Re: Consider whether the others Public Value input to (EC)DH deriveKey should be a Key object

Late +1


On Fri, Feb 28, 2014 at 4:10 PM, Ryan Sleevi <sleevi@google.com> wrote:

> In case it wasn't clear, +1 from me if we can make it work.
> On Feb 28, 2014 8:03 AM, "Mark Watson" <watsonm@netflix.com> wrote:
>
>> Two plus votes - any other opinions ?
>>
>> ...Mark
>>
>>
>> On Thu, Feb 27, 2014 at 12:59 PM, Vijay Bharadwaj <
>> Vijay.Bharadwaj@microsoft.com> wrote:
>>
>>> Agreed. I too think it's cleaner to always use Key objects.
>>>
>>> -----Original Message-----
>>> From: Jim Schaad [mailto:ietf@augustcellars.com]
>>> Sent: Thursday, February 27, 2014 12:54 PM
>>> To: 'Mark Watson'; public-webcrypto@w3.org
>>> Subject: RE: Consider whether the others Public Value input to (EC)DH
>>> deriveKey should be a Key object
>>>
>>>
>>>
>>> From: Mark Watson [mailto:watsonm@netflix.com]
>>> Sent: Wednesday, February 26, 2014 3:01 PM
>>> To: public-webcrypto@w3.org
>>> Subject: Consider whether the others Public Value input to (EC)DH
>>> deriveKey should be a Key object
>>>
>>> https://www.w3.org/Bugs/Public/show_bug.cgi?id=24830
>>>
>>> Presently, if the other's public value for the two DH algorithms is
>>> received in spki format, then it is necessary first to import this
>>> structure to obtain a Key object and then to export that in raw format to
>>> obtain an ArrayBuffer containing the Public Value. This may then be used
>>> with deriveKey.
>>>
>>> If this is the more common use-case it would make sense to change the
>>> type of the public property of (Ec)DhDeriveParams to have type Key.
>>>
>>> If we make that change, then the other use-case where the Public Value
>>> is received in some other form and extracted to an ArrayBuffer by the
>>> application would require that ArrayBuffer to be imported to obtain a Key.
>>> Thus the steps would be the same in both cases: import the received
>>> public value, provide this Key object to deriveKey.
>>>
>>> ...Mark
>>>
>>> I would feel much better if this was always using key objects rather
>>> than using an array of bytes as the second public key.  This would be true
>>> regardless of the number of keys used in the process (i.e. for 4 key DH key
>>> agreement methods).
>>>
>>> Jim
>>>
>>>
>>>
>>>
>>

Received on Monday, 3 March 2014 13:41:18 UTC