lucy-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mattmann, Chris A (388J)" <>
Subject Re: [lucy-dev] [DISCUSS] Graduation (was Re: [lucy-dev] All dependency licensing issues resolved)
Date Fri, 28 Oct 2011 13:50:31 GMT
Hi Torsten,

On Oct 28, 2011, at 12:34 AM, Torsten Curdt wrote:

>>> Right. So then there would be just no one committing?
>>> Please explain how that should work from your POV.
>> Committing isn't the only contribution.
> That doesn't answer the question and you know it.

Let me try and better explain then. What question are you asking?
Are you saying that if a community only has 1 major person writing 
code then it won't be a success? If so, then I'd disagree with you 
and my statement above holds, and I believe answers the question.

The Apache mantra is "community over code", and 3 +1s is what it 
is all about. On a project where one of those +1s is a coder/hacker, 
another is a release manager, and another is someone that writes 
documentation, then no I don't see a problem with that. If the project 
can muster 3 +1s for new committers, for releases, for etc., then 
it's absolutely fine. 
>>> Exactly. So let's give it the time it needs. What's so bad if it takes 3 years?
>> That's not a sustainable model for the incubator.
> So far we have been OK. We had other podlings that also took their time needed.

There have been several discussions over the past 9 months about Incubator 
projects that have been alive for years and in all of those cases the projects 
are either being encouraged to graduate (e.g., RAT [1]), or to be retired (Alios [2], etc.).

Growing projects that live in the Incubator for years is not a sustainable model. 
Projects should have goals of graduating as soon as they are ready. That's why we 
have monthly Incubator reporting for the first 3 months, and quarterly after that. Think 
about what quarterly means. And think about what reporting means. It means that 
at worst, every 3 months, the IPMC chair reports to the board on the _current status 
of the podling towards *graduation*_. That's what reporting is about. And, that's 
what mentoring is about. Mentors should actively check with their podlings during 
those 3 month reporting cycles to make sure they are on path to meet their goals 
for graduating. 

>>> Quoting from the incubator guide at
>>> Snip...
>>> That's just not the case yet.
>> Says you with absolutely 0 merit in this community besides lurking.
> As a member of the ASF and a member of the incubator PMC I am voicing
> my concerns.
> You just call me troll and don't adress my concerns at all.

You are welcome to your concerns. I am also welcome to my statement that 
I think Lucy is graduating. Both of us are Incubator PMC members, and 
have a single VOTE towards Lucy's graduation when the time that the VOTE 
is cast, at which point it's at least 3 +1s from the PMC, and then simply more +1s 
than -1s at that point after that. 

I have a few real concerns right now too, which led me to start this thread:

1. Those doing the work in Lucy (and yes, it's more than Marvin, and to reduce it
to just Marvin is a joke) don't have binding VOTEs on the work they are doing right 
now. Joe Schaefer stepped up and rewrote the build system 
with Marvin's help. Peter just RM'ed the frickin' project (NICE Peter!). We've had 
new committers (e.g., David, Brad). We've had releases, multiple of them. Heck, 
even I RM'ed one of them. I want to get the people doing the work to have 
binding VOTEs, and the best way to do that is to make them a committee, that is 
self-directing and managing, make them a PMC. Have they met the criteria for 

2. In my mind, anyone that goes back, reads the Lucy proposal [3], and 
reads out monthly reports for the last 1 year and 3 months will see that 
the checklist has been checked off from what we originally set out to do. 

3. Not gating or tying graduation to a set of development goals, and thinking 
that we are still X development goals away from graduation. Guess what, 
development can still occur while you are graduating! In fact, I'd encourage it.
In Tika and OODT, we didn't shut down SVN when we started graduating. 
We didn't shut down the mailing lists. We just kept working. Same in Nutch.
Same in all the projects I've watched graduate, and mentor over the years, 
including some pretty huge ones at Apache. 

4. A TLP allows the Lucy community to continue doing what they are doing. 
It's not TLP and then everything is stopped, we pat ourselves on the make, 
and we're 1.0. TLP != 1.0. TLP simply says that the commumity has shown 
that it can: (a) self-govern; (b) elect new committers; (c) make releases of 
ALv2 licensed software, with proper license vetting, etc.; (d) show diversity 
in terms of organizations, etc. Lucy has shown all that. There is no requirement 
that you need to be at 1.0, or have met N development goals before you 
can graduate. 

I hope that clarifies my position.



Chris Mattmann, Ph.D.
Senior Computer Scientist
NASA Jet Propulsion Laboratory Pasadena, CA 91109 USA
Office: 171-266B, Mailstop: 171-246
Adjunct Assistant Professor, Computer Science Department
University of Southern California, Los Angeles, CA 90089 USA

View raw message