Cleaning up MicroProfile inactive committers

[John Clingan]

We've been an Eclipse Foundation project for a few years now. Since then we have had folks become committers but then "life" happens. Perhaps retirement or a role change. Not that a creating a MicroProfile working group is a forcing function per se, we should review our committer roll (see "Who's involved" and push vs pull vote list). Looking around other Eclipse projects, this happens occasionally (example) (example) (example).

According to Mike Milinkovich in a Working Group Charter Proposal draft comment:

The definition of active used by the Eclipse Foundation reporting systems is a minimum of 1 commit in the preceding 9 months. YMMV.

The Eclipse Development Process committer section says:

 Unless otherwise specified, "an extended period" is defined as "no activity for more than six months".

So, it looks like there is not a fixed requirement placed on projects on a time frame and we have some flexibility. Questions:

1) Do others think it is the right time to clean up the committer list?

2) What should the committer activity threshold be? 6 months? 9? 12? IMHO, anything beyond 12 is too long

3) How often should we review for inactive committers? Annually?

Note, an inactive committer can be re-added at a later time via a successful vote using the committer election process.

[Kevin Sutter]

John,

There are more things to consider.  First, the reference that Mike made when commenting on our MP WG Charter proposal was in response to electing a Committer rep and how to ensure that they are "active" enough to be considered for election and to be an elected member of the Working Group.  We could spell that out as well in the Charter, or we could assume that the community wouldn't vote for someone that is not active.  I prefer the latter.

Also, how are we going to define "active"?  Some of us committers (myself included) do not have many commits on a regular basis.  Lately, speaking for myself, I've only committed code when helping with one of releases.  But, I am still quite active on the Hangouts and the Mailing list and Voting.  So, I'm assuming that type of activity would count as well?

I do think an annual review of the committers is worthwhile.  After we define "activity", then I would say that we could check for activity on an annual basis for the past twelve months.  If no "activity", then we attempt to contact the person first and let them know what's happening.  If there were extenuating circumstances, then we give them another year grace period.  Otherwise, we retire them.

Those are my thoughts. 

-- Kevin

[Emily Jiang]

My understanding for the "active" context is same as Kevin's. The requirment is a criteria for a committer rep to be elected.

I don't feel there is a need to reveiw the comitter list to make sure they are active. We discussed this previously here, but did not take any action, because most of open source projects have lifelong committer status. Comitters could be dormant for some time and then come back to contribute again. I don't see there is a need to review the list. As long as we make sure the comitter reps are active (either based on Mike's definition or agreeing with Eclipse Foundation with a slightly different criteria), we should be ok.

Emily

[John Clingan]

The Eclipse Foundation (via the Eclipse Development Process[EDP]) does not advocate lifelong "inactive commiter" commit access to a project, and defines a set of responsibilities for committers, like:

Committers are required to monitor the mailing lists associated with the Project. This is a condition of being granted commit rights to the Project. It is mandatory because Committers must participate in votes (which in some cases require a certain minimum number of votes) and must respond to the mailing list in a timely fashion in order to facilitate the smooth operation of the Project. 

I don't plan to push this issue too hard right now given other higher-priority work (and this being a "check the pulse" post), but this will need to be addressed as it is clearly defined in the EDP.  As I mentioned, ex-committers can come back for a subsequent committer vote after showing recent activity in the project.

Open to additional feedback.

[John Clingan]

On Wednesday, April 15, 2020 at 7:21:31 PM UTC-7, Kevin Sutter wrote:

John,

There are more things to consider.  First, the reference that Mike made when commenting on our MP WG Charter proposal was in response to electing a Committer rep and how to ensure that they are "active" enough to be considered for election and to be an elected member of the Working Group.  We could spell that out as well in the Charter, or we could assume that the community wouldn't vote for someone that is not active.  I prefer the latter.

Also, how are we going to define "active"?  Some of us committers (myself included) do not have many commits on a regular basis.  Lately, speaking for myself, I've only committed code when helping with one of releases.  But, I am still quite active on the Hangouts and the Mailing list and Voting.  So, I'm assuming that type of activity would count as well?

