Re: PROPOSAL: Close ISSUE-26 - Should key generation be allowed to specify multi-origin shared access

CORS are for resource requests, not for scriptable objects. So no,
it's entirely inappropriate for CORS.

Just reading the CORS spec should highlight this - the requirement to
pre-flight resource requests shows that there's absolutely no way to
express a request for a scriptable object via a Request.

On Mon, Feb 25, 2013 at 4:55 PM, Mountie Lee <mountie.lee@mw2.or.kr> wrote:
> Hi.
> if Key A generated by the script loaded from origin-A,
> will be the Key-A accessable from origin-B with CORS headers?
>
> if CORS controls are deployed to generated Key resource,
> we found improvement.
>
>
> On Thu, Feb 7, 2013 at 11:25 PM, GALINDO Virginie
> <Virginie.GALINDO@gemalto.com> wrote:
>>
>> Hi Mountie, Ryand, and all,
>>
>>
>>
>> About the certificate management.
>>
>> This is a discussion I wanted to have during our last call : how to
>> address the secondary features. It seems to me that if this topic if of
>> importance for you, you should start considering creating a draft
>> specification to address it.
>>
>> As I have mentioned several time, the WG welcomes contributions, even on
>> secondary features. If there is a contribution, you will get some time to
>> present it, and discuss it, with a low priority in the agenda compared to
>> primary features, but at least, it will allow the group to understand why it
>> is important and how it is feasible to manage certificates.
>>
>>
>>
>> About the multiple-origin.
>>
>> I would suggest that if there are some interested parties, they come
>> together with a proposal to the editors, as I recommend for the ISSUE-19.
>>
>>
>>
>> Regards,
>>
>> Virginie
>>
>>
>>
>>
>>
>>
>>
>> From: mountie@paygate.net [mailto:mountie@paygate.net] On Behalf Of
>> Mountie Lee
>> Sent: samedi 2 février 2013 05:34
>> To: Ryan Sleevi
>> Cc: public-webcrypto@w3.org
>> Subject: Re: PROPOSAL: Close ISSUE-26 - Should key generation be allowed
>> to specify multi-origin shared access
>>
>>
>>
>> there are many discuss threads for multi-origin issues.
>>
>> and I feel the solution was narrowed to certificate association.
>>
>>
>>
>> but the certificate related issues were postponed because it was not in
>> primary features.
>>
>>
>>
>> my main concern is that
>>
>> closing this issue will cause final decision of dropping multi-origin
>> allowness.
>>
>>
>>
>> I will still research and review the alternatives like CORS or script-src.
>>
>> but I feel still we need to handle multi-origin and certificate both.
>>
>>
>>
>> regards
>>
>> mountie.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On Fri, Feb 1, 2013 at 10:24 AM, Ryan Sleevi <sleevi@google.com> wrote:
>>
>> On Thu, Jan 31, 2013 at 5:19 PM, Mountie Lee <mountie@paygate.net> wrote:
>> > from the previous discussions
>> >
>> > I remember we have two proposals for this issue.
>>
>> There were not proposals. There were use cases, but no proposals on
>> how to satisfy those use cases.
>>
>>
>> >
>> > one is allowing multi-origin shared acces for certificate associated
>> > case.
>> > second is allowing multi-origin shared access by user consent
>> >
>> > the reason why this issue is important is
>> >
>> > in the online banking usecases.
>> > users generate keypair at CA website and get certificate.
>> > and the certificate-private key pair should be used at other bank sites
>> > for
>> > signing document or verifying signature.
>> >
>> > as compared to TLS certificate usecases,
>> > it is also common sense.
>> > generating and getting certificate from CA site
>> > and using it at different site
>>
>> And in that thread, solutions were explored on how that use case can be
>> met.
>>
>> That said, we're talking more generally about credential enrollment,
>> as you just described, which is clearly placed in our secondary use
>> cases.
>>
>> I think that, absent any concrete proposals, and based on the
>> timelines set out, this should not be seen as an ISSUE for the current
>> spec. The WG has (by charter) agreed to look at this, but at a later
>> point (the non-normative "roadmap" document).
>>
>> I'm proposing that, based on the lack of concrete proposals, and based
>> on the timeline set forward, that we should consider this "not
>> implemented due to time constraints".
>>
>>
>> >
>> >
>> > On Fri, Feb 1, 2013 at 4:18 AM, Ryan Sleevi <sleevi@google.com> wrote:
>> >>
>> >> http://www.w3.org/2012/webcrypto/track/issues/26
>> >>
>> >> I would like to propose that we CLOSE Issue-26.
>> >>
>> >> There have been no proposals put forward on how to securely address
>> >> multi-origin shared access. Further, such provisioning opens up a host
>> >> of security concerns that the use cases used to justify such access
>> >> are not compatible with.
>> >>
>> >> In the current specification, multi-origin applications may make use
>> >> of secure messaging exchanges, such as postMessage, to transition
>> >> across security domains, without requiring the granting of a single
>> >> origin full access to either plaintext or to keying material.
>> >>
>> >> As such, absent both concrete use cases and proposals, I propose that
>> >> this issue be closed.
>> >>
>> >
>> >
>> >
>> > --
>> > Mountie Lee
>> >
>> > PayGate
>> > CTO, CISSP
>> > Tel : +82 2 2140 2700
>> > E-Mail : mountie@paygate.net
>> >
>>
>> > =======================================
>> > PayGate Inc.
>> > THE STANDARD FOR ONLINE PAYMENT
>> > for Korea, Japan, China, and the World
>> >
>>
>>
>>
>>
>>
>> --
>> Mountie Lee
>>
>> PayGate
>>
>> CTO, CISSP
>> Tel : +82 2 2140 2700
>> E-Mail : mountie@paygate.net
>>
>> =======================================
>>
>> PayGate Inc.
>>
>> THE STANDARD FOR ONLINE PAYMENT
>>
>> for Korea, Japan, China, and the World
>>
>>
>
>
>
>
> --
> Mountie Lee
>
> PayGate
> CTO, CISSP
> Tel : +82 2 2140 2700
> E-Mail : mountie@paygate.net
>
> =======================================
> PayGate Inc.
> THE STANDARD FOR ONLINE PAYMENT
> for Korea, Japan, China, and the World
>

Received on Tuesday, 26 February 2013 01:00:38 UTC