[OpenID board] The Specs Council and Process (WAS: Re: Executive Committee meeting 12/18/2008 ...)

David Recordon drecordon at sixapart.com
Thu Dec 18 00:40:01 UTC 2008


On Dec 17, 2008, at 4:10 PM, Dick Hardt wrote:

> On 17-Dec-08, at 2:29 PM, David Recordon wrote:
>
>> Or ideally the group would emerge with a draft of a spec along with
>> some form of consensus around it and thus be easier to write and have
>> a better informed scope statement.
>>
>> If we were to do this, what would it look like?
>>
>> 1) Some group of people decide they want to tackle a problem.  They
>> email the specs@ mailing list with some sort of description of what
>> they want to do and we set them up with a mailing list like drafting-
>> <name>@openid.net.  (I'd also have no problem in using Google Groups
>> for these since it's less maintenance and makes it easier for those
>> leading the work to manage.)
>
> I'm fine with Google groups -- but we need to enforce membership and  
> have IPR in place to participate

Good points.  We need a process for dealing with the IPR docs.  This  
is ideally something the ED will do.  They receive an IPR doc, upload  
it somewhere, and add the person's name, affiliation, and email to a  
web page which WG editors can refer to when processing membership.   
ASF's page is at http://people.apache.org/~jim/committers.html.


>> 2) Group is created and it tries to stay away from calling their
>> document "OpenID <foo>" for the time being.
>
> how about OpenID-Proposal-<foo>

Sure, just some clear naming convention that won't confuse adopters.


>> 3) To post to the list, you must agree to the existing IPR policy so
>> everything around withdrawal and review periods of later stage drafts
>> remain the same.
>
> Is that an email or a executed form faxed in?

I think the form currently needs to be faxed or scanned and emailed.


> The challenge for some people will be that they do not want what  
> they consider unrelated IPR to be brought into a WG -- and since the  
> scope is not really defined, they need to be able to opt out after  
> scope is defined. I forget if that is in the IPR statements now.

I'd think these people wouldn't participate in this earlier work if  
they're uncomfortable doing so without a scope, though also reading  
the IPR Policy the withdrawal provision might already be enough to  
work.  Basically, as long as these groups don't publish Implementor  
Drafts or a Final Specification, then contributors are allowed to  
withdraw given seven days written notice, would not have any  
obligations around patents, and would remain subject to the copyrights  
section.


>> 4) At some point, one or more drafts in, the group decides to
>> formalize their WG.
>> 5) They write a charter/scope to submit along with their draft and an
>> accurate list of authors/contributors.
>> 6) Specs Council / membership approves their WG or decides that the
>> draft they've produced really doesn't fit into OpenID and works with
>> the group on either how to change that (e.g. more reuse) or helps  
>> them
>> move to another organization to finish their work.
>
> currently we require a membership vote to approve a WG do we not?

Yes, oversight not including that versus explicitly looking to not  
include the step.  That said, shortening the ~30 days here would be  
nice.  I also wonder if the membership vote is actually effective  
given how it is currently designed with quorum and a simple majority.


>> 7) They use the rest of the process to publish an Implementor's Draft
>> and then in the end a Final Draft.
>
> and then a vote by the membership

Yes, I don't propose changing the process for publishing Implementors  
Drafts or the Final Specification at this time.  As far as I can tell,  
this is:

Implementors Draft:
  - WG comes to consensus to publish an Implementors Draft
  - 45 day IPR review period starts aimed at contributors
  - The OIDF board has 30 days (within the 45) to make sure the  
Implementors Draft won't "create untenable legal liability for OIDF or  
the Board" and that it is not outside of the WG's scope.
  - (It seems there is also written in a membership vote here, though  
it doesn't make sense and I'm guessing is an extra copy/paste.)

Final Specification:
  - WG comes to consensus to publish the Final Specification after at  
least one Implementors Draft
  - 60 day IPR review period starts aimed at contributors
  - The OIDF board has 30 days (within the 45) to make sure the  
Implementors Draft won't "create untenable legal liability for OIDF or  
the Board" and that it is not outside of the WG's scope.
  - 14 day notice period of an OIDF member vote to approve the Final  
Specification

