[OpenLayers-Dev] Are there enough OpenLayers committers? (Was Re: Different lon and lat map resolution (and other issues))

[Christopher Schmidt]

On Fri, Nov 02, 2007 at 06:37:43PM +0100, Jachym Cepicky wrote:
> But the discussion went too concrete now. General question is: Do you
> think, current system is all right? Are there enough reviewers, who are
> checking the tracker regularly and communicating with the contributors?
> 
> As you probably understand: I do not think so.

First, let me state that I feel that no project can *ever* have enough
committers. There are always going to be more people writing patches
than there are people working to integrate them into the trunk codebase,
because writing patches is easy, and writing *good* patches -- or taking
patches and making them good patches -- is hard.

Currently, the OpenLayers 'trunk' committers list is a list of people
who have demonstrated, via their code, documentation, and bug management
to be competent developers on the OpenLayers codebase.

Those committers -- myself, Erik, Eric, Tim, Schuyler, Paul -- are the
ones who have routinely demonstrated a deep knowledge of OpenLayers, as
well as communicating regularly with the community what they are doing,
and why they are doing it.

Each of these contributors -- with the exception of myself, Erik, and
Schuyler, as co-founders of the project -- went through exactly what you
are going through. They wrote patches, did work, communicated with the
developer community, and established whether there was a reason for the
work they had done to be integrated into trunk.

Tim worked for several months writing patches and code before commit
access was offered to him. Eric has made more than a dozen patches, and
offered commentary on other existing patches (which is a large part of
what why I proposed he originally be given commit access). Building up a
relationship of understanding of the community as well as trust in 
judgement is an important factor in whether someone is given commit
access to the trunk of OpenLayers.

At this time, there are relatively few other members of the OpenLayers
community who have demonstrated that level of communication, and the
level of judgement, neccesary to make me personally feel comfortable
with granting commit access to them. (There are a couple who are very
close, but have a couple things that would cause me to hesitate.)

At this point in time, I think that many of us are recovering from 2.5.
It was a long, hard release on me personally, and since then, I've been
getting back into other things more so than OpenLayers. (I've been
developing some stuff for OpenStreetMap, for example, or putting
together things like http://hypercube.telascience.org/fire/ .) I know
that Erik has been working on a release here at MC, and I'm assuming
that Eric has been working at least in part on MapFish. 

We all have jobs. None of us work on OpenLayers full time, and lately
even the 20% time that some of us got is shrinking.

I think that it's a shame that there aren't more organizations who have
the ability to invest in OpenLayers. If we had more funds, we might be
able to free up some developer time to work on maintainence. (Might -- 
right now, I know that even if someone tossed a bag of cash at me, we
still wouldn't have a solution to the problem. Well, depending on the
size of the bag :)) However, that's the case. We have no paid staff, and
I've seen no one else who has enough experience with the project to
trust them with commit access popping up who is causing me to wish I had
the bags of money.

With that in mind, all I can say is that we're doing our best. I've
pushed a lot of development forward, and right now, I'm taking a break
because I need to recharge. At some point, I will pick it back up, and
when I do, I will review other patches. When others catch my eye --
others who have an intimate knowledge of the project dedicated through
long-term interest demonstrated by continued conversation with the
community and effort across various bugs to review, help fix, and so on
-- I will absolutely take on a goal of helping them hone their skills to
become trunk comitters.

Anyone who is only pursuing their own patches is unlikely to become a
core committer. Not because I have anything against that type of
development -- it's very important to have patches come back, and it is
unfortunate that we don't have unlimited resources to integrate them
all. They will not become a core committer because by only working with
your own code, it is not possible to gain a thorough understanding of
the OpenLayers code base. That takes a lot of time, and a lot of effort
on the part of a contributor, and to build up a lot of trust.

Trunk is pretty sacred. We try very hard to not put anything in there
that might break other people's code -- either with regressions or
anything else. That means that committers to trunk are trusted to a
great extent to know what changes they're making that might have affects
on other parts of the code base, and that knowledge can only be gained
through practice.

I'm hopeful that as time goes on, more widespread participation
throughout the  OpenLayers developer community will lead to more trunk
committers, and as that happens, we will have a more healthy and rapid
development project. Until that happens, we continue to move along --
reviewing the patches that are highest gain for the project, in an
attempt to build something that everyone can use with the resources we
have.

Regards,
-- 
Christopher Schmidt
MetaCarta