Phil, Before we get too far into a solution, I think we need to more clearly define the problem. First, what do subcommittees really do? They act on behalf of the TSC to make recommendations to projects, and then the projects need to decide whether or not to accept the recommendation. Subcommittee advice is not normative. For example, if the architecture subcommittee were to say that the top priority for the Dublin release is Interplanetary Internet (https://en.wikipedia.org/wiki/Interplanetary_Internet), many projects would rightly say, "good for you – it's not OUR top priority, and if you want it, pony up developers." Likewise, the TSC would likely still approve any projects not implementing IPN. ARC has no enforcement mechanism (other than the M3 reviews, which are advisory to the TSC and the projects). So how much structure do we need for a non-normative advisory body? And what does a subcommittee decide? To accept or reject a proposal, or to select between two (or more) competing proposals.
We tried to address this when we created the TSC Charter. We vested all voting power in the TSC, since it was the defined entity in the ONAP Charter. As with code developers, we wanted to encourage input from anyone who is interested, so specified that subcommittees operate under "rough consensus", rather than by voting. The only thing subcommittees should be voting on is the leadership. Standards bodies have a lot of experience working on a consensus basis, and I think we can learn a lot from them, particularly the IETF. From RFC 2418: Working groups make decisions through a "rough consensus" process. IETF consensus does not require that all participants agree although this is, of course, preferred. In general, the dominant view of the working group shall prevail. (However, "dominance" is not to be determined on the basis of volume or persistence, but rather a more general sense of agreement). Consensus can be determined by a show of hands, humming, or any other means on which the WG agrees (by rough consensus, of course). Note that 51% of the working group does not qualify as "rough consensus" and 99% is better than rough. It is up to the Chair to determine if rough consensus has been reached (IETF Working Group Guidelines and Procedures). Note that the Chair is empowered to make the determination of rough consensus. This is done through measurement at meetings or on the mailing list. The IETF doesn't typically ask for quorum – it's the preponderance of opinion of the people who participate in regularly scheduled meetings (or on the mailing list). If people have an issue with the Chair determination, they can always escalate (in our case, to the TSC). Speaking just for the ARC, I've noticed that results of the TSC votes closely mirror the rough sense of the community that I measure prior to bringing the release architecture to the TSC. You do raise good questions about how to prevent organizations from gaming the system by sending a bunch of people to a meeting, and implicitly how to balance between organizations who send a lot of resources and those who send fewer. Let's start with the latter question. I think our current system addresses it by allowing anyone to participate in the subcommittees (which may favor companies with more resources) but then vests decision-making authority with the TSC, which only allows one vote per company. I think that's a reasonable approach. Could we handle the former simply by freezing the membership list at some interval prior to an election (2 weeks, a month, ???)? Beyond that, I'm concerned that the more formality you introduce, (1) the fewer voices get to participate, (2) the greater the delay, and (3) the more we're enforcing waterfall vs. agile processes (which feeds the delay). Chris From: <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> on behalf of Phil Robb <pr...@linuxfoundation.org<mailto:pr...@linuxfoundation.org>> Reply-To: "ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>" <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Date: Wednesday, August 29, 2018 at 10:27 AM To: onap-tsc <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Subject: Re: [onap-tsc] Definition of Subcommittee Membership Hello Chris, Steve, and other ONAP TSC Members: In one sense, I think Kenny has raised some symptoms of issues that he is seeing with regard to setting up for the upcoming subcommittee elections. From discussions with Kenny, I have come to believe that we need to put a little more structure and formalization around the membership of subcommittees given the nature of this project. What Kenny and I have observed: 1) There is significant difference of opinion on several topics within the different subcommittees. At present, the Modeling subcommittee is exhibiting the highest amount of discord. Others have had it in the past. Discord is not necessarily a problem as long as coming to conclusion on a given issue/topic is well-understood, transparent, and perceived as fair by stakeholders. However.... 