» Eclipse Community Awards
Eclipse Community Awards
Update: I had some feedback that I should have separated the positive congrats to Benjamin from the criticism about the process. Apologies if he or others felt offended, this definitely wasn't my intention, and again Benjamin has some merit that should make him a committer more than a contributor award winner.
My former colleague Benjamin Cabé has been nominated Top Eclipse contributor for 2009. Congrats!
Now although I'm very happy for him, this awards thing makes the community guy in me a bit uncomfortable. I know Benjamin does great things and has a lot of merit, but don't other contributors also have a lot of merit? Does he really deserve this award more than the dozens of other active contributors? Has he been chosen because of his work, or because this work happens on more visible projects, or because he's more vocal than other people and thus more visible?
This introduces a kind of competition or ranking among contributors which I don't think is really healthy. Open source organizations are meritocracies, but your merit is defined by what you do, and not by how you compare to others. Becoming a committer on a project doesn't mean that because you were elected, other won't be elected if they also deserve it.
This award system goes against that spririt and can have bad side effects such attracting people that just want to compete for the award, or on the contrary leaving behind people that do a lot of less visible but necessary grunt work.
- Login to post comments
I mean no disrespect but I find your post dogmatic and misguided. A meritocracy is based on individual ability and achievement. Benjamin Cabé getting an award based on his contributions will not necessarily introduce competition or ill-will among his team-mates. If it did, then there is probably something else fishy going on.
Denying the value of individual merit is incompatible with the very idea of meritocracy. You can't have it both ways.
I think you have misunderstood Ceki, or maybe I wasn't clear. I don't deny the value of individual merit, rather the contrary.
Meritocracy is all about individual merit that allows someone to climb the various steps in an organization (contributor, committer, member, etc).
What I don't like here is that a choice is made among individuals, and thus requires comparing them rather than only considering their individual contributions independently from others.
> Meritocracy is all about individual merit that allows someone to climb
> the various steps in an organization (contributor, committer, member,
> etc).
Precisely, there is no etc. After you are a committer, becoming an ASF
member entitles you subscribe to certain member-restricted mailing
lists and to electing the board. You'd agree that is a pretty limited
"career" path circumscribed to promotion within the ASF as an
organization. What about promotion within the project the developer is
directly interested in? In the ASF, all PMC members of a project are
peers. Consequently, you will have some PMC members which have
contributed more and sometimes much more than other PMC members. As
long as every body plays nice, this will not have any negative
consequences, most committers will tend to sustain the opinions of the
most active developers. However, if certain committers stop playing
nice, the project will quickly decay into continuous bickering and
eventually into chaos.
And that's where meritocracy as practiced within the ASF shows its
limits. The current ASF dogma refuses to recognize that there are more
active developers than others. Those more active developers should
somehow carry more weight than less active developers. If they did,
then the project would be governed as a meritocracy. However, at
present time, if a committer will be committing political suicide if he
or she dares to declare that he or or she is more active than
others. That is not a meritocracy.
Another case in point. Some projects pat themselves on the back for
removing author tags from source code, a practice supposedly
encouraged by the ASF board, that is yet another practice in
contradiction with the notion of meritocracy. Given that peer
recognition is an important motivation for many individual
contributors to open-source, removing author tags is tantamount to
day-light robbery.
I understand that the ASF does not do any of these things out of
evil. It means to emphasize community over code. However, there should
be some recognition that by preferring community over code, it is also
in opposition with the principles of meritocracy.
You have some good points, although I don't think the opposition is as strong as you say.
There is at Apache, as in every human community an implicit non-written classification of people according to their actual actitivy, personality and achievements, that will lead some people's opinion to be more important than other's even if they have the same formal position in the organization.
Now what I think is not good is to start for formally rank individuals in a volunteer organization. One award for one person isn't that harmful, as arbitrary as it could be, but imagine if we wanted to start defining a "merit scale" for all people. This would very likely lead the entire organization to blow up.
> There is at Apache, as in every human community an implicit
> non-written classification of people according to their actual
> activity, personality and achievements, that will lead some people's
> opinion to be more important than other's even if they have the same
> formal position in the organization.
As long as those implicit non-written classification is aligned with the magnitude of contributions, the organization will be run as a meritocracy. However, bringing in other factors such as personality introduces a new dimension which might or might not be meritocratic, but admittedly I am merely nitpicking.
> Now what I think is not good is to start for formally rank individuals
> in a volunteer organization. One award for one person isn't that
> harmful, as arbitrary as it could be, but imagine if we wanted to
> start defining a "merit scale" for all people. This would very likely
> lead the entire organization to blow up.
Indeed, it is a difficult problem. Any metric for ranking contributions would be an earthshaking change, with the potential to destabilize the ASF. In a recent blog entry [1], I asked if one could imagine such a ranking mechanism.
As most metrics, any formal ranking metric could have unintended and
perverse results.
[1] http://ceki.blogspot.com/2008/12/is-apache-meritocracy.html