Re: [opnfv-tech-discuss] [opnfv-tsc] Process Overview - How to Bring a Project into OPNFV
Thanks, Manuel. I would like to encourage others to chime in too, as I hope OPNFV will continue to have new project proposals :) I really like the idea of Git based comments. I too have seen this done in other projects, and it’s a great way of being able to track history and comments. This method of collaboration does have a proven track record of being inclusive across all time zones and allows for time to formulate a response when English is not the native language. Git based reviews also can help with automation, reducing workload for the Helpdesk, as eventually JIRA project, Git repo, INFO.yaml, Confluence space creation, etc, could all be automated. Even the TSC vote could be done via Gerrit +1/-1 voting on the change, leaving the +2/submit -2/reject to either the TSC chair or LFN helpdesk, which then triggers the creation of the project. Regards, Mark From: on behalf of "Manuel Buil via Lists.Opnfv.Org" Reply-To: "mb...@suse.com" Date: Tuesday, August 27, 2019 at 07:13 To: "HU, BIN" , "opnfv-tech-discuss@lists.opnfv.org" , "opnfv-...@lists.opnfv.org" Cc: "opnfv-...@lists.opnfv.org" Subject: Re: [opnfv-tsc] Process Overview - How to Bring a Project into OPNFV Hi TSC, First of all, thanks Bin for the description. I personally did not know how this meeting got created. I think we can all agree that currently the only way for new projects to get the stamp "you have community consensus" is through the decision of the weekly technical discussion. Therefore, projects must go to the weekly technical discussion and get the approval from it to reach the next level, the TSC. I see some potential improvements in this process because there are some problems that I personally see: * We have a geographically widespread community. The time for this meeting is never good for everyone and most of the times not even half of the TSC participates, thus I don't think this meeting is a good way to measure the "community consensus" because we are indirectly excluding TSC members and other community members and we would like to hear their opinions too, right? * Even if the project has been up for review in the wiki for 2 weeks, most of the opinions, feedback, reviews are first time raised in that meeting and not in the comments of the wiki. Therefore, most of the times, there is not time to confront the points raised thus delaying everything * Things can only move once a week, when this meeting takes place. This adds an extra delay which could be avoided, or?. Let alone holidays, etc. * Non native speakers might have difficulties to respond to feedback, especially when comments are not raised before the meeting * How the meeting decides if a project gets the approval is not clearly settled. Currently, we don't vote and if there are no further questions, the moderator of the meeting decides that the project is good to go to the TSC. I'd like to suggest a new way for this which I think servers better the purpose of "foster community review and fulfill the metric of “Proposal has been socialized with potentially interested or affected projects and/or parties”". Actually I have two similar alternatives, just the tool changes: ## Alternative 1 ## (preferred) Upload the proposal to github (or gerrit or gitlab) and then allow TSC members to comment on this and do the reviews online. Same procedure that CNTT is following for its documents and a very common procedure in other open source communities. We give 2 weeks and during that time TSC members must write their concerns, questions, feedback, etc in the comments. Our PM or the person proposing the project should actively take care of reminding it to the TSC members, this way we will hopefully get reviews from the whole TSC. Things could be of course delayed in case of need, e.g. a TSC person is on vacation, comments were not addressed, etc... but in general, after 2 weeks, the project is voted in the TSC and everyone had a chance to review it and raise concerns. ## Alternative 2 ## Same as alternative 1 but uploading it to the wiki instead of github and review the proposal in the comments. The benefits I see: * Everyone in the TSC can participate in the review 24h a day, 7 days a week * We get opinions and comments from most of the TSC thus reaching what we want: "community consensus" * The person proposing the project will have space to think about the concerns and reviews that were raised * This eases a lot the participation of non-native speakers * The project could still go and present in the weekly technical discussion and of course clarify comments that were raised. However, the decision that it has community consensus will be made by checking that TSC members reviewed it and are fine with it * From process perspective, I think this is clearer and it scales better What you guys think? Any feedback is welcome! Regards, Manuel
