Re: [rtcweb] Which hashes are valid for the fingerprint attribute?

The recommended algorithms to support probably needs to appear in the
security document; JSEP should also mention or point to the recommended
algorithms to use when generating the digest.

On Fri, Jul 11, 2014 at 8:05 AM, Alexandre GOUAILLARD <agouaillard@gmail.com
> wrote:

> Inaki,
>
> I checked a little bit, following our previous exchange.
>
> There was a thread back in november last year in discuss-webrtc (how to
> specify desired hash function(s) (sha-1, sha-256, sha-512) for DTLS-SRTP?)
>
> Justin said "Chrome supports various hash functions in the remote
> description (SHA-1, 256, 512). Local description currently only supports
> SHA-256, no plans to change that (although we will probably support
> longer hashes in the future for hash agility). Asymmetric hash functions
> should not be an issue. If multiple fingerprints are specified in the
> remote description, I think Chrome will only use the first one."
>
> then in ORTC there was this thread in april:
> "Issue 64: Section 2.5.1 Fingerprint attribute"
> in which bernard proposed:
> "
> dictionary RTCDtlsParameters {
>     RTCDtlsRole                  role = "auto";
>     sequence<RTCDtlsFingerprint> fingerprint;
> };
> dictionary RTCDtlsFingerprint {
>     RTCDtlsCertificateHashAlgorithm algorithm;
>     ArrayBuffer                     value;
> };
> enum RTCDtlsCertificateHashAlgorithm {
>     "sha-1",
>     "sha-224",
>     "sha-256",
>     "sha-384",
>     "sha-512"
> };
> "
>
>
>
>
>
>
>
> On Fri, Jul 11, 2014 at 8:26 PM, Iñaki Baz Castillo <ibc@aliax.net> wrote:
>
>> Hi,
>>
>> Hi, RFC 5763 and RFC 5764 (DTLS-SRTP) do not mandate a specific hash
>> for the fingerprint attribute in the SDP. RFC 5763 refers to RFC 4572
>> "Connection-Oriented Media Transport over TLS in SDP”. Its section 5
>> clearly opens the door to multiple hash functions:
>>
>>    hash-func    =  "sha-1" / "sha-224" / "sha-256" /
>>                          "sha-384" / "sha-512" /
>>                          "md5" / "md2" / token
>>                          ; Additional hash functions can only come
>>                          ; from updates to RFC 3279
>>
>>
>> I'm pretty sure that WebRTC implementations are not ready for all
>> those hash functions. Is there any WebRTC related draft constraining
>> the hash functions that can be used?
>>
>> Thanks a lot.
>>
>>
>> PS: Not sure if this question should be placed here or in public-webrtc
>> ML.
>>
>> --
>> Iñaki Baz Castillo
>> <ibc@aliax.net>
>>
>> _______________________________________________
>> rtcweb mailing list
>> rtcweb@ietf.org
>> https://www.ietf.org/mailman/listinfo/rtcweb
>>
>
>
>
> --
> Alex. Gouaillard, PhD, PhD, MBA
>
> ------------------------------------------------------------------------------------
>  CTO - Temasys Communications, S'pore / Mountain View
> President - CoSMo Software, Cambridge, MA
>
> ------------------------------------------------------------------------------------
> sg.linkedin.com/agouaillard
>
>    -
>
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>
>