>
>
>>
>>
>> This thus makes it much easier to get started and far more concrete
>> when the specs council is looking to recommend if work becomes a
>> working group or not.
>>
>> It doesn't fix the 4 months to finalize a spec, but I guess we tackle
>> that problem as groups start to hit it.  (Which PAPE is already
>> starting to.)
>>
>> --David
>>
>> On Dec 17, 2008, at 12:52 PM, Dick Hardt wrote:
>>
>>> Sounds like a lighter way of getting a WG started is what is  
>>> needed so
>>> that people can get together to discuss the problem without having  
>>> to
>>> write up a scope document -- since the scope often shifts as people
>>> get together and talk about it.
>>>
>>> So how about we add an earlier stage to the WG -- the formation  
>>> stage
>>> -- at the end of that stage there may or may not be a scope that  
>>> has been created. If there is a scope document, it is put up for  
>>> approval
>>> per the current process. Once approved, the WG is in the  
>>> specification
>>> stage.
>>>
>>> -- Dick
>>>
>>> On 17-Dec-08, at 12:23 PM, Eran Hammer-Lahav wrote:
>>>
>>>> Well, it doesn't go all the way to the book-end approach we're
>>>> taking with OWF. This is just a slight simplification of the  
>>>> current
>>>> process.
>>>>
>>>> The OIDF requires upfront scope approved by the foundation to  
>>>> create
>>>> a WG. The approval process is taking too long and meanwhile, people
>>>> are writing specs elsewhere. Those specs are in IPR limbo and needs
>>>> cleanup if they to eventually enter a WG or have a different IPR
>>>> policy attached.
>>>>
>>>> So my suggestion is simple. Follow the same IPR policy as you have
>>>> today for pre-WG work, meaning, write a clear scope and have some
>>>> form of discussion among those interested in participation.  
>>>> Create a
>>>> mailing list (or designate an existing one) for that work, and  
>>>> apply
>>>> the IPR policy *as-if* this is an official WG. Once the WG is ready
>>>> to publish its first draft, that draft + scope (with possible
>>>> changes) is submitted for an actual WG creation.
>>>>
>>>> If a WG is created, the work continues and the IPR license is
>>>> already in place. If the WG is not created, the parties involved  
>>>> can
>>>> continue as they choose.
>>>>
>>>> All I am really suggesting is to move the WG approval to after the
>>>> first draft, but other than that, keep everything else the same.
>>>>
>>>> EHL
>>>>
>>>>> -----Original Message-----
>>>>> From: board-bounces at openid.net [mailto:board-bounces at openid.net]  
>>>>> On
>>>>> Behalf Of David Recordon
>>>>> Sent: Wednesday, December 17, 2008 11:52 AM
>>>>> To: board at openid.net
>>>>> Subject: Re: [OpenID board] The Specs Council and Process (WAS:  
>>>>> Re:
>>>>> Executive Committee meeting 12/18/2008 ...)
>>>>>
>>>>> Yeah, I don't think IPR per se is the roadblock, but the process
>>>>> that
>>>>> we've created chosen to ensure that IPR isn't an issue is.  Mart  
>>>>> is
>>>>> however correct that most of the current working group proposals  
>>>>> are
>>>>> more or less taking a spec draft that is already written,  
>>>>> turning it
>>>>> into a WG, and then having the non-asserts happen at the end
>>>>> implicitly with the review periods by the WG members versus
>>>>> explicitly
>>>>> as was done by OpenID 2.0 and OAuth 1.0.
>>>>>
>>>>> So, I think that Mart, Eran, and Dick are all correct in what
>>>>> they've
>>>>> said in this thread.
>>>>>
>>>>> Eran, I'm intrigued by your pre-WG idea.  How would you see it
>>>>> actually work?  Sounds a bit like what we've been talking about  
>>>>> for
>>>>> the Open Web Foundation.
>>>>>
>>>>> --David
>>>>>
>>>>> On Dec 17, 2008, at 11:43 AM, Dick Hardt wrote:
>>>>>
>>>>>>
>>>>>> On 17-Dec-08, at 11:28 AM, Martin Atkins wrote:
>>>>>>
>>>>>>> Eran Hammer-Lahav wrote:
>>>>>>>> I take it you didn't have to personally "figure out the IPR
>>>>>>>> afterwards"...
>>>>>>>>
>>>>>>>
>>>>>>> That's actually my point. There are lots of folks for whom the  
>>>>>>> IPR
>>>>>>> stuff
>>>>>>> isn't a concern for one reason or another. Those folks shouldn't
>>>>>>> be
>>>>>>> prevented from getting on with stuff while those who *do* care
>>>>>>> about
>>>>>>> IPR
>>>>>>> are figuring it out.
>>>>>>>
>>>>>>> That's exactly what happened with OpenID 2.0. Lots of folks  
>>>>>>> had it
>>>>>>> implemented long before the IPR was done.
>>>>>>>
>>>>>>> If I author a spec then I'm quite happy to sign an IPR non- 
>>>>>>> assert
>>>>>>> where
>>>>>>> necessary, but the current process is far heavier than that and
>>>>> isn't
>>>>>>> really helping anyone because folks are just writing and
>>>>> implementing
>>>>>>> specs outside of the IPR framework because the IPR framework  
>>>>>>> stops
>>>>>>> them
>>>>>>> actually getting any work done.
>>>>>>
>>>>>> It is MUCH more effort to figure out the IPR afterwards.
>>>>>>
>>>>>> IPR is NOT the roadblock in creating WGs. As David mentions, the
>>>>>> process is currently far to heavy. We need to make it simpler and
>>>>>> easily understood.
>>>>>>
>>>>>> -- Dick
>>>>>> _______________________________________________
>>>>>> board mailing list
>>>>>> board at openid.net
>>>>>> http://openid.net/mailman/listinfo/board
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> board mailing list
>>>>> board at openid.net
>>>>> http://openid.net/mailman/listinfo/board
>>>> _______________________________________________
>>>> board mailing list
>>>> board at openid.net
>>>> http://openid.net/mailman/listinfo/board
>>>
>>> _______________________________________________
>>> board mailing list
>>> board at openid.net
>>> http://openid.net/mailman/listinfo/board
>>
>>
>> _______________________________________________
>> board mailing list
>> board at openid.net
>> http://openid.net/mailman/listinfo/board
>





More information about the board mailing list