md5 encryption

[Robert Moskowitz]

Note my day job IS data communications and security.  I am NOT a 
cryptographer.  I am in a different select group that often refer to 
ourselves as the crypto-plumbers.  We know how to carefully use the 
crypto blocks to build whole systems.

On 08/28/2014 11:34 AM, dustin kempter wrote:
> hi all, I just had a question. so I have been hearing that md5 has 
> been compromised, how much of a security threat does this impose? if 
> it is something that poses a large concern is it possible to change 
> the encryption method from md5 to crypt? or another method that is not 
> compromised?

md5 is a message digest / hash.  It is an earlier effort by Ron Rivest.  
It was paired with Ron's rc5 streaming encryption cipher, but lived by 
itself as one of the major hash functions in 90s (md5, md160 (do I have 
that right :) ), and sha1).  All have been weakened to different 
levels.  md5 has real attacks where with crafted change to a message you 
can still get the same hash value.

Today you might STILL use sha1; it has had tremendous resiliency. NIST 
was expecting it to fall as badly as md5 by this point.   Most use at 
least sha256, and sha3 is now out there.  Choose your poison.

data confidentiality is another subject.

Hope this helps.