2) Within ONAP, we are still working to ensure that our community processes are perceived as fair and transparent. We had an issue last year around the Chairperson election of the UseCase Subcommittee that resulted in us changing the rules for that particular election. We modified and documented those new rules to create that well-understood, transparent, and fair process for that event. As Kenny and I look at the significantly informal processes currently documented for the Subcommittees, the following questions arise: 1) If a decision/consensus is made in a subcommittee, even if that decision is to make a recommendation to the TSC, who made that consensus/decision? Is it that the people's names on the Subcommittee wiki form the formal membership and hence more than 50% of them need to be present in any given meeting for consensus to be established? If that's the case, it is not currently being followed. If it isn't the case, gaining quorum and consensus will be very difficult for the subcommittees. 2) Given that there is significant difference of opinion across our community on a variety of topics that are discussed deeply in the subcommittees, what prevents one or a group of organizations from gaming the vote by putting a very large number of members on the subcommittee wiki? For the Chair elections we said there was general openness for subcommittee membership, but there was only one voting member from each company who can vote for the Chair. In retrospect, I think Kenny and I have come to believe that we should have such a rule for all meetings and consensus gathering activities. So to avoid bad actor or simple mistakes in Subcommittee management, some form of policy around the following should be considered by the TSC on behalf of the subcommittees: 1) There should be formal set of defined [voting] members of a subcommittee. Similar to the TSC, all voices are welcome, but when it comes to a decision, be it via consensus or vote, it is well understood who needs to participate in such an activity. 2) Voting criteria should be similar to voting for the Chair... ie one vote per company to ensure appropriate voting representation - This goes for even gaining a consensus... ie when the chair states "Does anyone disagree with this recommendation to the TSC?"... The only people that should be able to say "I object" are the designated voting members. This stops a group of participants from the same organization from all raising objections, that result in rough consensus not being gained. 3) Quorum and proxy criteria should be established so the subcommittee knows when it has participation from enough representatives of the stakeholders to establish a decision/consensus. As we have seen multiple times, it is easy to accidentally schedule a meeting during a holiday in some geography. If the meeting occurs, and no Quorum is needed, decisions/recommendations can be made during that meeting that do not take into account a large portion of the stakeholders. While Kenny and I expect that our ONAP participants are all well-intended, the lack of this formality in the structure and processes of the subcommittees fall short on transparency, and well-understood documented processes. That in turn leads to the potential perception of unfair and/or poorly-enforced rules of engagement which wastes energy and effort across the project. Best, Phil. On Mon, Aug 27, 2018 at 5:07 PM, Christopher Donley <christopher.don...@huawei.com<mailto:christopher.don...@huawei.com>> wrote: I agree with Steve. I don't think we want to be too prescriptive about subcommittee membership, and we want to be open and welcoming to people who will contribute. We set it up so that subcommittees work on a rough consensus basis, and decisions needing a formal vote get raised to the TSC. You mentioned the architecture subcommittee. We usually have about 50 people show up on the weekly calls (depending on topics), and additional people participate in various task forces or working groups, so I think the membership really is in the 72-108 person you cited. It is worth cleaning up the list, as some people have changed companies or have been reassigned, but in general, I am comfortable with the idea of self-selected membership. For the purposes of electing the leadership, I think one "finger guard" we could consider is temporarily freezing the membership list prior to the commencement of the vote (e.g., voters are those listed on the wiki as of two weeks prior to the date of the vote, or whatever interval we decide). This would eliminate a last-minute rush to sign up, and would more accurately reflect the people who have been active in the group without discouraging future participants. Chris From: <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> on behalf of Stephen Terrill <stephen.terr...@ericsson.com<mailto:stephen.terr...@ericsson.com>> Reply-To: "ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>" <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Date: Monday, August 27, 2018 at 1:48 PM To: "ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>" <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Subject: Re: [onap-tsc] Definition of Subcommittee Membership Hi Kenny, I appreciate the effort that you are putting in regarding the elections and encoring communication via email list membership. I wonder, though, whether we are looking at this from the optimal perspective and context. I’d be happy to hear from other sub-committee chairs here. My reflection is that we had to bring items back to the TSC for decision, where formal membership was required. So for me, working on a consensus basis was based upon the views both expressed in the meetings and what I considered may have