So, you want to crypto • Hacker News Board

Re: So, you want to crypto

#71

Post by eliteraspberrie » Wed, Dec 25, 2013, 6:46 PM UTC

haberman wrote: post #40

I'm curious to hear people's thoughts about git. Git is "crypto" to some extent, Linus does not appear to have tons of crypto expertise, and it uses SHA1 as a MAC AFAICT (which according to tptacek's earlier comment is invalid). And yet I've never heard about attacks on its crypto. This was interesting for me to think about because it seems like a counterpoint to the article, in that it is a very successful project t…

A Git commit history is a Merkle tree. The hash is not used for authentication.

Re: So, you want to crypto

#72

Post by dobbsbob » Wed, Dec 25, 2013, 7:03 PM UTC

milhous wrote: post #26

I'm taking an Intro to Crypto course this spring. What's interesting is that it's offered through the Math department, and assumed it was a CS class. We'll be using this text: http://www.amazon.com/Introduction-Cryptography-Coding-Theor... Is this any good? Apparently a best seller in the "Software Coding Theory" category on Amazon.

Most of Matthew Green's course is avail online http://spar.isi.jhu.edu/~mgreen/650.445/650.445__Main.html

Re: So, you want to crypto

#73

Post by tptacek » Fri, Dec 27, 2013, 12:45 AM UTC

andrewcooke wrote: post #68

tptacek wrote: post #23

Earlier quoted context omitted.

You're right. I was snippy, and in a particularly dumb way. Sorry about that.

eh, no need to apologise. i think we're used to you here and there's a level of tolerance given the contributions you make. also, did you see http://www.jakoblell.com/blog/2013/12/22/practical-malleabil... which i think could be added to an early crypto challenge? (but maybe you already cover the idea).

CBC bitflipping is in set #2. Incidentally, somewhere in the mists of time is a very weird blog post ("If You're Typing The Letters A-E-S Into Your Code, You're Doing It Wrong") I wrote describing how to implement this attack --- though not against LUKS.

Re: So, you want to crypto

#74

Post by haberman » Fri, Jan 03, 2014, 3:01 AM UTC

alinajaf wrote: post #52

haberman wrote: post #50

Earlier quoted context omitted.

Thanks for this, you answered my questions thoroughly. I'm not entirely convinced by this though: > [A signed commit] doesn't guarantee anything about the contents of that commit if the repository has been tampered with. I think most people would intuitively expect the signed commit to guarantee the contents of the tree being signed. The idea that you could "git pull" a repo from a compromised machine, verify the sig…

I did a bit of quick reading on this and at first glance my description of how git tagging works appears to be on point, i.e. all it guarantees is that a particular user asserts that tag X points to commit with SHA1 Y. I'm not sure that it says anywhere in the documentation that it guarantees anything more than that, but I agree that a significant proportion of developers would intuitively expect that the entire cont…

> I'm not sure that it says anywhere in the documentation that it guarantees anything more than that, but I agree that a significant proportion of developers would intuitively expect that the entire content of the tree to be signed rather than just the SHA1.

Further evidence that they do assume that: https://news.ycombinator.com/item?id=7003900

Re: So, you want to crypto

#75

Post by alinajaf » Tue, Jan 07, 2014, 10:38 AM UTC

haberman wrote: post #74

alinajaf wrote: post #52

Earlier quoted context omitted.

I did a bit of quick reading on this and at first glance my description of how git tagging works appears to be on point, i.e. all it guarantees is that a particular user asserts that tag X points to commit with SHA1 Y. I'm not sure that it says anywhere in the documentation that it guarantees anything more than that, but I agree that a significant proportion of developers would intuitively expect that the entire cont…

> I'm not sure that it says anywhere in the documentation that it guarantees anything more than that, but I agree that a significant proportion of developers would intuitively expect that the entire content of the tree to be signed rather than just the SHA1. Further evidence that they do assume that: https://news.ycombinator.com/item?id=7003900

I agree that they do assume that, but fail to see what connection that has to the actual workings of git. AFAIK the behaviour of software doesn't change in accordance with how developers think it works.