I do agree that we need to discuss what "active" means, which implies "clarifying" what committer means in the context of MicroProfile. I don't know if we have that ability, to be honest. The EF, from what I gather, puts a lot of emphasis on commits (a-la the name "committer"). I do need to R&D this more across other Eclipse projects to see if this is a historical emphasis with more recent modifications/perspectives. A background task for now given other priorities. I may bring this up on the next Live Hangout to see what guidance the EF folks may have.

[Amelia Eiras]

This topic is interesting as it applies to setting up the bare baseline on current WG work.

My suggestion is to add it to the topics for discussion. 

Something that I continue to be asked by active Microprofilers who are not committers yet is how long is the bare minimum to qualify for nomination. I don't have the answer for that but certainly is not 3 months. 

Informal-Guidance on what to expect when coming to Contribute and prioritize collaborating into MP is a +1 on my book. 

Historical Committers is a thing at EF Projects. 

THIS needs discussion without rushing, I am most interested on hearing from active Contributors. FYI- bias, 

[Mike Croft]

+1

 

Something that I continue to be asked by active Microprofilers who are not committers yet is how long is the bare minimum to qualify for nomination.

 

This is a big part of the reason I would like to be removed as a committer. From the start of this year I've been able to follow discussions better so I feel informed enough to contribute to votes etc but, since there are more active people who are not yet committers, it doesn't really feel right for me to vote anything other than abstain at the moment, so my committer status is somewhat moot.

I think the last time this was brought up someone (maybe Heiko?) pointed out that they did not want to lose their committer status since they expected to prioritise work on MP again in the near future.

Perhaps the main problem which needs to be resolved is that there is a high bar to clear to become a committer. As far as I know, there aren't regular scheduled votes for new committers and no definition of what qualifies someone to be a candidate. If this sort of thing was in place, then it's not a big leap to assume that someone in the same position as the person I mentioned would have some kind of fast-track to be re-elected as a committer. There is just one historical committer at the moment - John Ament. If he wanted to return, I would expect no-one would object.

Of course, this suggestion glosses over a *lot* of important detail (and we've all seen recently how the devil is most certainly in the detail) such as our definition of "active", how regular to make election votes etc, but I expect that it is very possible that we could end up with a much more dynamic group of committers and, therefore, a higher percentage of committers registering their votes.

[Werner Keil]

Guess the whole move towards a WG with a split into smaller parts like most TLDs might make that a lot easier.

Now https://projects.eclipse.org/projects/technology.microprofile/who is not very readable with a large number of committers where it isn't even clear who worked on what unless you go directly to Github.

As mentioned only John was declared a historical committer but guess there should be others.

Of course if some committers got elected into a WG committee, their activity might shift to other duties like voting and some could not always commit actual code at the same pace, while they still play a vital role for the project and ecosystem.

Werner

[Mark Struberg]

With emphasis on pure committs we do not honour all contributions equally.

Many people are active on the organisational level. When did you and Kevin for example commit code?

But of course you both do an excellent job stearing the community!

And the same happens for others wo are very active in the community, write documentation, review patches, give feedback on chats and meetings.

'Active Contribution' is not just counting committs.

LieGrue,

strub

[Amelia Eiras]

Dear Mark, miss you, so sending you hugs virtually! 

+1 to Mark on MicroProfilers' active contributions are not just counting commits. 

Happy June MicroProfilers

The Standalone WG Chapter draft doesn't talk at all about policing this ecosystem Committers good standing. 

Though at the beginning that section was there it got removed b/c we don't want to limit earning a commit by "just code", it is too limiting and not reflective of what we have achieved in the last 4yrs. 

Clarification on Werner's June 12th exchange. John Ament's historical status.  John was never declared a "Historical Committer" by MicroProfile. 

He changed employers so his email associated with MicroProfile changed. He will update his internal profile when he sees it fit and become once again active when he deems it appropriate. 

The coolest thing about Mike Croft's follow up is that as usual, you beautiful wonder about your commit access & share it with everyone in this forum. 

