Re: [gentoo-dev] Glep 48 update (as nominated for next meeting)

[Jeroen Roovers]

On Sun, 30 Jan 2011 20:39:15 -0600
Donnie Berkholz <dberkholz@g.o> wrote:

> We're here to create an awesome source-based distribution, not
> pretend we're United Nations and the U.S. government all rolled into
> one. =)

I've been observing for a few years now how sometimes a developer leans
toward a more corporate style over time.

We're an open source project, most of us are volunteers, and normal
corporate policy doesn't hold here, and yet now and again I see
proposals to technically or socially improve the project in a way that
comes down to three actions:

1) invest - money, time and/or other means; and/or

2) regroup; and/or

3) grant privileges for a special subset of people to act upon another
   subset of people.


= 1. The Leg Up =

Of course, investing means you have, well, the means to invest. Money
and goods usually aren't the means to invest in a project like ours,
but many requests seem to say that a particular ailment will
magically disappear after we throw another five person-weeks at it. We
do not have these resources, and we have no accounting for them, and we
do not set strict targets that way, so that's the entire corporate
model out the window for us. Scarcity of resources is exactly what makes
an open source project progress - you leave all the problematic bits
open to see, and some hapless user will ultimately come along and fix
it for you, or you finally find the time to do it yourself, or you
talk enthusiastically about it to someone else and she does it for you.


= 2. The Escalator =

Arranging developers in new hierarchies is not going to fix any
problems. In a corporate environment, you can simply cut out some
middlemen, put some new coordinators in place to oversee some team or
teams, and generally fire a dozen here and hire a dozen there as
needed, and nobody hurts (for very long) as it's usually all a big
shuffle among the same population of workers.

It's a good thing to have an escalator for conflicts (of interest or of
a technical nature), and we've luckily had that for years. Sometimes
the escalator needs fixing, but every time you fix it, some people are
going to walk away, some others will retreat into the safety of toiling
in some badly lit backroom where nobody bothers them, and a few will
simply decide that publicly discussing new ideas will just get them cut
down by their peers again. Regrouping is a kind of investing, and since
you can't even tell people what to do, it's not very wise to start
arguing to change the hierarchy - it's probably a better idea to set up
a new project, outline what needs to be done, and try to simply attract
the person-hours you need until the project has achieved its goal.


= 3. The Elevator =

Rearranging power in the sense of giving special privileges to
particular people (a direct line to the president, personal notification
when our team scores, an order to go out and buy some new socks, size
12, the ability to quickly decide when someone else's privileges
should be revoked) is another very good way to not treat volunteers. It
works very well in a corporate setting, because you can fire people or
even sue for damages when they abuse their privileges, so most works
wouldn't abuse them just like that or at least make sure their use of
the privileges is somehow accountable. Similarly instilling fear in
volunteers has the opposite effect.


The ideas that do come through in volunteer projects usually excel in
simplicity. They use the same or fewer resources than the conflict they
intend to end. And nobody ends up with better privileges than others,
or executive powers greater than those of their peers. You just see more
people volunteering more work.


     jer