Re: [jose] Signature algorithm "none"

You didn't answer my question: When I put a JWS with "alg":"none" into bool
JOSE::verify(), what do I get?

The consistency you assert is illusory.

On Wed, Jul 31, 2013 at 5:24 PM, John Bradley <ve7jtb@ve7jtb.com> wrote:

> Applications need to define what signature algorithms they accept.   In
> some cases over some communication channels the signature may not be
> required.
>
> Applications processing JWT like Connect want to process tokens
> consistently.  Receiving a JWT with a alg of none is fine under some
> circumstances.
> In general you would restrict the library from accepting it.
>
> John B.
>
> On 2013-07-31, at 3:44 PM, Richard Barnes <rlb@ipv.sx> wrote:
>
> Ok. That seems like a bug in OpenID Connect.  They should be switching the
> content type (JWS vs. bare request) or using detached signatures.
>
> What's the result of JWS verification when "alg" == "none"?  It seems like
> it has to be either "True" or "False".  If you pick "true", there's an easy
> attack where you just change the algorithm to "none" and delete the
> signature.  If you pick "false"... well it seems silly to have a signature
> algorithm that never verifies.
>
>
>
>
>
> On Wed, Jul 31, 2013 at 2:48 PM, Mike Jones <Michael.Jones@microsoft.com>wrote:
>
>>  It’s optional to sign lots of content.  For instance, OpenID Connect
>> requests can be signed or unsigned, depending upon the security properties
>> desired.  “alg”:”none” is used for such unsigned requests.****
>>
>> ** **
>>
>>                                                             -- Mike****
>>
>> ** **
>>
>> *From:* jose-bounces@ietf.org [mailto:jose-bounces@ietf.org] *On Behalf
>> Of *Richard Barnes
>> *Sent:* Wednesday, July 31, 2013 5:46 AM
>> *To:* jose@ietf.org
>> *Subject:* [jose] Signature algorithm "none"****
>>
>> ** **
>>
>> What's the use case for this?  Can we delete it?****
>>
>
> _______________________________________________
> jose mailing list
> jose@ietf.org
> https://www.ietf.org/mailman/listinfo/jose
>
>
>