I think we need more of that candid feedback loop when people like Mark or you or me can say to Croft, you are valuable & needed here. There are so many ways where you can help MicroProfile, choose & pic, it is not only code. :) 

Lastly, let's not forget that we have on purpose avoided the removal of commit access to those who were appointed, by the founding Partners via  direct Committer level in Q1 of 2017 when we moved MP to the Foundation. WHY?  b/c the expectation then and now is for each individual to have a choice to continue to be involved in this project, even if not a part of the Partner's employment records.  

Merit in MP today is earned. 

When the WG is set into place and a company chooses to become a Corporate member, entity will have access to seat in the Steering/Spec committee that is 1 body. 

We- in Tomitribe,  pushed for that Corporate member to have at least 1 Committer in the project before being able to join formally/officially the WG group as an entity.   That was shot down during the working group calls... Still seems to be pending THIS community attention. 

- As a co-founder of this project and the one who actually was tasked with the transfer of MP to the EF, document and transfer completed in January 2017, I would recommend for everyone in this ecosystem to pay close attention to how you want this community to grow. 

I believe that merit via actions helps develop respect, trust & enhances open source collaboration. I believe that MicroProfile's success has been possible because of how it has managed to protect such core belief. I am against anything that will cut or deviate for such mentality. The "pay to play" doesn't work. The "I am your boss here" doesn't work.  If anyone here earns a commit status, that she/he stands on its own feet as a MicroProfiler that first and foremost protects its own OSS signature beyond the nice employer who hopefully approves for time to invest during work hours to MicroProfile. 

--- 

Cheers,   

[Werner Keil]

It seems fine to wait till the whole WG thing is established, and that should require separating individual specs and features into subprojects (at least I don't know of any WG be it Science, IoT or Jakarta EE that throws all the projects into a single basked like the current MP setup) 

Whether each of these new projects keeps listing everyone as "active" committer or lists those who have not committed in a year or more (like Mark) as historical committers, could be up to the invidual project as long as the charter does not say otherwise, but it seems fair and more honest to also show who is actively committing. Take Jakarta NoSQL: https://projects.eclipse.org/projects/ee4j.nosql/who

Leonardo must have self-declared as historical committer as well but that does not mean his account or write access is blocked, because he just made a few commits, so why not be honest in cases where a committer is unable to contribute anything.  

Should someone be restricted to input to the spec or via this list on a regular basis then why not refrain from the "historical committer" move in some cases, but just as a vanity badge keeping inactive committers for years would not send a good sign.

I am contributor to several projects without working on the actual code much, MP Metrics here is probably the best example, STEM which I helped only to localize the GUI many years ago but presented the project long before COVID 19 around the world was another way of contributing, which is why I am also listed in the STEM team on the project wiki.

So it may be a case by case question but I would not keep everyone forever.

Werner

[Wayne Beaton]

Whether or not the project team decides to split the project into multiple projects, and how we configure that is a separate issue from the working group. There is no requirement inherent in the working group process to do so. I can argue both sides of "should we split up Eclipse Microprofile?" But... like I said, this is a separate issue.

HTH,

Wayne

--
You received this message because you are subscribed to the Google Groups "Eclipse MicroProfile" group.
To unsubscribe from this group and stop receiving emails from it, send an email to microprofile...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/56c70769-7e0b-45f1-9ff4-8d1b58cef7c1%40googlegroups.com.

--

Wayne Beaton

Director of Open Source Projects | Eclipse Foundation, Inc.

Join us at our virtual event: EclipseCon 2020 - October 20-22

[Werner Keil]

It would certainly make this project view https://projects.eclipse.org/projects/technology.microprofile/who

a lot less messy and easier to read.

EE4J https://projects.eclipse.org/projects/ee4j/who still has the big picture, but you can drill down, something that only seems to come with a separation, same as the platform view https://projects.eclipse.org/jakartaee/releases/8 that show the ingredients quite well, but first let the WG work out, then the WG may find out, if it wants to keep the monolith or be more "micro" also in that respect.

Werner

To unsubscribe from this group and stop receiving emails from it, send an email to microp...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/microprofile/56c70769-7e0b-45f1-9ff4-8d1b58cef7c1%40googlegroups.com.