference between revisions of "Talk:XTEA Strong Encryption Implementation"
m (New section: So does it work?) |
|||
| Line 10: | Line 10: | ||
Also is the XTEA version stronger than the TEA version? | Also is the XTEA version stronger than the TEA version? | ||
I've tested the Optimized version and it failed to start. Couldn't even figure out what's not trigging the listen event. :/ --[[User:Vincent Nacon|Vincent Nacon]] 19:01, 29 April 2008 (PDT) | |||
Revision as of 18:01, 29 April 2008
Discuss amongst yourselves:
From a perspective of security, using the XTEA encryption like this in Electronic CodeBook mode is very insecure. There is no authenticity and patterns could be recognized. Even using RC4 would be more secure. I recommend developing at least one of the operation modes like OFB and using that together with XTEA and using a nonce that is used really once for each pass and never used again. Also I recommend using llMD5String to add a hash, that is encrypted together with the data for authentication. A reference can be found at the wikipedia article about Block_cipher_modes_of_operation. Most of time when you are looking for encrytion, you are really looking for authentification of messages. There llMD5String comes in handy. Thomas Shikami 02:38, 20 October 2007 (PDT)
It should be noted that XTEA's chr & ord functions are wrong. -- Strife Onizuka 23:53, 6 December 2007 (PST)
So does it work?
Does this XTEA encryption work?(I don't want to have to go throught the math and testing myself=P)
Also is the XTEA version stronger than the TEA version?
I've tested the Optimized version and it failed to start. Couldn't even figure out what's not trigging the listen event. :/ --Vincent Nacon 19:01, 29 April 2008 (PDT)