[arch-dev-public] Can we trust our mirrors?

[Thomas Bächler]

Pierre Schmitz schrieb:
> On Sat, 29 Nov 2008 15:00:20 +0100, Thomas Bächler <thomas at archlinux.org>
> wrote:
>> If this is to provide any security, we need to stop using md5! md5 is 
>> okay when trying to detect corrupted downloads, however it is possible 
>> to find collisions and thus build a "bad" package that has the same md5 
>> as the good package.
> 
> Well, it should be quite easy to use sha instead. I am not an expert but
> how easy is it to produce a valid package with the same md5sum? I know that
> creating "some" file is not hard.

You can append arbitrary data at the end of a gzip file (this was used 
in the construction of the "forbidden prime number"). Thus you could 
create a relatively small malicious package and then try to construct a 
collision by appending data to the file. It should not be much harder 
than finding any collision, but I am no expert either.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
URL: <http://archlinux.org/pipermail/arch-dev-public/attachments/20081129/15817220/attachment.pgp>