Git 3.0 will make SHA-256 the new default content hashing algorithm and it will be an incomprehensibly expensive and ultimately valueless and avoidable global nightmare.
I meant that NIST makes a pretty strong move with the disappearance of SHA-1 in the future FIPS 180-5 (in constrast to 180-4). Of course, it never specified MD5 but the fact is that SHA-1 will not be deprecated: it will be un-standardized and that’s pretty uncommon.
It lists the transition and accepted usecases more clearly than the links in my previous comment.
As far as I can tell, SHA-1 was treated much more leniently than MD5: its deprecation process was far slower and gentler. That’s partly because MD5 is so broken that attacks became cheap and practical in 2005 so there was an undeniable urgency to doing the re-engineering work that was necessary to get rid of it; SHA-1 is not yet cheap to attack.
SHA-1 started the deprecation process 20 years ago. In 2011, SHA-1 was deprecated for digital signatures and was supposed to be unused for them after 2013, but it was still acceptable in HMAC https://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-131a.pdf
score 1 · canonical
I meant that NIST makes a pretty strong move with the disappearance of SHA-1 in the future FIPS 180-5 (in constrast to 180-4). Of course, it never specified MD5 but the fact is that SHA-1 will not be deprecated: it will be un-standardized and that’s pretty uncommon.
score 1 · canonical
Here’s a better summary of NIST’s SHA-1 deprecation timeline: https://csrc.nist.gov/Projects/Hash-Functions/NIST-Policy-on-Hash-Functions
It lists the transition and accepted usecases more clearly than the links in my previous comment.
As far as I can tell, SHA-1 was treated much more leniently than MD5: its deprecation process was far slower and gentler. That’s partly because MD5 is so broken that attacks became cheap and practical in 2005 so there was an undeniable urgency to doing the re-engineering work that was necessary to get rid of it; SHA-1 is not yet cheap to attack.
score 1 · canonical