Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [ee4j-pmc] [ee4j-build] EE4J Parent pom.xml update

The solution is to "heal" projects by replacing inactive committers by active contributors. If there are not even active contributors, a solution could be that committers from healthy projects chime in for a limited time at will of the project lead (if active) or PMC (otherwise). But in no cases I see any good reason for removing branch protection. It would simply lead to anarchy and open doors for bad code.

-Markus

 

From: ee4j-pmc-bounces@xxxxxxxxxxx [mailto:ee4j-pmc-bounces@xxxxxxxxxxx] On Behalf Of Christian Kaltepoth
Sent: Samstag, 3. November 2018 08:41
To: ee4j-build@xxxxxxxxxxx
Cc: EE4J PMC Discussions; stuart.w.douglas@xxxxxxxxx
Subject: Re: [ee4j-pmc] [ee4j-build] EE4J Parent pom.xml update

 

Hi all,

 

I understand the problem. It happens than projects are not healthy. It indicates that committers and project leads are not interesting in projects and not doing their job well. The problem is not in branch protection. The problem is that people are not responding. With branch protection on, all PRs are visible and committers are free to approve or reject them. Code review is definitely a good practice. Possibly you don’t realise how many bad code were fixed during the review process. Reviews must be required for all committers or completely disabled. I am voting against direct commits. It may add other kind of problems. Also, please remember that EE4J_8 branch protection is added to filter out all changes except critical bug fixes

 

I fully agree on this. Code reviews and branch protection are great. But they only work for healthy projects. Unfortunately many projects contain a huge number of inactive committers. Even for JAX-RS, which seems to be one of the most active EE4J projects, there are about 14(!) committers (>50%) who don't vote on pull requests or answer to any mails on the mailing list. I'm actually not sure if they receive these mails and notifications or if they are simply ignored. I don't know why they were added to the initial committer list. JAX-RS can handle this, because there are enough active committers, but other projects are running into huge problems because of this.

 

BTW, the EDP states in section 4.7:

 

Committers are required to monitor the mailing lists associated with the project. This is a condition of being granted commit rights to the project. [...] A committer must not be unsubscribed from a developer mailing list unless their associated commit privileges are also revoked.

 

And:

 

A committer who is disruptive, does not participate actively, or has been inactive for an extended period may have his or her commit status revoked by the project leads. Unless otherwise specified, "an extended period" is defined as "no activity for more than six months".

 

So these committers are not following the EDP and it would be absolutely fine to remove them from the corresponding projects (although it is the decision of the project lead).

 

Christian

 

--


Back to the top