[Openid-specs-heart] Flip the question of “Vanilla" OAuth vs. UMA

Sun Jul 12 21:14:37 UTC 2015

I think we're at a tipping point with respect to cybersecurity. The flood
of million-person breaches suggests to me that our perspective has to
change toward greater transparency to the individual resource owner for
every transaction relevant to them in all cases other than some law
enforcement. The pretense that institutional transactions are secure even
if the individual they pertain to is unaware of them is obviously failing.

We've reached a tipping point where the added friction of having to connect
with Alice's AS is now desirable from a cybersecurity perspective. Computer
and network costs are now trivial compared to the value of our data. Let's
just bite this bullet and focus on profiling and coding UMA.

Adrian

On Sun, Jul 12, 2015 at 4:10 PM, Justin Richer <jricher at mit.edu> wrote:

>  > So ... seems to me that client flows could (would) be designed
> differently anyway  based on whether Bob or Alice is making the request.
>
> Exactly: how is the client supposed to know that it's representing Alice
> or Bob? You'd expect a software client to be able to speak to an API and
> have a fairly clear path on how to fulfill its security obligations.
> Unfortunately, there's no "is_alice" flag in the protocol stack that we can
> count on. :)
>
> An UMA client needs to be able to be pointed to an AS, take in a ticket
> (as a JSON value, regardless of what encoding the API it's speaking to
> uses), talk to the "requesting party" endpoint to trade the ticket for a
> token, and then it needs to be able to gather the claims that the AS gives
> it hints for. As Eve keeps pointing out, those claims could be absolutely
> anything, fulfilled by the client or by someone else or just because it's
> Tuesday, so the client is really going to need to be written to a very
> specific profile of UMA for it to have any chance of doing something
> useful. Then it needs to come back with the ticket, again, and try to get a
> token, potentially repeating the claims gathering cycle if it guessed wrong
> on the last step.
>
> Exactly *none* of these actions are required for an OAuth client. The wire
> protocol is a very, very different thing, and the code paths are almost
> completely separate with OAuth 2.0 and UMA 1.0.
>
> This is to say nothing of the differences in the protected resource, which
> is also pretty substantial. It's very, very difficult to write something to
> be protected by both OAuth and UMA simultaneously without very careful
> delineation of how that's to happen.
>
>  -- Justin
>
>
> On 7/12/2015 11:16 AM, Debbie Bucci wrote:
>
>   To answer the reverse - I'd like to better understand:
>
>  ATT interactions and dependencies  - .   asynchronous authorizations and
> on the wire or calculus that happens  - flow is different get that ...
>  PAT - how that aligns/replaces/supplements the traditional Access
> Control  (PIP, PEP ABAC/RBAC) stuff  most likely will be required -
>  OAuth 2.0 likeness  ... I don't necessarily see where there will be an
> entirely different rewrite.   Example:  - authentications may be assumed
> via a standard OAUTH flow - yet some implementations will choose to add the
> additional id_token (OIDC).    If an RPT token is handled programatically
> as an access token  ...what has to change?   So ... seems to me that client
> flows could (would) be designed differently anyway  based on whether Bob or
> Alice is making the request.
>
>
>
>
>
>
> On Thu, Jul 9, 2015 at 12:55 PM, Aaron Seib <aaron.seib at nate-trust.org>
> wrote:
>
>>  If you get  a life be sure to leave a bread crumb trail so the rest of
>> us can follow you out.