Re: [opnfv-tech-discuss] [opnfv-tsc] Process Overview - How to Bring a Project into OPNFV
Jim, Thank you for the good questions. Regarding 1 "Step 2, third bullet", in the past practice, this situation hasn’t happened. Once there were comments, the author of project proposals would actively work with those who made comments, revised proposals and brought it back to community discussion. It may take several rounds as shown in examples. Eventually, everyone would agree and consensus achieved as shown in examples. Regarding 2 "Step 3", community discussion is never a decision making body. The essence of process is to foster community review and fulfill the metric of “Proposal has been socialized with potentially interested or affected projects and/or parties” [1] in a more realizable and achievable way. The final decision body is TSC. Please also note that many TSC members have been actively engaged in weekly technical discussion. To some extent, you can view it as a very informal, mini TSC discussion dedicated to a specific topic, but with broader participation from community. Regarding 3, I don’t know where it is defined. So appreciate if you can let everyone know where it is defined or required. As a good practice in process, if any community member is interested in promoting a topic, a full community discussion is expected. The community discussion can be started in mailing list, or in weekly technical discussion where many TSC members also participate. Once it is fully discussed and debated, then it is more reasonable to decide the next step, including bring it to and request for a vote in TSC if absolutely needed. The essence is to allow for full community participation and collaboration. Hope it helps. Thanks Bin From: opnfv-...@lists.opnfv.org On Behalf Of Jim Baker Sent: Monday, August 26, 2019 12:08 PM To: HU, BIN Cc: opnfv-tech-discuss@lists.opnfv.org; opnfv-...@lists.opnfv.org Subject: Re: [opnfv-tsc] Process Overview - How to Bring a Project into OPNFV Greetings Bin, Thanks for your sharing your thinking on how the process works. There are a couple items I wanted to ask you to address: 1. "Step 2, third bullet" uses the word "consensus", without a clear definition of what that means. Can one objection hold up a project request? 2. In "Step 3", the process defines the tech-discuss meeting as a gate OR decision making body? My understanding of the charter and founding docs is only the TSC has been granted decision making powers. I looked for documentation on the tech-discuss forum and was unable to find their specific responsibilities/operational policy? I believe it is a weekly forum for discussion of technical matters relating to OPNFV. 3. It is my understanding that anyone can request an item for vote by the TSC at any time. Is that your understanding as well? Jim On Sat, Aug 24, 2019 at 10:54 PM HU, BIN mailto:bh5...@att.com>> wrote: Hello community, Fresh blood is joining us. New project proposals are coming. I want to take this opportunity to give an overview of the process of bringing a project to OPNFV. 1. History of Process General Project Lifecycle is documented in [1]. Project promotion, and demotion, across states can only be done by TSC review and voting. During the reviews, the candidate projects are evaluated based on predefined metrics and KPIs. The target numbers may vary for the different levels. In order for creation review in TSC, proposals need to be emailed to TSC mailing list, and posted for two weeks. Various review criteria are available in [1]. In practice, after OPNFV was founded, we started to discuss new project proposals in TSC meeting on 10/21/2014 [1]. TSC spent 2 hours to discuss several projects. None could be agreed. TSC realized that a separate community discussion was needed to discuss project proposals, help proposals get mature before it could go to TSC for creation review. See minutes [1] regarding the topic “Recurring weekly calls on general technical content” and the action to create a separate weekly meeting on wiki dedicated to discussing project proposals. The first weekly technical discussion started on 10/30/2014, and history maintained in [3]. Project proposals were reviewed and discussed on 10/30/2014, 11/06/2014, 11/13/2014 and 11/20/2014. And the first project was approved by TSC on 11/25/2014 after that project got community feedback, made best effort to resolve community comments, made several revisions and discussed in those 4 weekly meetings. * Note: due to wiki upgrade in March 2016, all details of records of agenda and minutes of weekly technical discussion [3] before 03/24/2016 were lost. 1. Process Overview Step 1: create your proposal in wiki [4], and email to TSC and Tech-Discussion mailing list Step 2: schedule a community review and discussion in weekly technical discussion. The review process is documented in [3]. * This community review is the major part of the process, and it may take several weeks and iterate several rounds. *
