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

Re: [xmpp] SCRAM as MTI?



Fabio Forno <fabio.forno at gmail.com> writes:

> 2009/9/14 Peter Saint-Andre <stpeter at stpeter.im>:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> The SASL SCRAM mechanism is currently in IETF Last Call:
>>
>> http://www.ietf.org/internet-drafts/draft-ietf-sasl-scram-07.txt
>>
>> This mechanism is intended to replace DIGEST-MD5. Does it make sense for
>> the XMPP WG to specify SCRAM as mandatory-to-implement?
>
> Does SCRAM support a digest based auth like DIGEST-MD5 or, better,
> does it allow secure authentication over an unsecure channel?

Yes.  (Of course it depend on your definition of "secure"..)

> In that case definitively yes, since there many case in which I don't
> need TLS, but at least i want to protect my credentials.

You are still vulnerable to active attackers that hijack the connection
after you have authenticated yourself.  But I agree in some situations
that is an acceptable risk.

To use SCRAM (or CRAM-MD5 or DIGEST-MD5) you need a secure channel.  If
you used DIGEST-MD5 with security layers you will need SCRAM+TLS
instead.  SCRAM does not provide any SASL security layers.  I hope you
weren't using DIGEST-MD5 security layers, because they weren't that
good...

/Simon

Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.