[IAEP] maintaining activities (was Re: Too many open fronts?)

[David Farning]

>>>> We have to focus on polish, on getting something useful in the hands
>>>> of children. On making a difference there. We cannot leave XO-1.5
>>>> unfinished. We have to hear the SoaS crowd when they point out that
>>>> very few activities are shippable.
>>>
>>> I agree in that focusing is very important. But paradoxically, you can
>>> be in a situation in which individuals are focused on their work but
>>> the organization as a whole isn't focused at all.
>>
>> One possible approach is for Sugar Labs to focus on being an
>> innovative upstream while downstreams like Activity Central, SoaS, and
>> deployments focus on the polish. Activity Central is pulling
>> deplolyment level polish into Sugar.
>
> Will all those downstreams work on polish separately or will they
> cooperate and pool resources in a common place?

The best answer I can give is "It depends."  Typically, upstreams and
downstreams tend to branch and merge.

A good example is the kernel. Over the last 5-10 years many projects
and companies have tired to branch... and maintain that branch.  They
realized that maintaining a branch is _really_ expensive.  The
remaining (successful) projects and companies work very hard to stay
in sync with the upstream kernel.

FWIW, at Activity Central we have no infrastructure of our own.  (This
is not completely true, bernie does send around his "hot bugs" list
for members of the Deployment Support Network. Once the bugs have been
assigned, all review happens on the lists or irc.)    Everything thing
happens in either Sugar Labs, OLPC, Debian, Ubuntu, or Fedora.  This
is intentional.  Rather than ship anything custom, we must push it
upstream.

The key is that we, and others, search for common goals and push them
forward.  A successful open source ecosystem depends on finding
intersections rather than forcing unions.

david