I think my position was clear in my long
message ... especially after Wayne's clarifications ... but the short version
I am quiet sure Buildship is a fine
project, in content, and the way it is ran, but some things are not well
documented, for a graduating project, and more important, I really do think
you need coaching to "become a full Eclipse Citizen". I'd
hate to see Buildship be mostly an "external project" that just
happens to host their code at Eclipse -- and even that might be ok, if
that was the intent, and stated explicitly. (Not sure that's really possible,
or what it would mean ... just saying I am more flexible than I may sound.
Flexible on everything except openness and transparency. :)
Etienne Studer <etienne@xxxxxxxxxx>
Tools PMC mailing list
06/22/2015 01:34 AM
Request for approval of Buildship 1.0
Whether we graduate now or remain in incubation until
Buildship is included in Mars with SR1 in fall is not critical to us.
But, we really want this to be 1.0 - not 0.9. We feel
the project is mature from a technical perspective. We have added certain
types of high-level functionality early and since then we have improved,
polished, and fine-tuned that functionality based on our own experience
and based on the feedback from the community. This is reflected in new
Gradle versions, new Tooling Commons versions, and new Buildship versions.
Personal feedback/input/reports came from companies/people like XText,
Vogella, RedHat who I respect highly. We are ready for 1.0.
I hope other PMC members take a look
at this request, and chime in, but I'll make some high level comments.
I certainly have no objections to the project releasing, but given the
version number of "1.0", I assume this review is not only for
release, but also to graduate from incubation.
Do you, (project lead and mentors) really think the project is ready to
be a graduated, "stand alone" project, without continued mentorship?
>From what I can see in the write-up, I don't really see it.
I will admit I am biased, since I'm not sure how any project can prove
"ready to graduate" after only 2 months at Eclipse.
And, this project is obviously a special case, and perhaps does not fit
into any Eclipse criteria (reference points) that I allows me to make a
Here's some things I found anomalous:
- 69 bugs open and many of them fixed (that's good) but seemed that most
were opened by 2 or 3 people (not strong evidence of "a community",
in other words).
[I know in the past, you've claimed to have a community "outside of
Eclipse" ... and if that's part of the justification, then details
should be given in the documentation,
such as pointers to another bug tracking system? (Not sure that's good,
to have another build tracking system ... just saying, if there's something
"outside" of Eclipse that demonstrates,
readiness to graduate, then I think should be included.]
- You had 4 "milestone builds" but they were a week apart. That's
not normally what we would call a "milestone" at Eclipse, so
wonder what your definition is. Do you plan to continue to have weekly
If so, would be good to cover that, and define what that means, in your
Given our history, about the release train, I was afraid to say anything,
for fear of giving the wrong impression of being "not supportive"
or "in the way".
But, I do really do mean these remarks as constructive.
And, I feel a responsibility to "treat all projects equally"
and these are questions and things I would look at for any graduating project.
And, since no other PMC members said anything ... thought I should say
Hope you find my comments at least a little helpful, and again, would welcome
comments from other PMC members.