>>
>>
>>
>> Aaron
>>
>>
>>
>> Aaron Seib, CEO
>>
>> @CaptBlueButton
>>
>>  (o) 301-540-2311
>>
>> (m) 301-326-6843
>>
>> <http://nate-trust.org>
>>
>>
>>
>> *From:* Openid-specs-heart [mailto:
>> openid-specs-heart-bounces at lists.openid.net] *On Behalf Of *Eve Maler
>> *Sent:* Thursday, July 09, 2015 11:18 AM
>> *To:* Moehrke, John (GE Healthcare)
>>
>> *Cc:* openid-specs-heart at lists.openid.net
>> *Subject:* Re: [Openid-specs-heart] Flip the question of “Vanilla" OAuth
>> vs. UMA
>>
>>
>>
>> John, no doubt you're 200% overbooked, but if we could attract you to the
>> UMA meetings, you'd probably find our current workstream fascinating. :-)
>> The Kantara Leadership Council members laugh at me for getting excited
>> about the "BLT sandwich" (business-legal-technical) progress updates I give
>> them... We are literally mapping specific bits of UMA flows to the norms of
>> sociotechnical systems being identified in cutting-edge academic research.
>>
>>
>>
>> Then again, maybe I need to get a life.
>>
>>
>>
>> *Eve Maler *ForgeRock Office of the CTO | VP Innovation & Emerging
>> Technology
>> Cell +1 425.345.6756 <%2B1%20425.345.6756> | Skype: xmlgrrl | Twitter:
>> @xmlgrrl
>> Join our ForgeRock.org OpenUMA <http://forgerock.org/openuma/> community!
>>
>>
>>
>> On Thu, Jul 9, 2015 at 8:07 AM, Moehrke, John (GE Healthcare) <
>> <John.Moehrke at med.ge.com>John.Moehrke at med.ge.com> wrote:
>>
>> Eve,
>>
>>
>>
>> You are right on…
>>
>>
>>
>> As one of the co-chairs of the HL7 Security wg, I have struggled with
>> this. However in Healthcare we must keep moving forward. So the work in HL7
>> is to incrementally improve the ability to capture anything we can related
>> to the act of consent, and the meaning of that consent. We recognize these
>> as difficult domains. Historic precedent in healthcare is embarrassing.
>>
>>
>>
>> In HL7 we have two different encodings. One is based on the format of a
>> document used for all other medical-fact documentation (CDA). The format is
>> not important, but it is important to indicate that it is a form that
>> healthcare handles for almost everything related to the patient, so it is
>> easy for them to handle it for this purpose too. It is a form however that
>> is totally foreign to any access control decision system. We thus recognize
>> that the rules might be ‘equivalently’ encoded in some access control
>> decision language (common is XACML today). The equivalency is left to
>> humans to agree… Thus the bridging of the divide between them is not
>> computable, but we are trying.
>>
>>
>>
>> The HL7 FHIR effort is just getting going, and for the most part will
>> start with the same capability, just with different technical encoding.
>> Still not necessarily fixing the big issue… But getting more processable
>> (FHIR is predominantly HTTP/REST and encodes things in more simple XML or
>> JSON; where as CDA is ugly XML).
>>
>>
>>
>> The actual systems-design might have the Access Control rules interjected
>> at the time that the patient gives consent; or might interpret the consent
>> at the time of needing to make a decision. This is a systems-design
>> responsibility that is not the role of HL7. There are good reasons to do
>> either, especially since healthcare is a lifelong effort (so we think in
>> 100+ years).
>>
>>
>>
>> So, look to HL7 to help with capturing the ceremony of Consent. But they
>> definitely need help with the Enforcement, we hope HEART can help with. AND
>> we all need to work on everything in-between.
>>
>>
>>
>> John
>>
>>
>>
>> *From:* Openid-specs-heart [mailto:
>> <openid-specs-heart-bounces at lists.openid.net>
>> openid-specs-heart-bounces at lists.openid.net] *On Behalf Of *Eve Maler
>> *Sent:* Thursday, July 09, 2015 9:47 AM
>> *To:* Kinsley, William
>> *Cc:* openid-specs-heart at lists.openid.net
>>
>>
>> *Subject:* Re: [Openid-specs-heart] Flip the question of “Vanilla" OAuth
>> vs. UMA
>>
>>
>>
>> Hi Bill-- I'm not sure what the HL7 consent/contract standards are about,
>> but based on what I know of "notice and consent" (privacy) theory, and
>> "contract" (legal) theory (IANAL but I have collected lawyers for many
>> years as part of my UMA work), there is not very much in the technical
>> underpinnings of almost any technical or machine-readable consent flow used
>> on the web today, to support a true non-adhesion contract for Alice.