been expressed if active members were absent. This was to say, that the actual operations of the sub-committee isn’t missing the concept of membership and instead appreciates views and contribution. The need arises when it comes to selecting the chairperson (as that is the only real formal voting). Lets not let that need complicate the day-to-day operations. Normally I would side with formality as that tends to play out better in the long run, but I am urging that we introduce formality to solve the necessary issues and see if we can use the evolving culture here. BR, Steve PS – I will admit I have not been good at keeping a record of sub-committee meeting attendance. From:ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org> <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> On Behalf Of Kenny Paul Sent: Monday, August 27, 2018 5:25 PM To: ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org> Subject: Re: [onap-tsc] Definition of Subcommittee Membership Hi Steve, The points you've raised are the same ones that prompted my email. Anyone is always welcome at any Subcommittee meeting either as an active participant, accidental tourist, interested party or concerned onlooker. They do not have to be on the wiki membership roster or the mailing list to attend a meeting. However, even if they passively attend a meeting and do perform work behind the scenes, they would be missing any and all of the email-based discussions for the subcommittee. As mentioned, I haven't really looked into Seccom yet so I don't know its status, but in a more general sense attendance records for the subcommittee meetings seem to be sparse or non-existent. Without such records is not possible to identify people on the membership roster who may be passively listening to the meetings at a minimum. There is no way to differentiate them from someone on the roster in name only that neither attends meetings or receives email. A governance model where "Membership" has no real definition or meaning is not governance. The question is, "Do we care about that?" Speaking on a purely personal level as Kenny the community member and not Kenny the PM, yes, I do care great deal. I care because in less than a minute anyone in the world can benefit themselves while doing absolutely nothing whatsoever to contribute to our community. All it takes is adding their name to a roster and then they can put "ONAP XYZ Subcommittee Member" on their resume or list of accomplishments. While using what we produce without contributing may be unfortunate, but perfectly acceptable in open source, using us to benefit just ain't right. End personal opinion. One year after the launch of our project there were still dozens of approved Committers that had never even received their credentials for to perform a commit. Just as we've been cleaning that up, based upon what I've discovered, it seems that the subcommittee membership is the next area to address. Thanks! -kenny From: <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> on behalf of Stephen Terrill <stephen.terr...@ericsson.com<mailto:stephen.terr...@ericsson.com>> Reply-To: <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Date: Monday, August 27, 2018 at 12:21 AM To: "ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>" <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Subject: Re: [onap-tsc] Definition of Subcommittee Membership Hi Kenny, I’ve struggled with this a little from the security sub-committee and decided not to focus on it, the reason comes down to what does it mean to be a member. * We have people actively participating (which is great) – and that is important irrespective of whether they are identified as members. I do hope that all active participants can identify themselves as members. * We have people listening in to the calls that may be normally quite. I do not, however know, what they are doing behind the scenes to connect the dots – and I wouldn’t want to project any view that they are not welcome join, or participate, or listen whether or not they identify themselves as a member or not. * We can have people that are formally identified as members, that don’t do the above. We could go down the path of “active membership”, however and have criteria like meeting attendance, wiki updates etc; but do we clearly gain when anyway at the end of the day the sub-committeess are advisory and work on rough consensus. BR, Steve From:ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org> <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> On Behalf Of Alla Goldner Sent: Monday, August 27, 2018 8:48 AM To: ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org> Subject: Re: [onap-tsc] Definition of Subcommittee Membership Hi Kenny, As you mention below: As defined in 4.4.1.3 Subcommittee Chair / Vice Chair Elections<https://wiki.onap.org/display/DW/ONAP+Technical+Community+Document#ONAPTechnicalCommunityDocument-4.4.1.3SubcommitteeChair/ViceChairElections> : The Chair or Vice-Chair will be elected by members of the subcommittee as of the date the nomination process starts for the election. Section 4.4.1.4 Subcommittee Voter Eligibility<https://wiki.onap.org/display/DW/ONAP+Technical+Community+Document#ONAPTechnicalCommunityDocument-4.4.1.4SubcommitteeVoterEligibility>the criteria only defines: Voting for a Chair or Vice-Chair is not limited to ONAP member