|RE: [eclipse.org-project-leadership] Time to trim inactive committers|
Here’s an example (for the MDT project).
The portal shows this:
Based on this information, one might think that Nick Dowler, Miguel Garcia, A. Jibran Shidiqie, and Yves YANG are inactive, but this is because the components on which they are committers have just recently been provisioned.
Dash shows this (http://dash.eclipse.org/dash/commits/web-app/active-committers.cgi?project=modeling.mdt):
Here, we can see that only some of the cells beside the committers mentioned above are shaded; as far as I can tell, this corresponds to the fact that they’ve only been committers for that long. So, as project lead, I would consider only gxie, mli, and pyue as truly inactive because all of the cells beside them are shaded.
The “inactive” status isn’t entirely accurate for projects/components that are new (i.e. it’s possible that committers haven’t committed anything yet because the project /component was only recently provisioned). It seems that Dash has the information to make this distinction (based on the shaded cells that appear in the active committers report); could the new portal box be modified to reflect this distinction?
Could you point me to an example project so that I can
understand what you're suggesting? I agree that the source code commits
activity information is only an indication, a hint, at the activity or
inactivity of a committer - as project lead, you know much more about the
project than any automated tool could.
Another observation is that the election participation information isn’t as useful for projects with components since not every committer is eligible to vote in every election (mind you, this will be addressed if/when the proposed changes to the development process are approved; when do you expect that might happen, by the way?).
Again, you are correct and you
are the final arbitrator of committer activity. The current Foundation database
(as you point out), does not have enough information to determine the committer