Re: [opnfv-tech-discuss] [opnfv-tsc] Process Overview - How to Bring a Project into OPNFV
Greetings Bin, Thanks for your sharing your thinking on how the process works. There are a couple items I wanted to ask you to address: 1. "Step 2, third bullet" uses the word "consensus", without a clear definition of what that means. Can one objection hold up a project request? 2. In "Step 3", the process defines the tech-discuss meeting as a gate OR decision making body? My understanding of the charter and founding docs is only the TSC has been granted decision making powers. I looked for documentation on the tech-discuss forum and was unable to find their specific responsibilities/operational policy? I believe it is a weekly forum for discussion of technical matters relating to OPNFV. 3. It is my understanding that anyone can request an item for vote by the TSC at any time. Is that your understanding as well? Jim On Sat, Aug 24, 2019 at 10:54 PM HU, BIN wrote: > Hello community, > > > > Fresh blood is joining us. New project proposals are coming. I want to > take this opportunity to give an overview of the process of bringing a > project to OPNFV. > > > >1. History of Process > > > > General Project Lifecycle is documented in [1]. Project promotion, and > demotion, across states can only be done by TSC review and voting. During > the reviews, the candidate projects are evaluated based on predefined > metrics and KPIs. The target numbers may vary for the different levels. In > order for creation review in TSC, proposals need to be emailed to TSC > mailing list, and posted for two weeks. Various review criteria are > available in [1]. > > > > In practice, after OPNFV was founded, we started to discuss new project > proposals in TSC meeting on 10/21/2014 [1]. TSC spent 2 hours to discuss > several projects. None could be agreed. TSC realized that a separate > community discussion was needed to discuss project proposals, help > proposals get mature before it could go to TSC for creation review. See > minutes [1] regarding the topic “Recurring weekly calls on general > technical content” and the action to create a separate weekly meeting on > wiki dedicated to discussing project proposals. > > > > The first weekly technical discussion started on 10/30/2014, and history > maintained in [3]. Project proposals were reviewed and discussed on > 10/30/2014, 11/06/2014, 11/13/2014 and 11/20/2014. And the first project > was approved by TSC on 11/25/2014 after that project got community > feedback, made best effort to resolve community comments, made several > revisions and discussed in those 4 weekly meetings. > >- Note: due to wiki upgrade in March 2016, all details of records of >agenda and minutes of weekly technical discussion [3] before 03/24/2016 >were lost. > > > >1. Process Overview > > > > Step 1: create your proposal in wiki [4], and email to TSC and > Tech-Discussion mailing list > > Step 2: schedule a community review and discussion in weekly technical > discussion. The review process is documented in [3]. > >- This community review is the major part of the process, and it may >take several weeks and iterate several rounds. >- 2-week review period is the minimum period, i.e. a “necessary >condition”, but NOT a “sufficient condition” to bring to TSC for creation >review >- A consensus from community review is normally expected to recommend >the project proposal for TSC creation review. >- This process was designed by Founding TSC and its leadership, >started with the 1st project (reviewed for 4 rounds based on available >record), and a case study below. >- All approved projects have followed this process. > > Step 3: once TSC receives the recommendation from community review in > weekly technical discussion, TSC will schedule a creation review, and vote > to approve or disapprove this proposal. > > > >1. A Case Study > > > > Let us look at a case study based on available details of wiki record [3]. > The examples are 2 projects proposed in May 2016 – the earliest time that > details of wiki record are available [3]: > >- VNF Event Stream > - Announced in early May 2016 > - Discussed in weekly technical discussions on 05/12/2016, > 05/19/2016 and 05/26/2016 (3 rounds). > - Founders worked with community, revised proposal based on > feedback, resolved all comments and get consensus from community for TSC > creation review > - “The group reached the consensus to recommend it for TSC > Creation Review.” [5] > - TSC (2015-2016) did Creation Review on 05/31/2016 and approved it > [6]. >- Daisy > - Announced in early May 2016 > - Discussed in weekly technical discussions on 05/12/2016, > 05/19/2016 and 06/02/2016 (3 rounds). > - Founders worked with community, revised proposal based on > feedback, resolved all comments and get consensus from community for TSC > creation review > - “The group reached the consensus to