companies. However only 1 Subcommittee member from each company, or group of related companies may vote in the election. Therefore, at least my reading is that the problem you describe below may exist only if at least one of assigned voting members (max 1 per company or group of related companies) is not fully identified as a subcommittee member, as only they should be getting a ballot. Is it the case? Best regards, Alla Goldner Open Network Division Amdocs Technology [cid:image001.png@01D43DE7.54281640] From:ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org> [mailto:ONAP-TSC@lists.onap.org] On Behalf Of Kenny Paul Sent: Sunday, August 26, 2018 11:37 PM To: onap-tsc <ONAP-TSC@lists.onap.org<mailto:ONAP-TSC@lists.onap.org>> Subject: [onap-tsc] Definition of Subcommittee Membership TSC Members, I have run into an issue I have a concern with the Subcommittee elections. I spent all day Saturday working this and I was going to make the call myself but after sleeping on it I decided that this really requires a decision from the TSC, not from the PM. I am going to hold off on balloting until I have received clear guidance. The primary membership criteria for most Subcommittees were established with an overly broad, “If you want to be a member, just add your name to the wiki” model. The result of this is basically unmanageable and has introduced a grey area that calls into question both the definition and validity of membership. In the crush to get ONAP off the ground this resulted in a flood of people adding themselves and then failing to participate. The membership lists have never been curated as far as I am aware, so there are a lot of "members" that have never been active. I have not yet looked into Security, Open Lab or University yet and Control-Loop isn't due for another couple months. However, for the "big three" here is the situation… Architecture Subcommittee: 108 members listed, 70 are actually subscribed to the Onap-arc mailing list Use case Subcommittee: 98 members listed, 49 are actually subscribed to the Onap-usecasesub mailing list Modeling Subcommittee: 84 members listed. It does not have its own mailing list and instead uses onap-discuss. Out of the 35 company representatives (as per section 4.4.1.4 of the Community Document), 9 of them are not subscribed to the mailing list. A few of these deltas can be attributed to the fact people did not provide the same email address on the wiki page that they used to subscribe to the list. It is the member's responsibility to reconcile that, but it represents only a small percentage of the cases. This raises the issue as to whether someone can legitimately be considered of a subcommittee "member" if they are not even subscribed to the subcommittee's mailing list? My perspective is that they should not be. They may be an interested party, but since they are not able to participate in list-based discussion and they are not in a position to make decisions relative to the subcommittee. As defined in 4.4.1.3 Subcommittee Chair / Vice Chair Elections<https://wiki.onap.org/display/DW/ONAP+Technical+Community+Document#ONAPTechnicalCommunityDocument-4.4.1.3SubcommitteeChair/ViceChairElections> : The Chair or Vice-Chair will be elected by members of the subcommittee as of the date the nomination process starts for the election. Section 4.4.1.4 Subcommittee Voter Eligibility<https://wiki.onap.org/display/DW/ONAP+Technical+Community+Document#ONAPTechnicalCommunityDocument-4.4.1.4SubcommitteeVoterEligibility>the criteria only defines: Voting for a Chair or Vice-Chair is not limited to ONAP member companies. However only 1 Subcommittee member from each company, or group of related companies may vote in the election. Based on the Community Document ballots should be sent to anyone listed on the wiki regardless of their actual involvement. Doing that makes me uncomfortable as it runs counter to promoting a vibrant community. Instead implies that we are basically apathetic in that regard. So, my questions to the TSC Members are: Should subcommittee election ballots only be distributed to individuals that are subscribed to the subcommittee's mailing list? Should subcommittee members that are not subscribed to the subcommittee's mailing list be dropped from the roster? (Assuming yes to the above) What is a reasonable period of time to allow subcommittee members to correct their information before proceeding? I will add this topic to the TSC meeting agenda, however discussions should take place here in advance. Thanks! Best Regards, -kenny Kenny Paul, Technical Program Manager, The Linux Foundation kp...@linuxfoundation.org<mailto:kp...@linuxfoundation.org>, 510.766.5945 San Francisco Bay Area, Pacific Time Zone This message and the information contained herein is proprietary and confidential and subject to the Amdocs policy statement, you may review at https://www.amdocs.com/about/email-disclaimer -- Phil Robb VP Operations - Networking & Orchestration, The Linux Foundation (O) 970-229-5949 (M) 970-420-4292 Skype: Phil.Robb -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#3656): https://lists.onap.org/g/ONAP-TSC/message/3656 Mute This Topic: https://lists.onap.org/mt/24966937/21656 Group Owner: onap-tsc+ow...@lists.onap.org Unsubscribe: https://lists.onap.org/g/ONAP-TSC/leave/2743226/1412191262/xyzzy [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
