Re: [Mops] Confirming adoption consensus Re: draft-ietf-mops-streaming-opcons

["Leslie Daigle" <[e-mail]>]

Hi,


On 5 Feb 2020, at 10:40, Kyle Rose wrote:

> I think ownership might be covered, at least from a policy 
> perspective, by
> some text from the git draft (notwithstanding the weaselly use of the
> passive voice):
>
>    When an area director or working group chair makes a request to
>    create a GitHub organization, the following process would be
>    initiated:
>
>    1.  Create a GitHub organization for the working group.
>
>    2.  Name the organization as ietf-wg-<wgname>
>
>    3.  Initialize the organization by designating the IETF Secretariat
>        and the area directors in the working group's area as owners.  
> If
>        the responsible AD for the working group is from another area,
>        that AD will be an owner as well.
>
>    4.  Initialize the organization with a team that has administrator
>        access.  This team will consist of the working group chairs and
>        working group secretary, if one exists.
>
>    After the organization is created, the URL for the organization 
> would
>    be added to the working group's page in the Datatracker.
>
>    Steps 3 and 4 above imply that the GitHub identities of the
>    organization owners and administrators are known.  Recording GitHub
>    identities in the Datatracker (see
>    <https://trac.tools.ietf.org/tools/ietfdb/ticket/2548>) would
>    facilitate this.  The person requesting the organization would need
>    to be notified if the GitHub identities of any of the people meant 
> to
>    be owners or administrators were not available.

Yeah, I think you’re right, that’s probably the best coverage we can 
reasonably implement.

>
> Maybe less inspiring is the wording from the following section, which 
> uses
> something very close to RFC 6919 "REALLY OUGHT TO" language:
>
>    If a working group already has an organization, it would be useful 
> to
>    be able to make it have the same management as one would get with
>    going through the steps in Section 2.1
> <https://tools.ietf.org/html/draft-ietf-git-github-wg-configuration-05#section-2.1>.
> That is, it would be good to
>    be able to run steps 3 and 4 from Section 2.1
> <https://tools.ietf.org/html/draft-ietf-git-github-wg-configuration-05#section-2.1>
> so that the rest of the
>    activities in this section such as personnel work the same for the
>    organizations that were created on their own.
>
> I'm inclined to go forward with creating a mops-wg organization on the
> assumption that, at worst, we're in the same boat with other WGs using
> GitHub, but maybe we should have a consensus call on whether 
> participants

I think that’s right — I think we’ve heard support for it for 
documents.  If we want to do more than that with the git repo, we can 
(should) do a consensus call.

As I said — I don’t want to get in the way of implementing tools 
people want, and at this point any remaining concerns I have are not 
MOPS-specific.

Leslie.

> want to incorporate an official GitHub org into their workflows. 
> Otherwise,
> individual document teams are still free use unofficial GitHub repos 
> for
> revision control and manageability of collaboration, but we would 
> probably
> need to be more strict about conducting official work on the mailing 
> list,
> which could negate some of the advantages of using GitHub.
>
> Kyle
>
> On Tue, Feb 4, 2020 at 2:29 PM Leslie Daigle <[e-mail]>
> wrote:
>
>> Hi Spencer,
>>
>> On 4 Feb 2020, at 11:04, Spencer Dawkins at IETF wrote:
>>
>> Hi, Leslie,
>>
>> On Mon, Feb 3, 2020 at 10:40 PM Leslie Daigle 
>> <[e-mail]>
>> wrote:
>>
>> Spencer,
>>
>> One thing I’ve been trying to find out is: how to tie a git repo to 
>> the
>> IETF. Looking at the CELLAR example, it seems like one just makes the
>> assertion that it is the WG git repo? And, it seems odd that there 
>> are no
>> visible members, let alone owner, of the git repo?
>>
>> Apologies if I’m missing the obvious.
>>
>> I don't think you are.
>>
>> I believe the current state of play is described in
>> https://tools.ietf.org/html/draft-ietf-git-github-wg-configuration-05,
>> especially in Section 2.1 and following sections (note that this 
>> draft is
>> in PUBREQ state, so it's probably mostly right, but hasn't even been
>> through AD review yet). Some of the text is input to the tools team, 
>> but
>> most (selecting organization names, etc.) isn't dependent on 
>> datatracker
>> support.
>>
>> Yep, I had read through that.
>>
>> I believe you are correct that in Github, that's an assertion. In the
>> datatracker, co-chairs can add the URL for the organization (I think 
>> you
>> can just stick this on the working group page now, but apparently 
>> there's
>> development underway that would allow the datatracker to know that 
>> URL is a
>> GitHub repository).
>>
>> That’s a good point — it leaves it within the IETF’s purview to 
>> identify
>> any “authentic” repo.
>>
>> I've only been watching CELLAR for a few weeks, so don't know the 
>> history
>> on defined users, etc. - sorry!
>>
>> I hope that's helpful.
>>
>> I’m confident GitHub isn’t going away any time soon, but my 
>> general
>> concern follows on from the IETF’s experience when everybody hosted 
>> their
>> own WG mailing list (federated responsibility! No single point of 
>> failure!)
>> and archives were lost when people got irritated with the IETF or 
>> companies
>> went belly-up.
>>
>> I think it would be good if the IETF had a more persistent role in 
>> the
>> administration of the git repos, if a group’s work is going to 
>> depend on it.
>>
>> As a co-chair of MOPS, I don’t want to get in the way of providing 
>> a tool
>> that current participants actually want to use! But, I also want to 
>> make
>> sure we use it in a way that is going to stand up to the test of 
>> time.
>>
>> Leslie.
>>
> <quote trimmed>


> -- 
> Mops mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mops

-- 

-------------------------------------------------------------------
Leslie Daigle
Principal, ThinkingCat Enterprises
[e-mail]
-------------------------------------------------------------------