>> (Perhaps the closest thing on the OAuth side would be requiring an OAuth
>> scope unchecking facility, but that's still not a formal mapping to a
>> contract model -- has someone done that mapping?)
>>
>>
>>
>> In the UMA group, with the help of Tim Reiniger (the primary drafter of
>> the new Virginia digital identity law), we have recently started applying a
>> "license" vs. a "contract" paradigm to the UMA flows. In fact, we're
>> planning to discuss this on today's UMA call in a little over an hour.
>>
>>
>>
>> I may be on the wrong track here, but this is what your note put me in
>> mind of...
>>
>>
>>
>> *Eve Maler *ForgeRock Office of the CTO | VP Innovation & Emerging
>> Technology
>> Cell +1 425.345.6756 <%2B1%20425.345.6756> | Skype: xmlgrrl | Twitter:
>> @xmlgrrl
>> Join our ForgeRock.org OpenUMA
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__forgerock.org_openuma_&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=N6U-VffSODQM0Eo_2TrZgSe_XSZepWfJ14eAp4h6qGY&e=>
>> community!
>>
>>
>>
>> On Wed, Jul 8, 2015 at 8:08 AM, Kinsley, William <BKinsley at nextgen.com>
>> wrote:
>>
>> I need Eva to verify that what I am saying is correct, I am still coming
>> up to speed on UMA … HL7 has/is developing FHIR consent(contract) standards
>> and I believe the oAuth and UMA profiles are in scope to support these and
>> possibly others.
>>
>>
>>
>> To easily support interoperability, is it possible (or in scope) to
>> develop profile standards or profile taxonomy that would be flexible and
>> adaptable so that the RS would be able to process FHIR consents even if
>> they are modified by a specific implementation without  special coding.
>> This could be too idealistic, out of scope or just not supportable with
>> oAuth and/or UMA profiles.
>>
>>
>>
>> For example: A client could connect to any FHIR resource and process any
>> Heart Profile that is provided, even if the client is another RS running in
>> a “Client” role to another RS .
>>
>>
>>
>> Bill
>>
>>
>>
>> *From:* Eve Maler [mailto: <eve.maler at forgerock.com>
>> eve.maler at forgerock.com]
>> *Sent:* Tuesday, July 07, 2015 10:00 PM
>> *To:* Aaron Seib
>> *Cc:* Kinsley, William; openid-specs-heart at lists.openid.net
>> *Subject:* Re: [Openid-specs-heart] Flip the question of “Vanilla" OAuth
>> vs. UMA
>>
>>
>>
>> Commenting on one of Aaron's bulleted items:
>>
>>
>>
>> Profiling "a standard way to label assets managed by the RS" is one of
>> the tasks I was contemplating to be potentially in scope for our semantic
>> UMA profile. This is because it is possible for resource set descriptions
>> (these are things that an UMA RS registers at an AS, to put resources under
>> protection) to include resource types. As I understand it, the FHIR API
>> conveys data structures that reflect quite a lot of HL7 standardization
>> work already done, which amounts to "resource typing".
>>
>>
>>
>> While a lot of companies with proprietary APIs might not want to
>> standardize their resource assets, there's a lot of power in standardized
>> resource labeling in open ecosystems like the one we're working on. For
>> starters, there's a security consideration
>> <https://urldefense.proofpoint.com/v2/url?u=https-3A__docs.kantarainitiative.org_uma_rec-2Doauth-2Dresource-2Dreg-2Dv1-5F0.html-23rfc.section.4&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=XZvNdZZ9gPcuIcTPhPiiE_QBSahOHLBvTzItUZRdivI&e=> that
>> is mitigated by the use of "well-known and standardized" description
>> elements. (See this UMA issue
>> <https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_xmlgrrl_UMA-2DSpecifications_issues_151&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=71o1AnmsYBp0TU9vvpx_TuaYsMN1txaagwxyDCElFnc&e=>
>> for some background.) For another example, standard types could drive
>> automated policy workflows in interesting ways.
