« first day (4768 days earlier)      last day (204 days later) » 

avpaderno
avpaderno
06:17
I saw PHP 5.6 code used from a CMS that takes some data about the database connections used for a site, it creates a string, and obtains its SHA256 hash. (That value is usually considered a salt from the CMS.) When it needs a HMAC, it uses that SHA256 hash concatenated with the inode change time and the inode value (for the file containing the executed code).

I know that an SHA256 hash is not distinguishable from random output, but is there a better way to obtain a HMAC key that is used for short periods (for example, when running tests)?
Crypto HNQ
Crypto HNQ
06:41
3
Q: Efficiently implementing Speck96

eofAre there any efficient implementations of Speck96? The problem is how to efficiently do 48-bit arithmetic on 64-bit words, the rotates in particular. I've been trying to implement it and have written the following in Rust: #[inline(always)] fn rotate_right_48(x: u64, r: usize) -> u64 { ((x >>...

block-cipher implementation speck
 
3 hours later…
Maarten Bodewes
Maarten Bodewes
09:27
@kelalaka I agree that our policy on programming related questions probably requires an update. But I do not think that questions about cryptographic optimizations would be a good fit for SO or codereview as the expertise is simply not there. If you don't mind I'll remove the comments below fgrieu's answer for now.
kelalaka
kelalaka
@MaartenBodewes go on.
Maarten Bodewes
Maarten Bodewes
@kelalaka Thanks. Removed the comments. I took another look at your question on CR and it is similar. But I think that in the end the answers were more Python related than anything else.
I do think we need to allow more programming questions when they are directly related to cryptographic algorithms. I've been planning a meta question about this for years now, but I'm always a bit hesitant because of 1. what to include / exclude and 2. because there has been quite a lot of dislike towards SO and the way that site works.
Inner workings of algorithms would be included though, not just "side channels" which always was a bit of an odd one out.
kelalaka
kelalaka
I did not want to dive into the closed and old questions, however, we need to draw a line for this tag. Otherwise, many will come with a fuzzy border. I still consider this Q is not on topic here. It is the programming aspect of cryptographic algorithms. It doesn't ask how this works, rather asks a basic optimization. So, will we accept questions about code optimization of cryptographic algorithms?
Maarten Bodewes
Maarten Bodewes
@kelalaka Hmm, I agree that the current question is also quite a lot asking about code optimization wrt register use which I think is a programming topic (on second read). I'll discuss with fgrieu to see if he supports migration.
When I read the first part I thought it was more about optimization techniques of the cipher itself. Then again, that's also a grey area (or, as you indicated, a "fuzzy border" :) ).

« first day (4768 days earlier)      last day (204 days later) »