Jump to content

Talk:Xor–encrypt–xor

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Picture/Text discrepancy

[edit]

The explanation here and on the Disk encryption theory page both show a picture with another keyed block cipher sandwiched between the XORs, but the algorithm description and, in fact, it's definition, do not have this middle step, but instead only rely on a (random) permutation. A keyed block cipher does qualify as such permutation, but is not required (for example, AES with a fixed key should do). Or in other words, they "Key" is not required and does not exist in the algorithm, only Key1 and Key2 are. Or am I totally cofnused? 194.126.175.154 (talk) 06:30, 21 December 2018 (UTC)[reply]

Ok, I understand now , I am confused by the History section which at length describes a different block cipher (Even-Mansour) and then switches to XEX without any explanation of any differences, or why Even-Mansour is even mentioned. I think this could be more clearer. 194.126.175.154 (talk) 06:34, 21 December 2018 (UTC)[reply]

I agree. I feel that the Even-Mansour scheme is very helpful in understanding block ciphers in general (whether or not it is actually practical), and so should be mentioned *somewhere* on Wikipedia. Perhaps in some other article -- but which one? Or perhaps a clearly separate section of this article, separating "xor; unkeyed permutation; xor" of Even Mansour vs. the "xor; encrypt (with a secret key); xor" of the Rogaway scheme of tweakable block cipher? --DavidCary (talk) 00:01, 27 April 2021 (UTC)[reply]

Where Key1/Key2 comes from?

[edit]

It does not explain where Key1 and Key2 is comes from. (One possibility might be to be derived from a combination of a fixed key and an address, although I don't know if it actually does this or if it does something else.) --Zzo38 (talk) 05:59, 4 September 2021 (UTC)[reply]

The text is confusing, as it tries to simultaneously explain two things that are related, but different: a xor-enc-xor technique used in, say, key whitening and XEX encryption mode. For the second, the Key1 and Key2 are derived based on the "location" (for example, a block address in storage). For the first, Key, Key1, and Key2 are simply part of a (longer) symmetric key. The text should be rewritten (or article split) to reflect this dichotomy. Dimawik (talk) 04:08, 13 April 2023 (UTC)[reply]