human sheparding ( was package shepherding )

[Jef Spaleta]

Alexander Larsson wrote:
> Also, there is all this talk about "Management" with I just don't
> understand. Redhat is a company with managers, but Fedora is about free
> software development. People are supposed to work on it because they
> like to, to scratch itches, to learn stuff, to have fun. If all you want
> is to have a Redhat fix bugs for you the best way do to that is to spend
> money on Redhat Enterprise Linux, that means we'll get more paid
> developers and will be able to fix more bugs (in fedora too, fedora is
> what will become the next RHEL version)

You've read the leadership draft document right? Now talking about
micro-management of developer time maybe isn't something productive to
talk about (unless i knew who your manager was and had a way to bribe
them into telling you to do exactly what i wanted partitioned into 15
minute blocks of time). But certainly there is room to talk about Red
Hat's need to manage in broad strokes, the large human capital resources
represented not only in red hat employees, but also the enthusiastic but
maybe somewhat out-of-touch volunteers. This is not going to be a very
successful community experiment, if Red Hat contains full control of the
high level policy and decision making that directs the distro...but
doesn't do a very good job of communicating and directing volunteers.  

Everyone involved in a process gets paid in different ways. Volunteers,
get paid either through control to shape the direction of the project or
through access to information, and both forms of metaphysical volunteer
payment seems to be lost in the mail when it comes to the Fedora
project. Personally I'd say Red Hat misjudged the cost of sending out
those checks, and didn't spend enough money on stamps. Or dropping the
cutesy allegory... I don't think Red Hat has allocated enough internal
resources aimed directly at the problem of building and managing the
volunteer community aspects of the project. A few very motivated and
extremely technically knowledgable people have broken through the barren
wastelands of communication separating the nation-states of Fedora users
and Fedora developers, but I wouldn't call that sort of vision quest
something a majority of volunteers are skilled enough to do. There is a
need here to manage aspects of communication with volunteers, to take
the day-to-day burden of dealing with common volunteer/developer
communication off the shoulders of the developers. 

And I'm not talking about individual developer time, I'm talking about
someone specifically whose job it is to act as a volunteer coordinator.
If the Fedora project, as a community project, is really important to
Red Hat corporate, they will allocate manhours specifically to address
the horrible task of figuring out how to recruit and retain community
volunteer interest in a project whose overall goals are set not by
community..but by Red Hat. If Red Hat does tap someone to be a volunteer
coordinator, and they make the mistake of not hiring someone with
training and experience dealing with volunteer husbandry in the brick
and mortar world, might I suggest they look at the first 18 or so pages
of this volunteer management handbook pdf:
http://tinyurl.com/33ban

The section outlining on how to start off on the wrong foot, in terms of
volunteer organization..reads like a history of the Fedora project.
Read the first 18 pages or so and grimace. Actually...everyone involved
in Fedora development should read the first 18 or so pages of that
volunteer workbook. The majority of the pdf, concerns itself with
template tools as example tools to address many of the issues raised in
the first 18 or so summary pages.

In the darker moments of my day (when I'm calibrating optical beam paths
in the dead of night) i sometimes wonder what exactly is the point of
opening this process up to the 'community.' Is it all about the binary
bits? Or is the real potential value not in the bits at all but in the
building of a vibrant community process, where an average fedora user
(who lets be honest is going to be less technically inclined than users
of some of the more prominent community development model distros) can
learn how to contribute. People just wandering through, scratching their
own itch, in my mind, isn't a really useful definition of
community...its sort of like calling the mass of people standing in line
at the DMV a community.  Or too but it another way,
idle technical proficient manpower at this point is probably in short
supply, since anyone who has an itch to scratch already has a
sourceforge project listing for some sort of mp3 playing something or
other.  Fedora either has to tap into existing volunteer manpower that
is being used for other things by offering those people something of
value inherent in the fedora contribution process, or by developing and
nurturing less technically inclined user so that they can become active
contributors.  

-jef"Communities are built through proactive leadership reaching out and
building a process by which people feel empowered to be responsible for
the project. I'm still very hopeful such a community can be built from
the people standing in line at the Fedora's DMV"spaleta