| 1 |
On Thursday 21 September 2006 08:54, Hanno Böck wrote: |
| 2 |
> I think sha256/512 is the only thing that makes sense at the moment, as it |
| 3 |
> most probably will stay secure for quite a while and we don't have real |
| 4 |
> alternatives. So imho use sha256, get rid of everything else, because that |
| 5 |
> rarely improves security, and wait for the nist to define something new |
| 6 |
> (which will happen, but probably take some years from now). |
| 7 |
|
| 8 |
Well, the problem that occurs here is the verification process. With MD5, you |
| 9 |
can hit most upstream sites, and they'll have an MD5SUM avaliable that you |
| 10 |
can authenticate against. With SHA256, you would need an upstream that |
| 11 |
actually implements them as hashes for release notifications. Without this |
| 12 |
sort of verification, there's a better chance of someone putting out some |
| 13 |
kind of exploit tarball, us hashing it as per the usual, and the whole |
| 14 |
purpose gets defeated. Yes, you can consider that developers should be going |
| 15 |
in and checking the changes, etc., but the problem it's something a lot of |
| 16 |
devs would be less likely to do versus an easy md5sum lookup. |
| 17 |
|
| 18 |
-- |
| 19 |
Chris White |
| 20 |
Gentoo Developer aka: |
| 21 |
xxxxxx (Scissors Were Here) xxxxxx |
Archives