[Openid-specs-ab] ABSTAIN on the vote to approve Implementer’s Draft of OpenID Connect Federation Specification

Phil Hunt phil.hunt at oracle.com
Tue Jul 31 19:41:37 UTC 2018


Inline

Phil

> On Jul 31, 2018, at 1:11 PM, Mike Jones <Michael.Jones at microsoft.com> wrote:
> 
> Phil – you can contribute as a working group member at any time.  There’s no need to wait for one of the Implementer’s Draft votes to submit comments on specifications you care about.  Attend calls.  File issues.  Actively participate in the working group if you want to see specific changes made.

You miss my point entirely. 

We are submitting comments which are largely ignored. No attempt is made to reconcile these comments!
>  
> The Foundation-wide votes are about IPR and a final check-and-balance on the working group output.  They’re not where the work happens.

Why did you claim it meant consensus on the work then?

>  
> The consensus sought was to seek IPR protections for implementers.  No consensus was sought or assumed that the spec was in any way close to final.  (Indeed, the spec itself makes it clear that it’s not at https://openid.net/specs/openid-connect-federation-1_0-04.html#rfc.appendix.C.)  The two things shouldn’t be confused.

This is not what happens in practice. Once specs are implementer major changes are refused in practice. 

You need to have consensus on the approach so that only minor changes are needed. 

For example on backchannel I pleaded with you to await the SECEVENT delivery draft. You ignored me. 

Now RISC used this as leverage against the common standard. 

Mike S raised foundational issues that went largely undiscussed. 
>  
> We want implementations to inform our understanding of what works well and what doesn’t so we can iterate.

Yes. That is true. When there is consensus on a proposal. There often is not in my experience. 
>  
> And yes, once the editors work on Mike’s and Filip’s and other’s comment on the Federation spec, of course we’ll check back to see whether the submitter’s concerns were resolved.

Jumping to implementer is inappropriate given the type of feedback received. 

Making people write code that may be yanked is always wrong and informs nothing. 
>  
> I appreciate that you’re trying to improve things and you’re being listened to.  Hopefully the “How do working groups work?” doc that you’d requested will help make things clearer and easier for everyone.
>  
>                                                                 Thanks,
>                                                                 -- Mike
>  
> From: Phil Hunt <phil.hunt at oracle.com> 
> Sent: Tuesday, July 31, 2018 11:44 AM
> To: Mike Jones <Michael.Jones at microsoft.com>
> Cc: Prateek Mishra <Prateek.Mishra at oracle.com>; openid-specs-ab at lists.openid.net; Mike Schwartz <mike at gluu.org>; Don Thibeau <don at oidf.org>
> Subject: Re: ABSTAIN on the vote to approve Implementer’s Draft of OpenID Connect Federation Specification
>  
> Mike
>  
> The issue is checking back with an issue submitter whether their concern was resolved. Too often it feels like “Your concern was heard and we do not agree and have ignored it”. This seems to fly in the face of building consensus from the issuers perspective. 
>  
> You have also stated to me that the IPR vote has nothing to do with the evolution/consensus of the spec except legal standing to implement. But then why did you just presume consensus with with a 100% vote to implement?
>  
> This is why I have asked many times for more work on specs that seem to lack consensus before the IPR vote. For example backchannel was voter yes against my concerns. 
>  
> This process has been negative for newcomers and Mike S’s concerns were well founded in my experience. 
>  
> My point in commenting here is to improve OpenId process. 
>  
> As a corporate employee member I cannot vote. In the absence of consensus calls I effectively have no say in the spec process and am strongly demotivated by OIDF practices from contributing. 
>  
> Phil
> 
> On Jul 31, 2018, at 11:04 AM, Mike Jones <Michael.Jones at microsoft.com> wrote:
> 
> Hi Prateek,
>  
> We wrote the “How do working groups work?” doc in response to feedback from you and Phil earlier on wanting to understand the working group processes better.  Thanks for asking us to write that.  If there are points that you believe still need to be described better, please let us know.  We want everyone to understand how to effectively contribute to OpenID working groups.  (My apologies that it took me a while to get to it – lots going on!)
>  
> As has been documented earlier on this list, no comments are being ignored and none will be.  In fact, the editors responded to Mike Schwartz and Filip Skokan, who provided editorial feedback, saying that their suggested improvements will be incorporated in the next draft (see http://lists.openid.net/pipermail/openid-specs-ab/Week-of-Mon-20180611/006794.html).  This will occur following the current IPR vote.
>  
> Yes, we can take the feedback into account that you’d like more explicit notification prior to all foundation-wide votes, and do this more clearly in the future.  That being said, given that there was a unanimous decision to seek IPR protections for implementers at the in-person working group meeting that you hosted at Oracle, I don’t understand why you’re saying now that you were surprised.  If you thought there was anything to discuss when the decision was made, you could have spoken up then and the working group would had a conversation about it on the spot.
>  
> Finally, please read what the “How do working groups work?” doc says about Implementer’s Drafts and the value of implementation.  An Implementer’s Draft is no different than any other draft, other than that the working group has decided that enough new content is there that it’s worth locking in IPR commitments to protect implementers who are producing experimental implementations so as to be able to provide feedback on the draft.  Changes can and will be made in the future – many of them informed by what’s learned from the experimental implementations that the working group wants to encourage as a core part of how we produce high-quality, implementable, deployable specifications.
>  
> We value your input and do seek to have effective working group experiences for all participants.
>  
>                                                        Thanks again,
>                                                        -- Mike
>  
> -----Original Message-----
> From: Prateek Mishra <Prateek.Mishra at oracle.com> 
> Sent: Tuesday, July 31, 2018 9:35 AM
> To: openid-specs-ab at lists.openid.net
> Cc: Mike Schwartz <mike at gluu.org>; Don Thibeau <don at oidf.org>; Mike Jones <Michael.Jones at microsoft.com>; Phil Hunt <phil.hunt at oracle.com>
> Subject: ABSTAIN on the vote to approve Implementer’s Draft of OpenID Connect Federation Specification
>  
> I have voted to abstain. I did not find a way to add a comment to the vote, hence this message.
>  
> This reflects our perception that there are WG consensus issues where feedback on issues is received and apparently ignored in the perception of the submitter. 
>  
> Editors need to confirm agreement and chairs should be calling for consensus on items which might not yet be resolved. 
>  
> There should be a notice prior to major voting calling for all participants to review the document before the vote. This should be a “consensus call”.”
>  
>  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openid.net/pipermail/openid-specs-ab/attachments/20180731/2479f0f3/attachment-0001.html>


More information about the Openid-specs-ab mailing list