>>
>>
>>
>> *Eve Maler *ForgeRock Office of the CTO | VP Innovation & Emerging
>> Technology
>> Cell +1 425.345.6756 | Skype: xmlgrrl | Twitter: @xmlgrrl
>> Join our ForgeRock.org OpenUMA
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__forgerock.org_openuma_&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=N6U-VffSODQM0Eo_2TrZgSe_XSZepWfJ14eAp4h6qGY&e=>
>> community!
>>
>>
>>
>> On Tue, Jul 7, 2015 at 2:56 PM, Aaron Seib < <aaron.seib at nate-trust.org>
>> aaron.seib at nate-trust.org> wrote:
>>
>>   Thanks for starting this new thread.
>>
>>
>>
>> I am not expert in this space (yet) but let me see if I can repeat back
>> what I think you are proposing.
>>
>>
>>
>> Are you suggesting that for Resource Server (RS) be able to accept a
>> standard profile authorization assertion (based on the UMA profile) from a
>> standard (UMA-based) Authorization Server (AS)?
>>
>>
>>
>> I maybe out of date but I seem to remember reading that the UMA profile
>> states that the Authorization Policy service capabilities (as required to
>> implement an AS) are out of scope for the UMA profile - as are the specific
>> policies for how you label assets (network, applications, data) managed by
>> the RS with access tokens that are registered with and managed by the AS.
>>
>>
>>
>> To echo back your language is your suggestion that it ^would^ be simpler
>> to have consistent patterns (libraries) implemented that would address
>> what the UMA profile has intentionally said is out of scope?  I.e.,
>>
>> ·        addressing the need for a standard way to label assets managed
>> by the RS; and (?)
>>
>> ·        a standard way to represent the inputs to an Authorization
>> Policy Service
>>
>>
>>
>> In my mind this would allow us to not only solve the simple cases but
>> also enable us to develop libraries that represent the applicable policy of
>> a given Federal Reg or libraries of applicable state law that could be
>> re-used by everyone.  It might also enable the different associations to
>> provide recommended policies to be adopted by their members and plugged
>> into the solution following a period of local policy tweaking by a given
>> institution or Agency.
>>
>>
>>
>> Am I getting this right?
>>
>>
>>
>> Aaron Seib, CEO
>>
>> @CaptBlueButton
>>
>>  (o) 301-540-2311
>>
>> (m) 301-326-6843
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__nate-2Dtrust.org&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=tPRDdAyGDknsaFwvwbIPyKu9NkZpJ-UX0L62WfJTfPk&e=>
>>
>>
>>
>> *From:* Openid-specs-heart [mailto:
>> <openid-specs-heart-bounces at lists.openid.net>
>> openid-specs-heart-bounces at lists.openid.net] *On Behalf Of *Kinsley,
>> William
>> *Sent:* Monday, July 06, 2015 8:45 PM
>> *To:* <openid-specs-heart at lists.openid.net>
>> openid-specs-heart at lists.openid.net
>> *Subject:* [Openid-specs-heart] Flip the question of “Vanilla" OAuth vs.
>> UMA
>>
>>
>>
>> I am starting a new thread …  I think we need to flip the question of
>> “Vanilla" OAuth vs. UMA”. I feel confident that we are going to discover
>> use cases that cannot be supported by “Vanilla” OAuth or would be greatly
>> simplified by using UMA.
>>
>>
>>
>> Maybe the real question to ask is:  Are there any augments (use case,
>> technology restriction, cost, etc.) that justifies NOT using (requiring)
>> UMA?
