Joerg Schilling wrote:
> An ARC needs to work in a way that makes it accepted by all communities.


If the ARC was made up of core contributers from each of code producing
communities (or consolidations or ...), such that the ARC was a
common gathering place for the intersection of these cross-community
issues, then the ARC would be the place to work out any differences.
This leads directly to ARC decisions that are respected and honored by
all, because the decisions were made by all.

If you instead balkanize things by saying sub-communities can selectively
veto ARC decisions if they don't like them or add constraints without
associated participation and negotiation, the process quickly becomes
meaningless.

As I've said elsewhen, the architecture of OpenSolaris is a shared commons,
one that we /ALL/ need to take responsibility for managing.  There is no such
thing as an isolated sub-community that effects nobody else - because we all
exist as part of the OpenSolaris.org community, at some time our work
product will impact other parts of the community's shared architecture.

The truism of software development is thinking about things earlier is
better than reacting to surprises later.  The ARC exists in a large part
to cause that thinking to happen.  Even so, it can't cure the problem of
people refusing to think.

   -John

Reply via email to