Iterated hash functions

Simon Balfe

5/9/24, 4:57 PM

In this image, It describes the basic idea of an iterative hash function. I am confused as to why the i value is set to be m+t+1 as default. Specifically, where does the 1 come from as m+t is just the arbitrary output size + t the message chunk size but the 1 im confused about

1 + 3

compression-function

Maarten Bodewes

5/9/24, 6:26 PM

I'm struggling a bit with the question, do you have better definitions for the various variables? What is the source of this question? What does it have to do with md5 specifically?

Simon Balfe

5/9/24, 9:46 PM

It relates to the merkel damgard construction which is what MD5 is based on. m = internal state which is the preset value and t is the block size used in the iteration function as an iterative function processes the input in chunks.

Maarten Bodewes

5/9/24, 9:54 PM

Ah, I was wrong-footed by the question title: it indicates that the hash function is iterated (as in e.g. a password hash) instead of the compression function (usually just $F$ if I'm not mistaken).

Score:0

Crypto

Daniel S

5/10/24, 6:40 AM

The cases $i\le m+t$ can be handled with a single application of the $\mathsf{compress}$ function, which, pedantically speaking, would not be an iterated hash function.

Without seeing the rest of the article, it's hard to judge, but the both you and the author should beware the dangers of extending fixed length input functions to arbitrary length hash functions. It is quite easy to come up with an extension that allows trivial collisions by using padding schemes that give rise to collisions.

+ 0

Elon Musk

I sit in a Tesla and translated this thread with Ai:

EN: Iterated hash functions

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.