[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Signature strength of .dsc



Hello Dimitri

On Fri, 2023-12-01 at 00:20 +0000, Dimitri John Ledkov wrote:
> This makes me wonder if signatures on uploaded or published .dsc have
> any value at all.

Cryptographically speaking, 160-bit hash algorithms are vulnerable to
collision attacks but not to preimage attacks. Even today, no one can
create a fake package that matches an *existing* hash. However, you can
create *two* new packages that result in the same hash.

I checked a random sample from your list and they all appear to be ten
years or older. As long as we have additional proof (e.g. signatures
from the release team) that these signatures were actually created at
the time of upload (when collision attacks were not yet feasible), they
are still meaningful.

For comparison:
Note that issuing of new x.509 certificates with SHA-1 was deprecated a
while ago, but old CA certificates with SHA-1 remained valid for the
same reason.

Regards
Stephan

Attachment: signature.asc
Description: This is a digitally signed message part


Reply to: