April 1st, 2009

Graphs are sooo very pretty

I think the effort Fabian and Dimitris efforts at trying to relate how we are growing as a participatory community are really great.

Both of them have shown that our packaging and translation communities are growing at a somewhat constant rate.
http://fabaff.blogspot.com/2009/04/verfugbare-pakete-fur-fedora-april-2009.html
http://dimitris.glezos.com/weblog/2009/03/08/flp-advancement-study

That's great to see.  I'll take steady growth over impulsive growth any day, its much easier to evolve processes under steady growth pressure than it is to go through large impulsive shocks that stress the system.

But what would be really interesting for me is to see if we are doing a better job of using that growing pool of human capital more effectively over time. That's a much harder question than just tracking growth, but its a much more valuable question from planning purposes.  If our processes are reaching an inherent carrying capacity (that we may not know about), we should start to see a relative drop in effectiveness of adding additional contributors. But what's a good marginal effectiveness metric?

Fabian's other graph hints at something for package maintainers.  The package collection in the development tree is growing faster than the maintainer pool. Is there a non-linear relationship there?  How are we doing in terms of co-maintainership and team building? As we add more manpower are we growing partnerships and teams inside our packager collective?  Is our package growth more related to a few 'hero' packages or is it more related to the summation of a lot of maintainers with small package workloads?

What would a metric for marginal effectiveness metric be for translators? That's a question I can't even hint at suggestions about, as I'm not involved in the process at all. But its something I think the translation guys should think about.  The growth of language coverage graph Dimitris already revealed might be turned into marginal effectiveness measurement for the transifex process.  Though there might be something else that goes to the heart of trying to measure the effectivness of individual language team.  As language teams get larger will each language team reach a carrying capacity?


Hopefully will see the graph meme bite Adam and he'll start to produce some trending associated with participation some of the QA and testing efforts as well.

-jef