>>
>>
>>
>> From a interoperability, quality, security and development perspective,
>> would it be simpler to have consistent patterns (libraries) implemented
>> that are more likely to be “drop-in compatible” without source changes.
>> While the standard itself would be considered rigid, it would be flexible
>> by the use and implementation of the UMA profiles.
>>
>>
>>
>> The caveat here is the resource server (RS) would need to be able to
>> accept/process a UMA profile without developing custom code to interpret
>> it.  Would this require resource servers to adhere to a standard set of UMA
>> profiles or a defined UMA profile taxonomy that could describe the
>> healthcare consent models (if one exists)?
>>
>>
>>
>> Bill
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> *William Kinsley *Enterprise Architect, Ambulatory
>>
>>
>> *NEXTGEN HEALTHCARE *Solutions for: Ambulatory, Inpatient and Community
>> Connectivity
>> 795 Horsham Road, Horsham, PA 19044
>> (215) 657-7010 x21128 <%28215%29%20657-7010%20x21128> [o]
>>  <BKinsley at nextgen.com>BKinsley at nextgen.com
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>> *Be ready for MU and ICD-10 in 2015. Start your EHR version 5.8 and KBM
>> version 8.3 upgrade today. Get the resources you need at
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>www.nextgen.com/upgradecentral
>> <http://www.nextgen.com/upgradecentral>*
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>> This message, and any documents attached hereto, may contain confidential
>> or proprietary information intended only for the use of the addressee(s)
>> named above or may contain information that is legally privileged. If you
>> are not the intended addressee, or the person responsible for delivering it
>> to the intended addressee, you are hereby notified that reading,
>> disseminating, distributing or copying this message is strictly prohibited.
>> If you have received this message by mistake, please immediately notify us
>> by replying to the message and delete the original message and any copies
>> immediately thereafter. Thank you for your cooperation.
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>>
>> _______________________________________________
>> Openid-specs-heart mailing list
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>> Openid-specs-heart at lists.openid.net
>> http://lists.openid.net/mailman/listinfo/openid-specs-heart
>>
>>
>> <https://urldefense.proofpoint.com/v2/url?u=http-3A__www.oneugm.com&d=AwMFaQ&c=IV_clAzoPDE253xZdHuilRgztyh_RiV3wUrLrDQYWSI&r=B4hg7NQHul-cxfpT_e9Lh49ujUftqzJ6q17C2t3eI64&m=G5SLFEapI55EmVv6qepixkuIjAo8MIS-sDAtWeTMWBY&s=bzR8BQA3rz7axUDSwHpXrotdgE261lSqW-LBhoq1GWs&e=>
>>
>>
>>
>>
>>
>> _______________________________________________
>> Openid-specs-heart mailing list
>> Openid-specs-heart at lists.openid.net
>> http://lists.openid.net/mailman/listinfo/openid-specs-heart
>>
>>
>
>
> _______________________________________________
> Openid-specs-heart mailing listOpenid-specs-heart at lists.openid.nethttp://lists.openid.net/mailman/listinfo/openid-specs-heart
>
>
>
> _______________________________________________
> Openid-specs-heart mailing list
> Openid-specs-heart at lists.openid.net
> http://lists.openid.net/mailman/listinfo/openid-specs-heart
>
>

-- 
Adrian Gropper MD
Ensure Health Information Privacy. Support Patient Privacy Rights.
*http://patientprivacyrights.org/donate-2/*
<http://patientprivacyrights.org/donate-2/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openid.net/pipermail/openid-specs-heart/attachments/20150712/fc98b430/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 3142 bytes
Desc: not available
URL: <http://lists.openid.net/pipermail/openid-specs-heart/attachments/20150712/fc98b430/attachment-0002.jpe>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/jpeg
Size: 3141 bytes
Desc: not available
URL: <http://lists.openid.net/pipermail/openid-specs-heart/attachments/20150712/fc98b430/attachment-0003.jpe>